KUBY
Immunology
Icons Used in This Book
Antigenic
peptide
T cell
receptor
Antibody
CD3
Immature
thymocyte
TH cell
CD8
TC cell
Plasma cell
B cell
CD4
Class I MHC
Cytokine
Class II MHC
Cytokine
receptor
Cytotoxic T cell
Bone marrow
stromal cell
Neutrophil
Basophil
Eosinophil
Dendritic cell
Monocyte
Macrophage
Class I MHC
Erythrocyte
Antigen-presenting cell
CD4
Altered self cell
B cell
Platelets
Mast cell
Class II MHC
CD8
TC cell
Natural killer cell
TH cell
KUBY
Immunology
Judith A. Owen
Haverford College
Jenni Punt
Haverford College
Sharon A. Stranford
Mount Holyoke College
with contributions by
Patricia P. Jones
Stanford University
Seventh Edition
W. H. Freeman and Company • New York
Publisher: Susan Winslow
Senior Acquisitions Editor: Lauren Schultz
Associate Director of Marketing: Debbie Clare
Marketing Assistant: Lindsay Neff
Developmental Editor: Erica Champion
Developmental Editor: Irene Pech
Developmental Coordinator: Sara Ruth Blake
Associate Media Editor: Allison Michael
Library of Congress Control Number: 2012950797
North American Edition
Cover image:
©2009 Pflicke and Sixt. Originally published in
The Journal of Experimental Medicine. 206:2925-2935.
doi:10.1084/jem.20091739.
Image provided by Holger Pflicke and Michael Sixt.
International Edition
Cover design: Dirk Kaufman
Cover image: Nastco/iStockphoto.com
Supplements Editor: Yassamine Ebadat
Senior Project Manager at Aptara: Sherrill Redd
Photo Editor: Christine Buese
Photo Researcher: Elyse Reider
Art Director: Diana Blume
Text Designer: Marsha Cohen
Illustrations: Imagineering
Illustration Coordinator: Janice Donnola
North American Edition
ISBN-13: 978-14292-1919-8
ISBN-10: 1-4292-1919-X
International Edition
ISBN-13: 978-14641-3784-6
ISBN-10: 1-4641-3784-6
© 1992, 1994, 1997, 2000, 2003, 2007, 2013
by W. H. Freeman and Company
All rights reserved
Production Coordinator: Lawrence Guerra
Composition: Aptara®, Inc.
Printing and Binding: RR Donnelley
Printed in the United States of America
First printing
North American Edition
W. H. Freeman and Company
41 Madison Avenue
New York, NY 10010
www.whfreeman.com
International Edition
Macmillan Higher Education
Houndmills, Basingstoke
RG21 6XS, England
www.macmillanhighered.com/international
To all our students, fellows, and colleagues who have
made our careers in immunology a source of joy
and excitement, and to our families who made these
careers possible. We hope that future generations of
immunology students will find this subject as fascinating and rewarding as we have.
About the Authors
All four authors are active scholars and teachers who have been/are recipients of research
grants from the NIH and the NSF. We have all served in various capacities as grant proposal
reviewers for NSF, NIH, HHMI, and other funding bodies as well as evaluating manuscripts
submitted for publication in immunological journals. In addition, we are all active members
of the American Association of Immunologists and have served our national organization in
a variety of ways.
Judy Owen holds B.A. and M.A. (Hons) degrees from Cambridge University. She pursued
her Ph.D. at the University of Pennsylvania with the late Dr. Norman Klinman and her postdoctoral fellowship with Dr. Peter Doherty in viral immunology. She was appointed to the
faculty of Haverford College, one of the first undergraduate colleges to offer a course in immunology, in 1981. She teaches numerous laboratory and lecture courses in biochemistry and
immunology and has received several teaching and mentorship awards. She is a participant
in the First Year Writing Program and has been involved in curriculum development across
the College.
Jenni Punt received her A.B. from Bryn Mawr College (magna cum laude) majoring in
Biology at Haverford College, She received her VMD (summa cum laude) and Ph.D. in immunology from the University of Pennsylvania and was a Damon Runyon-Walter Winchell
Physician-Scientist fellow with Dr. Alfred Singer at the National Institutes of Health. She was
appointed to the faculty of Haverford College in 1996 where she teaches cell biology and immunology and performs research in T cell development and hematopoiesis. She has received
several teaching awards and has contributed to the development of college-wide curricular
initiatives.
Together, Jenni Punt and Judy Owen developed and ran the first AAI Introductory Immunology course, which is now offered on an annual basis.
Sharon Stranford obtained her B.A. with Honors in Biology from Arcadia University
and her Ph.D. in Microbiology and Immunology from Hahnemann (now Drexel) University,
where she studied autoimmunity with funding from the Multiple Sclerosis Foundation. She
pursued postdoctoral studies in transplantation immunology at Oxford University in England,
followed by a fellowship at the University of California, San Francisco, working on HIV/AIDS
with Dr. Jay Levy. From 1999 to 2001, Sharon was a Visiting Assistant Professor of Biology at
Amherst College, and in 2001 joined the faculty of Mount Holyoke College as a Clare Boothe
Luce Assistant Professor. She teaches courses in introductory biology, cell biology, immunology, and infectious disease, as well as a new interdisciplinary course called Controversies in
Public Health.
Pat Jones graduated from Oberlin College in Ohio with Highest Honors in Biology and
obtained her Ph.D. in Biology with Distinction from the Johns Hopkins University. She was a
postdoctoral fellow of the Arthritis Foundation for two years in the Department of Biochemistry and Biophysics at the University of California, San Francisco, Medical School, followed
by two years as an NSF postdoctoral fellow in the Departments of Genetics and Medicine/
Immunology at Stanford University School of Medicine. In 1978 she was appointed Assistant
Professor of Biology at Stanford and is now a full professor. Pat has received several undergraduate teaching awards, was the founding Director of the Ph.D. Program in Immunology,
and in July, 2011, she assumed the position of Director of Stanford Immunology, a position
that coordinates activities in immunology across the university.
Contents
Chapter 1
Overview of the Immune System
A Historical Perspective of Immunology
SUMMARY
23
REFERENCES
23
1
USEFUL WEB SITES
23
2
STUDY QUESTIONS
24
Early vaccination studies led the way to immunology
2
Vaccination is an ongoing, worldwide enterprise
3
Chapter 2
Immunology is about more than just vaccines
and infectious disease
4
Immunity involves both humoral and cellular
components
6
How are foreign substances recognized by the
immune system?
Cells, Organs, and Microenvironments of the Immune
System
9
Important Concepts for Understanding
the Mammalian Immune Response
Pathogens come in many forms and must first
breach natural barriers
11
12
The immune response quickly becomes tailored
to suit the assault
12
Pathogen recognition molecules can be encoded
in the germline or randomly generated
14
Tolerance ensures that the immune system avoids
destroying the host
15
The immune response is composed of two
interconnected arms: innate immunity and
adaptive immunity
16
Adaptive immune responses typically generate
memory
17
The Good, Bad, and Ugly of the Immune
System
19
Cells of the Immune System
27
27
Hematopoietic stem cells have the ability to
differentiate into many types of blood cells
28
Hematopoeisis is the process by which hematopoietic
stem cells develop into mature blood cells
32
Cells of the myeloid lineage are the first responders
to infection
32
Cells of the lymphoid lineage regulate the adaptive
immune response
37
Primary Lymphoid Organs—
Where Immune Cells Develop
41
The bone marrow provides niches for hematopoietic
stem cells to self-renew and differentiate into myeloid
cells and B lymphocytes
41
The thymus is a primary lymphoid organ where
T cells mature
41
Secondary Lymphoid Organs—
Where the Immune Response Is Initiated
48
Secondary lymphoid organs are distributed throughout the body and share some anatomical features
48
22
Lymphoid organs are connected to each other and
to infected tissue by two different circulatory
systems: blood and lymphatics
48
22
The lymph node is a highly specialized secondary
lymphoid organ
50
Inappropriate or dysfunctional immune
responses can result in a range of disorders
19
The immune response renders tissue transplantation
challenging
Cancer presents a unique challenge to the immune
response
viii
Contents
The spleen organizes the immune response against
blood-borne pathogens
53
Signal-induced PIP2 breakdown by PLC causes an
increase in cytoplasmic calcium ion concentration
75
MALT organizes the response to antigen that
enters mucosal tissues
53
Ubiquitination may inhibit or enhance signal
transduction
76
The skin is an innate immune barrier and also
includes lymphoid tissue
56
Tertiary lymphoid tissues also organize and maintain
an immune response
57
SUMMARY
60
REFERENCES
Frequently Encountered Signaling
Pathways
77
The PLC pathway induces calcium release and
PKC activation
77
60
The Ras/Map kinase cascade activates transcription
through AP-1
78
USEFUL WEB SITES
61
PKC activates the NF-κB transcription factor
79
STUDY QUESTIONS
61
Receptor-Ligand Interactions
Receptor-ligand binding occurs via multiple
noncovalent bonds
80
Antibodies share a common structure of two light
chains and two heavy chains
81
There are two major classes of antibody light chains
85
There are five major classes of antibody heavy chains
85
66
Antibodies and antibody fragments can serve as
antigens
86
66
Each of the domains of the antibody heavy and
light chains mediate specific functions
88
X-ray crystallography has been used to define
the structural basis of antigen-antibody
binding
90
65
How do we quantitate the strength of receptorligand interactions?
66
Interactions between receptors and ligands can be
multivalent
67
Receptor and ligand expression can vary during the
course of an immune response
Local concentrations of cytokines and other ligands
may be extremely high
Common Strategies Used in Many Signaling
Pathways
Ligand binding can induce conformational changes
in, and/or clustering of, the receptor
80
Antibodies are made up of multiple
immunoglobulin domains
Chapter 3
Receptors and Signaling: B and
T-Cell Receptors
The Structure of Antibodies
Signal Transduction in B Cells
91
68
Antigen binding results in docking of adapter
molecules and enzymes into the BCR-Igα/Igβ
membrane complex
91
68
B cells use many of the downstream signaling
pathways described above
92
69
B cells also receive signals through co-receptors
94
T-Cell Receptors and Signaling
71
95
The T-cell receptor is a heterodimer with variable
and constant regions
95
Some receptors require receptor-associated
molecules to signal cell activation
71
The T-cell signal transduction complex includes CD3
98
Ligand-induced receptor clustering can alter
receptor location
71
The T cell co-receptors CD4 and CD8 also bind
the MHC
99
Tyrosine phosphorylation is an early step in many
signaling pathways
73
Lck is the first tyrosine kinase activated in T cell
signaling
100
Adapter proteins gather members of signaling
pathways
74
T cells use downstream signaling strategies similar
to those of B cells
100
Phosphorylation on serine and threonine residues
is also a common step in signaling pathways
SUMMARY
101
74
REFERENCES
102
USEFUL WEB SITES
102
STUDY QUESTIONS
103
Phosphorylation of membrane phospholipids
recruits PH domain-containing proteins to the cell
membrane
75
Contents
Cytokine storms may have caused many deaths in
the 1918 Spanish influenza
Chapter 4
Receptors and Signaling:
Cytokines and Chemokines
General Properties of Cytokines and
Chemokines
Cytokine-Based Therapies
105
106
REFERENCES
138
USEFUL WEB SITES
139
STUDY QUESTIONS
140
Cytokines have numerous biological functions
107
Chapter 5
Cytokines can elicit and support the activation of
specific T-cell subpopulations
107
Innate Immunity
Cell activation may alter the expression of receptors
and adhesion molecules
109
Signaling through multiple receptors can fine tune
a cellular response
Six Families of Cytokines and Associated
Receptor Molecules
Cytokines of the IL-1 family promote proinflammatory
signals
110
113
Hematopoietin (Class I) family cytokines share
three-dimensional structural motifs, but induce a
diversity of functions in target cells
116
The Interferon (Class II) cytokine family was the
first to be discovered
119
Members of the TNF cytokine family can signal
development, activation, or death
123
The IL-17 family is a recently discovered,
proinflammatory cytokine cluster
127
Chemokines direct the migration of leukocytes
through the body
129
Cytokine Antagonists
Anatomical Barriers to Infection
143
143
Antimicrobial proteins and peptides kill would-be
invaders
145
147
Microbes are recognized by receptors on
phagocytic cells
147
Phagocytosed microbes are killed by multiple
mechanisms
151
Phagocytosis contributes to cell turnover and the
clearance of dead cells
152
Induced Cellular Innate Responses
133
141
Epithelial barriers prevent pathogen entry into the
body’s interior
Phagocytosis
111
137
138
107
110
137
SUMMARY
Cytokines mediate the activation, proliferation,
and differentiation of target cells
Cytokines are concentrated between secreting and
target cells
ix
152
Cellular pattern recognition receptors activate
responses to microbes and cell damage
153
Toll-like receptors recognize many types of
pathogen molecules
153
C-type lectin receptors bind carbohydrates on the
surfaces of extracellular pathogens
158
Retinoic acid-inducible gene-I-like receptors bind
viral RNA in the cytosol of infected cells
160
The IL-1 receptor antagonist blocks the IL-1
cytokine receptor
133
Nod-like receptors are activated by a variety of
PAMPs, DAMPs, and other harmful substances
160
Cytokine antagonists can be derived from cleavage
of the cytokine receptor
134
Expression of innate immunity proteins is induced
by PRR signaling
160
Some viruses have developed strategies to exploit
cytokine activity
134
Cytokine-Related Diseases
Inflammatory Responses
134
166
Inflammation results from innate responses
triggered by infection, tissue damage, or harmful
substances
167
Proteins of the acute phase response contribute
to innate immunity and inflammation
168
Septic shock is relatively common and potentially
lethal
135
Bacterial toxic shock is caused by superantigen
induction of T-cell cytokine secretion
135
Natural Killer Cells
168
Cytokine activity is implicated in lymphoid and
myeloid cancers
137
Regulation and Evasion of Innate and
Inflammatory Responses
169
x
Contents
Innate and inflammatory responses can be harmful
169
Innate and inflammatory responses are regulated
both positively and negatively
172
Complement activity is passively regulated by protein
stability and cell surface composition
210
Pathogens have evolved mechanisms to evade
innate and inflammatory responses
173
The C1 inhibitor, C1INH, promotes dissociation
of C1 components
211
Decay Accelerating Factors promote decay of C3
convertases
211
Interactions Between the Innate and
Adaptive Immune Systems
173
The Regulation of Complement Activity
210
The innate immune system activates and regulates
adaptive immune responses
Factor I degrades C3b and C4b
212
174
Protectin inhibits the MAC attack
213
Adjuvants activate innate immune responses to
increase the effectiveness of immunizations
175
Carboxypeptidases can inactivate the anaphylatoxins,
C3a and C5a
213
Some pathogen clearance mechanisms are common
to both innate and adaptive immune responses
176
Ubiquity of Innate Immunity
176
Complement Deficiencies
213
Microbial Complement Evasion Strategies
214
Plants rely on innate immune responses to combat
infections
177
Some pathogens interfere with the first step of
immunoglobulin-mediated complement activation
215
Invertebrate and vertebrate innate immune
responses show both similarities and differences
177
Microbial proteins bind and inactivate complement
proteins
215
SUMMARY
180
Microbial proteases destroy complement proteins
215
REFERENCES
181
USEFUL WEB SITES
182
Some microbes mimic or bind complement
regulatory proteins
215
STUDY QUESTIONS
182
Chapter 6
The Complement System
187
The Evolutionary Origins of the Complement
System
215
SUMMARY
219
REFERENCES
220
USEFUL WEB SITES
220
STUDY QUESTIONS
221
The Major Pathways of Complement Activation 189
The classical pathway is initiated by antibody binding
190
The lectin pathway is initiated when soluble proteins
recognize microbial antigens
195
The alternative pathway is initiated in three
distinct ways
196
The three complement pathways converge at the
formation of the C5 convertase
The Organization and Expression
of Lymphocyte Receptor Genes 225
200
The Puzzle of Immunoglobulin Gene Structure
C5 initiates the generation of the MAC
200
The Diverse Functions of Complement
201
Complement receptors connect complementtagged pathogens to effector cells
201
Complement enhances host defense against infection
204
Complement mediates the interface between innate
and adaptive immunities
207
Complement aids in the contraction phase of the
immune response
207
Complement mediates CNS synapse elimination
210
Chapter 7
226
Investigators proposed two early theoretical
models of antibody genetics
226
Breakthrough experiments revealed that multiple
gene segments encode the light chain
227
Multigene Organization of Ig Genes
231
Kappa light-chain genes include V, J, and C segments
231
Lambda light-chain genes pair each J segment
with a particular C segment
231
Heavy-chain gene organization includes VH, D, JJ,
and CH segments
232
Contents
The Mechanism of V(D)J Recombination
Recombination is directed by signal sequences
Gene segments are joined by the RAG1/2
recombinase combination
V(D)J recombination results in a functional Ig variable
region gene
V(D)J recombination can occur between
segments transcribed in either the same or opposite
directions
Five mechanisms generate antibody diversity in
naïve B cells
B-Cell Receptor Expression
Allelic exclusion ensures that each B cell synthesizes
only one heavy chain and one light chain
232
233
234
235
239
239
242
242
Receptor editing of potentially autoreactive
receptors occurs in light chains
243
Ig gene transcription is tightly regulated
244
Mature B cells express both IgM and IgD antibodies
by a process that involves mRNA splicing
246
T-Cell Receptor Genes and Expression
247
Understanding the protein structure of the TCR
was critical to the process of discovering the genes
247
The β-chain gene was discovered simultaneously
in two different laboratories
249
A search for the α-chain gene led to the γ-chain
gene instead
250
TCR genes undergo a process of rearrangement
very similar to that of Ig genes
251
TCR expression is controlled by allelic exclusion
253
TCR gene expression is tightly regulated
253
SUMMARY
255
REFERENCES
256
USEFUL WEB SITES
257
STUDY QUESTIONS
258
The Major Histocompatibility
Complex and Antigen
Presentation
The Structure and Function of MHC Molecules
Class I molecules have a glycoprotein heavy chain
and a small protein light chain
Class II molecules have two non-identical
glycoprotein chains
262
Class I and II molecules exhibit polymorphism in
the region that binds to peptides
263
General Organization and Inheritance of
the MHC
261
262
262
267
The MHC locus encodes three major classes of
molecules
268
The exon/intron arrangement of class I and II genes
reflects their domain structure
270
Allelic forms of MHC genes are inherited in linked
groups called haplotypes
270
MHC molecules are codominantly expressed
271
Class I and class II molecules exhibit diversity at
both the individual and species levels
273
MHC polymorphism has functional relevance
276
The Role of the MHC and Expression
Patterns
277
MHC molecules present both intracellular and
extracellular antigens
278
MHC class I expression is found throughout the
body
278
Expression of MHC class II molecules is primarily
restricted to antigen-presenting cells
279
MHC expression can change with changing
conditions
279
T cells are restricted to recognizing peptides
presented in the context of self-MHC alleles
281
Evidence suggests different antigen processing
and presentation pathways
284
The Endogenous Pathway of Antigen
Processing and Presentation
285
Peptides are generated by protease complexes
called proteasomes
285
Peptides are transported from the cytosol to the RER
285
Chaperones aid peptide assembly with MHC
class I molecules
286
The Exogenous Pathway of Antigen
Processing and Presentation
Chapter 8
xi
288
Peptides are generated from internalized molecules
in endocytic vesicles
288
The invariant chain guides transport of class II
MHC molecules to endocytic vesicles
289
Peptides assemble with class II MHC molecules
by displacing CLIP
289
Cross-Presentation of Exogenous Antigens
291
xii
Contents
Dendritic cells appear to be the primary crosspresenting cell type
292
Mechanisms and Functions of Cross-Presentation
292
Presentation of Nonpeptide Antigens
293
SUMMARY
295
REFERENCES
295
USEFUL WEB SITES
296
STUDY QUESTIONS
296
Chapter 9
T-Cell Development
Early Thymocyte Development
299
Apoptosis allows cells to die without triggering
an inflammatory response
318
Different stimuli initiate apoptosis, but all activate
caspases
318
Apoptosis of peripheral T cells is mediated by the
extrinsic (Fas) pathway
320
TCR-mediated negative selection in the thymus
induces the intrinsic (mitochondria-mediated)
apoptotic pathway
321
Bcl-2 family members can inhibit or induce apoptosis
321
SUMMARY
324
REFERENCES
325
USEFUL WEB SITES
326
STUDY QUESTIONS
327
301
Thymocytes progress through four double-negative
stages
301
Chapter 10
Thymocytes can express either TCRαβ or TCRγδ
receptors
302
B-Cell Development
DN thymocytes undergo β-selection, which results
in proliferation and differentiation
303
Positive and Negative Selection
The Site of Hematopoiesis
304
Thymocytes “learn” MHC restriction in the thymus
305
T cells undergo positive and negative selection
305
Positive selection ensures MHC restriction
307
Negative selection (central tolerance) ensures
self-tolerance
310
The selection paradox: Why don’t we delete all cells
we positively select?
312
An alternative model can explain the thymic
selection paradox
313
Do positive and negative selection occur at the
same stage of development, or in sequence?
314
Lineage Commitment
Several models have been proposed to explain
lineage commitment
Double-positive thymocytes may commit to other
types of lymphocytes
Exit from the Thymus and Final Maturation
329
330
The site of B-cell generation changes during gestation
330
Hematopoiesis in the fetal liver differs from that
in the adult bone marrow
332
B-Cell Development in the Bone Marrow
The stages of hematopoiesis are defined by cellsurface markers, transcription-factor expression,
and immunoglobulin gene rearrangements
332
334
The earliest steps in lymphocyte differentiation culminate
in the generation of a common lymphoid progenitor
337
The later steps of B-cell development result in
commitment to the B-cell phenotype
339
Immature B cells in the bone marrow are
exquisitely sensitive to tolerance induction
344
Many, but not all, self-reactive B cells are deleted
within the bone marrow
345
314
B cells exported from the bone marrow are still
functionally immature
345
316
Mature, primary B-2 B cells migrate to the lymphoid
follicles
349
314
316
Other Mechanisms That Maintain Self-Tolerance 316
The Development of B-1 and Marginal-Zone
B Cells
351
TREG cells negatively regulate immune responses
317
B-1 B cells are derived from a separate developmental
lineage
351
Peripheral mechanisms of tolerance also protect
against autoreactive thymocytes
318
Marginal-zone cells share phenotypic and functional
characteristics with B-1 B cells and arise at the T2 stage
352
Apoptosis
318
Comparison of B- and T-Cell Development
352
Contents
SUMMARY
354
REFERENCES
355
USEFUL WEB SITES
355
STUDY QUESTIONS
356
Chapter 12
B-Cell Activation, Differentiation,
and Memory Generation
385
T-Dependent B-Cell Responses
Chapter 11
T-Cell Activation, Differentiation,
and Memory
357
T-Cell Activation and the Two-Signal Hypothesis
358
Costimulatory signals are required for optimal T-cell
activation and proliferation
359
Clonal anergy results if a costimulatory signal is absent
363
Cytokines provide Signal 3
364
Antigen-presenting cells have characteristic
costimulatory properties
365
Superantigens are a special class of T-cell activators
366
T-Cell Differentiation
368
Helper T cells can be divided into distinct subsets
370
The differentiation of T helper cell subsets is regulated
by polarizing cytokines
371
Effector T helper cell subsets are distinguished by
three properties
372
Helper T cells may not be irrevocably committed
to a lineage
378
Helper T-cell subsets play critical roles in immune
health and disease
378
T-Cell Memory
379
xiii
388
T-dependent antigens require T-cell help to
generate an antibody response
388
Antigen recognition by mature B cells provides a
survival signal
389
B cells encounter antigen in the lymph nodes and
spleen
390
B-cell recognition of cell-bound antigen results in
membrane spreading
391
What causes the clustering of the B-cell receptors
upon antigen binding?
392
Antigen receptor clustering induces internalization
and antigen presentation by the B cell
393
Activated B cells migrate to find antigen-specific
T cells
393
Activated B cells move either into the extrafollicular space or into the follicles to form germinal
centers
395
Plasma cells form within the primary focus
395
Other activated B cells move into the follicles and
initiate a germinal center response
396
Somatic hypermutation and affinity selection occur
within the germinal center
398
Class switch recombination occurs within the
germinal center after antigen contact
401
Most newly generated B cells are lost at the end of
the primary immune response
403
Naïve, effector, and memory T cells display broad
differences in surface protein expression
379
Some germinal center cells complete their
maturation as plasma cells
403
TCM and TEM are distinguished by their locale and
commitment to effector function
380
B-cell memory provides a rapid and strong
response to secondary infection
404
How and when do memory cells arise?
380
What signals induce memory cell commitment?
381
Do memory cells reflect the heterogeneity of
effector cells generated during a primary response?
381
Are there differences between CD4+ and CD8+
memory T cells?
381
How are memory cells maintained over many years?
381
SUMMARY
381
REFERENCES
382
USEFUL WEB SITES
383
STUDY QUESTIONS
383
T-Independent B-Cell Responses
406
T-independent antigens stimulate antibody
production without the need for T-cell help
406
Two novel subclasses of B cells mediate the response
to T-independent antigens
407
Negative Regulation of B Cells
411
Negative signaling through CD22 shuts down
unnecessary BCR signaling
411
Negative signaling through the FcγRIIb receptor
inhibits B-cell activation
411
B-10 B cells act as negative regulators by
secreting IL-10
411
xiv
Contents
SUMMARY
412
REFERENCES
413
USEFUL WEB SITES
414
STUDY QUESTIONS
414
Naïve lymphocytes sample stromal cells in the
lymph nodes
461
Naïve lymphocytes browse for antigen along
reticular networks in the lymph node
461
Immune Cell Behavior during the Innate
Immune Response
464
Chapter 13
Antigen-presenting cells travel to lymph nodes
and present processed antigen to T cells
465
Effector Responses: Cell-and
Antibody-Mediated Immunity
Unprocessed antigen also gains access to lymphnode B cells
465
Antibody-Mediated Effector Functions
Antibodies mediate the clearance and destruction
of pathogen in a variety of ways
Antibody isotypes mediate different effector functions
Fc receptors mediate many effector functions of
antibodies
Cell-Mediated Effector Responses
415
416
Immune Cell Behavior during the Adaptive
Immune Response
467
+
416
419
423
427
Naïve CD4 T cells arrest their movements after
engaging antigens
468
+
B cells seek help from CD4 T cells at the border
between the follicle and paracortex of the Lymph Node
468
Dynamic imaging approaches have been used to
address a controversy about B-cell behavior in
germinal centers
470
+
Cytotoxic T lymphocytes recognize and kill infected
or tumor cells via T-cell receptor activation
428
CD8 T cells are activated in the lymph node via a
multicellular interaction
471
Natural killer cells recognize and kill infected cells and
tumor cells by their absence of MHC class I
435
Activated lymphocytes exit the lymph node and
recirculate
472
A summary of our current understanding
472
The immune response contracts within 10 to 14 days
474
NKT cells bridge the innate and adaptive immune
systems
441
Experimental Assessment of Cell-Mediated
Cytotoxicity
444
Co-culturing T cells with foreign cells stimulates
the mixed-lymphocyte reaction
444
Chemokine receptors and integrins regulate homing
of effector lymphocytes to peripheral tissues
474
CTL activity can be demonstrated by cell-mediated
lympholysis
445
Effector lymphocytes respond to antigen in
multiple tissues
475
The graft-versus-host reaction is an in vivo indication
of cell-mediated cytotoxicity
446
SUMMARY
480
REFERENCES
481
SUMMARY
446
USEFUL WEB SITES
482
REFERENCES
447
STUDY QUESTIONS
482
USEFUL WEB SITES
448
STUDY QUESTIONS
448
Immune Cell Behavior in Peripheral
Tissues
474
Chapter 15
Chapter 14
The Immune Response
in Space and Time
Immune Cell Behavior before Antigen
Is Introduced
Naïve lymphocytes circulate between secondary
and tertiary lymphoid tissues
451
455
455
Allergy, Hypersensitivities, and
Chronic Inflammation
485
Allergy: A Type I Hypersensitivity Reaction
486
IgE antibodies are responsible for type I hypersensitivity
487
Many allergens can elicit a type I response
487
IgE antibodies act by cross-linking Fcε receptors on
the surfaces of innate immune cells
487
Contents
IgE receptor signaling is tightly regulated
491
Chapter 16
Innate immune cells produce molecules responsible
for type I hypersensitivity symptoms
491
Type I hypersensitivities are characterized by both
early and late responses
494
Tolerance, Autoimmunity, and
Transplantation
There are several categories of type I hypersensitivity
reactions
494
There is a genetic basis for type I hypersensitivity
497
Diagnostic tests and treatments are available for
type I hypersensitivity reactions
The hygiene hypothesis has been advanced to
explain increases in allergy incidence
Antibody-Mediated (Type II) Hypersensitivity
Reactions
Transfusion reactions are an example of type II
hypersensitivity
498
501
501
Hemolytic disease of the newborn is caused by
type II reactions
503
Hemolytic anemia can be drug induced
504
Immune Complex-Mediated (Type III)
Hypersensitivity
Immune complexes can damage various tissues
Immune complex-mediated hypersensitivity can
resolve spontaneously
505
505
505
517
Establishment and Maintenance of Tolerance
518
Antigen sequestration is one means to protect
self antigens from attack
519
Central tolerance limits development of autoreactive
T cells and B cells
520
Peripheral tolerance regulates autoreactive cells
in the circulation
520
Autoimmunity
501
xv
525
Some autoimmune diseases target specific organs
526
Some autoimmune diseases are systemic
529
Both intrinsic and extrinsic factors can favor
susceptibility to autoimmune disease
531
Several possible mechanisms have been proposed
for the induction of autoimmunity
533
Autoimmune diseases can be treated by general or
pathway-specific immunosuppression
534
Transplantation Immunology
536
Graft rejection occurs based on immunologic
principles
536
Graft rejection follows a predictable clinical course
541
Autoantigens can be involved in immune complexmediated reactions
506
Immunosuppressive therapy can be either general
or target-specific
543
Arthus reactions are localized type III hypersensitivity
reactions
506
Immune tolerance to allografts is favored in certain
instances
545
Delayed-Type (Type IV) Hypersensitivity (DTH)
506
The initiation of a type IV DTH response involves
sensitization by antigen
Some organs are more amenable to clinical
transplantation than others
546
507
SUMMARY
549
The effector phase of a classical DTH response is
induced by second exposure to a sensitizing antigen
507
REFERENCES
550
USEFUL WEB SITES
551
The DTH reaction can be detected by a skin test
508
STUDY QUESTIONS
551
Contact dermatitis is a type IV hypersensitivity response
508
Chronic Inflammation
Infections can cause chronic inflammation
509
509
Chapter 17
There are noninfectious causes of chronic inflammation 510
Infectious Diseases and Vaccines 553
Obesity is associated with chronic inflammation
510
Chronic inflammation can cause systemic disease
510
The Importance of Barriers to Infection and
the Innate Response
554
SUMMARY
513
Viral Infections
555
REFERENCES
515
Many viruses are neutralized by antibodies
556
USEFUL WEB SITES
515
STUDY QUESTIONS
516
Cell-mediated immunity is important for viral
control and clearance
556
xvi
Contents
Viruses employ several different strategies to evade
host defense mechanisms
556
B-cell immunodeficiencies exhibit depressed
production of one or more antibody isotypes
601
Influenza has been responsible for some of the
worst pandemics in history
557
Disruptions to innate components may also impact
adaptive responses
601
Complement deficiencies are relatively common
603
560
Immunodeficiency that disrupts immune regulation
can manifest as autoimmunity
603
Bacteria can evade host defense mechanisms at
several different stages
563
Immunodeficiency disorders are treated by
replacement therapy
604
Tuberculosis is primarily controlled by CD4+ T cells
564
Animal models of immunodeficiency have been
used to study basic immune function
604
Diphtheria can be controlled by immunization with
inactivated toxoid
565
Bacterial Infections
Immune responses to extracellular and intracellular
bacteria can differ
Parasitic Infections
Protozoan parasites account for huge worldwide
disease burdens
A variety of diseases are caused by parasitic
worms (helminths)
Fungal Infections
560
565
565
567
569
Innate immunity controls most fungal infections
569
Immunity against fungal pathogens can be acquired
571
Emerging and Re-emerging Infectious Diseases 571
Some noteworthy new infectious diseases have
appeared recently
572
Diseases may re-emerge for various reasons
573
Vaccines
Protective immunity can be achieved by active
or passive immunization
574
574
Secondary Immunodeficiencies
606
HIV/AIDS has claimed millions of lives worldwide
607
The retrovirus HIV-1 is the causative agent of AIDS
608
HIV-1 is spread by intimate contact with infected
body fluids
610
In vitro studies have revealed the structure and life
cycle of HIV-1
612
Infection with HIV-1 leads to gradual impairment
of immune function
615
Active research investigates the mechanism of
progression to AIDS
616
Therapeutic agents inhibit retrovirus replication
619
A vaccine may be the only way to stop the
HIV/AIDS epidemic
621
SUMMARY
623
REFERENCES
623
USEFUL WEB SITES
624
STUDY QUESTIONS
624
There are several vaccine strategies, each with
unique advantages and challenges
578
Conjugate or multivalent vaccines can improve
immunogenicity and outcome
583
Chapter 19
Adjuvants are included to enhance the immune
response to a vaccine
585
Cancer and the Immune System
627
SUMMARY
586
Terminology and Common Types of Cancer
627
REFERENCES
587
Malignant Transformation of Cells
628
USEFUL WEB SITES
588
DNA alterations can induce malignant transformation
629
STUDY QUESTIONS
588
The discovery of oncogenes paved the way for our
understanding of cancer induction
629
Genes associated with cancer control cell
proliferation and survival
630
Malignant transformation involves multiple steps
633
Chapter 18
Immunodeficiency Disorders
Primary Immunodeficiencies
Combined immunodeficiencies disrupt adaptive
immunity
593
593
597
Tumor Antigens
634
Tumor-specific antigens are unique to tumor cells
636
Tumor-associated antigens are normal cellular
proteins with unique expression patterns
636
Contents
The Immune Response to Cancer
Immunoediting both protects against and
promotes tumor growth
638
639
Key immunologic pathways mediating tumor
eradication have been identified
639
Some inflammatory responses can promote cancer
642
Some tumor cells evade immune recognition
and activation
643
Cancer Immunotherapy
644
xvii
Hemagglutination inhibition reactions are used to detect
the presence of viruses and of antiviral antibodies
658
Bacterial agglutination can be used to detect
antibodies to bacteria
659
Antibody Assays Based on Antigen Binding to
Solid-Phase Supports
659
Radioimmunoassays are used to measure the
concentrations of biologically relevant proteins
and hormones in bodily fluids
659
Monoclonal antibodies can be targeted to tumor cells
644
ELISA assays use antibodies or antigens covalently
bound to enzymes
660
Cytokines can be used to augment the immune
response to tumors
646
The design of an ELISA assay must consider various
methodological options
662
Tumor-specific T cells can be expanded and
reintroduced into patients
647
ELISPOT assays measure molecules secreted by
individual cells
663
New therapeutic vaccines may enhance the anti-tumor
immune response
647
Western blotting can identify a specific protein
in a complex protein mixture
664
Methods to Determine the Affinity of AntigenAntibody Interactions
664
Manipulation of costimulatory signals can improve
cancer immunity
647
Combination cancer therapies are yielding
surprising results
648
SUMMARY
649
REFERENCES
650
USEFUL WEB SITES
650
STUDY QUESTIONS
651
Chapter 20
Experimental Systems
and Methods
Antibody Generation
Polyclonal antibodies are secreted by multiple clones
of antigen-specific B cells
A monoclonal antibody is the product of a single
stimulated B cell
Monoclonal antibodies can be modified for use in
the laboratory or the clinic
Immunoprecipitation- Based Techniques
653
654
654
654
655
656
Immunoprecipitation can be performed in solution
656
Immunoprecipitation of soluble antigens can be
performed in gel matrices
656
Immunoprecipitation allows characterization of
cell-bound molecules
657
Agglutination Reactions
Hemagglutination reactions can be used to detect
antigen conjugated to the surface of red blood cells
658
any
658
Equilibrium dialysis can be used to measure
antibody affinity for antigen
665
Surface plasmon resonance is commonly used
for measurements of antibody affinity
667
Microscopic Visualization of Cells and
Subcellular Structures
668
Immunocytochemistry and immunohistochemistry
use enzyme-conjugated antibodies to create images
of fixed tissues
668
Immunoelectron microscopy uses gold beads
to visualize antibody-bound antigens
669
Immunofluorescence-Based Imaging
Techniques
669
Fluorescence can be used to visualize cells
and molecules
669
Immunofl uorescence microscopy uses antibodies
conjugated with fluorescent dyes
669
Confocal fluorescence microscopy provides threedimensional images of extraordinary clarity
670
Multiphoton fluorescence microscopy is a variation
of confocal microscopy
670
Intravital imaging allows observation of immune
responses in vivo
671
Flow Cytometry
672
Magnetic Activated Cell Sorting
677
Cell Cycle Analysis
678
Tritiated (3H) thymidine uptake was one of the
first methods used to assess cell division
678
xviii
Contents
Colorimetric assays for cell division are rapid and
eliminate the use of radioactive isotopes
678
Bromodeoxyuridine-based assays for cell division use
antibodies to detect newly synthesized DNA
678
Propidium iodide enables analysis of the cell cycle
status of cell populations
Carboxyfluorescein succinimidyl ester can be used
to follow cell division
Assays of Cell Death
678
679
679
The 51Cr release assay was the first assay used
to measure cell death
679
Fluorescently labeled annexin V measures phosphatidyl
serine in the outer lipid envelope of apoptotic cells
680
684
Knock-in and knockout technologies replace an
endogenous with a nonfunctional or engineered
gene copy
685
The cre/lox system enables inducible gene deletion in
selected tissues
687
SUMMARY
689
REFERENCES
690
USEFUL WEB SITES
690
STUDY QUESTIONS
691
Appendix I
The TUNEL assay measures apoptotically generated
DNA fragmentation
680
Caspase assays measure the activity of enzymes
involved in apoptosis
681
Biochemical Approaches Used to Elucidate
Signal Transduction Pathways
681
Biochemical inhibitors are often used to identify
intermediates in signaling pathways
681
Many methods are used to identify proteins
that interact with molecules of interest
682
Whole Animal Experimental Systems
Transgenic animals carry genes that have been
artificially introduced
CD Antigens
A-1
Appendix II
Cytokines
B-1
Appendix III
682
Animal research is subject to federal guidelines
that protect nonhuman research subjects
682
Inbred strains can reduce experimental variation
683
Congenic resistant strains are used to study the
effects of particular gene loci on immune
responses
684
Adoptive transfer experiments allow in vivo
examination of isolated cell populations
684
Chemokines and Chemokine
Receptors
Glossary
Answers to Study Questions
Index
C-1
G-1
AN-1
I-1
Feature Boxes in Kuby 7e
Clinical Focus
Classic Experiment
Box 1.1
Box 2.1
Box 2.3
Box 3.1
Box 3.3
Box 6.1
Box 7.1
Box 1.2
Box 1.3
Box 2.2
Box 3.2
Box 4.2
Box 4.4
Box 5.2
Box 6.2
Box 7.3
Box 8.2
Box 8.4
Box 9.2
Box 9.3
Box 10.1
Box 11.2
Box 11.4
Box 13.1
Box 15.2
Box 15.3
Box 16.1
Box 16.2
Box 16.4
Box 17.1
Box 18.1
Box 19.1
Vaccine Controversy: What’s Truth and What’s
Myth? p. 5
Passive Antibodies and the Iditarod p. 8
The Hygiene Hypothesis p. 20
Stem Cells—Clinical Uses and Potential p. 42
Defects in the B-Cell Signaling Protein Btk Lead
to X-Linked Agammaglobulinemia p. 93
Therapy with Interferons p. 120
Cytokines and Obesity p. 136
Genetic Defects in Components of Innate
and Inflammatory Responses Associated with
Disease p. 170
The Complement System as a Therapeutic
Target p. 208
Some Immunodeficiencies Result from Impaired
Receptor Gene Recombination p. 255
MHC Alleles and Susceptibility to Certain
Diseases p. 277
Deficiencies in TAP Can Lead to Bare
Lymphocyte Syndrome p. 287
How Do T Cells That Cause Type 1 Diabetes
Escape Negative Selection? p. 311
Failure of Apoptosis Causes Defective
Lymphocyte Homeostasis p. 322
B-Cell Development in the Aging
Individual p. 333
Costimulatory Blockade p. 364
What a Disease Reveals about the Physiological
Role of TH17 Cells p. 376
Monoclonal Antibodies in the Treatment of
Cancer p. 420
The Genetics of Asthma and Allergy p. 498
Type 2 Diabetes, Obesity, and
Inflammation p. 511
It Takes Guts to Be Tolerant p. 523
Why Are Women More Susceptible Than Men
to Autoimmunity? Gender Differences in
Autoimmune Disease p. 528
Is There a Clinical Future for
Xenotransplantation? p. 548
The 1918 Pandemic Influenza Virus: Should It
Publish or Perish? p. 557
Prevention of Infant HIV Infection by AntiRetroviral Treatment p. 610
A Vaccine to Prevent Cervical Cancer, and
More p. 637
Evolution
Box 2.4
Box 5.3
Box 7.2
Box 8.1
Variations on Anatomical Themes p. 57
Plant Innate Immune Responses p. 178
Evolution of Recombined Lymphocyte
Receptors p. 240
The Sweet Smell of Diversity p. 275
Box 8.3
Box 9.1
Box 10.3
Box 11.1
Box 12.1
Box 13.2
Box 15.1
Box 16.3
Isolating Hematopoietic Stem Cells p. 29
The Discovery of a Thymus—and Two p. 46
The Elucidation of Antibody Structure p. 82
The Discovery of the T-Cell Receptor p. 96
The Discovery of Properdin p. 198
Hozumi and Tonegawa’s Experiment: DNA
Recombination Occurs in immunoglobulin
Genes in Somatic Cells p. 227
Demonstration of the Self-MHC Restriction of
CD8⫹ T Cells p. 282
Insights about Thymic Selection from the First
TCR Transgenic Mouse Have Stood the Test of
Time p. 308
The Stages of B-Cell Development:
Characterization of the Hardy Fractions p. 342
Discovery of the First Costimulatory Receptor:
CD28 p. 362
Experimental Proof That Somatic
Hypermutation and Antigen- Induced Selection
Occurred Within the Germinal Centers p. 399
Rethinking Immunological Memory: NK Cells
Join Lymphocytes as Memory-Capable
Cells p. 442
The Discovery and Identification of IgE as the
Carrier of Allergic Hypersensitivity p. 488
Early Life Exposure to Antigens Favors Tolerance
Induction p. 546
Advances
Box 4.1
Box 4.3
Box 5.1
Box 6.3
Box 10.2
Box 11.3
Box 12.2
Box 14.1
Box 14.2
Box 17.2
Box 20.1
Methods Used to Map the Secretome p. 111
How Does Chemokine Binding to a Cell-Surface
Receptor Result in Cellular Movement Along the
Chemokine Gradient? p. 130
Inflammasomes p. 162
Staphylococcus aureus Employs Diverse Methods
to Evade Destruction by the Complement
System p. 216
The Role of miRNAs in the Control of B-Cell
Development p. 336
How Many TCR Complexes Must Be Engaged to
Trigger T-Cell Activation? p. 368
New Ideas on B-Cell Help: Not All Cells That Help
B Cells Make Antibodies Are T Cells p. 408
Dynamic Imaging Techniques p. 452
Molecular Regulation of Cell Migration Between
and Within Tissues p. 456
A Prime and Pull Vaccine Strategy for Preventing
Sexually Transmitted Diseases p. 586
Flow Cytometry Under the Hood p. 674
Preface
Like all of the previous authors of this book, we are dedicated
to the concept that immunology is best taught and learned in
an experimentally-based manner, and we have retained that
emphasis with this edition. It is our goal that students should
complete an immunology course not only with a firm grasp
of content, but also with a clear sense of how key discoveries
were made, what interesting questions remain, and how they
might best be answered. We believe that this approach ensures
that students both master fundamental immunological concepts and internalize a vision of immunology as an active and
ongoing process. Guided by this vision, the new edition has
been extensively updated to reflect the recent advances in all
aspects of our discipline.
New Authorship
As a brand-new team of authors, we bring experience in both
research and undergraduate teaching to the development of this
new edition, which continues to reflect a dedication to pedagogical excellence originally modeled by Janis Kuby. We remain
deeply respectful of Kuby’s unique contribution to the teaching
of immunology and hope and trust that this new manifestation
of her creation will simply add to her considerable legacy.
2a
1
Lymph
node
P
3
P
P
B
B
T
T
5a
T
B
4
5b
Memory
2b
N
B
T
OVERVIEW FIGURE 1-9 Collaboration between innate and
adaptive immunity in resolving an infection.
A new capstone chapter (Chapter 14) integrates the events
of an immune response into a complete story, with particular reference to the advanced imaging techniques that have
become available since the writing of the previous edition.
In this way, the molecular and cellular details presented in
Chapters 2-13 are portrayed in context, a moving landscape
of immune response events in time and space (Figure 14-5).
Understanding
Immunology As a Whole
We recognize that the immune system is an integrated network
of cells, molecules, and organs, and that each component relies
on the rest to function properly. This presents a pedagogical
challenge because to understand the whole, we must attain
working knowledge of many related pieces of information,
and these do not always build upon each other in simple linear fashion. In acknowledgment of this challenge, this edition
presents the “big picture” twice; first as an introductory overview to immunity, then, thirteen chapters later, as an integration of the details students have learned in the intervening text.
Specifically, Chapter 1 has been revised to make it more
approachable for students who are new to immunology. The
chapter provides a short historical background to the field and
an introduction to some of the key players and their roles in
the immune response, keeping an eye on fundamental concepts (Overview Figure 1-9). A new section directly addresses
some of the biggest conceptual hurdles, but leaves the cellular
and molecular details for later chapters.
FIGURE 14-5 A T cell (blue) on a fibroblastic reticular network
(red and green) in the lymph node.
xxi
Preface
Focus on the Fundamentals
The order of chapters in the seventh edition has been revised
to better reflect the sequence of events that occurs naturally
during an immune response in vivo. This offers instructors
the opportunity to lead their students through the steps of
an immune response in a logical sequence, once they have
learned the essential features of the tissues, cells, molecular
structures, ligand-receptor binding interactions, and signaling pathways necessary for the functioning of the immune
system. The placement of innate immunity at the forefront
of the immune response enables it to take its rightful place
as the first, and often the only, aspect of immunity that an
organism needs to counter an immune insult. Similarly, the
chapter on complement is located within the sequence in a
place that highlights its function as a bridge between innate
and adaptive immune processes. However, we recognize that
a course in immunology is approached differently by each
instructor. Therefore, as much as possible, we have designed
each of the chapters so that it can stand alone and be offered
in an alternative order.
signaling, as well as to specific molecules and pathways
involved in signaling through antigen receptors. Chapter
4 includes a more thorough introduction to the roles of
cytokines and chemokines in the immune response.
• An expanded and updated treatment of innate immunity
(Chapter 5), which now includes comprehensive coverage of the many physical, chemical, and cellular defenses
that constitute the innate immune system, as well as
the ways in which it activates and regulates adaptive
immunity.
• Substantial rewriting of chapters concerned with
complement (Chapter 6) and antigen receptor gene
rearrangement (Chapter 7). These chapters have been extensively revised for clarity in both text and figures. The
description of the complement system has been updated
to include the involvement of complement proteins in
both innate and adaptive aspects of immunity.
• A restructured presentation of the MHC, with the addition of new information relevant to cross-presentation
pathways (Chapter 8) (Figure 8-22b).
(b) DC cross-presentation and activation of CTL
Challenging All Levels
While this book is written as a text for students new to immunology, it is also our intent to challenge students to reach
deeply into the field and to appreciate the connections with
other aspects of biology. Instead of reducing difficult topics to vague and simplistic forms, we instead present them
with the level of detail and clarity necessary to allow the
beginning student to find and understand information they
may need in the future. This offers the upper level student
a foundation from which they can progress to the investigation of advances and controversies within the current
immunological literature. Supplementary focus boxes have
been used to add nuance or detail to discussions of particular experiments or ideas without detracting from the flow
of information. These boxes, which address experimental
approaches, evolutionary connections, clinical aspects, or
advanced material, also allow instructors to tailor their use
appropriately for individual courses. They provide excellent
launching points for more intensive in class discussions
relevant to the material.
Some of the most visible changes and improvements
include:
• A rewritten chapter on the cells and organs of the immune system (Chapter 2) that includes up to date images
reflecting our new understanding of the microenvironments where the host immune system develops and
responds.
• The consolidation of signaling pathways into two
chapters: Chapter 3 includes a basic introduction to
ligand:receptor interactions and principles of receptor
Cross-presenting
dendritic cell
Exogenous
antigen
TLR
Crossover
pathway
Class I MHC
CD8
CD3
CD80/CD86
CD28
IL-2
Naïve
TC cell
FIGURE 8-22b Exogenous antigen activation of naïve Tc cells requires DC licensing and cross-presentation
• The dedication of specialized chapters concerned with
T cell development and T cell activation (Chapters 9
and Chapter 11, respectively). Chapter 11 now includes
current descriptions of the multiple helper T cell subsets
that regulate the adaptive immune response.
• Substantially rewritten chapters on B cell development
and B cell activation (Chapters 10 and 12, respectively)
that address the physiological locations as well as the nature of the interacting cells implicated in these processes.
• An updated discussion of the role of effector cells and
molecules in clearing infection (Chapter 13), including a
more thorough treatment of NK and NKT cells.
xxii
Preface
• A new chapter that describes advances in understanding and visualizing the dynamic behavior and activities of immune cells in secondary and tertiary tissue
(Chapter 14).
• Substantial revision and updating of the clinical chapters
(Chapters 15-19) including the addition of several new
clinically relevant focus boxes.
• Revised and updated versions of the final methods chapter (Chapter 20), and the appendices of CD antigens,
chemokines, and cytokines and their receptors.
Throughout the book, we attempt to provide a “big picture”
context for necessary details in a way that facilitates greater
student understanding.
Recent Advances and Other Additions
Immunology is a rapidly growing field, with new discoveries,
advances in techniques, and previously unappreciated connections coming to light every day. The 7th edition has been
thoroughly updated throughout, and now integrates the following new material and concepts:
• New immune cell types and subtypes, as well as the
phenotypic plasticity that is possible between certain
subtypes of immune cells.
• A greater appreciation for the wide range of mechanisms
responsible for innate immunity and the nature and roles
of innate responses in sensing danger, inducing inflammation, and shaping the adaptive response (Figure 5-18).
2
T cell
Inhibitory
cytokines
T cell
TGF
1
Cytokine
deprivation
IL–2R
FoxP3
3
TCR
APC
Inhibiting antigen
presenting cells
MHC
4
Cytotoxicity
T cell
T cell
FIGURE 9-10 How regulatory T cells inactivate traditional T cells.
• The roles of the microbiome and commensal organisms in
the development and function of immunity, as well as the
connections between these and many chronic diseases.
• A new appreciation for the micro environmental
substructures that guide immune cell interactions with
antigen and with one another (Figure 14-11a).
Antigen delivery to T cells
Subcapsular
sinus (SCS)
Lymph node
DC presenting
antigen
Afferent
lymphatic
Antigen
B cell follicle
Bacteria
Naïve
TLR4 or
TLR5
Dectin-1
TH1
IFN-γ
IL-12
TLR3, 7, 9
Fungi
IL-6
IL-23
Virus
Naïve
TH17
IL-17
Tricellular complex
(CD8+ T cell, CD4+
T cell, and DC)
FRC network
T cell zone
(paracortex)
FIGURE 14-11a How antigen travels into a lymph node.
IL-10
TLR2/1
Naïve
Helminth
TH2
IL-4
IL-5
IL-13
TLR2/6
Fungi
IL-10
RA
TGF-β
CD28
CD80/86
TCR
MHC II with peptide
Naïve
Treg
IL-10
TGF-β
FIGURE 5-18 Differential signaling through dendritic cell PRRs
influences helper T cell functions.
• Regulation of immunity, including new regulatory cell
types, immunosuppressive chemical messengers and
the roles these play, for example, in tolerance and in
the nature of responses to different types of antigens
(Figure 9-10).
• Many technical advances, especially in the areas of imaging and sequencing, which have collectively enhanced
our understanding of immune function and cellular
interactions, allowing us to view the immune response
in its natural anatomical context, and in real time (see
Figure 14-5).
Connections to the Bench,
the Clinic, and Beyond
We have made a concerted effort in the 7th edition to integrate
experimental and clinical aspects of immunology into the
text. In Chapter 2, illustrations of immune cells and tissues are
shown alongside histological sections or, where possible, electron
Preface
micrographs, so students can see what they actually look like.
Throughout the text, experimental data are used to demonstrate the bases for our knowledge (Figure 3-4b), and the
clinical chapters at the end of the book (Chapters 15 through
19) describe new advances, new challenges, and newly appreciated connections between the immune system and disease.
FIGURE 3-4b Targeted delivery of cytokines (pink).
Featured Boxes
Associated with each chapter are additional boxed materials
that provide specialized information on historically-important
studies (Classic Experiments) that changed the way immunologists viewed the field, noteworthy new breakthroughs
(Advances) that have occurred since the last edition, the
clinical relevance of particular topics (Clinical Focus) and
the evolution of aspects of immune functioning (Evolution).
Examples of such boxes are “The Prime and Pull Vaccine
strategy,” “Genetic defects in components of innate and inflammatory responses associated with disease,” “The role of
miRNAs in the control of B cell development” and an updated
“Stem cells: Clinical uses and potential.” We have involved our
own undergraduate students in the creation of some of these
boxes, which we believe have greatly benefitted from their
perspective on how to present interesting material effectively
to their fellow students.
Critical Thinking and Data Analysis
Integration of experimental evidence throughout the book
keeps students focused on the how and why. Detailed and
clear descriptions of the current state of the field provide
students with the knowledge, skills, and vocabulary to read
critically in the primary literature. Updated and revised study
questions at the end of the chapter range from simple recall of
information to analyzing original data or proposing hypotheses to explain remaining questions in the field. Classic Experiment boxes throughout the text help students to appreciate
the seminal experiments in immunology and how they were
conducted, providing a bridge to the primary research articles
and emphasizing data analysis at every step.
Media and Supplements
NEW! ImmunoPortal (courses.bfwpub.com/immunology7e)
This comprehensive and robust online teaching and
learning tool combines a wealth of media resources, vigorous
xxiii
assessment, and helpful course management features into one
convenient, fully customizable space.
ImmunoPortal Features:
NEW! Kuby Immunology Seventh Edition e-Book—also
available as a standalone resource (ebooks.bfwpub.com/
immunology7e)
This online version of the textbook combines the contents
of the printed book, electronic study tools, and a full complement of student media, including animations and videos.
Students can personalize their e-Book with highlighting,
bookmarking, and note-taking features. Instructors can customize the e-Book to focus on specific sections, and add their
own notes and files to share with their class.
NEW! LearningCurve—A Formative
Quizzing Engine
With powerful adaptive quizzing, a game-like format, and the
promise of significantly better grades, LearningCurve gives
instructors a quickly implemented, highly effective new way
to get students more deeply involved in the classroom. Developed by experienced teachers and experts in educational
technology, LearningCurve offers a series of brief, engaging
activities specific to your course. These activities put the concept of “testing to learn” into action with adaptive quizzing
that treats each student as an individual with specific needs:
• Students work through LearningCurve activities one
question at a time.
• With each question, students get immediate feedback.
Responses to incorrect answers include links to book
sections and other resources to help students focus on
what they need to learn.
• As they proceed toward completion of the activity, the
level of questioning adapts to the level of performance.
The questions become easier, harder, or the same depending on how the student is doing.
• And with a more confident understanding of assigned
material, students will be more actively engaged during
classtime.
Resources
The Resources center provides quick access to all instructor
and student resources for Kuby Immunology.
For Instructors—
All instructor media are available in the ImmunoPortal and on
the Instructor Resource DVD.
NEW! test bank—over 500 dynamic questions in PDF and
editable Word formats include multiple-choice and shortanswer problems, rated by level of difficulty and Bloom’s
Taxonomy level.
xxiv
Acknowledgements
Fully optimized JPEG files of every figure, photo, and table
in the text, featuring enhanced color, higher resolution, and
enlarged fonts. Images are also offered in PowerPoint® format
for each chapter.
Animations of complex text concepts and figures help students better understand key immunological processes.
Videos specially chosen by the authors to complement and
supplement text concepts.
For Students—
All of these resources are also available in the ImmunoPortal.
• Student versions of the Animations and Videos, to help
students understand key mechanisms and techniques at
their own pace.
• Flashcards test student mastery of vocabulary and allow
students to tag the terms they’ve already learned.
• Immunology on the Web weblinks introduce students to
a world of online immunology resources and references.
Assignments
In this convenient space, ImmunoPortal provides instructors with the ability to assign any resource, as well as
e-Book readings, discussion board posts, and their own
materials. A gradebook tracks all student scores and can
be easily exported to Excel or a campus Course Management System.
Acknowledgements
We owe special thanks to individuals who offered insightful
ideas, who provided detailed reviews that led to major improvements, and who provided the support that made writing this
text possible. These notable contributors include Dr. Stephen
Emerson, Dr. David Allman, Dr. Susan Saidman, Dr. Nan
Wang, Nicole Cunningham, and the many undergraduates who
provided invaluable students’ perspectives on our chapters. We
hope that the final product reflects the high quality of the input
from these experts and colleagues and from all those listed
below who provided critical analysis and guidance.
We are also grateful to the previous authors of Kuby’s Immunology, whose valiant efforts we now appreciate even more
deeply. Their commitment to clarity, to providing the most
current material in a fast moving discipline, and to maintaining the experimental focus of the discussions set the standard
that is the basis for the best of this text.
We also acknowledge that this book represents the work
not only of its authors and editors, but also of all those whose
experiments and writing provided us with ideas, inspiration
and information. We thank you and stress that all errors and
inconsistencies of interpretation are ours alone.
We thank the following reviewers for their comments and
suggestions about the manuscript during preparation of this
seventh edition. Their expertise and insights have contributed
greatly to the book.
Lawrence R. Aaronson, Utica College
Jeffrey K. Actor, University of Texas Medical School at Houston
Richard Adler, University of Michigan-Dearborn
Emily Agard, York University, North York
Karthik Aghoram, Meredith College
Rita Wearren Alisauskas, Rutgers University
John Allsteadt, Virginia Intermont College
Gaylene Altman, University of Washington
Angelika Antoni, Kutztown University
Jorge N. Artaza, Charles R. Drew University of Medicine
and Science
Patricia S. Astry, SUNY Fredonia
Roberta Attanasio, Georgia State University
Elizabeth Auger, Saint Joseph’s College of Maine
Avery August, Penn State University
Rajeev Aurora, Saint Louis University Hospital
Christine A. Bacon, Bay Path College
Jason C. Baker, Missouri Western State College
Kenneth Balazovich, University of Michigan-Dearborn
Jennifer L. Bankers-Fulbright, Augsburg College
Amorette Barber, Longwood University
Brianne Barker, Hamilton College
Scott R. Barnum, University of Alabama at Birmingham
Laura Baugh, University of Dallas
Marlee B. Marsh, Columbia College
Rachel Venn Beecham, Mississippi Valley State University
Fabian Benencia, Ohio University Main Campus
Charlie Garnett Benson, Georgia State University
Daniel Bergey, Black Hills State University
Carolyn A. Bergman, Georgian Court College
Elke Bergmann-Leitner, WRAIR/Uniformed Services
University of Health Services
Acknowledgements
Brian P. Bergstrom, Muskingum College
Susan Bjerke, Washburn University of Topeka
Earl F. Bloch, Howard University
Elliott J. Blumenthal, Indiana University–Purdue University
Fort Wayne
Kathleen Bode, Flint Hills Technical College
Dennis Bogyo, Valdosta State University
Mark Bolyard, Union University
Lisa Borghesi, University of Pittsburgh
Phyllis C. Braun, Fairfield University
Jay H. Bream, Johns Hopkins University School of Medicine
Heather A. Bruns, Ball State University
Walter J. Bruyninckx, Hanover College
Eric L. Buckles, Dillard University
Sandra H. Burnett, Brigham Young University
Peter Burrows, University of Alabama at Birmingham
Ralph Butkowski, Augsburg College
Jean A. Cardinale, Alfred University
Edward A. Chaperon, Creighton University
Stephen K. Chapes, Kansas State University
Christopher Chase, South Dakota State University
Thomas Chiles, Boston College
Harold Chittum, Pikeville College
Peter A. Chung, Pittsburg State University
Felicia L. Cianciarulo, Carlow University
Bret A. Clark, Newberry College
Patricia A. Compagnone-Post, Albertus Magnus College
Yasemin Kaya Congleton, Bluegrass Community and
Technical College
Vincent A. Connors, University of South CarolinaSpartanburg
Conway-Klaassen, University of Minnesota
Lisa Cuchara, Quinnipiac University
Tanya R. Da Sylva, York University, North York
Kelley L. Davis, Nova Southeastern University
Jeffrey Dawson, Duke University
Joseph DeMasi, Massachusetts College of Pharmacy & Allied
Health
Stephanie E. Dew, Centre College
Joyce E. S. Doan, Bethel University
Diane Dorsett, Georgia Gwinnett College
James R. Drake, Albany Medical College
Erastus C. Dudley, Huntingdon College
Jeannine M. Durdik, University of Arkansas Fayetteville
Karen M. Duus, Albany Medical College
Christina K. Eddy, North Greenville University
Anthony Ejiofor, Tennessee State University
Jennifer Ellington, Belmont Abbey College
Samantha L. Elliott, Saint Mary’s College of Maryland
Lehman L. Ellis, Our Lady of Holy Cross College
Sherine F. Elsawa, Northern Illinois University
xxv
Uthayashanker Ezekiel, Saint Louis University Medical Center
Diana L. Fagan, Youngstown State University
Rebecca V. Ferrell, Metropolitan State College of Denver
Ken Field, Bucknell University
Krista Fischer-Stenger, University of Richmond
Howard B. Fleit, SUNY at Stony Brook
Sherry D. Fleming, Kansas State University
Marie-dominique Franco, Regis University
Joel Gaikwad, Oral Roberts University
D. L. Gibson, University of British Columbia-Okanagan
Laura Glasscock, Winthrop University
David Glick, Kings College
Elizabeth Godrick, Boston University
Karen Golemboski, Bellarmine University
Sandra O. Gollnick, SUNY Buffalo
James F. Graves, University of Detroit-Mercy
Demetrius Peter Gravis, Beloit College
Anjali D. Gray, Lourdes University
Valery Z. Grdzelishvili, University of North Carolina-Charlotte
Carla Guthridge, Cameron University
David J. Hall, Lawrence University
Sandra K. Halonen, Montana State University
Michael C. Hanna, Texas A & M-Commerce
Kristian M. Hargadon, Hampden-Sydney College
JL Henriksen, Bellevue University
Michelle L. Herdman, University of Charleston
Jennifer L. Hess, Aquinas College
Edward M. Hoffmann, University of Florida
Kristin Hogquist, University of Minnesota
Jane E. Huffman, East Stroudsburg University of Pennsylvania
Lisa A. Humphries, University of California, Los Angeles
Judith Humphries, Lawrence University
Mo Hunsen, Kenyon College
Vijaya Iragavarapu-Charyulu, Florida Atlantic University
Vida R. Irani, Indiana University of Pennsylvania
Christopher D. Jarvis, Hampshire College
Eleanor Jator, Austin Peay State University
Stephen R. Jennings, Drexel University College of Medicine
Robert Jonas, Texas Lutheran University
Vandana Kalia, Penn State University- Main Campus
Azad K. Kaushik, University of Guelph
George Keller, Samford University
Kevin S. Kinney, De Pauw University
Edward C. Kisailus, Canisius College
David J. Kittlesen, University of Virginia
Dennis J. Kitz, Southern Illinois University-Edwardsville
Janet Kluftinger, University of British Columbia -Okanagan
Rolf König, University of Texas Medical Branch at Galveston
Kristine Krafts, University of Minnesota-Duluth
Ruhul Kuddus, Utah Valley University
Narendra Kumar, Texas A&M Health Science Center
xxvi
Acknowledgements
N. M. Kumbaraci, Stevens Institute of Technology
Jesse J. Kwiek, The Ohio State University Main Camp
John M. Lammert, Gustavus Aldolphus College
Courtney Lappas, Lebanon Valley College
Christopher S. Lassiter, Roanoke College
Jennifer Kraft Leavey, Georgia Institute of Technology
Melanie J. Lee-Brown, Guilford College
Vicky M. Lentz, SUNY College at Oneonta
Joseph Lin, Sonoma State University
Joshua Loomis, Nova Southeastern University
Jennifer Louten, Southern Polytechnic State University
Jon H. Lowrance, Lipscomb University
Milson J. Luce, West Virginia University Institute of Technology
Phillip J. Lucido, Northwest Missouri State University
M.E. MacKay, Thompson Rivers University
Andrew P. Makrigiannis, University of Ottawa
Greg Maniero, Stonehill College
David Markwardt, Ohio Wesleyan University
John Martinko, Southern Illinois University
Andrea M. Mastro, Penn State University-Main Campus
Ann H. McDonald, Concordia University
Lisa N. McKernan, Chestnut Hill College
Catherine S. McVay, Auburn University
Daniel Meer, Cardinal Stritch University
JoAnn Meerschaert, Saint Cloud State University
Brian J. Merkel, University Wisconsin-Green Bay
Jiri Mestecky, University of Alabama at Birmingham
Dennis W. Metzger, Albany Medical College
Jennifer A. Metzler, Ball State University
John A. Meyers, Boston University Medical School
Yuko J. Miyamoto, Elon College
Jody M. Modarelli, Hiram College
Devonna Sue Morra, Saint Francis University
Rita B. Moyes, Texas A&M
Annette Muckerheide, College of Mount Saint Joseph
Sue Mungre, Northeastern Illinois University
Kari L. Murad, College of Saint Rose
Karen Grandel Nakaoka, Weber State University
Rajkumar Nathaniel, Nicholls State University
David Nemazee, University of California, San Diego
Hamida Rahim Nusrat, San Francisco State University
Tracy O’Connor, Mount Royal College
Marcos Oliveira, University of the Incarnate Word
Donald Ourth, University of Memphis
Deborah Palliser, Albert Einstein College of Medicine
Shawn Phippen, Valdosta State University
Melinda J. Pomeroy-Black, La Grange College
Edith Porter, California State University, Los Angeles
Michael F. Princiotta, SUNY Upstate Medical University
Gerry A Prody, Western Washington University
Robyn A. Puffenbarger, Bridgewater College
Aimee Pugh-Bernard, University of Colorado at Denver
Pattle Pun, Wheaton College
Sheila Reilly, Belmont Abbey College
Karen A. Reiner, Andrews University
Margaret Reinhart, University of the Sciences in Philadelphia
Stephanie Richards, Bates College
Sarah M. Richart, Azusa Pacific University
James E. Riggs, Rider University
Vanessa Rivera-Amill, Ponce School of Medicine
Katherine Robertson, Westminster College
James L. Rooney, Lincoln University
Robin S. Salter, Oberlin College
Sophia Sarafova, Davidson College
Surojit Sarkar, Penn State University-Main Campus
Perry M. Scanlan, Austin Peay State University
Ralph Seelke, University Wisconsin-Superior
Diane L. Sewell, University Wisconsin-La Crosse
Anding Shen, Calvin College
Penny Shockett, Southeastern Louisiana University
Michael Sikes, North Carolina State University
Maryanne C. Simurda, Washington and Lee University
Paul K. Small, Eureka College
Jonathan Snow, Williams College
Ralph A. Sorensen, Gettysburg College
Andrew W. Stadnyk, Dalhousie University Faculty of Medicine
Douglas A. Steeber, University Wisconsin-Milwaukee
Viktor Steimle, University of Sherbrooke, Sherbrooke
Douglas J. Stemke, University of Indianapolis
Carolyn R. Stenbak, Seattle University
Jennifer Ripley Stueckle, West Virginia University
Kathleen Sullivan, Louisiana Technical College Alexandria
Susmit Suvas, Oakland University
Gabor Szalai, University of South Carolina
Seetha M Tamma, Long Island University-C.W. Post
Matthew J. Temple, Nazareth College
Kent R. Thomas, Wichita State University
Diane G. Tice, SUNY Morrisville
Sara Sybesma Tolsma, Northwestern College
Clara Tóth, Saint Thomas Aquinas College
Bebhinn Treanor, University of Toronto Scarborough
Allen W. Tsang, Bowman Gray Medical School
Amar S. Tung, Lincoln University
Lloyd Turtinen, University Wisconsin-Eau Claire
Timothy M VanWagoner, Oklahoma Christian/ University of
Oklahoma HSC
Evros Vassiliou, Kean University
Vishwanath Venketaraman, Western University of Health
Sciences
Kathleen Verville, Washington College
Katherine A. Wall, University of Toledo
Helen Walter, Mills College
Acknowledgements
Christopher Ward, University of Alberta
Benjamin S. Weeks, Adelphi University
Ben B. Whitlock, University of Saint Francis
Robert Winn, Northern Michigan University
Candace R. Winstead, California Polytechnic State UniversitySan Luis Obispo
Dorothy M. Wrigley, Minnesota State University
Jodi L. Yorty, Elizabethtown College
Sheryl Zajdowicz, Metropolitan State University of Denver
Mary Katherine Zanin, The Citadel The Military College of
South Carolina
Gary Zieve, SUNY at Stony Brook
Michael I. Zimmer, Purdue Calumet
Gilbert L. Zink, University of the Sciences in Philadelphia
Patty Zwollo, College of William & Mary
xxvii
Finally, we thank our experienced and talented colleagues
at W. H. Freeman and Company. Particular thanks to the
production team members Philip McCaffrey, Sherrill Redd,
Heath Lynn Silberfeld, Diana Blume, Lawrence Guerra, Janice
Donnola, Christine Buese, and Elyse Reider. Thanks are also
due to the editorial team of Lauren Schultz, Susan Winslow,
Allison Michael, Yassamine Ebadat, and Irene Pech.
However, a very special thanks go to our developmental
editor, Erica Champion, and our developmental coordinator,
Sara Ruth Blake. Erica has guided us from the beginning with
a probing vision, endless patience, and keen eye for narrative
and clarity. Sara kept us organized and true to deadlines with
heroic resolve. The involvement of these two extraordinarily
talented team members has made this edition, and its ambitious aspirations, possible.
This page lelt intentionally blank.
1
Overview of the
Immune System
T
he immune system evolved to protect
multicellular organisms from pathogens. Highly
adaptable, it defends the body against invaders
as diverse as the tiny (~30 nm), intracellular
virus that causes polio and as large as the giant parasitic
kidney worm Dioctophyme renale, which can grow to
over 100 cm in length and 10 mm in width. This
diversity of potential pathogens requires a range of
recognition and destruction mechanisms to match the
multitude of invaders. To accomplish this feat,
vertebrates have evolved a complicated and dynamic
network of cells, molecules, and pathways. Although
elements of these networks can be found throughout the
plant and animal kingdoms, the focus of this book will
be on the highly evolved mammalian immune system.
The fully functional immune system involves so many
organs, molecules, cells, and pathways in such an
interconnected and sometimes circular process that it is
often difficult to know where to start! Recent advances in
cell imaging, genetics, bioinformatics, as well as cell and
molecular biology, have helped us to understand many of
the individual players in great molecular detail. However,
a focus on the details (and there are many) can make
taking a step back to see the bigger picture challenging,
and it is often the bigger picture that motivates us to study
immunology. Indeed, the field of immunology can be
credited with the vaccine that eradicated smallpox, the
ability to transplant organs between humans, and the
drugs used today to treat asthma. Our goal in this chapter
is therefore to present the background and concepts in
immunology that will help bridge the gap between the
cellular and molecular detail presented in subsequent
chapters and the complete picture of an immune response.
A clear understanding of each of the many players involved
will help one appreciate the intricate coordination of an
immune system that makes all of this possible.
The study of immunology has produced amazing and
fascinating stories (some of which you will see in this
book), where host and microbe engage in battles waged
over both minutes and millennia. But the immune system
is also much more than an isolated component of the
body, merely responsible for search-and-destroy missions.
In fact, it interleaves with many of the other body systems,
A phagocytic cell (macrophage, green)
engulfing the bacteria that cause
tuberculosis (orange). Max Planck Institute for
Infection Biology/Dr. Volker Brinkmann
■
■
■
A Historical Perspective of Immunology
Important Concepts for Understanding the
Mammalian Immune Response
The Good, Bad, and Ugly of the Immune System
including the endocrine, nervous, and metabolic systems,
with more connections undoubtedly to be discovered in
time. Finally, it has become increasingly clear that elements
of immunity play key roles in regulating homeostasis in the
body for a healthy balance. Information gleaned from the
study of the immune system, as well as its connections
with other systems, will likely have resounding
repercussions across many basic science and biomedical
fields, not to mention in the future of clinical medicine.
This chapter begins with a historical perspective,
charting the beginnings of the study of immunology, largely
driven by the human desire to survive major outbreaks of
infectious disease. This is followed by presentation of a few
key concepts that are important hallmarks of the
mammalian immune response, many of which may not
have been encountered elsewhere in basic biology. This is
not meant as a comprehensive overview of the mammalian
immune system but rather as a means for jumping the large
conceptual hurdles frequently encountered as one begins to
describe the complexity and interconnected nature of the
immune response. We hope this will whet the appetite and
prepare the reader for a more thorough discussion of the
specific components of immunity presented in the
1
2
PA R T I
|
Introduction
following chapters. We conclude with a few challenging
clinical situations, such as instances in which the immune
system fails to act or becomes the aggressor, turning its
awesome powers against the host. More in-depth coverage
of these and other medical aspects of immunology can be
found in the final chapters of this book.
A Historical Perspective
of Immunology
The discipline of immunology grew out of the observation
that individuals who had recovered from certain infectious
diseases were thereafter protected from the disease. The
Latin term immunis, meaning “exempt,” is the source of the
English word immunity, a state of protection from infectious disease. Perhaps the earliest written reference to the
phenomenon of immunity can be traced back to Thucydides, the great historian of the Peloponnesian War. In
describing a plague in Athens, he wrote in 430 bc that only
those who had recovered from the plague could nurse the
sick because they would not contract the disease a second
time. Although early societies recognized the phenomenon
of immunity, almost 2000 years passed before the concept
was successfully converted into medically effective practice.
Early Vaccination Studies Led
the Way to Immunology
The first recorded attempts to deliberately induce immunity
were performed by the Chinese and Turks in the fifteenth
century. They were attempting to prevent smallpox, a disease
that is fatal in about 30% of cases and that leaves survivors
disfigured for life (Figure 1-1). Reports suggest that the dried
crusts derived from smallpox pustules were either inhaled or
inserted into small cuts in the skin (a technique called variolation) in order to prevent this dreaded disease. In 1718,
Lady Mary Wortley Montagu, the wife of the British ambassador in Constantinople, observed the positive effects of
variolation on the native Turkish population and had the
technique performed on her own children.
The English physician Edward Jenner later made a giant
advance in the deliberate development of immunity, again
targeting smallpox. In 1798, intrigued by the fact that milkmaids who had contracted the mild disease cowpox were subsequently immune to the much more severe smallpox, Jenner
reasoned that introducing fluid from a cowpox pustule into
people (i.e., inoculating them) might protect them from smallpox. To test this idea, he inoculated an eight-year-old boy with
fluid from a cowpox pustule and later intentionally infected the
child with smallpox. As predicted, the child did not develop
smallpox. Although this represented a major breakthrough, as
one might imagine, these sorts of human studies could not be
conducted under current standards of medical ethics.
Jenner’s technique of inoculating with cowpox to protect
against smallpox spread quickly through Europe. However, it
FIGURE 1-1 African child with rash typical of smallpox on
face, chest, and arms. Smallpox, caused by the virus Variola major,
has a 30% mortality rate. Survivors are often left with disfiguring scars.
[Centers for Disease Control.]
was nearly a hundred years before this technique was applied
to other diseases. As so often happens in science, serendipity
combined with astute observation led to the next major
advance in immunology: the induction of immunity to cholera. Louis Pasteur had succeeded in growing the bacterium
that causes fowl cholera in culture, and confirmed this by
injecting it into chickens that then developed fatal cholera.
After returning from a summer vacation, he and colleagues
resumed their experiments, injecting some chickens with an
old bacterial culture. The chickens became ill, but to Pasteur’s
surprise, they recovered. Interested, Pasteur then grew a fresh
culture of the bacterium with the intention of injecting this
lethal brew into some fresh, unexposed chickens. But as the
story is told, his supply of fresh chickens was limited, and
therefore he used a mixture of previously injected chickens
and unexposed birds. Unexpectedly, only the fresh chickens
died, while the chickens previously exposed to the older
bacterial culture were completely protected from the disease.
Pasteur hypothesized and later showed that aging had weakened the virulence of the pathogen and that such a weakened
or attenuated strain could be administered to provide immunity against the disease. He called this attenuated strain a
Overview of the Immune System
vaccine (from the Latin vacca, meaning “cow”), in honor of
Jenner’s work with cowpox inoculation.
Pasteur extended these findings to other diseases, demonstrating that it was possible to attenuate a pathogen and administer the attenuated strain as a vaccine. In a now classic
experiment performed in the small village of Pouilly-le-Fort in
1881, Pasteur first vaccinated one group of sheep with anthrax
bacteria (Bacillus anthracis) that were attenuated by heat treatment. He then challenged the vaccinated sheep, along with
some unvaccinated sheep, with a virulent culture of the anthrax
bacillus. All the vaccinated sheep lived and all the unvaccinated
animals died. These experiments marked the beginnings of the
discipline of immunology. In 1885, Pasteur administered his
first vaccine to a human, a young boy who had been bitten
repeatedly by a rabid dog (Figure 1-2). The boy, Joseph Meister,
was inoculated with a series of attenuated rabies virus preparations. The rabies vaccine is one of very few that can be successful
when administered shortly after exposure, as long as the virus
has not yet reached the central nervous system and begun to
induce neurologic symptoms. Joseph lived, and later became a
caretaker at the Pasteur Institute, which was opened in 1887 to
treat the many rabies victims that began to flood in when word
of Pasteur’s success spread; it remains to this day an institute
dedicated to the prevention and treatment of infectious disease.
FIGURE 1-2 Wood engraving of Louis Pasteur watching
Joseph Meister receive the rabies vaccine. [Source: From
Harper’s Weekly 29:836; courtesy of the National Library of Medicine.]
|
CHAPTER 1
3
Vaccination Is an Ongoing,
Worldwide Enterprise
The emergence of the study of immunology and the discovery
of vaccines are tightly linked. The development of effective
vaccines for some pathogens is still a major challenge, discussed in greater detail in Chapter 17. However, despite many
biological and social hurdles, vaccination has yielded some of
the most profound success stories in terms of improving mortality rates worldwide, especially in very young children.
In 1977, the last known case of naturally acquired smallpox was seen in Somalia. This dreaded disease was eradicated by universal application of a vaccine similar to that
used by Jenner in the 1790s. One consequence of eradication
is that universal vaccination becomes unnecessary. This is a
tremendous benefit, as most vaccines carry at least a slight
risk to persons vaccinated. And yet in many cases every
individual does not need to be immune in order to protect
most of the population. As a critical mass of people acquire
protective immunity, either through vaccination or infection,
they can serve as a buffer for the rest. This principle, called
herd immunity, works by decreasing the number of individuals who can harbor and spread an infectious agent, significantly decreasing the chances that susceptible individuals
will become infected. This presents an important altruistic
consideration: although many of us could survive infectious
diseases for which we receive a vaccine (such as the flu), this
is not true for everyone. Some individuals cannot receive the
vaccine (e.g., the very young or immune compromised), and
vaccination is never 100% effective. In other words, the susceptible, nonimmune individuals among us can benefit from
the pervasive immunity of their neighbors.
However, there is a darker side to eradication and the end
of universal vaccination. Over time, the number of people
with no immunity to the disease will begin to rise, ending
herd immunity. Vaccination for smallpox largely ended by
the early to mid-1970s, leaving well over half of the current
world population susceptible to the disease. This means that
smallpox, or a weaponized version, is now considered a
potential bioterrorism threat. In response, new and safer
vaccines against smallpox are still being developed today,
most of which go toward vaccinating U.S. military personnel
thought to be at greatest risk of possible exposure.
In the United States and other industrialized nations, vaccines have eliminated a host of childhood diseases that were
the cause of death for many young children just 50 years ago.
Measles, mumps, chickenpox, whooping cough (pertussis),
tetanus, diphtheria, and polio, once thought of as an inevitable part of childhood are now extremely rare or nonexistent in the United States because of current vaccination
practices (Table 1-1). One can hardly estimate the savings to
society resulting from the prevention of these diseases. Aside
from suffering and mortality, the cost to treat these illnesses
and their aftereffects or sequelae (such as paralysis, deafness,
blindness, and mental retardation) is immense and dwarfs
the costs of immunization. In fact, recent estimates suggest
PA R T I
4
TABLE 1-1
|
Introduction
Cases of selected infectious disease in the United States before and after the
introduction of effective vaccines
Disease
Smallpox
ANNUAL CASES/YR
CASES IN 2010
Prevaccine
Postvaccine
48,164
Reduction (%)
0
100
100
Diphtheria
175,885
0
Rubeola (measles)
503,282
26
99.99
Mumps
152,209
2,612
98.28
Pertussis (“whooping cough”)
147,271
27,550
81.29
Paralytic polio
16,316
0
Rubella (German measles)
47,745
5
Tetanus (“lockjaw”)
Invasive Haemophilus influenzae
1,314 (deaths)
20,000
26 (cases)
3,151
100
99.99
98.02
84.25
SOURCE: Adapted from W. A. Orenstein et al., 2005. Health Affairs 24:599 and CDC statistics of Notifiable Diseases.
that significant economic and human life benefits could be
realized by simply scaling up the use of a few childhood vaccines in the poorest nations, which currently bear the brunt
of the impact of these childhood infectious diseases. For
example, it is estimated that childhood pneumonia alone,
caused primarily by vaccine-preventable Streptococcus pneumoniae (aka, pneumococcus) and Haemophilus influenzae
type b (aka, Hib), will account for 2.7 million childhood
deaths in developing nations over the next decade if vaccine
strategies in these regions remain unchanged.
Despite the many successes of vaccine programs, such as
the eradication of smallpox, many vaccine challenges still
remain. Perhaps the greatest current challenge is the design of
effective vaccines for major killers such as malaria and AIDS.
Using our increased understanding of the immune system,
plus the tools of molecular and cellular biology, genomics, and
proteomics, scientists will be better positioned to make progress toward preventing these and other emerging infectious
diseases. A further issue is the fact that millions of children in
developing countries die from diseases that are fully preventable by available, safe vaccines. High manufacturing costs,
instability of the products, and cumbersome delivery problems
keep these vaccines from reaching those who might benefit the
most. This problem could be alleviated in many cases by development of future-generation vaccines that are inexpensive,
heat stable, and administered without a needle. Finally, misinformation and myth surrounding vaccine efficacy and side
effects continues to hamper many potentially life-saving vaccination programs (see Clinical Focus Box on p. 5).
Immunology Is About More Than Just Vaccines
and Infectious Disease
For some diseases, immunization programs may be the best or
even the only effective defense. At the top of this list are infec-
tious diseases that can cause serious illness or even death in
unvaccinated individuals, especially those transmitted by
microbes that also spread rapidly between hosts. However, vaccination is not the only way to prevent or treat infectious disease. First and foremost is preventing infection, where access to
clean water, good hygiene practices, and nutrient-rich diets can
all inhibit transmission of infectious agents. Second, some
infectious diseases are self-limiting, easily treatable, and nonlethal for most individuals, making them unlikely targets for
costly vaccination programs. These include the common cold,
caused by the Rhinovirus, and cold sores that result from Herpes Simplex Virus infection. Finally, some infectious agents are
just not amenable to vaccination. This could be due to a range
of factors, such as the number of different molecular variants
of the organism, the complexity of the regimen required to
generate protective immunity, or an inability to establish the
needed immunologic memory responses (more on this later).
One major breakthrough in the treatment of infectious
disease came when the first antibiotics were introduced in
the 1920s. Currently there are more than a hundred different
antibiotics on the market, although most fall into just six or
seven categories based on their mode of action. Antibiotics
are chemical agents designed to destroy certain types of bacteria. They are ineffective against other types of infectious
agents, as well as some bacterial species. One particularly
worrying trend is the steady rise in antibiotic resistance
among strains traditionally amenable to these drugs, making
the design of next-generation antibiotics and new classes of
drugs increasingly important. Although antiviral drugs are
also available, most are not effective against many of the most
common viruses, including influenza. This makes preventive
vaccination the only real recourse against many debilitating
infectious agents, even those that rarely cause mortality in
healthy adults. For instance, because of the high mutation
rate of the influenza virus, each year a new flu vaccine must
Overview of the Immune System
|
CHAPTER 1
5
BOX 1-1
CLINICAL FOCUS
Vaccine Controversy: What’s Truth and What’s Myth?
Despite the record of worldwide success of vaccines in improving public health,
some opponents claim that vaccines do
more harm than good, pressing for elimination or curtailment of childhood vaccination programs. There is no dispute that
vaccines represent unique safety issues,
since they are administered to people who
are healthy. Furthermore, there is general
agreement that vaccines must be rigorously tested and regulated, and that the
public must have access to clear and complete information about them. Although
the claims of vaccine critics must be evaluated, many can be answered by careful and
objective examination of records.
A recent example is the claim that vaccines given to infants and very young
children may contribute to the rising incidence of autism. This began with the suggestion that thimerosal, a mercury-based
additive used to inhibit bacterial growth in
some vaccine preparations since the
1930s, was causing autism in children. In
1999 the U.S. Centers for Disease Control
and Prevention (CDC) and the American
Association of Pediatricians (AAP) released
a joint recommendation that vaccine
manufacturers begin to gradually phase
out thimerosal use in vaccines. This recommendation was based on the increase in
the number of vaccines given to infants
and was aimed at keeping children at or
below Environmental Protection Agency
(EPA)–recommended maximums in mercury exposure. However, with the release
of this recommendation, parent-led public
advocacy groups began a media-fueled
campaign to build a case demonstrating
what they believed was a link between
vaccines and an epidemic of autism. These
AAP recommendations and public fears
led to a dramatic decline in the latter half
of 1999 in U.S. newborns vaccinated for
hepatitis B. To date, no credible study has
shown a scientific link between thimerosal
and autism. In fact, cases of autism in children have continued to rise since thimerosal was removed from all childhood
vaccines in 2001. Despite evidence to the
contrary, some still believe this claim.
A 1998 study appearing in The Lancet, a
reputable British medical journal, further
fueled these parent advocacy groups and
anti-vaccine organizations. The article,
published by Andrew Wakefield, claimed
the measles-mumps-rubella (MMR) vaccine caused pervasive developmental disorders in children, including autism
spectrum disorder. More than a decade of
subsequent research has been unable to
substantiate these claims, and 10 of the
original 12 authors on the paper later withdrew their support for the conclusions of
the study. In 2010, The Lancet retracted the
original article when it was shown that the
data in the study had been falsified to
reach desired conclusions. Nonetheless, in
the years between the original publication
of the Lancet article and its retraction, this
case is credited with decreasing rates of
MMR vaccination from a high of 92% to a
low of almost 60% in certain areas of the
United Kingdom. The resulting expansion
in the population of susceptible individuals led to endemic rates of measles and
mumps infection, especially in several
areas of Europe, and is credited with thou-
be prepared based on a prediction of the prominent genotypes likely to be encountered in the next season. Some years
this vaccine is more effective than others. If and when a more
lethal and unexpected pandemic strain arises, there will be a
race between its spread and the manufacture and administration of a new vaccine. With the current ease of worldwide
travel, present-day emergence of a pandemic strain of influenza could dwarf the devastation wrought by the 1918 flu
pandemic, which left up to 50 million dead.
sands of extended hospitalizations and
several deaths in infected children.
Why has there been such a strong urge
to cling to the belief that childhood vaccines
are linked with developmental disorders in
children despite much scientific evidence to
the contrary? One possibility lies in the timing of the two events. Based on current AAP
recommendations, in the United States
most children receive 14 different vaccines
and a total of up to 26 shots by the age of 2.
In 1983, children received less than half this
number of vaccinations. Couple this with
the onset of the first signs of autism and
other developmental disorders in children,
which can appear quite suddenly and peak
around 2 years of age. This sharp rise in the
number of vaccinations young children
receive today and coincidence in timing of
initial autism symptoms is credited with
sparking these fears about childhood vaccines. Add to this the increasing drop in
basic scientific literacy by the general public
and the overabundance of ways to gather
such information (accurate or not). As concerned parents search for answers, one can
begin to see how even scientifically unsupported links could begin to take hold as
families grapple with how to make intelligent public health risk assessments.
The notion that vaccines cause autism
was rejected long ago by most scientists.
Despite this, more work clearly needs to be
done to bridge the gap between public
perception and scientific understanding.
Gross, L. 2009. A broken trust: Lessons from the
vaccine–autism wars. PLoS Biology 7:e1000114.
Larson, H.J., et al. 2011. Addressing the vaccine
confidence gap. Lancet 378:526.
However, the eradication of infectious disease is not the
only worthy goal of immunology research. As we will see later,
exposure to infectious agents is part of our evolutionary history, and wiping out all of these creatures could potentially
cause more harm than good, both for the host and the environment. Thanks to many technical advances allowing scientific
discoveries to move efficiently from the bench to the bedside,
clinicians can now manipulate the immune response in ways
never before possible. For example, treatments to boost, inhibit,
6
PA R T I
|
Introduction
or redirect the specific efforts of immune cells are being applied
to treat autoimmune disease, cancer, and allergy, as well as
other chronic disorders. These efforts are already extending
and saving lives. Likewise, a clearer understanding of immunity has highlighted the interconnected nature of body systems, providing unique insights into areas such as cell biology,
human genetics, and metabolism. While a cure for AIDS and a
vaccine to prevent HIV infection are still the primary targets
for many scientists who study this disease, a great deal of basic
science knowledge has been gleaned from the study of just this
one virus and its interaction with the human immune system.
TABLE 1-2
Immunity Involves Both Humoral
and Cellular Components
Pasteur showed that vaccination worked, but he did not
understand how. Some scientists believed that immune protection in vaccinated individuals was mediated by cells, while
others postulated that a soluble agent delivered protection.
The experimental work of Emil von Behring and Shibasaburo
Kitasato in 1890 gave the first insights into the mechanism
of immunity, earning von Behring the Nobel Prize in Physiology or Medicine in 1901 (Table 1-2). Von Behring and
Nobel Prizes for immunologic research
Year
Recipient
Country
Research
1901
Emil von Behring
Germany
Serum antitoxins
1905
Robert Koch
Germany
Cellular immunity to tuberculosis
1908
Elie Metchnikoff
Paul Ehrlich
Russia
Germany
Role of phagocytosis (Metchnikoff )
and antitoxins (Ehrlich) in immunity
1913
Charles Richet
France
Anaphylaxis
1919
Jules Bordet
Belgium
Complement-mediated bacteriolysis
1930
Karl Landsteiner
United States
Discovery of human blood groups
1951
Max Theiler
South Africa
Development of yellow fever vaccine
1957
Daniel Bovet
Switzerland
Antihistamines
1960
F. Macfarlane Burnet
Peter Medawar
Australia
Great Britain
Discovery of acquired immunological
tolerance
1972
Rodney R. Porter
Gerald M. Edelman
Great Britain
United States
Chemical structure of antibodies
1977
Rosalyn R. Yalow
United States
Development of radioimmunoassay
1980
George Snell
Jean Dausset
Baruj Benacerraf
United States
France
United States
Major histocompatibility complex
1984
Niels K. Jerne
Cesar Milstein
Georges E. Köhler
Denmark
Great Britain
Germany
Immune regulatory theories (Jerne) and
technological advances in the development
of monoclonal antibodies (Milstein and Köhler)
1987
Susumu Tonegawa
Japan
Gene rearrangement in antibody production
1991
E. Donnall Thomas
Joseph Murray
United States
United States
Transplantation immunology
1996
Peter C. Doherty
Rolf M. Zinkernagel
Australia
Switzerland
Role of major histocompatibility complex in
antigen recognition by T cells
2002
Sydney Brenner
H. Robert Horvitz
J. E. Sulston
South Africa
United States
Great Britain
Genetic regulation of organ development
and cell death (apoptosis)
2008
Harald zur Hausen
Françoise Barré-Sinoussi
Luc Montagnier
Germany
France
France
Role of HPV in causing cervical cancer
(Hausen) and the discovery of HIV
(Barré-Sinoussi and Montagnier)
2011
Jules Hoffman
Bruce Beutler
Ralph Steinman
France
United States
United States
Discovery of activating principles of innate
immunity (Hoffman and Beutler) and role of
dendritic cells in adaptive immunity (Steinman)
Overview of the Immune System
|
CHAPTER 1
7
FIGURE 1-3 Drawing by Elie Metchnikoff of phagocytic cells surrounding a foreign particle (left) and modern image of a
phagocyte engulfing the bacteria that cause tuberculosis (right). Metchnikoff first described and named the process of phagocytosis,
or ingestion of foreign matter by white blood cells. Today, phagocytic cells can be imaged in great detail using advanced microscopy techniques.
[Drawing reproduced by permission of The British Library:7616.h.19, Lectures on the Comparative Pathology of Inflammation delivered at the Pasteur Institute in
1891, translated by F. A. Starling and E. H. Starling, with plates by II’ya II’ich Mechnikov, 1893, p. 64, fig. 32. Photo courtesy Dr. Volker Brinkmann/Visuals Unlimited, Inc.]
Kitasato demonstrated that serum—the liquid, noncellular
component recovered from coagulated blood—from animals previously immunized with diphtheria could transfer
the immune state to unimmunized animals.
In 1883, even before the discovery that a serum component could transfer immunity, Elie Metchnikoff, another
Nobel Prize winner, demonstrated that cells also contribute
to the immune state of an animal. He observed that certain
white blood cells, which he termed phagocytes, ingested
(phagocytosed) microorganisms and other foreign material (Figure 1-3, left). Noting that these phagocytic cells
were more active in animals that had been immunized,
Metchnikoff hypothesized that cells, rather than serum
components, were the major effectors of immunity. The
active phagocytic cells identified by Metchnikoff were
likely blood monocytes and neutrophils (see Chapter 2),
which can now be imaged using very sophisticated microscopic techniques (Figure 1-3, right).
Humoral Immunity
The debate over cells versus soluble mediators of immunity raged for decades. In search of the protective agent of
immunity, various researchers in the early 1900s helped
characterize the active immune component in blood
serum. This soluble component could neutralize or precipitate toxins and could agglutinate (clump) bacteria. In
each case, the component was named for the activity it
exhibited: antitoxin, precipitin, and agglutinin, respectively. Initially, different serum components were thought
to be responsible for each activity, but during the 1930s,
mainly through the efforts of Elvin Kabat, a fraction of
serum first called gamma globulin (now immunoglobulin) was shown to be responsible for all these activities.
The soluble active molecules in the immunoglobulin fraction of serum are now commonly referred to as antibodies. Because these antibodies were contained in body
fluids (known at that time as the body humors), the immunologic events they participated in was called humoral
immunity.
The observation of von Behring and Kitasato was quickly
applied to clinical practice. Antiserum, the antibodycontaining serum fraction from a pathogen-exposed individual, derived in this case from horses, was given to patients
suffering from diphtheria and tetanus. A dramatic vignette
of this application is described in the Clinical Focus box on
page 8. Today there are still therapies that rely on transfer of
immunoglobulins to protect susceptible individuals. For
example, emergency use of immune serum, containing antibodies against snake or scorpion venoms, is a common
practice for treating bite victims. This form of immune protection that is transferred between individuals is called passive immunity because the individual receiving it did not
make his or her own immune response against the pathogen.
Newborn infants benefit from passive immunity by the presence of maternal antibodies in their circulation. Passive
immunity may also be used as a preventive (prophylaxis) to
boost the immune potential of those with compromised
immunity or who anticipate future exposure to a particular
microbe.
While passive immunity can supply a quick solution, it is
short-lived and limited, as the cells that produce these antibodies are not being transferred. On the other hand, administration
8
PA R T I
|
Introduction
CLINICAL FOCUS
Passive Antibodies and the Iditarod
In 1890, immunologists Emil Behring
and Shibasaburo Kitasato, working
together in Berlin, reported an extraordinary experiment. After immunizing rabbits
with tetanus and then collecting blood
serum from these animals, they injected a
small amount of immune serum (cell-free
fluid) into the abdominal cavity of six mice.
Twenty-four hours later, they infected the
treated mice and untreated controls with
live, virulent tetanus bacteria. All of the control mice died within 48 hours of infection,
whereas the treated mice not only survived
but showed no effects of infection. This
landmark experiment demonstrated two
important points. One, it showed that substances that could protect an animal
against pathogens appeared in serum following immunization. Two, this work demonstrated that immunity could be passively
acquired, or transferred from one animal to
another by taking serum from an immune
animal and injecting it into a nonimmune
one. These and subsequent experiments
did not go unnoticed. Both men eventually
received titles (Behring became von Behring and Kitasato became Baron Kitasato). A
few years later, in 1901, von Behring was
awarded the first Nobel Prize in Physiology
or Medicine (see Table 1-2).
These early observations, and others,
paved the way for the introduction of passive immunization into clinical practice.
During the 1930s and 1940s, passive
immunotherapy, the endowment of
resistance to pathogens by transfer of
antibodies from an immunized donor to
an unimmunized recipient, was used to
prevent or modify the course of measles
and hepatitis A. Subsequently, clinical
experience and advances in the technology of immunoglobulin preparation have
made this approach a standard medical
practice. Passive immunization based on
the transfer of antibodies is widely used in
the treatment of immunodeficiency and
some autoimmune diseases. It is also used
to protect individuals against anticipated
exposure to infectious and toxic agents
against which they have no immunity.
Finally, passive immunization can be lifesaving during episodes of certain types of
acute infection, such as following exposure to rabies virus.
Immunoglobulin for passive immunization is prepared from the pooled
of a vaccine or natural infection is said to engender active
immunity in the host: the production of one’s own immunity.
The induction of active immunity can supply the individual
with a renewable, long-lived protection from the specific infectious organism. As we discuss further below, this long-lived
protection comes from memory cells, which provide protection for years or even decades after the initial exposure.
Cell-Mediated Immunity
A controversy developed between those who held to the
concept of humoral immunity and those who agreed with
Metchnikoff ’s concept of immunity imparted by specific
cells, or cell-mediated immunity. The relative contributions
of the two were widely debated at the time. It is now obvious
that both are correct—the full immune response requires
both cellular and humoral (soluble) components. Early studies of immune cells were hindered by the lack of genetically
defined animal models and modern tissue culture tech-
plasma of thousands of donors. In effect,
recipients of these antibody preparations are receiving a sample of the antibodies produced by many people to a
broad diversity of pathogens—a gram of
intravenous immune globulin (IVIG) contains about 1018 molecules of antibody
and recognize more than 107 different
antigens. A product derived from the
blood of such a large number of donors
carries a risk of harboring pathogenic
agents, particularly viruses. This risk is
minimized by modern-day production
techniques. The manufacture of IVIG
involves treatment with solvents, such
as ethanol, and the use of detergents
that are highly effective in inactivating
viruses such as HIV and hepatitis. In
addition to treatment against infectious
disease, or acute situations, IVIG is also
used today for treating some chronic
diseases, including several forms of
immune deficiency. In all cases, the
transfer of passive immunity supplies
only temporary protection.
One of the most famous instances of
passive antibody therapy occurred in
1925, when an outbreak of diphtheria
niques, whereas early studies with serum took advantage of
the ready availability of blood and established biochemical
techniques to purify proteins. Information about cellular
immunity therefore lagged behind a characterization of
humoral immunity.
In a key experiment in the 1940s, Merrill Chase, working
at The Rockefeller Institute, succeeded in conferring immunity
against tuberculosis by transferring white blood cells
between guinea pigs. Until that point, attempts to develop an
effective vaccine or antibody therapy against tuberculosis
had met with failure. Thus, Chase’s demonstration helped to
rekindle interest in cellular immunity. With the emergence
of improved cell culture and transfer techniques in the
1950s, the lymphocyte was identified as the cell type responsible for both cellular and humoral immunity. Soon thereafter, experiments with chickens pioneered by Bruce Glick at
Mississippi State University indicated the existence of two
types of lymphocytes: T lymphocytes (T cells), derived from
Overview of the Immune System
|
CHAPTER 1
9
BOX 1-2
White
Mountain
Golovin Koyuk
Galena
Nulato
Nome
Ruby
Safety
Elim
Kaktag Cripple
Shaktoolik
Nenana
Unalakleet
Nikolai
Ophir
Takotna
Rohn
McGrath
Skwentna
Rainy Pass
Yentna
Finger Lake
Willow
Anchorage
Campell
Airstrip
FIGURE 1
(left) Leonhard Seppala, the Norwegian who led a team of sled dogs in the 1925 diphtheria antibody run from Nenana to Nome, Alaska. (right) Map of the
current route of the Iditarod Race, which commemorates this historic delivery of lifesaving antibody. [Source: Underwood & Underwood/Corbis.]
was diagnosed in what was then the
remote outpost of Nome, Alaska. Lifesaving diphtheria-specific antibodies were
available in Anchorage, but no roads
were open and the weather was too dangerous for flight. History tells us that 20
mushers set up a dogsled relay to cover
the almost 700 miles between Nenana,
the end of the railroad run, and remote
Nome. In this relay, two Norwegians and
their dogs covered particularly critical territory and withstood blizzard conditions:
Leonhard Seppala (Figure 1, left), who
covered the most treacherous territory,
and Gunnar Kaasen, who drove the final
two legs in whiteout conditions, behind
his lead dog Balto. Kaasen and Balto
arrived in time to save many of the chil-
the thymus, and B lymphocytes (B cells), derived from the
bursa of Fabricius in birds (an outgrowth of the cloaca). In a
convenient twist of nomenclature that makes B and T cell
origins easier to remember, the mammalian equivalent of
the bursa of Fabricius is bone marrow, the home of developing B cells in mammals. We now know that cellular immunity
is imparted by T cells and that the antibodies produced by B
cells confer humoral immunity. The real controversy about
the roles of humoral versus cellular immunity was resolved
when the two systems were shown to be intertwined and it
became clear that both are necessary for a complete immune
response against most pathogens.
How Are Foreign Substances Recognized
by the Immune System?
One of the great enigmas confronting early immunologists
was what determines the specificity of the immune response
dren in the town. To commemorate this
heroic event, later that same year a statue
of Balto was placed in Central Park, New
York City, where it still stands today. This
journey is memorialized every year in the
running of the Iditarod sled dog race. A
map showing the current route of this
more than 1000-mile trek is shown in
Figure 1, right.
for a particular foreign material, or antigen, the general
term for any substance that elicits a specific response by B
or T lymphocytes. Around 1900, Jules Bordet at the Pasteur Institute expanded the concept of immunity beyond
infectious diseases, demonstrating that nonpathogenic
substances, such as red blood cells from other species,
could also serve as antigens. Serum from an animal that
had been inoculated with noninfectious but otherwise
foreign (nonself) material would nevertheless react with
the injected material in a specific manner. The work of
Karl Landsteiner and those who followed him showed that
injecting an animal with almost any nonself organic chemical could induce production of antibodies that would bind
specifically to the chemical. These studies demonstrated
that antibodies have a capacity for an almost unlimited
range of reactivity, including responses to compounds
that had only recently been synthesized in the laboratory
and are otherwise not found in nature! In addition, it was
10
PA R T I
|
Introduction
shown that molecules differing in the smallest detail, such
as a single amino acid, could be distinguished by their
reactivity with different antibodies. Two major theories
were proposed to account for this specificity: the selective
theory and the instructional theory.
The earliest conception of the selective theory dates to
Paul Ehrlich in 1900. In an attempt to explain the origin
of serum antibody, Ehrlich proposed that cells in the
blood expressed a variety of receptors, which he called
side-chain receptors, that could bind to infectious agents
and inactivate them. Borrowing a concept used by Emil
Fischer in 1894 to explain the interaction between an
enzyme and its substrate, Ehrlich proposed that binding
of the receptor to an infectious agent was like the fit
between a lock and key. Ehrlich suggested that interaction
between an infectious agent and a cell-bound receptor
would induce the cell to produce and release more receptors with the same specificity (Figure 1-4). In Ehrlich’s
mind, the cells were pluripotent, expressing a number of
FIGURE 1-4 Representation of Paul Ehrlich’s side chain
theory to explain antibody formation. In Ehrlich’s initial theory,
the cell is pluripotent in that it expresses a number of different receptors or side chains, all with different specificities. If an antigen
encounters this cell and has a good fit with one of its side chains,
synthesis of that receptor is triggered and the receptor will be
released. [From Ehrlich’s Croonian lecture of 1900 to the Royal Society.]
different receptors, each of which could be individually
“selected.” According to Ehrlich’s theory, the specificity of
the receptor was determined in the host before its exposure to the foreign antigen, and therefore the antigen
selected the appropriate receptor. Ultimately, most aspects
of Ehrlich’s theory would be proven correct, with the following minor refinement: instead of one cell making
many receptors, each cell makes many copies of just one
membrane-bound receptor (one specificity). An army of
cells, each with a different antigen specificity, is therefore
required. The selected B cell can be triggered to proliferate
and to secrete many copies of these receptors in soluble form
(now called antibodies) once it has been selected by antigen
binding.
In the 1930s and 1940s, the selective theory was challenged by various instructional theories. These theories held
that antigen played a central role in determining the specificity of the antibody molecule. According to the instructional
theorists, a particular antigen would serve as a template
around which antibody would fold—sort of like an impression mold. The antibody molecule would thereby assume a
configuration complementary to that of the antigen template. This concept was first postulated by Friedrich Breinl
and Felix Haurowitz in about 1930 and redefined in the
1940s in terms of protein folding by Linus Pauling.
In the 1950s, selective theories resurfaced as a result of
new experimental data. Through the insights of F. Macfarlane Burnet, Niels Jerne, and David Talmadge, this model
was refined into a hypothesis that came to be known as the
clonal selection theory. This hypothesis has been further
refined and is now accepted as an underlying paradigm of
modern immunology. According to this theory, an individual B or T lymphocyte expresses many copies of a membrane
receptor that is specific for a single, distinct antigen. This
unique receptor specificity is determined in the lymphocyte
before it is exposed to the antigen. Binding of antigen to its
specific receptor activates the cell, causing it to proliferate
into a clone of daughter cells that have the same receptor
specificity as the parent cell. The instructional theories were
formally disproved in the 1960s, by which time information
was emerging about the structure of protein, RNA, and DNA
that would offer new insights into the vexing problem of how
an individual could make antibodies against almost anything, sight unseen.
Overview Figure 1-5 presents a very basic scheme of
clonal selection in the humoral (B cell) and cellular (T cell)
branches of immunity. We now know that B cells produce
antibodies, a soluble version of their receptor protein, which
bind to foreign proteins, flagging them for destruction. T cells,
which come in several different forms, also use their surfacebound T-cell receptors to sense antigen. These cells can
perform a range of different functions once selected by antigen encounter, including the secretion of soluble compounds
to aid other white blood cells (such as B lymphocytes) and
the destruction of infected host cells.
Overview of the Immune System
|
CHAPTER 1
11
1-5
OVERVIEW FIGURE
An Outline for the Humoral and Cell-Mediated (Cellular)
Branches of the Immune System
Foreign proteins or infectious agents
Vertebrate body
Humoral response
(B lymphocytes)
Cell-mediated response
(T lymphocytes)
T-cell
receptor
B cell
B
T-cell
receptor
+
T
B-cell
receptor
T
+ Antigen
+ Antigen
Antigen
T
Antigen-selected
antibodysecreting
B cell
T
Killing of
infected
cells
B
Antibody
Antigenselected
T cells
Cytokine
secretion
Antigen
elimination
The humoral response involves interaction of B cells with foreign
proteins, called antigens, and their differentiation into antibodysecreting cells. The secreted antibody binds to foreign proteins or
infectious agents, helping to clear them from the body. The cell-
Important Concepts for
Understanding the
Mammalian Immune Response
Today, more than ever, we are beginning to understand on
a molecular and cellular level how a vaccine or infection
leads to the development of immunity. As highlighted by
mediated response involves various subpopulations of T lymphocytes, which can perform many functions, including the secretion
of soluble messengers that help direct other cells of the immune
system and direct killing of infected cells.
the historical studies described above, this involves a complex system of cells and soluble compounds that have
evolved to protect us against an enormous range of invaders
of all shapes, sizes, and chemical structures. In this section,
we cover the range of organisms that challenge the immune
system and several of the important new concepts that are
unique hallmarks of how the immune system carries out
this task.
12
PA R T I
|
Introduction
Pathogens Come in Many Forms and Must
First Breach Natural Barriers
Organisms causing disease are termed pathogens, and the
process by which they induce illness in the host is called
pathogenesis. The human pathogens can be grouped into
four major categories based on shared characteristics:
viruses, fungi, parasites, and bacteria (Table 1-3). Some
example organisms from each category can be found in
Figure 1-6. As we will see in the next section, some of the
shared characteristics that are common to groups of pathogens, but not to the host, can be exploited by the immune
system for recognition and destruction.
The microenvironment in which the immune response
begins to emerge can also influence the outcome; the same
pathogen may be treated differently depending on the context in which it is encountered. Some areas of the body, such
as the central nervous system, are virtually “off limits” for
the immune system because the immune response could do
more damage than the pathogen. In other cases, the environment may come with inherent directional cues for
immune cells. For instance, some foreign compounds that
enter via the digestive tract, including the commensal
microbes that help us digest food, are tolerated by the
immune system. However, when these same foreigners
enter the bloodstream they are typically treated much more
aggressively. Each encounter with pathogen thus engages a
distinct set of strategies that depends on the nature of the
invader and on the microenvironment in which engagement occurs.
It is worth noting that immune pathways do not become
engaged until foreign organisms first breach the physical
barriers of the body. Obvious barriers include the skin and
TABLE 1-3
the mucous membranes. The acidity of the stomach contents, of the vagina, and of perspiration poses a further barrier to many organisms, which are unable to grow in low pH
conditions. The importance of these barriers becomes obvious when they are surmounted. Animal bites can communicate rabies or tetanus, whereas insect puncture wounds can
transmit the causative agents of such diseases as malaria
(mosquitoes), plague (fleas), and Lyme disease (ticks). A
dramatic example is seen in burn victims, who lose the protective skin at the burn site and must be treated aggressively
with drugs to prevent the rampant bacterial and fungal
infections that often follow.
The Immune Response Quickly Becomes
Tailored to Suit the Assault
With the above in mind, an effective defense relies heavily on
the nature of the invading pathogen offense. The cells and
molecules that become activated in a given immune response
depend on the chemical structures present on the pathogen,
whether it resides inside or outside of host cells, and the
location of the response. This means that different chemical
structures and microenvironmental cues need to be detected
and appropriately evaluated, initiating the most effective
response strategy. The process of pathogen recognition involves
an interaction between the foreign organism and a recognition
molecule (or molecules) expressed by host cells. Although
these recognition molecules are frequently membranebound receptors, soluble receptors or secreted recognition
molecules can also be engaged. Ligands for these recognition
molecules can include whole pathogens, antigenic fragments
of pathogens, or products secreted by these foreign organisms. The outcome of this ligand binding is an intracellular
Major categories of human pathogens
Major groups of human pathogens
Specific examples
Disease
Viruses
Poliovirus
Variola Virus
Human Immunodeficiency Virus
Rubeola Virus
Poliomyelitis (Polio)
Smallpox
AIDS
Measles
Fungi
Candida albicans
Tinea corporis
Cryptococcus neoformans
Candidiasis (Thrush)
Ringworm
Cryptococcal meningitis
Parasites
Plasmodium species
Leishmania major
Entamoeba histolytica
Malaria
Leishmaniasis
Amoebic colitis
Bacteria
Mycobacterium tuberculosis
Bordetella pertussis
Vibrio cholerae
Borrelia burgdorferi
Tuberculosis
Whooping cough (pertussis)
Cholera
Lyme disease
Overview of the Immune System
|
CHAPTER 1
13
(a) Virus: Rotavirus
(b) Fungus: Candida albicans
(c) Parasite: Filaria
(d) Bacterium: Mycobacterium tuberculosis
FIGURE 1-6 Pathogens representing the major categories
of microorganisms causing human disease. (a) Viruses: Transmission electron micrograph of rotavirus, a major cause of infant
diarrhea. Rotavirus accounts for approximately 1 million infant deaths
per year in developing countries and hospitalization of about 50,000
infants per year in the United States. (b) Fungi: Candida albicans, a
yeast inhabiting human mouth, throat, intestines, and genitourinary
tract; C. albicans commonly causes an oral rash (thrush) or vaginitis in
immunosuppressed individuals or in those taking antibiotics that kill
normal bacterial flora. (c) Parasites: The larval form of filaria, a parasitic
worm, being attacked by macrophages. Approximately 120 million
persons worldwide have some form of filariasis. (d) Bacteria: Mycobacterium tuberculosis, the bacterium that causes tuberculosis, being
ingested by a human macrophage. [(a) Dr. Gary Gaugler/Getty Images;
or extracellular cascade of events that ultimately leads to the
labeling and destruction of the pathogen—simply referred to
as the immune response. The entirety of this response is
actually engagement of a complex system of cells that can
recognize and kill or engulf a pathogen (cellular immunity),
as well as a myriad of soluble proteins that help to orchestrate labeling and destruction of foreign invaders (humoral
immunity).
The nature of the immune response will vary depending
on the number and type of recognition molecules engaged.
For instance, all viruses are tiny, obligate, intracellular patho-
(b) SPL/Photo Researchers; (c) Oliver Meckes/Nicole Ottawa/Eye of Science/
Photo Researchers; (d) Max Planck Institute for Infection Biology/Dr. Volker
Brinkmann.]
gens that spend the majority of their life cycle residing inside
host cells. An effective defense strategy must therefore
involve identification of infected host cells along with recognition of the surface of the pathogen. This means that some
immune cells must be capable of detecting changes that occur
in a host cell after it becomes infected. This is achieved by a
range of cytotoxic cells but especially cytotoxic T lymphocytes (aka CTLs, or Tc cells), a part of the cellular arm of
immunity. In this case, recognition molecules positioned
inside cells are key to the initial response. These intracellular
receptors bind to viral proteins present in the cytosol and
14
PA R T I
|
Introduction
initiate an early warning system, alerting the cell to the presence of an invader.
Sacrifice of virally infected cells often becomes the only
way to truly eradicate this type of pathogen. In general, this
sacrifice is for the good of the whole organism, although in
some instances it can cause disruptions to normal function.
For example, the Human Immunodeficiency Virus (HIV)
infects a type of T cell called a T helper cell (TH cell). These
cells are called helpers because they guide the behavior of
other immune cells, including B cells, and are therefore pivotal for selecting the pathway taken by the immune response.
Once too many of these cells are destroyed or otherwise rendered nonfunctional, many of the directional cues needed for
a healthy immune response are missing and fighting all types
of infections becomes problematic. As we discuss later in this
chapter, the resulting immunodeficiency allows opportunistic infections to take hold and potentially kill the patient.
Similar but distinct immune mechanisms are deployed to
mediate the discovery of extracellular pathogens, such as
fungi, most bacteria, and some parasites. These rely primarily on cell surface or soluble recognition molecules that
probe the extracellular spaces of the body. In this case, B cells
and the antibodies they produce as a part of humoral immunity play major roles. For instance, antibodies can squeeze
into spaces in the body where B cells themselves may not be
able to reach, helping to identify pathogens hiding in these
out-of-reach places. Large parasites present yet another problem; they are too big for phagocytic cells to envelop. In this
case, cells that can deposit toxic substances or that can
secrete products that induce expulsion (e.g., sneezing,
coughing, vomiting) become a better strategy.
As we study the complexities of the mammalian immune
response, it is worth remembering that a single solution does
not exist for all pathogens. At the same time, these various
immune pathways carry out their jobs with considerable
overlap in structure and in function.
Pathogen Recognition Molecules
Can Be Encoded in the Germline
or Randomly Generated
As one might imagine, most pathogens express at least a few
chemical structures that are not typically found in mammals.
Pathogen-associated molecular patterns (or PAMPs) are
common foreign structures that characterize whole groups of
pathogens. It is these unique antigenic structures that the
immune system frequently recognizes first. Animals, both
invertebrates and vertebrates, have evolved to express several
types of cell surface and soluble proteins that quickly recognize many of these PAMPs; a form of pathogen profiling.
For example, encapsulated bacteria possess a polysaccharide
coat with a unique chemical structure that is not found on
other bacterial or human cells. White blood cells naturally
express a variety of receptors, collectively referred to as pattern recognition receptors (PRRs), that specifically recognize
these sugar residues, as well as other common foreign struc-
tures. When PRRs detect these chemical structures, a cascade
of events labels the target pathogen for destruction. PPRs are
proteins encoded in the genomic DNA and are always
expressed by many different immune cells. These conserved,
germline-encoded recognition molecules are thus a first line of
defense for the quick detection of many of the typical chemical identifiers carried by the most common invaders.
A significant and powerful corollary to this is that it allows
early categorizing or profiling of the sort of pathogen of concern. This is key to the subsequent immune response routes
that will be followed, and therefore the fine tailoring of the
immune response as it develops. For example, viruses frequently expose unique chemical structures only during their
replication inside host cells. Many of these can be detected via
intracellular receptors that bind exposed chemical moieties
while still inside the host cell. This can trigger an immediate
antiviral response in the infected cell that blocks further virus
replication. At the same time, this initiates the secretion of
chemical warning signals sent to nearby cells to help them
guard against infection (a neighborhood watch system!). This
early categorizing happens via a subtle tracking system that
allows the immune response to make note of which recognition molecules were involved in the initial detection event
and relay that information to subsequent responding immune
cells, allowing the follow-up response to begin to focus attention on the likely type of assault underway.
Host-pathogen interactions are an ongoing arms race;
pathogens evolve to express unique structures that avoid host
detection, and the host germline-encoded recognition system
co-evolves to match these new challenges. However, because
pathogens generally have much shorter life cycles than their
vertebrate hosts, and some utilize error-prone DNA polymerases to replicate their genomes, pathogens can evolve rapidly to
evade host encoded recognition systems. If this were our only
defense, the host immune response would quickly become
obsolete thanks to these real-time pathogen avoidance strategies. How can the immune system prepare for this? How can our
DNA encode a recognition system for things that change in
random ways over time? Better yet, how do we build a system to
recognize new chemical structures that may arise in the future?
Thankfully, the vertebrate immune system has evolved a
clever, albeit resource intensive, response to this dilemma: to
favor randomness in the design of some recognition molecules. This strategy, called generation of diversity, is
employed only by developing B and T lymphocytes. The result
is a group of B and T cells where each expresses many copies of
one unique recognition molecule, resulting in a population
with the theoretical potential to respond to any antigen that
may come along (Figure 1-7). This feat is accomplished by
rearranging and editing the genomic DNA that encodes the
antigen receptors expressed by each B or T lymphocyte. Not
unlike the error-prone DNA replication method employed by
pathogens, this system allows chance to play a role in generating a menu of responding recognition molecules.
As one might imagine, however, this cutting and splicing of
chromosomes is not without risk. Many B and T cells do not
|
Overview of the Immune System
Generation of
diversity
Deletion
CHAPTER 1
15
Clonal selection and expansion
2
1
2
Antigen
2
2
2
2
Stem
cell
3
2
3
2
4
Primary lymphoid organs
FIGURE 1-7 Generation of diversity and clonal selection in
T and B lymphocytes. Maturation in T and B cells, which occurs in
primary lymphoid organs (bone marrow for B cells and thymus for
T cells) in the absence of antigen, produces cells with a committed
antigenic specificity, each of which expresses many copies of surface receptor that binds to one particular antigen. Different clones
of B cells (1, 2, 3, and 4) are illustrated in this figure. Cells that do not
die or become deleted during this maturation and weeding-out process move into the circulation of the body and are available to interact with antigen. There, clonal selection occurs when one of these cells
survive this DNA surgery or the quality control processes that
follow, all of which take place in primary lymphoid organs: the
thymus for T cells and bone marrow for B cells. Surviving cells
move into the circulation of the body, where they are available
if their specific, or cognate, antigen is encountered. When
antigens bind to the surface receptors on these cells, they trigger clonal selection (see Figure 1-7). The ensuing proliferation
of the selected clone of cells creates an army of cells, all with
the same receptor and responsible for binding more of the
same antigen, with the ultimate goal of destroying the pathogen in question. In B lymphocytes, these recognition molecules are B-cell receptors when they are surface structures
and antibodies in their secreted form. In T lymphocytes,
where no soluble form exists, they are T-cell receptors. In
1976 Susumu Tonegawa, then at The Basel Institute for Immunology in Switzerland, discovered the molecular mechanism
behind the DNA recombination events that generate B-cell
receptors and antibodies (Chapter 7 covers this in detail). This
Circulation through the body
encounters its cognate or specific antigen. Clonal proliferation of an
antigen-activated cell (number 2 or pink in this example) leads to
many cells that can engage with and destroy the antigen, plus
memory cells that can be called upon during a subsequent exposure.
The B cells secrete antibody, a soluble form of the receptor, reactive with the activating antigen. Similar processes take place in the
T-lymphocyte population, resulting in clones of memory T cells and
effector T cells; the latter include activated TH cells, which secrete
cytokines that aid in the further development of adaptive immunity,
and cytotoxic T lymphocytes (CTLs), which can kill infected host cells.
was a true turning point in immunologic understanding; for
this discovery he received widespread recognition, including
the 1987 Nobel Prize in Physiology or Medicine (see Table 1-2).
Tolerance Ensures That the Immune System
Avoids Destroying the Host
One consequence of generating random recognition receptors is that some could recognize and target the host. In
order for this strategy to work effectively, the immune system must somehow avoid accidentally recognizing and
destroying host tissues. This principle, which relies on self/
nonself discrimination, is called tolerance, another hallmark
of the immune response. The work credited with its illumination also resulted in a Nobel Prize in Physiology or Medicine, awarded to F. Macfarlane Burnet and Peter Medawar in
1960. Burnet was the first to propose that exposure to nonself antigens during certain stages of life could result in an
16
PA R T I
|
Introduction
immune system that ignored these antigens later. Medawar
later proved the validity of this theory by exposing mouse
embryos to foreign antigens and showing that these mice
developed the ability to tolerate these antigens later in life.
To establish tolerance, the antigen receptors present on
developing B and T cells must first pass a test of nonresponsiveness against host structures. This process, which begins
shortly after these randomly generated receptors are produced, is achieved by the destruction or inhibition of any
cells that have inadvertently generated receptors with the
ability to harm the host. Successful maintenance of tolerance
ensures that the host always knows the difference between self
and nonself (usually referred to as foreign).
One recent re-envisioning of how tolerance is operationally maintained is called the danger hypothesis. This hypothesis suggests that the immune system constantly evaluates
each new encounter more for its potential to be dangerous to
the host than for whether it is self or not. For instance, cell
death can have many causes, including natural homeostatic
processes, mechanical damage, or infection. The former is a
normal part of the everyday biological events in the body,
and only requires a cleanup response to remove debris. The
latter two, however, come with warning signs that include
the release of intracellular contents, expression of cellular
stress proteins, or pathogen-specific products. These pathogen or cell-associated stress compounds, sometimes referred
to as danger signals, can engage specific host recognition
molecules (e.g., PRRs) that deliver a signal to immune cells to
get involved during these unnatural causes of cellular death.
One unintended consequence of robust self-tolerance is
that the immune system frequently ignores cancerous cells
that arise in the body, as long as these cells continue to
express self structures that the immune system has been
trained to ignore. Dysfunctional tolerance is at the root of
most autoimmune diseases, discussed further at the end of
this chapter and in greater detail in Chapter 16. As one
might imagine, failures in the establishment or maintenance
of tolerance can have devastating clinical outcomes.
The Immune Response Is Composed of Two
Interconnected Arms: Innate Immunity
and Adaptive Immunity
Although reference is made to “the immune system,” it is
important to appreciate that there are really two interconnected systems of immunity: innate and adaptive. These two
systems collaborate to protect the body against foreign
invaders. Innate immunity includes built-in molecular and
cellular mechanisms that are encoded in the germline and
are evolutionarily more primitive, aimed at preventing infection or quickly eliminating common invaders (Chapter 5).
This includes physical and chemical barriers to infection, as
well as the DNA-encoded receptors recognizing common
chemical structures of many pathogens (see PRRs, above). In
this case, rapid recognition and phagocytosis or destruction
of the pathogen is the outcome. Innate immunity also
includes a series of preexisting serum proteins, collectively
referred to as complement, that bind common pathogenassociated structures and initiate a cascade of labeling and
destruction events (Chapter 6). This highly effective first line
of defense prevents most pathogens from taking hold, or
eliminates infectious agents within hours of encounter. The
recognition elements of the innate immune system are fast,
some occurring within seconds of a barrier breach, but they
are not very specific and are therefore unable to distinguish
between small differences in foreign antigens.
A second form of immunity, known as adaptive immunity, is much more attuned to subtle molecular differences.
This part of the system, which relies on B and T lymphocytes,
takes longer to come on board but is much more antigen
specific. Typically, there is an adaptive immune response
against a pathogen within 5 or 6 days after the barrier breach
and initial exposure, followed by a gradual resolution of the
infection. Adaptive immunity is slower partly because fewer
cells possess the perfect receptor for the job: the antigenspecific, randomly generated receptors found on B and T
cells. It is also slower because parts of the adaptive response
rely on prior encounter and “categorizing” of antigens undertaken by innate processes. After antigen encounter, T and B
lymphocytes undergo selection and proliferation, described
earlier in the clonal selection theory of antigen specificity.
Although slow to act, once these B and T cells have been
selected and have honed their attack strategy, they become
formidable opponents that can typically resolve the infection.
The adaptive arm of the immune response evolves in real
time in response to infection and adapts (thus the name) to
better recognize, eliminate, and remember the invading
pathogen. Adaptive responses involve a complex and interconnected system of cells and chemical signals that come
together to finish the job initiated during the innate immune
response. The goal of all vaccines against infectious disease
is to elicit the development of specific and long-lived adaptive responses, so that the vaccinated individual will be protected in the future when the real pathogen comes along.
This arm of immunity is orchestrated mainly via B and T
lymphocytes following engagement of their randomly generated antigen recognition receptors. How these receptors are
generated is a fascinating story, covered in detail in Chapter 7
of this book. An explanation of how these cells develop to
maturity (Chapters 9 and 10) and then work in the body to
protect us from infection (Chapters 11-14) or sometimes fail
us (Chapters 15-19) takes up the vast majority of this book.
The number of pages dedicated to discussing adaptive
responses should not give the impression that this arm of the
immune response is more important, or can work independently from, innate immunity. In fact, the full development of
the adaptive response is dependent upon earlier innate pathways. The intricacies of their interconnections remain an area
of intense study. The 2011 Nobel Prize in Physiology or
Medicine was awarded to three scientists who helped clarify
these two arms of the response: Bruce Beutler and Jules Hoffmann for discoveries related to the activation events important for innate immunity, and Ralph Steinman for his discovery
of the role of dendritic cells in activating adaptive immune
|
responses (see Table 1-2). Because innate pathways make first
contact with pathogens, the cells and molecules involved in
this arm of the response use information gathered from their
early encounter with pathogen to help direct the process of
adaptive immune development. Adaptive immunity thus provides a second and more comprehensive line of defense, informed
by the struggles undertaken by the innate system. It is worth
noting that some infections are, in fact, eliminated by innate
immune mechanisms alone, especially those that remain
localized and involve very low numbers of fairly benign foreign invaders. (Think of all those insect bites or splinters in
life that introduce bacteria under the skin!) Table 1-4 compares the major characteristics that distinguish innate and
adaptive immunity. Although for ease of discussion the
immune system is typically divided into these two arms of
the response, there is considerable overlap of the cells and
mechanisms involved in each of these arms of immunity.
For innate and adaptive immunity to work together, these
two systems must be able to communicate with one another.
This communication is achieved by both cell-cell contact
and by soluble messengers. Most of these soluble proteins
are growth factor–like molecules known by the general
name cytokines. Cytokines and cell surface ligands can bind
with receptors found on responding cells and signal these
cells to perform new functions, such as synthesis of other
soluble factors or differentiation to a new cell type. A subset
of these soluble signals are called chemokines because they
have chemotactic activity, meaning they can recruit specific
cells to the site. In this way, cytokines, chemokines, and
other soluble factors produced by immune cells recruit or
instruct cells and soluble proteins important for eradication
of the pathogen from within the infection site. We’ve probably all felt this convergence in the form of swelling, heat,
and tenderness at the site of exposure. These events are a part
of a larger process collectively referred to as an inflammatory
response, which is covered in detail in Chapter 15.
Adaptive Immune Responses
Typically Generate Memory
One particularly significant and unique attribute of the
adaptive arm of the immune response is immunologic
TABLE 1-4
Magnitude of immune response
Overview of the Immune System
CHAPTER 1
17
Repeat
Antigen A
Antigen A
Adaptive
Innate
0
14
28 0
14
28
Time, days
Primary response
Secondary response
FIGURE 1-8 Differences in the primary and secondary
response to injected antigen reflect the phenomenon of
immunologic memory. When an animal is injected with an antigen, it produces a primary antibody response (dark blue) of low magnitude and short duration, peaking at about 10 to 20 days. At some
later point, a second exposure to the same antigen results in a secondary response that is greater in magnitude, peaks in less time (1–4 days),
and is more antigen specific than the primary response. Innate
responses, which have no memory element and occur each time an
antigen is encountered, are unchanged regardless of how frequently
this antigen has been encountered in the past (light blue).
memory. This is the ability of the immune system to respond
much more swiftly and with greater efficiency during a second exposure to the same pathogen. Unlike almost any other
biological system, the vertebrate immune response has
evolved not only the ability to learn from (adapt to) its
encounters with foreign antigen in real time but also the
ability to store this information for future use. During a first
encounter with foreign antigen, adaptive immunity undergoes what is termed a primary response, during which the
key lymphocytes that will be used to eradicate the pathogen
are clonally selected, honed, and enlisted to resolve the
infection. As mentioned above, these cells incorporate messages received from the innate players into their tailored
response to the specific pathogen.
All subsequent encounters with the same antigen or pathogen are referred to as the secondary response (Figure 1-8).
Comparison of innate and adaptive immunity
Innate
Adaptive
Response time
Minutes to hours
Days
Specificity
Limited and fixed
Highly diverse; adapts to improve during the course of
immune response
Response to repeat infection
Same each time
More rapid and effective with each subsequent exposure
Major components
Barriers (e.g., skin); phagocytes;
pattern recognition molecules
T and B lymphocytes; antigen-specific receptors; antibodies
18
PA R T I
|
Introduction
1-9
OVERVIEW FIGURE
Collaboration Between Innate and Adaptive Immunity
in Resolving an Infection
2a
1
Lymph
node
P
3
P
P
B
B
T
T
5a
T
B
4
5b
Memory
2b
This very basic scheme shows the sequence of events that occurs
during an immune response, highlighting interactions between
innate and adaptive immunity. 1. Pathogens are introduced at a
mucosal surface or breach in skin (bacteria entering the throat, in this
case), where they are picked up by phagocytic cells (yellow). 2a. In this
innate stage of the response, the phagocytic cell undergoes changes
and carries pieces of bacteria to a local lymph node to help activate
adaptive immunity. 2b. Meanwhile, at the site of infection resident
phagocytes encountering antigen release chemokines and cytokines
(black dots) that cause fluid influx and help recruit other immune cells
to the site (inflammation). 3. In the lymph node, T (blue) and B (green)
cells with appropriate receptor specificity are clonally selected when
During a secondary response, memory cells, kin of the final
and most efficient B and T lymphocytes trained during the
primary response, are re-enlisted to fight again. These cells
begin almost immediately and pick up right where they left
off, continuing to learn and improve their eradication strategy
during each subsequent encounter with the same antigen.
Depending on the antigen in question, memory cells can
remain for decades after the conclusion of the primary
response. Memory lymphocytes provide the means for subsequent responses that are so rapid, antigen-specific, and effective that when the same pathogen infects the body a second
time, dispatch of the offending organism often occurs without
symptoms. It is the remarkable property of memory that prevents us from catching many diseases a second time. Immunologic memory harbored by residual B and T lymphocytes is
the foundation for vaccination, which uses crippled or killed
pathogens as a safe way to “educate” the immune system to
prepare it for later attacks by life-threatening pathogens.
Overview Figure 1-9 highlights the ways in which the
innate and adaptive immune responses work together to
N
B
T
their surface receptors bind antigen that has entered the system,
kicking off adaptive immunity. 4. Collaboration between T and B cells
and continued antigen encounter occurs in the lymph node, driving
lymphocyte proliferation and differentiation, generating cells that
can very specifically identify and eradicate the pathogen. For example: 5a. B cells secrete antibodies specific for the antigen, which travels to the site of infection to help label and eradicate the pathogen.
5b. In addition to the cells that will destroy the pathogen here,
memory T and B cells are generated in this primary response and
will be available at the initiation of a secondary response, which will
be much more rapid and antigen specific. (Abbreviations: T T cell,
B B cell, P = phagocyte; N = neutrophil, a type of immune cell.)
resolve an infection. In this example, bacteria breach the
mucosal lining of the throat, a skin or mucous barrier,
where it is recognized and engulfed by a local phagocytic
cell (step 1). As part of the innate immune system, the
phagocytic cell releases cytokines and chemokines that
attract other white blood cells to the site of infection, initiating inflammation (step 2b). That phagocytic cell may
then travel to a local lymph node, the tissue where antigen
and lymphocytes meet, carrying bacterial antigens to B and
T lymphocytes (step 2a). Those lymphocytes with receptors
that are specific for the antigen are selected, activated, and
begin the adaptive immune response by proliferating (step 3).
Activated TH cells help to activate B cells, and clonal
expansion of both types of lymphocyte occurs in the lymph
node (step 4). This results in many T and B cells specific for
the antigen, with the latter releasing antibodies that can
attach to the intruder and direct its destruction (step 5a).
The adaptive response leaves behind memory T and B cells
available for a future, secondary encounter with this antigen (step 5b). It is worth noting that memory is a unique
Overview of the Immune System
capacity that arises from adaptive responses; there is no
memory component of innate immunity.
Sometimes, as is the case for some vaccines, one round of
antigen encounter and adaptation is not enough to impart
protective immunity from the pathogen in question. In
many of these cases, immunity can develop after a second or
even a third round of exposure to an antigen. It is these sorts
of pathogens that necessitate the use of vaccine booster
shots. Booster shots are nothing more than a second or third
episode of exposure to the antigen, each driving a new round
of adaptive events (secondary response) and refinements in
the responding lymphocyte population. The aim is to hone
these responses to a sufficient level to afford protection
against the real pathogen at some future date.
The Good, Bad, and Ugly
of the Immune System
The picture we’ve presented so far depicts the immune
response as a multicomponent interactive system that
always protects the host from invasion by all sorts of
pathogens. However, failures of this system do occur.
They can be dramatic and often garner a great deal of
attention, despite the fact that they are generally rare.
Certain clinical situations also pose unique challenges to
the immune system, including tissue transplants between
individuals (probably not part of any evolutionary plan!)
and the development of cancer. In this section we briefly
describe some examples of common failures and challenges to the development of healthy immune responses.
Each of these clinical manifestations is covered in much
greater detail in the concluding chapters of this book
(Chapters 15–19).
Inappropriate or Dysfunctional
Immune Responses Can Result
in a Range of Disorders
Most instances of immune dysfunction or failure fall into
one of the following three broad categories:
•
Hypersensitivity (including allergy): overly zealous
attacks on common benign but foreign antigens
•
Autoimmune Disease: erroneous targeting of selfproteins or tissues by immune cells
•
Immune Deficiency: insufficiency of the immune
response to protect against infectious agents
A brief overview of these situations and some examples of
each are presented below. At its most basic level, immune
dysfunction occurs as a result of improper regulation that
allows the immune system to either attack something it
shouldn’t or fail to attack something it should. Hypersensitivities, including allergy, and autoimmune disease are cases
of the former, where the immune system attacks an improper
|
CHAPTER 1
19
target. As a result, the symptoms can manifest as pathological
inflammation—an influx of immune cells and molecules that
results in detrimental symptoms, including chronic inflammation and rampant tissue destruction. In contrast, immune
deficiencies, caused by a failure to properly deploy the
immune response, usually result in weakened or dysregulated immune responses that can allow pathogens to get the
upper hand.
This is a good time to mention that the healthy immune
response involves a balancing act between immune aggression and immune suppression pathways. While we rarely
fail to consider erroneous attacks (autoimmunity) or failures to engage (immune deficiency) as dysfunctional, we
sometimes forget to consider the significance of the suppressive side of the immune response. Imperfections in
the inhibitory arm of the immune response, present as a
check to balance all the immune attacks we constantly
initiate, can be equally profound. Healthy immune
responses must therefore be viewed as a delicate balance,
spending much of the time with one foot on the brakes
and one on the gas.
Hypersensitivity Reactions
Allergies and asthma are examples of hypersensitivity reactions. These result from inappropriate and overly active
immune responses to common innocuous environmental
antigens, such as pollen, food, or animal dander. The possibility that certain substances induce increased sensitivity
(hypersensitivity) rather than protection was recognized in
about 1902 by Charles Richet, who attempted to immunize
dogs against the toxins of a type of jellyfish. He and his colleague Paul Portier observed that dogs exposed to sublethal
doses of the toxin reacted almost instantly, and fatally, to a
later challenge with even minute amounts of the same
toxin. Richet concluded that a successful vaccination typically results in phylaxis (protection), whereas anaphylaxis
(anti-protection)—an extreme, rapid, and often lethal overreaction of the immune response to something it has
encountered before—can result in certain cases in which
exposure to antigen is repeated. Richet received the Nobel
Prize in 1913 for his discovery of the anaphylactic response
(see Table 1-2). The term is used today to describe a severe,
life-threatening, allergic response.
Fortunately, most hypersensitivity or allergic reactions in
humans are not rapidly fatal. There are several different types
of hypersensitivity reactions; some are caused by antibodies
and others are the result of T-cell activity (see Chapter 15).
However, most allergic or anaphylactic responses involve a
type of antibody called immunoglobulin E (IgE). Binding of
IgE to its specific antigen (allergen) induces the release of
substances that cause irritation and inflammation, or the
accumulation of cells and fluid at the site. When an allergic
individual is exposed to an allergen, symptoms may include
sneezing, wheezing (Figure 1-10), and difficulty in breathing
(asthma); dermatitis or skin eruptions (hives); and, in more
severe cases, strangulation due to constricted airways
20
PA R T I
|
Introduction
BOX 1-3
CLINICAL FOCUS
The Hygiene Hypothesis
Worldwide, 300 million people
suffer from asthma and approximately
250,000 people died from the disease in
2007 (see Chapter 15). As of 2009, in the
United States alone, approximately 1 in 12
people (8.2%) are diagnosed with asthma.
The most common reason for a trip to a
hospital emergency room (ER) is an
asthma attack, accounting for one-third of
all visits. In addition to those treated in the
ER, over 400,000 hospitalizations for
asthma occurred in the United States in
2006, with an average stay of 3 to 4 days.
In the past 25 years, the prevalence of
asthma in industrialized nations has doubled. This is coupled with an overall rise in
other types of allergic disease during the
same time frame. What accounts for this
climb in asthma and allergy in the last few
decades? One idea, called the hygiene
hypothesis, suggests that a decrease in
human exposure to environmental
microbes has had adverse effects on the
human immune system. The hypothesis
suggests that several categories of allergic
or inflammatory disease, all disorders
caused by excessive immune activation,
have become more prevalent in industrialized nations thanks to diminished exposure to particular classes of microbes
following the widespread use of antibiotics, immunization programs, and overall
hygienic practices in those countries. This
idea was first proposed by D. P. Strachan
and colleagues in an article published in
1989 suggesting a link between hay fever
and household hygiene. More recently, this
hypothesis has been expanded to include
the view by some that it may be a contributing factor in many allergic diseases, several autoimmune disorders, and, more
recently, inflammatory bowel disease.
What is the evidence supporting the
hygiene hypothesis? The primary clinical
support comes from studies that have
shown a positive correlation between
growing up under environmental conditions that favor microbe-rich (sometimes
called “dirty”) environments and a decreased
incidence of allergy, especially asthma. To
date, childhood exposure to cowsheds and
farm animals, having several older siblings,
attending day care early in life, or growing
up in a developing nation have all been
correlated with a decreased likelihood of
developing allergies later in life. While viral
exposures during childhood do not seem
to favor protection, exposure to certain
classes of bacteria and parasitic organisms
may. Of late, the primary focus of attention
has been on specific classes of parasitic
worms (called helminthes), spawning New
Age allergy therapies involving intentional
exposure. This gives whole new meaning to
the phrase “Go eat worms”!
What are the proposed immunologic
mechanisms that might underlie this link
between a lack of early-life microbial
exposure and allergic disease? Current
dogma supporting this hypothesis posits
that millions of years of coevolution of
microbes and humans has favored a system in which early exposure to a broad
range of common environmental bugs
helps set the immune system on a path of
homeostatic balance between aggression
and inhibition. Proponents of this immune
regulation argument, sometimes referred
to as the “old friends” hypothesis, suggest
that antigens present on microbial organisms that have played a longstanding role
in our evolutionary history (both pathogens and harmless microbes we ingest or
that make up our historical flora) may
engage with the pattern recognition
receptors (PPRs) present on cells of our
innate immune system, driving them to
warn cells involved in adaptive responses
to tone it down. This hypothesis posits
that without early and regular exposure of
our immune cells to these old friends and
their antigens, the development of “normal” immune regulatory or homeostatic
responses is thrown into disarray, setting
us up for an immune system poised to
overreact in the future.
Animal models of disease lend some
support to this hypothesis and have
helped immunologists probe this line of
thinking. For instance, certain animals
raised in partially or totally pathogen-free
environments are more prone to type 1,
or insulin-dependent, diabetes, an autoimmune disease caused by immune
attack of pancreatic cells (see Chapter 16).
The lower the infectious burden of exposure in these mice, the greater the incidence of diabetes. Animals specifically
bred to carry enhanced genetic susceptibility favoring spontaneous development
of diabetes (called NOD mice, for nonobese diabetic) and treated with a variety
of infectious agents can be protected
from diabetes. Meanwhile, NOD mice
maintained in pathogen-free housing
almost uniformly develop diabetes. Much
like this experimental model, susceptibility to asthma and most other allergies is
known to run in families. Although all the
genes linked to asthma have not yet been
characterized, it is known that you have a
30% chance of developing the disease if
one of your parents is a sufferer, and a 70%
chance if both parents have asthma. While
the jury may still be out concerning the
verdict behind the hygiene hypothesis,
animal and human studies clearly point to
strong roles for both genes and environment in susceptibility to allergy. As data in
support of this hypothesis continue to
grow, the old saying concerning a dirty
child—that “It’s good for their immune
system”—may actually hold true!
Centers for Disease Control and Prevention. 2012.
CDC: Preventing Chronic Disease 9: 110054.
Liu, A. H., and Murphy, J. R. 2003. Hygiene hypothesis: Fact or fiction? Journal of Allergy and Clinical Immunology 111(3):471–478.
Okada, H., et al. 2010. The hygiene hypothesis for
autoimmune and allergic diseases: An update.
Clinical and Experimental Immunology 160:1.
Sironi, M., and M. Clerici. 2010. The hygiene
hypothesis: An evolutionary perspective. Microbes
and Infection 12:421.
Overview of the Immune System
FIGURE 1-10 Patient suffering from hay fever as a result
of an allergic reaction. Such hypersensitivity reactions result from
sensitization caused by previous exposure to an antigen in some
individuals. In the allergic individual, histamines are released as a part
of the hypersensitivity response and cause sneezing, runny nose,
watery eyes, and such during each subsequent exposure to the antigen (now called an allergen) [Source: Chris Rout/Alamy.]
following extreme inflammation. A significant fraction of
our health resources is expended to care for those suffering
from allergies and asthma. One particularly interesting
rationale to explain the unexpected rise in allergic disease is
called the hygiene hypothesis and is discussed in the Clinical
Focus on page 20.
Autoimmune Disease
Sometimes the immune system malfunctions and a breakdown in self-tolerance occurs. This could be caused by a
sudden inability to distinguish between self and nonself or
by a misinterpretation of a self-component as dangerous,
causing an immune attack on host tissues. This condition,
called autoimmunity, can result in a number of chronic
debilitating diseases. The symptoms of autoimmunity differ,
depending on which tissues or organs are under attack. For
example, multiple sclerosis is due to an autoimmune attack
on a protein in nerve sheaths in the brain and central nervous system that results in neuromuscular dysfunction.
Crohn’s disease is an attack on intestinal tissues that leads to
destruction of gut epithelia and poor absorption of food.
One of the most common autoimmune disorders, rheumatoid arthritis, results from an immune attack on joints of the
hands, feet, arms, and legs.
Both genetic and environmental factors are likely involved
in the development of most autoimmune diseases. However,
the exact combination of genes and environmental exposures that favor the development of a particular autoimmune
disease are difficult to pin down, and constitute very active
areas of immunologic research. Recent discoveries in these
areas and the search for improved treatments are all covered
in greater detail in Chapter 16.
|
CHAPTER 1
21
Immune Deficiency
In most cases, when a component of innate or adaptive
immunity is absent or defective, the host suffers from some
form of immunodeficiency. Some of these deficiencies produce major clinical effects, including death, while others are
more minor or even difficult to detect. Immune deficiency
can arise due to inherited genetic factors (called primary
immunodeficiencies) or as a result of disruption/damage by
chemical, physical, or biological agents (termed secondary
immunodeficiencies). Both of these forms of immune deficiency are discussed in greater detail in Chapter 18.
The severity of the disease resulting from immune deficiency depends on the number and type of affected immune
response components. A common type of primary immunodeficiency in North America is a selective immunodeficiency in which only one type of antibody, called
Immunoglobulin A is lacking; the symptoms may be an
increase in certain types of infections, or the deficiency may
even go unnoticed. In contrast, a more rare but much more
extreme deficiency, called severe combined immunodeficiency (SCID), affects both B and T cells and basically wipes
out adaptive immunity. When untreated, SCID frequently
results in death from infection at an early age.
By far, the most common form of secondary immunodeficiency is Acquired Immune Deficiency Syndrome (AIDS),
resulting from infection with Human Immunodeficiency
Virus (HIV). As discussed further in Chapter 18, humans do
not effectively recognize and eradicate this virus. Instead, a
state of persistent infection occurs, with HIV hiding inside
the genomes of TH cells, its target cell type and the immune
cell type that is critical to guiding the direction of the adaptive immune response. As the immune attack on the virus
mounts, more and more of these TH cells are lost. When the
disease progresses to AIDS, so many TH cells have been
destroyed or otherwise rendered dysfunctional that a gradual collapse of the immune system occurs. It is estimated
that at the end of 2010, more than 34 million people worldwide suffered from this disease (for more current numbers,
see www.unaids.org), which if not treated can be fatal. For
patients with access, certain anti-retroviral treatments can
now prolong life with HIV almost indefinitely. However,
there is neither a vaccine nor a cure for this disease.
It is important to note that many pervasive pathogens in
our environment cause no problem for healthy individuals
thanks to the immunity that develops following initial exposure. However, individuals with primary or secondary deficiencies in immune function become highly susceptible to
disease caused by these ubiquitous microbes. For example,
the fungus Candida albicans, present nearly everywhere and
a nonissue for most individuals, can cause an irritating rash
and a spreading infection in the mucosal surface of the
mouth and vagina in patients suffering from immune deficiency (see Figure 1-6b). The resulting rash, called thrush,
can sometimes be the first sign of immune dysfunction (Figure 1-11). If left unchecked, C. albicans can spread, causing
systemic candidiasis, a life-threatening condition. Such
22
PA R T I
|
Introduction
FIGURE 1-11 An immune deficient patient suffering from
oral thrush due to opportunistic infection by Candida
albicans. [Creative Commons <http://en.wikipedia.org/wiki/File: Thrush.JPG]
infections by ubiquitous microorganisms that cause no harm
in an immune competent host but which are often observed
only in cases of underlying immune deficiency, are termed
opportunistic infections. Several rarely seen opportunistic
infections identified in early AIDS patients were the first
signs that these patients had seriously compromised immune
systems, and helped scientists to identify the underlying
cause. Unchecked opportunistic infections are still the main
cause of death in AIDS patients.
The Immune Response Renders Tissue
Transplantation Challenging
Normally, when the immune system encounters foreign cells,
it responds strongly to rid the host of the presumed invader.
However, in the case of transplantation, these cells or tissues
from a donor may be the only possible treatment for lifethreatening disease. For example, it is estimated that more than
70,000 persons in the United States alone would benefit from a
kidney transplant. The fact that the immune system will attack
and reject any transplanted organ that is nonself, or not a
genetic match, raises a formidable barrier to this potentially
lifesaving treatment, presenting a particularly unique challenge
to clinicians who treat these patients. While the rejection of
this transplant by the recipient’s immune system may be seen
as a “failure,” in fact it is just a consequence of the immune
system functioning properly. Normal tolerance processes governing self/nonself discrimination and immune engagement
caused by danger signals (partially the result of the trauma
caused by surgical transplantation) lead to the rapid influx of
immune cells and coordinated attacks on the new resident
cells. Some of these transplant rejection responses can be suppressed using immune inhibitory drugs, but treatment with
these drugs also suppresses general immune function, leaving
the host susceptible to opportunistic infections.
Research related to transplantation studies has played a
major role in the development of the field of Immunology. A
Nobel Prize was awarded to Karl Landsteiner (mentioned
earlier for his contributions to the concept of immune specificity) in 1930 for the discovery of the human ABO blood
groups, a finding that allowed blood transfusions to be carried out safely. In 1980, G. Snell, J. Dausset, and B. Benacerraf
were recognized for discovery of the major histocompatibility
complex (MHC). These are the tissue antigens that differ
most between non-genetically identical individuals, and are
thus one of the primary targets of immune rejection of transplanted tissues. Finally, in 1991 E. D. Thomas and J. Murray
were awarded Nobel prizes for treatment advances that paved
the way for more clinically successful tissue transplants (see
Table 1-2). Development of procedures that would allow a
foreign organ or cells to be accepted without suppressing
immunity to all antigens still remains a major goal, and a
challenge, for immunologists today (see Chapter 16).
Cancer Presents a Unique Challenge
to the Immune Response
Cancer, or malignancy, occurs in host cells when they
begin to divide out of control. Since these cells are self in
origin, self-tolerance mechanisms can inhibit the development of an immune response, making the detection and
eradication of cancerous cells a continual challenge. That
said, it is clear that many tumor cells do express unique or
developmentally inappropriate proteins, making them
potential targets for immune cell recognition and elimination, as well as targets for therapeutic intervention. However, as with many microbial pathogens, the increased
genetic instability of these rapidly dividing cells gives them
an advantage in terms of evading immune detection and
elimination machinery.
We now know that the immune system actively participates in the detection and control of cancer in the body (see
Chapter 19). The number of malignant disorders that arise in
individuals with compromised immunity highlights the
degree to which the immune system normally controls the
development of cancer. Both innate and adaptive elements
have been shown to be involved in this process, although
adaptive immunity likely plays a more significant role. However, associations between inflammation and the development
of cancer, as well as the degree to which cancerous cells evolve
to become more aggressive and evasive under pressure from
the immune system, have demonstrated that the immune
response to cancer can have both healing and disease-inducing
characteristics. As the mechanics of these elements are
resolved in greater detail, there is hope that therapies can be
designed to boost or maximize the anti-tumor effects of
Overview of the Immune System
immune cells while dampening their tumor enhancing
activities.
Our understanding of the immune system has clearly
come a very long way in a fairly short time. Yet much still
remains to be learned about the mammalian immune
response and the ways in which this system interacts with
other body systems. With this enhanced knowledge, the
|
CHAPTER 1
23
hope is that we will be better poised to design ways to modulate these immune pathways through intervention. This
would allow us to develop more effective prevention and
treatment strategies for cancer and other diseases that plague
society today, not to mention preparing us to respond
quickly to the new diseases or infectious agents that will
undoubtedly arise in the future.
S U M M A R Y
■
■
■
■
■
■
■
■
■
Immunity is the state of protection against foreign pathogens or substances (antigens).
Vaccination is a means to prepare the immune system to
effectively eradicate an infectious agent before it can cause
disease, and its widespread use has saved many lives.
Humoral immunity involves combating pathogens via
antibodies, which are produced by B cells and can be found
in bodily fluids. Antibodies can be transferred between
individuals to provide passive immune protection.
Cell-mediated immunity involves primarily antigenspecific T lymphocytes, which act to eradicate pathogens
or otherwise aid other cells in inducing immunity.
Pathogens fall into four major categories and come in
many forms. The immune response quickly becomes tailored to the type of organism involved.
The immune response relies on recognition molecules that
can be germline encoded or randomly generated.
The process of self-tolerance ensures that the immune
system avoids destroying host tissue.
The vertebrate immune response can be divided into two
interconnected arms of immunity: innate and adaptive.
Innate responses are the first line of defense, utilizing
germline-encoded recognition molecules and phagocytic
■
■
■
■
■
■
cells. Innate immunity is faster but less specific than adaptive responses, which take several days but are highly
antigen specific.
Innate and adaptive immunity operate cooperatively; activation of the innate immune response produces signals that
stimulate and direct subsequent adaptive immune pathways.
Adaptive immunity relies upon surface receptors, called
B- and T-cell receptors, that are randomly generated by
DNA rearrangements in developing B and T cells.
Clonal selection is the process by which individual T and
B lymphocytes are engaged by antigen and cloned to create a population of antigen-reactive cells.
Memory cells are residual B and T cells that remain after
antigen exposure and that pick up where they left off during a subsequent, or secondary, response.
Dysfunctions of the immune system include common
maladies such as allergies, asthma, and autoimmune disease (overly active or misdirected immune responses) as
well as immune deficiency (insufficient immune responses).
Transplanted tissues and cancer present unique challenges to
clinicians, because the healthy immune system typically
rejects or destroys nonself proteins, such as those encountered
in most transplant situations, and tolerates self cells.
R E F E R E N C E S
Burnet, F. M. 1959. The Clonal Selection Theory of Acquired
Immunity. Cambridge University Press, Cambridge, England.
Paul, W., ed. 2003. Fundamental Immunology, 5th ed. Lippincott
Williams & Wilkins, Philadelphia.
Desour, L. 1922. Pasteur and His Work (translated by A. F. and
B. H. Wedd). T. Fisher Unwin, London.
Silverstein, A. M. 1979. History of immunology. Cellular versus
humoral immunity: determinants and consequences of an
epic 19th century battle. Cellular immunology. 48:208.
Kimbrell, D. A., and B. Beutler. 2001. The evolution and genetics
of innate immunity. Nature Reviews Genetics 2:256.
Kindt, T. J., and J. D. Capra. 1984. The Antibody Enigma. Plenum
Press, New York.
Landsteiner, K. 1947. The Specificity of Serologic Reactions.
Harvard University Press, Cambridge, MA.
Medawar, P. B. 1958. The Immunology of Transplantation: The
Harvey Lectures, 1956–1957. Academic Press, New York.
Metchnikoff, E. 1905. Immunity in the Infectious Diseases.
Macmillan, New York.
Useful Web Sites
www.aai.org The Web site of the American Association
of Immunologists contains a good deal of information of
interest to immunologists.
www.ncbi.nlm.nih.gov/PubMed PubMed, the National
Library of Medicine database of more than 9 million publications, is the world’s most comprehensive bibliographic
24
PA R T I
|
Introduction
database for biological and biomedical literature. It is also a
highly user-friendly site.
www.nobelprize.org/nobel_prizes/medicine/
laureates The official Web site of the Nobel Prize in Physi-
www.aaaai.org The American Academy of Allergy
ology or Medicine.
Asthma and Immunology site includes an extensive library
of information about allergic diseases.
www.historyofvaccines.org A Web site run by The
www.who.int/en The World Health Organization directs
and coordinates health-related initiatives and collects worldwide health statistics data on behalf of the United Nations
system.
www.cdc.gov Part of the United States Department of
Health and Human Services, the Centers for Disease Control
and Prevention coordinates health efforts in the United
States and provides statistics on U.S. health and disease.
S T U D Y
College of Physicians of Philadelphia with facts, articles, and
timelines related to vaccine developments.
www.niaid.nih.gov The National Institute of Allergy and
Infectious Disease is a branch of the U.S. National Institute
of Health that specifically deals with research, funding, and
statistics related to basic immunology, allergy, and infectious
disease threats.
Q U E S T I O N S
1. Why was Jenner’s vaccine superior to previous methods for
e. Innate immunity is deployed only during the primary
conferring resistance to smallpox?
2. Did the treatment for rabies used by Pasteur confer active
or passive immunity to the rabies virus? Is there any way to
test this?
f.
g.
3. Infants immediately after birth are often at risk for infec-
tion with group B Streptococcus. A vaccine is proposed for
administration to women of childbearing years. How can
immunizing the mothers help the babies?
4. Indicate to which branch(es) of the immune system the
following statements apply, using H for the humoral
branch and CM for the cell-mediated branch. Some statements may apply to both branches (B).
a. Involves B cells
b. Involves T cells
c. Responds to extracellular bacterial infection
d. Involves secreted antibody
e. Kills virus-infected self cells
5. Adaptive immunity exhibits several characteristic attributes,
which are mediated by lymphocytes. List four attributes of
adaptive immunity and briefly explain how they arise.
6. Name three features of a secondary immune response that
distinguish it from a primary immune response.
7. Give examples of mild and severe consequences of immune
dysfunction. What is the most common cause of immunodeficiency throughout the world today?
8. For each of the following statements, indicate whether the
statement is true or false. If you think the statement is false,
explain why.
a. Booster shots are required because repeated exposure to
an antigen builds a stronger immune response.
b. The gene for the T cell receptor must be cut and spliced
together before it can be expressed.
c. Our bodies face the greatest onslaught from foreign
invaders through our skin.
d. Increased production of antibody in the immune system is driven by the presence of antigen.
h.
i.
j.
response, and adaptive immunity begins during a secondary response.
Autoimmunity and immunodeficiency are two different
terms for the same set of general disorders.
If you receive intravenous immunoglobulin to treat a
snakebite, you will be protected from the venom of this
snake in the future, but not venom from other types of
snakes.
Innate and adaptive immunity work collaboratively to
mount an immune response against pathogens.
The genomic sequences in our circulating T cells for
encoding a T-cell receptor are the same as those our
parents carry in their T cells.
Both the innate and adaptive arms of the immune
response will be capable of responding more efficiently
during a secondary response.
9. What was the significance of the accidental re-inoculation
of some chickens that Pasteur had previously exposed to
the bacteria that causes cholera? Why do you think
these chickens did not die after the first exposure to this
bacterium?
10. Briefly describe the four major categories of pathogen.
Which are likely to be the most homogenous in form and
which the most diverse? Why?
11. Describe how the principle of herd immunity works to
protect unvaccinated individuals. What characteristics of
the pathogen or of the host do you think would most
impact the degree to which this principle begins to take
hold?
12. What is the difference between the discarded instructional
theory for lymphocyte specificity and the selection theory,
which is now the accepted explanation?
13. Compare and contrast innate and adaptive immunity by
matching the following characteristics with the correct arm
of immunity, using I for innate and A for adaptive:
a. Is the first to engage upon initial encounter with antigen
b. Is the most pathogen specific
Overview of the Immune System
c. Employs T and B lymphocytes
d. Adapts during the response
e. Responds identically during a first and second exposure
f.
g.
h.
i.
j.
k.
to the same antigen
Responds more effectively during a subsequent exposure
Includes a memory component
Is the target of vaccination
Can involve the use of PAMP receptors
Involves antigen-specific receptors binding to pathogens
Can be mediated by antibodies
|
CHAPTER 1
25
14. What is meant by the term tolerance? How do we become
tolerant to the structures in our own bodies?
15. What is an antigen? An antibody? What is their relation-
ship to one another?
16. How are PRRs different from B- or T-cell receptors? Which
is most likely to be involved in innate immunity and which
in adaptive immunity?
17. In general terms, what role do cytokines play in the develop-
ment of immunity? How does this compare to chemokines?
This page lelt intentionally blank.
2
Cells, Organs, and
Microenvironments
of the Immune System
A
successful immune response to a pathogen
depends on finely choreographed interactions
among diverse cell types (see Figure 1-9):
innate immune cells that mount a first line of
defense against pathogen, antigen-presenting cells that
communicate the infection to lymphoid cells, which
coordinate the adaptive response and generate the
memory cells, which prevent future infections. The
coordination required for the development of a full
immune response is made possible by the specialized
anatomy and microanatomy of the immune system,
which is dispersed throughout the body and organizes
cells in time and space. Primary lymphoid organs—
including the bone marrow and the thymus—regulate
the development of immune cells from immature
precursors. Secondary lymphoid organs—including the
spleen, lymph nodes, and specialized sites in the gut and
other mucosal tissues—coordinate the encounter of
antigen with antigen-specific lymphocytes and their
development into effector and memory cells. Blood
vessels and lymphatic systems connect these organs,
uniting them into a functional whole.
Remarkably, all functionally specialized, mature blood
cells (red blood cells, granulocytes, macrophages,
dendritic cells, and lymphocytes) arise from a single cell
type, the hematopoietic stem cell (HSC) (Figure 2-1).
The process by which HSCs differentiate into mature
blood cells is called hematopoiesis. Two primary
lymphoid organs are responsible for the development of
stem cells into mature immune cells: the bone marrow,
where HSCs reside and give rise to all cell types; and the
thymus, where T cells complete their maturation. We will
begin this chapter by describing the structure and
function of each cell type that arises from HSCs, and the
structure and function of both the bone marrow and
thymus in the context of hematopoiesis and
thymopoiesis. We will then describe the secondary
lymphoid organs, where the immune response is
initiated. The lymph nodes and spleen will be featured,
but lymphoid tissue associated with mucosal layers will
also be discussed. Four focused discussions are also
Scanning electron micrograph of blood
vessels in a lymph node. Susumu Nishinaga/
Photo Researchers
■
■
■
Cells of the Immune System
Primary Lymphoid Organs—Where Immune Cells
Develop
Secondary Lymphoid Organs—Where the Immune
Response Is Initiated
included in this chapter. Specifically, in two Classic
Experiment Boxes, we describe the discovery of a second
thymus and the history behind the identification of
hematopoietic stem cells. In a Clinical Focus Box, we
discuss the clinical use and promise of hematopoietic
stem cells, and finally, in an Evolution Box, we describe
some intriguing variations in the anatomy of the immune
system among our vertebrate relatives.
Cells of the Immune System
Stem cells are defined by two capacities: (1) the ability to
regenerate or “self-renew” and (2) the ability to differentiate
into all diverse cell types. Embryonic stem cells have the
capacity to generate every specialized cell type in an organism (in other words, they are pluripotent). Adult stem cells,
in contrast, have the capacity to give rise to the diverse cell
types that specify a particular tissue. Multiple adult organs
harbor stem cells (“adult stem cells”) that can give rise to
mature tissue-specific cells. The HSC is considered the paradigmatic adult stem cell because it can differentiate into all
the types of blood cells.
27
28
PA R T I
|
Introduction
Peripheral
tissues
Bone-marrow
Hematopoietic
stem cell
Peripheral
tissues
Selfrenewing
Dendritic cell
Myeloid
progenitor
Lymphoid
progenitor
Natural killer
(NK) cell
Thymus
Macrophage
Monocyte
TH helper cell
Neutrophil
Granulocytemonocyte
progenitor
T-cell
progenitor
TC cytotoxic cell
Eosinophil
Eosinophil
progenitor
B-cell
progenitor
Mast cell
Basophil
B cell
Basophil progenitor
Dendritic cell
Platelets
Megakaryocyte
Erythrocyte
Erythroid
progenitor
FIGURE 2-1 Hematopoiesis. Self-renewing hematopoietic stem cells give rise to lymphoid and myeloid progenitors. Most immune cells
mature in the bone marrow and then travel to peripheral organs via the blood. Some, including mast cells and macrophages, undergo further
maturation outside the bone marrow. T cells develop to maturity in the thymus.
Hematopoietic Stem Cells Have the Ability to
Differentiate into Many Types of Blood Cells
HSCs are rare—fewer than one HSC is present per 5 104
cells in the bone marrow—and their numbers are strictly controlled by a balance of cell division, death, and differentiation.
Under conditions where the immune system is not being
challenged by a pathogen (steady state or homeostatic conditions), most HSCs are quiescent. A small number divide,
generating daughter cells. Some daughter cells retain the
stem-cell characteristics of the mother cell—that is, they
remain self-renewing and able to give rise to all blood cell
types. Other daughter cells differentiate into progenitor cells
that lose their self-renewal capacity and become progressively
more committed to a particular blood cell lineage. As an
organism ages, the number of HSCs decreases, demonstrating that there are limits to an HSC’s self-renewal potential.
When there is an increased demand for hematopoiesis (e.g.,
during an infection or after chemotherapy), HSCs display an
enormous proliferative capacity. This can be demonstrated in
Cells, Organs, and Microenvironments of the Immune System
mice whose hematopoietic systems have been completely
destroyed by a lethal dose of x-rays (950 rads). Such irradiated
mice die within 10 days unless they are infused with normal
bone marrow cells from a genetically identical mouse. Although
a normal mouse has 3 108 bone marrow cells, infusion of
only 104 to 105 bone marrow cells from a donor is sufficient to
completely restore the hematopoietic system, which demonstrates the enormous capacity of HSCs for self-renewal. Our
ability to identify and purify this tiny subpopulation has
improved considerably, and investigators can now theoretically
rescue irradiated animals with just a few purified stem cells,
which give rise to progenitors that proliferate rapidly and
populate the blood system relatively quickly.
Because of the rarity of HSCs and the challenges of culturing them in vitro, investigators initially found it very difficult
to identify and isolate HSCs. The Classic Experiment Box on
|
CHAPTER 2
29
pages 29–31 describes experimental approaches that led to
successful enrichment of HSCs. Briefly, the first successful
efforts featured clever process-of-elimination strategies. Investigators reasoned that undifferentiated hematopoietic stem
cells would not express surface markers specific for mature
cells from the multiple blood lineages (“Lin” markers). They
used several approaches to eliminate cells in the bone marrow
that did express these markers (Lin cells) and then examined
the remaining (Lin) population for its potential to continually give rise to all blood cells over the long term. Lin cells
were, indeed, enriched for this potential. Other investigators
took advantage of two technological developments that revolutionized immunological research—monoclonal antibodies
and flow cytometry (see Chapter 20)—and identified surface
proteins, including CD34, Sca-1, and c-Kit, that distinguished
the rare hematopoietic stem cell population.
BOX 2-1
CLASSIC EXPERIMENT
Isolating Hematopoietic Stem Cells
By the 1960s researchers knew that
hematopoietic stem cells (HSCs) existed
and were a rare population in the bone
marrow. However, they did not know what
distinguished them from the other millions
of cells that crowded the bone marrow. It
was clear that in order to fully understand
how these remarkable cells give rise to all
other blood cells, and in order to harness
this potential for clinical use, investigators
would have to find a way to isolate HSCs.
But how do you find something that is
very rare, whose only distinctive feature is
its function—its ability to give rise to all
blood cells? Investigators devised a variety of strategies that evolved rapidly with
every technological advance, particularly
with the advent of monoclonal antibodies
and flow cytometry.
Regardless of what method one uses to
try to isolate a cell, it is critically important to
have a reliable experimental assay that can
tell you that the cells that you have teased
out are, indeed, the ones you are looking
for. Fundamentally, in order to prove that
you have enriched or purified an HSC, you
have to show that it can proliferate and give
rise to all blood cell types in an animal over
the long term. Many of the assays that were
originally established to show this are still in
use. They include colony formation assays,
where the ability of individual cells to proliferate (and differentiate) is determined by
looking for evidence of cell division either in
vitro (on plates) or in vivo (in the spleens of
irradiated mice). However, the best evidence for successful isolation of HSCs is the
demonstration that they can restore the
blood cells and immune system of a lethally
irradiated animal, preventing its death. This
can be done for mouse stem cells by injecting stem cell candidates into irradiated
mice and determining if they confer survival and repopulate all blood cell types.
The development of a mouse model that
accepts human hematopoietic stem cells
(the SCID-hu (man) mouse model) has
greatly enhanced investigators’ ability to
verify the pluripotentiality of candidate
human stem cell populations.
In the 1970s investigators did not have
the ability to easily compare differences in
protein and gene expression among single cells, so they had to try to distinguish
cell types based on other physical and
structural features. It wasn’t until monoclonal antibodies were introduced into
research repertoires that investigators
could seriously consider purifying a stem
cell. Monoclonal antibodies (described in
Chapter 20) can be raised to virtually any
protein, lipid, or carbohydrate. Monoclo-
nal antibodies can, themselves, be covalently modified with gold particles,
enzymes, or fluorochromes in order to
visualize their binding by microscopy.
In the early 1980s, investigators reasoned
that HSCs were unlikely to express proteins
specific for mature blood cells. Using monoclonal antibodies raised against multiple
mature cells, they trapped and removed
them from bone marrow cell suspensions,
first via a process called panning, where the
heterogeneous pool of cells was incubated
with antibodies bound to a plastic and then
those cells that did not stick were dislodged
and poured off. The cells that did not stick
to the antibodies were, indeed, enriched (in
some cases by several thousandfold) in cells
with HSC potential. This negative selection
strategy against mature cell lineages continues to be useful and is now referred to as
“lineage or Lin” selection; cells enriched by
this method are referred to as “Lin.” The
panning process that yielded one of the first
images of cells that included human hematopoietic stem cells is shown in Figure 1.
Investigators also worked to identify
surface molecules that were specific to
hematopoietic stem cells, so that they
could positively select them from the diverse
bone marrow cell types. The first protein
that identified human HSCs, now known as
(continued)
PA R T I
30
|
CLASSIC EXPERIMENT
Introduction
(continued)
B
P
E
N S
P L
Eo
P E
M
L
B
E
P
Add bone marrow
suspension
Coat plate with
lineage specific
antibodies
Culture for 1-2 h
S
P
E
Eo N
P
L
P
B
M
P
Discard what
does bind
E
Eo N
L
Keep what
doesn’t bind
B
P
P
S
Lin– cells
Stain and observe
under microscope
M
FIGURE 1
Panning for stem cells. Early approaches to isolate HSCs took advantage of antibodies that were raised against mature blood cells and a process called
panning, where cells are incubated in plastic plates that are coated with antibodies. Specifically, investigators layered a suspension of bone marrow onto
plastic plates coated with antibodies that could bind multiple different mature (“lineage positive”) blood cells. They waited 1 to 2 hours and then washed the
cells that did not stick (the immature, “lineage negative” blood cells) off the plate. Most cells—mature blood cells—did stick firmly to the plate. However, the
cells that did not stick were enriched for hematopoietic stem cells. This process led to the first image of human bone marrow cells enriched for hematopoietic
stem cells by panning. S = stem cell; P = progenitor cell; M monocyte; B basophil; N neutrophil; Eo eosinophil; L lymphocyte; E erythrocyte.
[Emerson, S.G., Colin, A.S., Wang, E.A., Wong, G.G., Clark, S.C., and Nathan, D.G. Purification and Demonstration of Fetal Hematopoietic Progenitors and Demonstration of Recombinant Multipotential Colony-stimulating Activity. J. Clin. Invest., Vol. 76, Sept. 1985, 1286-1290. ©The American Society for Clinical Investigation.]
CD34, was identified with a monoclonal
antibody raised against a tumor of primitive
blood cell types (acute myeloid leukemia).
Although one can positively select cells
from a diverse population using the panning procedures described above (or by
using its more current variant where cells
are applied to columns of resin-bound
monoclonal antibodies or equivalents), the
flow cytometer provides the most efficient
way to pull out a rare population from a
diverse group of cells. This machine,
invented by the Herzenberg laboratory and
its interdisciplinary team of scholars and
inventors, has revolutionized immunology
and clinical medicine. In a nutshell, it is a
machine that allows one to identify, separate, and recover individual cells from a
diverse pool of cells on the basis of the profiles of proteins and/or genes they express.
Chapter 20 provides you with an under-the-
hood introduction to the remarkable technology. Briefly, cells from a heterogeneous
suspension are tagged with (“stained with”)
monoclonal antibodies (or other molecules)
that bind to distinct features and are coupled to distinct fluorochromes. These cells
flow single file in front of lasers that excite
the several fluorochromes, and the intensities of the multiple wavelengths given off by
each individual cell are recorded. Cells that
express specific antigens at desired levels
(e.g., showing evidence for expression of
CD34) can be physically separated from
other cells and recovered for further studies.
In the late 1980s, Irv Weissman and his
laboratories discovered that differences in
expression of the Thy protein (a T-cell
marker), and later the expression of Sca protein, differentiated mouse hematopoietic
stem cells from more mature cells. His laboratory combined both negative and positive
selection techniques to develop one of the
most efficient approaches for hematopoietic stem cell enrichment (Figure 2). As other
surface molecules were identified, the
approach was honed. Currently, HSCs are
most frequently identified by their Lin Sca1 c-Kit (“LSK”) phenotype. It has become
clear that even this subgroup, which represents less than 1% of bone marrow cells, is
phenotypically and functionally heterogeneous. Other surface markers, including
SLAM proteins, that can distinguish among
these subpopulations continue to be identified. The synergy between technical developments and experimental strategies will
undoubtedly continue so that, ultimately,
we will be able to unambiguously identify,
isolate, and manipulate what remains the
holy grail of HSC investigations: the longterm stem cell that can both self-renew and
give rise to all blood cell types.
|
Cells, Organs, and Microenvironments of the Immune System
CHAPTER 2
31
BOX 2-1
(a)
Whole bone
marrow
Lethally irradiated
mouse (950 rads)
2 × 105
unenriched cells
L
P
Eo
L
P
E
S
B
N
E
N
M
P
React with
Fl-antibodies
to differentiation
antigens
Restore hematopoiesis,
mouse lives
Negative selection
against cells that
express mature
(lineage) markers
Lin- cells
1 × 103
partly enriched cells
S
M
P
P
N
E
P
L
B
E
React with
Fl-antibodies
against Sca-1
and c-Kit
Restore hematopoiesis,
mouse lives
Lin-Sca-1+c-Kit+ (LSK) cells
L
Eo
Differentiated
cells
N
Positive selection
for cells that
express stem cell
markers (e.g., Sca-1
and c-Kit)
30–100
fully enriched cells
P
P
Stem Progenitor
cell
cells
Restore hematopoiesis,
mouse lives
(b)
P
100
Survival rate, %
S
(Lin-Sca-1+c-Kit+ (LSK) cells
Fully
enriched
cells
Partly
enriched
cells
(Lin–)
Unenriched
cells
(Whole bone
marrow)
101
102
103
104
105
Number of cells injected into lethally irradiated mouse
FIGURE 2
Current approaches for enrichment of the pluripotent stem cells from bone marrow. A schematic of the type of stem cell enrichment now
routinely employed, but originated by Irv Weissman and colleagues. (a) Enrichment is accomplished by (1) removal (negative selection) of differentiated hematopoietic cells (white) from whole bone marrow after treatment with fluorescently labeled antibodies (Fl-antibodies) specific for membrane molecules expressed
on differentiated (mature) lineages but absent from the undifferentiated stem cells (S) and progenitor cells (P), followed by (2) retention (positive selection) of cells
within the resulting partly enriched preparation that bound to antibodies specific for Sca-1 and c-Kit, two early differentiation antigens. Cells that are enriched by
removal of differentiated cells are referred to as “lineage-minus” or Lin populations. Cells that are further enriched by positive selection according to Sca-1 and
c-Kit expression are referred to as LSK (for LinSca-1c-Kit) cells. M monocyte; B basophil; N neutrophil; Eo eosinophil; L lymphocyte; E erythrocyte.
(b) Enrichment of stem cell preparations is measured by their ability to restore hematopoiesis in lethally irradiated mice. Only animals in which hematopoiesis
occurs survive. Progressive enrichment of stem cells (from whole bone marrow, to Lin populations, to LSK populations) is indicated by the decrease in the number of injected cells needed to restore hematopoiesis. A total enrichment of about 1000-fold is possible by this procedure.
Emerson, S. G., C. A. Sieff, E. A. Wang, G. G. Wong,
S. C. Clark, and D. G. Nathan. 1985. Purification of fetal hematopoietic progenitors and
demonstration of recombinant multipotential colony-stimulating activity. Journal of
Clinical Investigation 76:1286.
Shizuru, J. A., R. S. Negrin, and I. L. Weissman.
2005. Hematopoietic stem and progenitor
cells: clinical and preclinical regeneration of
the hematolymphoid system. Annual Review
of Medicine 56:509.
Weissman, I. L. 2000. Translating stem and progenitor cell biology to the clinic: Barriers and
opportunities. Science 287:1442.
PA R T I
32
|
Introduction
Hematopoiesis Is the Process by Which
Hematopoietic Stem Cells Develop into
Mature Blood Cells
An HSC that is induced to differentiate (undergo hematopoiesis) loses its self-renewal capacity and makes one of two
broad lineage commitment choices (see Figure 2-1). It can
become a common myeloid-erythroid progenitor (CMP),
which gives rise to all red blood cells (the erythroid lineage), granulocytes, monocytes, and macrophages (the myeloid
lineage), or it can become a common lymphoid progenitor
(CLP), which gives rise to B lymphocytes, T lymphocytes,
and NK cells. Myeloid cells and NK cells are members of the
innate immune system, and are the first cells to respond to
infection or other insults. Lymphocytes are members of the
adaptive immune response and generate a refined antigenspecific immune response that also gives rise to immune
memory.
As HSCs progress along their chosen lineages, they lose
the capacity to contribute to other cellular lineages. Interestingly, both myeloid and lymphoid lineages give rise to dendritic cells, antigen-presenting cells with diverse features and
functions that play an important role in initiating adaptive
immune responses. The concentration and frequency of
immune cells in blood are listed in Table 2-1.
Regulation of Lineage Commitment
during Hematopoiesis
Each step a hematopoietic stem cell takes toward commitment
to a particular cellular lineage is accompanied by genetic
changes. Multiple genes that specify lineage commitment have
been identified. Many of these are transcriptional regulators.
For instance, the transcription factor GATA-2 is required for
the development of all hematopoietic lineages; in its absence
animals die during embryogenesis. Another transcriptional
regulator, Bmi-1, is required for the self-renewal capacity of
HSCs, and in its absence animals die within 2 months of birth
because of the failure to repopulate their red and white blood
TABLE 2-1
Concentration and frequency
of cells in human blood
Cell type
Cells/mm3
Red blood cells
5.0 106
Platelets
2.5 105
Leukocytes
7.3 103
Total leukocytes (%)
Neutrophil
3.7–5.1 103
50–70
Lymphocyte
1.5–3.0 103
20–40
Monocyte
1–4.4 102
1–6
Eosinophil
2
1–2.2 10
1–3
Basophil
1.3 102
1
cells. Ikaros and Notch are both families of transcriptional
regulators that have more specific effects on hematopoiesis.
Ikaros is required for lymphoid but not myeloid development; animals survive in its absence but cannot mount a full
immune response (i.e., they are severely immunocompromised). Notch1, one of four Notch family members, regulates
the choice between T and B lymphocyte lineages (see Chapter 9). More master regulators of lineage commitment during
hematopoiesis continue to be identified.
The rate of hematopoiesis, as well as the production and
release of specific cell lineages, is also responsive to environmental changes experienced by an organism. For instance,
infection can result in the release of cytokines that markedly
enhance the development of myeloid cells, including neutrophils. Investigators have also recently shown that the release
of mature cells from the bone marrow is responsive to circadian cycles and regulated by the sympathetic nervous system.
Distinguishing Blood Cells
Early investigators originally classified cells based on their
appearance under a microscope, often with the help of dyes.
Their observations were especially helpful in distinguishing
myeloid from lymphoid lineages, granulocytes from macrophages, neutrophils from basophils and eosinophils. Hematoxylin and eosin (H&E) stains are still commonly used in
combination to distinguish cell types in blood smears and
tissues. They highlight intracellular differences because of
their pH sensitivity and different affinities for charged macromolecules in a cell. Thus, the basic dye hematoxylin binds
basophilic nucleic acids, staining them blue, and the acidic
dye eosin binds eosinophilic proteins in granules and cytoplasm, staining them pink.
Microscopists drew astute inferences about cell function
by detailed examination of the structure stained cells, as well
as the behavior of live cells in solution. The advent of the flow
cytometer in the 1980s revolutionized our understanding of
cell subtypes by allowing us to evaluate multiple surface and
internal proteins expressed by individual cells simultaneously. The development of ever more sophisticated fluorescent microscopy approaches to observe live cells in vitro and
in vivo have allowed investigators to penetrate the complexities of the immune response in time and space. These
advances coupled with abilities to manipulate cell function
genetically have also revealed a remarkable diversity of cell
types among myeloid and lymphoid cells, and continue to
expose new functions and unexpected relationships among
hematopoietic cells. Therefore, while our understanding of
the cell subtypes is impressive, it is by no means complete.
Cells of the Myeloid Lineage Are the
First Responders to Infection
Cells that arise from a common myeloid progenitor (CMP)
include red blood cells (erythroid cells) as well as various
types of white blood cells (myeloid cells such as granulocytes,
monocytes, macrophages, and some dendritic cells). Myeloid
Cells, Organs, and Microenvironments of the Immune System
cells are the first to respond to the invasion of a pathogen and
communicate the presence of an insult to cells of the lymphoid lineage (below). As we will see in Chapter 15, they also
contribute to inflammatory diseases (asthma and allergy).
Granulocytes
Granulocytes are at the front lines of attack during an immune
response and are considered part of the innate immune system.
Granulocytes are white blood cells (leukocytes) that are classified as neutrophils, basophils, mast cells, or eosinophils on the
basis of differences in cellular morphology and the staining of
their characteristic cytoplasmic granules (Figure 2-2). All
granulocytes have multilobed nuclei that make them visually
distinctive and easily distinguishable from lymphocytes, whose
nuclei are round. The cytoplasm of all granulocytes is replete
with granules that are released in response to contact with
pathogens. These granules contain a variety of proteins with
distinct functions: Some damage pathogens directly; some
regulate trafficking and activity of other white blood cells,
including lymphocytes; and some contribute to the remodeling
of tissues at the site of infection. See Table 2-2 for a partial list
of granule proteins and their functions.
Neutrophils constitute the majority (50% to 70%) of circulating leukocytes (see Figure 2-2a) and are much more
numerous than eosinophils (1%–3%), basophils (1%), or
mast cells (1%). After differentiation in the bone marrow,
neutrophils are released into the peripheral blood and circulate for 7 to 10 hours before migrating into the tissues, where
they have a life span of only a few days. In response to many
types of infections, the number of circulating neutrophils
increases significantly and more are recruited to tissues, partially in response to cues the bone marrow receives to produce and release more myeloid cells. The resulting transient
increase in the number of circulating neutrophils, called
leukocytosis, is used medically as an indication of infection.
Neutrophils are recruited to the site of infection in response
to inflammatory molecules (e.g., chemokines) generated by
innate cells (including other neutrophils) that have engaged a
pathogen. Once in tissues, neutrophils phagocytose (engulf)
bacteria very effectively, and also secrete a range of proteins
that have antimicrobial effects and tissue remodeling potential. Neutrophils are the dominant first responders to infection
and the main cellular components of pus, where they accumulate at the end of their short lives. Although once considered a
simple and “disposable” effector cell, the neutrophil has
recently inspired renewed interest from investigations indicating that it may also regulate the adaptive immune response.
Basophils are nonphagocytic granulocytes (see Figure 2-2b)
that contain large granules filled with basophilic proteins
(i.e., they stain blue in standard H&E staining protocols).
Basophils are relatively rare in the circulation, but can be
very potent. In response to binding of circulating antibodies,
basophils release the contents of their granules. Histamine,
one of the best known proteins in basophilic granules,
increases blood vessel permeability and smooth muscle
activity. Basophils (and eosinophils, below) are critical to
|
CHAPTER 2
33
our response to parasites, particularly helminths (worms),
but in areas where worm infection is less prevalent, histamines are best appreciated as the cause of allergy symptoms.
Like neutrophils, basophils may also secrete cytokines that
modulate the adaptive immune response.
Mast cells (see Figure 2-2c) are released from the bone
marrow into the blood as undifferentiated cells; they mature
only after they leave the blood. Mast cells can be found in a
wide variety of tissues, including the skin, connective tissues
of various organs, and mucosal epithelial tissue of the respiratory, genitourinary, and digestive tracts. Like circulating
basophils, these cells have large numbers of cytoplasmic
granules that contain histamine and other pharmacologically active substances. Mast cells also play an important role
in the development of allergies.
Basophils and mast cells share many features and their
relationship is not unequivocally understood. Some speculate
that basophils are the blood-borne version of mast cells; others speculate that they have distinct origins and functions.
Eosinophils, like neutrophils, are motile phagocytic cells
(see Figure 2-2d) that can migrate from the blood into the
tissue spaces. Their phagocytic role is significantly less
important than that of neutrophils, and it is thought that
they play their most important role in the defense against
multicellular parasitic organisms, including worms. They
can be found clustering around invading worms, whose
membranes are damaged by the activity of proteins released
from eosinophilic granules. Like neutrophils and basophils,
eosinophils may also secrete cytokines that regulate B and T
lymphocytes, thereby influencing the adaptive immune
response. In areas where parasites are less of a health problem, eosinophils are better appreciated as contributors to
asthma and allergy symptoms.
Myeloid Antigen-Presenting Cells
Myeloid progenitors also give rise to a group of phagocytic
cells (monocytes, macrophages, and dendritic cells) that
have professional antigen-presenting cell (APC) function
(Figure 2-3). Myeloid APCs are considered cellular bridges
between the innate and adaptive immune systems because
they make contact with a pathogen at the site of infection
and communicate this encounter to T lymphocytes in the
lymph node (“antigen presentation”). Each APC can respond
to pathogens and secrete proteins that attract and activate
other immune cells. Each can ingest pathogens via phagocytosis, digest pathogenic proteins into peptides, then present
these peptide antigens on their membrane surfaces. Each
can be induced to express a set of costimulatory molecules
required for optimal activation of T lymphocytes. However,
it is likely that each plays a distinct role during the immune
response, depending on its locale and its ability to respond
to pathogens. Dendritic cells, in particular, play a primary
role in presenting antigen to—and activating—naïve T cells.
Macrophages and neutrophils are especially efficient in
removing both pathogen and damaged host cells, and can
provide a first line of defense against pathogens.
34
PA R T I
|
Introduction
(a) Neutrophil
Multilobed
nucleus
Granules
Phagosome
(b) Basophil
Glycogen
Granule
(c) Mast cell
Granule
(d) Eosinophil
Granule
FIGURE 2-2 Examples of granulocytes. (a, b, c, d) Hematoxylin and eosin (H&E) stains of indicated cells in blood smears. (a, middle)
Neutrophil engulfing bacteria visualized by scanning electron microscopy (SEM) and colorized digitally. (b, middle) SEM of activated granulocytes (colorized). Each image is accompanied by a cartoon depicting
the typical morphology of the indicated granulocyte. Note differences
in the shape of the nucleus and in the number, color, and shape of the
cytoplasmic granules. [2-2a, left: Science Source/Getty Images; 2-2a, right:
Creative Commons, http://es.wikipedia.org/wiki/Archivo:Neutrophil_with_
anthrax_copy.jpg; 2-2b, left: Dr. Gladden Willis/Visuals Unlimited, Inc.; 2-2b, right:
Steve Gschmeissner/Photo Researchers; 2-2c, left: Courtesy Gwen V. Childs,
Ph.D., University of Arkansas for Medical Sciences; 2-2d, left; Pathpedia.com.]
Cells, Organs, and Microenvironments of the Immune System
TABLE 2-2
|
CHAPTER 2
35
Examples of proteins contained in neutrophil, eosinophil, and basophil granules
Cell type
Molecule in granule
Examples
Function
Neutrophil
Proteases
Antimicrobial proteins
Protease inhibitors
Histamine
Elastase, Collagenase
Defensins, lysozyme
1-anti-trypsin
Tissue remodeling
Direct harm to pathogens
Regulation of proteases
Vasodilation, inflammation
Eosinophil
Cationic proteins
Ribonucleases
Cytokines
Chemokines
EPO
MBP
ECP, EDN
IL-4, IL-10, IL-13, TNF
RANTES, MIP-1
Induces formation of ROS
Vasodilation, basophil degranulation
Antiviral activity
Modulation of adaptive immune responses
Attract leukocytes
Basophil/Mast Cell
Cytokines
Lipid mediators
Histamine
IL-4, IL-13
Leukotrienes
Modulation of adaptive immune response
Regulation of inflammation
Vasodilation, smooth muscle activation
Monocytes make up about 5% to 10% of white blood cells
and are a heterogeneous group of cells that migrate into tissues and differentiate into a diverse array of tissue-resident
phagocytic cells, including macrophages and dendritic cells
(see Figure 2-3a). During hematopoiesis in the bone marrow,
granulocyte-monocyte progenitor cells differentiate into promonocytes, which leave the bone marrow and enter the
blood, where they further differentiate into mature monocytes. Two broad categories of monocytes have recently been
identified. Inflammatory monocytes enter tissues quickly in
response to infection. Patrolling monocytes, a smaller group
of cells that crawl slowly along blood vessels, provide a reservoir for tissue-resident monocytes in the absence of infection,
and may quell rather than initiate immune responses.
Monocytes that migrate into tissues in response to infection can differentiate into specific tissue macrophages. Like
monocytes, macrophages can play several different roles.
Some macrophages are long-term residents in tissues and
play an important role in regulating their repair and regeneration. Other macrophages participate in the innate
immune response and undergo a number of key changes
when they are stimulated by encounters with pathogens or
tissue damage. These are referred to as inflammatory macrophages and play a dual role in the immune system as effective
phagocytes that can contribute to the clearance of pathogens
from a tissue, as well as antigen-presenting cells that can
activate T lymphocytes. Osteoclasts in the bone, microglial
cells in the central nervous system, and alveolar macrophages in the lung are tissue-specific examples of macrophages with these properties.
Activated, inflammatory macrophages are more effective
than resting ones in eliminating potential pathogens for several reasons: They exhibit greater phagocytic activity, an
increased ability to kill ingested microbes, increased secretion of inflammatory and cytotoxic mediators, and the abil-
ity to activate T cells. More will be said about the
antimicrobial activities of macrophages in Chapter 5. Activated macrophages also function more effectively as antigenpresenting cells for helper T cells (TH cells), which, in turn
regulate and enhance macrophage activity. Thus, macrophages and TH cells facilitate each other’s activation during
the immune response.
Many macrophages also express receptors for certain
classes of antibody. If an antigen (e.g., a bacterium) is coated
with the appropriate antibody, the complex of antigen and
antibody binds to antibody receptors on the macrophage
membrane more readily than antigen alone and phagocytosis
is enhanced. In one study, for example, the rate of phagocytosis of an antigen was 4000-fold higher in the presence of
specific antibody to the antigen than in its absence. Thus, an
antibody is an example of an opsonin, a molecule that binds
an antigen marking it for recognition by immune cells. The
modification of particulate antigens with opsonins (which
come in a variety of forms) is called opsonization, a term
from the Greek that literally means “to supply food” or “make
tasty.” Opsonization is traditionally described as a process
that enhances phagocytosis of an antigen, but it serves multiple purposes that will be discussed in subsequent chapters.
Although most of the antigen ingested by macrophages is
degraded and eliminated, early experiments with radiolabeled
antigens demonstrated the presence of antigen peptides on the
macrophage membrane. Although the macrophage is a very
capable antigen-presenting cell, the dendritic cell is considered
the most efficient activator of naïve T cells.
The discovery of the dendritic cell (DC) by Ralph Steinman in the mid 1970s resulted in awarding of the Nobel Prize
in 2011. Dendritic cells are critical for the initiation of the
immune response and acquired their name because they are
covered with long membranous extensions that resemble the
dendrites of nerve cells and extend and retract dynamically,
(a) Monocyte
Lysosome
Nucleus
Phagosome
(b) Macrophage
Phagosome
Pseudopodia
Phagosome
Phagolysosome
Lysosome
(c) Dendritic cell
Phagosome
Processes
Phagosome
Lysosome
(d) Megakaryocyte
Platelets
Megakaryocyte
FIGURE 2-3 Examples of monocytes, macrophages, dendritic cells, and megakaryocytes. (a, d) H&E stain of blood
smear. (b) H&E stain of tissue section. (b, middle) SEM of macrophage engulfing mycobacteria (colorized). (c) SEM micrograph.
Each image is accompanied by a cartoon depicting the typical
morphology of the indicated cell. Note that macrophages are
five- to tenfold larger than monocytes and contain more organelles, especially lysosomes. [2-3a, left: Pathpedia.com; 2-3b, left: Courtesy Dr. Thomas Caceci, Virginia-Maryland Regional College of Veterinary
Medicine; 2-3b, middle: SPL/Photo Researchers; 2-3c, left: David Scharf/
Photo Researchers; 2-3d, left: Carolina Biological Supply Co./Visuals
Unlimited.]
Cells, Organs, and Microenvironments of the Immune System
increasing the surface area available for browsing lymphocytes. They are more diverse a population of cells than once
was thought, and seem to arise from both the myeloid and
lymphoid lineages of hematopoietic cells. The functional
distinctions among these diverse cells are still being clarified
and are likely critically important in tailoring immune
responses to distinct pathogens and targeting responding
cells to distinct tissues.
Dendritic cells perform the distinct functions of antigen
capture in one location and antigen presentation in another.
Outside lymph nodes, immature forms of these cells monitor the body for signs of invasion by pathogens and capture
intruding or foreign antigens. They process these antigens,
then migrate to lymph nodes, where they present the antigen
to naïve T cells, initiating the adaptive immune response.
When acting as sentinels in the periphery, immature dendritic cells take on their cargo of antigen in three ways. They
engulf it by phagocytosis, internalize it by receptor-mediated
endocytosis, or imbibe it by pinocytosis. Indeed, immature
dendritic cells pinocytose fluid volumes of 1000 to 1500 m3
per hour, a volume that rivals that of the cell itself. Through
a process of maturation, they shift from an antigen-capturing
phenotype to one that is specialized for presentation of
antigen to T cells. In making the transition, some attributes
are lost and others are gained. Lost is the capacity for phagocytosis and large-scale pinocytosis. However, the ability to
present antigen increases significantly, as does the expression
of costimulatory molecules that are essential for the activation of naïve T cells. After activation, dendritic cells abandon
residency in peripheral tissues, enter the blood or lymphatic
circulation, and migrate to regions of the lymphoid organs,
where T cells reside, and present antigen.
It is important to note that, although they share a name,
follicular dendritic cells do not arise in bone marrow and
have completely different functions from those described for
the dendritic cells discussed above. Follicular dendritic cells
do not function as antigen-presenting cells for TH-cell activation. These dendritic cells were named for their exclusive
location in organized structures of the lymph node called
lymph follicles, which are rich in B cells. As discussed in
Chapters 12 and 14, the interaction of B cells with follicular
dendritic cells is an important step in the maturation and
diversification of B cells.
It is clear that myeloid cells are not the only cells that can
present antigen efficiently. As mentioned above, lymphoidderived dendritic cells are fully capable APCs. In addition,
activated B lymphocytes can act as professional antigenpresenting cells. B cells can internalize antigen very efficiently via their antigen-specific receptor, and can process
and present antigenic peptides at the cell surface. Activated
B cells also express the full complement of costimulatory
molecules that are required to activate T cells. By presenting
antigen directly to T cells, B cells efficiently solicit help, in
the form of cytokines, that induces their differentiation into
memory cells, as well as into antibody-producing cells
(plasma cells).
|
CHAPTER 2
37
Erythroid Cells
Cells of the erythroid lineage—erythrocytes, or red blood
cells—also arise from a common myeloid precursor (sometimes referred to as a common myeloid-erythroid precursor). They contain high concentrations of hemoglobin, and
circulate through blood vessels and capillaries delivering
oxygen to surrounding cells and tissues. Damaged red blood
cells can also release signals (free radicals) that induce innate
immune activity. In mammals, erythrocytes are anuclear;
their nucleated precursors (erythroblasts) extrude their
nuclei in the bone marrow. However, the erythrocytes of
almost all nonmammalian vertebrates (birds, fish, amphibians, and reptiles) retain their nuclei. Erythrocyte size and
shape vary considerably across the animal kingdom—the
largest red blood cells can be found among some amphibians, and the smallest among some deer species.
Megakaryocytes
Megakaryocytes are large myeloid cells that reside in the
bone marrow and give rise to thousands of platelets, very
small cells (or cell fragments) that circulate in the blood and
participate in the formation of blood clots. Although platelets have some of the properties of independent cells, they do
not have their own nuclei.
Cells of the Lymphoid Lineage Regulate
the Adaptive Immune Response
Lymphocytes (Figure 2-4) are the principal cell players in the
adaptive immune response. They represent 20% to 40% of
circulating white blood cells and 99% of cells in the lymph.
Lymphocytes can be broadly subdivided into three major
populations on the basis of functional and phenotypic differences: B lymphocytes (B cells), T lymphocytes (T cells), and
natural killer (NK) cells. In humans, approximately a trillion
(1012) lymphocytes circulate continuously through the blood
and lymph and migrate into the tissue spaces and lymphoid
organs. We briefly review the general characteristics and
functions of each lymphocyte group and its subsets below.
Lymphocytes are relatively nondescript cells that are very
difficult to distinguish morphologically. T and B cells, in
particular, appear identical under a microscope. We therefore rely heavily on the signature of surface proteins they
express to differentiate among lymphocyte subpopulations.
Surface proteins expressed by immune cells are often
referred to by the cluster of differentiation (CD or cluster of
designation) nomenclature. This nomenclature was established
in 1982 by an international group of investigators who recognized that many of the new antibodies produced by laboratories all over the world (largely in response to the advent of
monoclonal antibody technology) were seeing the same proteins. They therefore defined clusters of antibodies that
appeared to be seeing the same protein and assigned a name—
a cluster of differentiation or CD—to each group. Although
originally designed to categorize the multiple antibodies, the
38
PA R T I
|
Introduction
(b) Lymphocyte with red
blood cells
(a) Lymphocyte
TH helper cell
TC cytotoxic T cell
B cell
(c) Plasma cell
Plasma cell
(d) NK cell
(e)
CLP
NK
T
B
TH
Natural killer
(NK) cell
FIGURE 2-4 Examples of lymphocytes. (a, c, d) H&E stain of
blood smear showing typical lymphocyte. Note that naïve B cells
and T cells look identical by microscopy. (b) Scanning electron
micrograph of lymphocytes and red blood cells. Cartoons depicting
the typical morphology of the cells indicated accompany each
image (including three different lymphocytes that would all have
the same appearance). Note the enlarged area of cytoplasm of the
plasma cell, which is occupied by an extensive network of endoplasmic reticulum and Golgi—an indication of the cell’s dedication to
CD nomenclature is now firmly associated with specific surface proteins found on cells of many types. Table 2-3 lists some
common CD molecules found on human and mouse lymphocytes. Note that the shift from use of a “common” name to the
more standard “CD” name can take place slowly. (For example,
investigators often still refer to the pan-T cell marker as “Thy1” rather than CD90, and the costimulatory molecules as “B7-1”
and “B7-2,” rather than CD80 and CD86.) Appendix 1 lists over
three hundred CD markers expressed by immune cells.
In addition to their CD surface signatures, each B or T
cell also expresses an antigen-specific receptor (the B cell
receptor (BCR) or the T cell receptor (TCR), respectively)
on its surface. Although the populations of B cells and T cells
express a remarkable diversity of antigen receptors (more
than a billion), all receptors on an individual cell’s surface
have identical structures and therefore have identical specificities for antigen. If a given lymphocyte divides to form two
daughter cells, both daughters bear antigen receptors with
antigen specificities identical to each other and to the parental cell from which they arose, and so will any descendants
they produce. The resulting population of lymphocytes, all
arising from the same founding lymphocyte, is a clone.
TH1
TC
T H2
TREG TH17
antibody secretion. The NK cell also has more cytoplasm than a
naïve lymphocyte; this is full of granules that are used to kill target
cells. (e) A branch diagram that depicts the basic relationship among
the lymphocyte subsets described in the text. [2-4a, left: Fred Hossler/
Visuals Unlimited; 2-4b: Creative Commons, http://commons.wikimedia.org/
wiki/File:SEM_blood_cells.jpg; 2-4c, left: Benjamin Koziner/Phototake; 2-4d:
Courtesy Ira Ames, Ph.D., Dept. Cell & Developmental Biology, SUNY-Upstate
Medical University.]
At any given moment, a human or a mouse will contain
tens of thousands, perhaps a hundred thousand, distinct
mature T- and B-cell clones, each distinguished by its own
unique and identical cohort of antigen receptors. Mature B
cells and T cells are ready to encounter antigen, but they are
considered naïve until they do so. Contact with antigen
induces naïve lymphocytes to proliferate and differentiate
into both effector cells and memory cells. Effector cells carry
out specific functions to combat the pathogen, while the
memory cells persist in the host, and upon rechallenge with
the same antigen mediate a response that is both quicker and
greater in magnitude. The first encounter with antigen is
termed a primary response, and the re-encounter a secondary response.
B Lymphocytes
The B lymphocyte (B cell) derived its letter designation
from its site of maturation, in the bursa of Fabricius in birds;
the name turned out to be apt, as bone marrow is its major
site of maturation in humans, mice, and many other mammals. Mature B cells are definitively distinguished from
other lymphocytes and all other cells by their synthesis and
Cells, Organs, and Microenvironments of the Immune System
TABLE 2-3
|
CHAPTER 2
39
Common CD markers used to distinguish functional lymphocyte subpopulations
CD designation
Function
B cell
TH
TC
NK cell
CD2
Adhesion molecule;
signal transduction
CD3
Signal transduction element
of T-cell receptor
CD4
Adhesion molecule that binds
to class II MHC molecules;
signal transduction
(usually)
(usually)
CD5
Unknown
(subset)
CD8
Adhesion molecule that binds
to class I MHC molecules;
signal transduction
(usually)
(usually)
(variable)
CD16 (Fc RIII)
Low-affinity receptor for Fc
region of IgG
CD19
Signal transduction; CD21
co-receptor
CD21 (CR2)
Receptor for complement (C3d
and Epstein-Barr virus)
CD28
Receptor for costimulatory B7
molecule on antigen-presenting cells
CD32 (Fc RII)
Receptor for Fc region of IgG
CD35 (CR1)
Receptor for complement (C3b)
CD40
Signal transduction
CD45
Signal transduction
CD56
Adhesion molecule
Synonyms are shown in parentheses.
display of the B-cell receptor (BCR), a membrane-bound
immunoglobulin (antibody) molecule that binds to antigen.
Each B cell expresses a surface antibody with a unique
specificity, and each of the approximately 1.5–3 105 molecules of surface antibody has identical binding sites for
antigen. B lymphocytes also can improve their ability to bind
antigen through a process known as somatic hypermutation
and can generate antibodies of several different functional
classes through a process known as class switching. Somatic
hypermutation and class switching are covered in detail in
Chapter 12.
Ultimately, activated B cells differentiate into effector
cells known as plasma cells (see Figure 2-4c). Plasma cells
lose expression of surface immunoglobulin and become
highly specialized for secretion of antibody. A single cell is
capable of secreting from a few hundred to more than a
thousand molecules of antibody per second. Plasma cells
do not divide and, although some long-lived populations
of plasma cells are found in bone marrow, many die within
1 or 2 weeks.
T Lymphocytes
T lymphocytes (T cells) derive their letter designation
from their site of maturation in the thymus. Like the B cell,
the T cell expresses a unique antigen-binding receptor called
the T-cell receptor. However, unlike membrane-bound antibodies on B cells, which can recognize soluble or particulate
antigen, T-cell receptors only recognize processed pieces of
antigen (typically peptides) bound to cell membrane proteins called major histocompatibility complex (MHC)
molecules. MHC molecules are genetically diverse glycoproteins found on cell membranes (their structure and function are covered in detail in Chapter 8). The ability of MHC
molecules to form complexes with antigen allows cells to
decorate their surfaces with internal (foreign and self) proteins, exposing them to browsing T cells. MHC comes in two
versions: class I MHC molecules, which are expressed by
nearly all nucleated cells of vertebrate species, and class II
MHC molecules, which are expressed by professional
antigen-presenting cells and a few other cell types during
inflammation.
40
PA R T I
|
Introduction
T lymphocytes are divided into two major cell types—T
helper (TH) cells and T cytotoxic (TC) cells—that can be
distinguished from one another by the presence of either
CD4 or CD8 membrane glycoproteins on their surfaces. T
cells displaying CD4 generally function as TH cells and recognize antigen in complex with MHC class II, whereas those
displaying CD8 generally function as TC cells and recognize
antigen in complex with MHC class I. The ratio of CD4 to
CD8 T cells is approximately 2:1 in normal mouse and
human peripheral blood. A change in this ratio is often an
indicator of immunodeficiency disease (e.g., HIV), autoimmune diseases, and other disorders.
Naïve CD8 T cells browse the surfaces of antigenpresenting cells with their T-cell receptors. If and when they
bind to an MHC-peptide complex, they become activated,
proliferate, and differentiate into an effector cell called a cytotoxic T lymphocyte (CTL). The CTL has a vital function in
monitoring the cells of the body and eliminating any cells
that display foreign antigen complexed with class I MHC,
such as virus-infected cells, tumor cells, and cells of a foreign
tissue graft. To proliferate and differentiate optimally, naïve
CD8 T cells also need help from mature CD4 T cells.
Naïve CD4 T cells also browse the surfaces of antigenpresenting cells with their T-cell receptors. If and when they
recognize an MHC-peptide complex, they can become activated and proliferate and differentiate into one of a variety of
effector T cell subsets (see Figure 2-4e). T helper type 1
(TH1) cells regulate the immune response to intracellular
pathogens, and T helper type 2 (TH2) cells regulate the
response to many extracellular pathogens. Two additional TH
cell subsets have been recently identified. T helper type 17
cells (TH17), so named because they secrete IL-17, play an
important role in cell-mediated immunity and may help the
defense against fungi. T follicular helper cells (TFH) play an
important role in humoral immunity and regulate B-cell
development in germinal centers. Which helper subtype
dominates a response depends largely on what type of pathogen (intracellular versus extracellular, viral, bacterial, fungal,
helminth) has infected an animal. Each of these CD4 T-cell
subtypes produces a different set of cytokines that enable or
“help” the activation of B cells, TC cells, macrophages, and
various other cells that participate in the immune response.
The network of cytokines that regulate and are produced by
these effector cells is described in detail in Chapter 11.
Another type of CD4 T cell, the regulatory T cell (TREG),
has the unique capacity to inhibit an immune response. These
cells can arise during maturation in the thymus from autoreactive cells (natural TREG), but also can be induced at the site of
an immune response in an antigen-dependent manner
(induced TREG). They are identified by the presence of CD4 and
CD25 on their surfaces, as well as the expression of the internal
transcription factor FoxP3. TREG cells are critical in helping us
to quell autoreactive responses that have not been avoided via
other mechanisms. In fact, mice depleted of TREG cells are
afflicted with a constellation of destructive self-reactive inflammatory reactions. However, TREG cells may also play a role in
limiting our normal T-cell response to a pathogen. CD4 and
CD8 T-cell subpopulations may be even more diverse than
currently described, and the field should expect identification
of additional functional subtypes in the future.
Natural Killer Cells
Natural killer (NK) cells are lymphoid cells that are closely
related to B and T cells. However, they do not express antigenspecific receptors and are considered part of the innate
immune system. They are distinguished by the expression of
a surface marker known as NK1.1, as well as the presence of
cytotoxic granules. Once referred to as “large granular lymphocytes” because of their appearance under a microscope,
NK cells constitute 5% to 10% of lymphocytes in human
peripheral blood. They are efficient cell killers and attack a
variety of abnormal cells, including some tumor cells and
some cells infected with virus. They distinguish cells that
should be killed from normal cells in a very clever way: by
“recognizing” the absence of MHC class I, which is expressed
by almost all normal cells, but is specifically down-regulated
by some tumors and in response to some viral infections.
How can cells recognize an absence? NK cells express a variety of receptors for self MHC class I that, when engaged,
inhibit their ability to kill other cells. When NK cells encounter cells that have lost their MHC class I, these receptors are
no longer engaged and can no longer inhibit the potent cytotoxic tendencies of the NK cell, which then releases its cytolytic granules and kills the abnormal target cell.
NK cells also express receptors for immunoglobulins and
can therefore decorate themselves with antibodies that bind
pathogens or proteins from pathogens on the surface of infected
cells. This allows an NK cell to make a connection with a variety
of target cells (independently of their MHC class I expression).
Once the antibodies bring the NK cell in contact with target
cells, the NK cell releases its granules and induces cell death.
The mechanism of NK-cell cytotoxicity, the focus of much current experimental study, is described further in Chapter 13.
NKT Cells
Another type of cell in the lymphoid lineage, NKT cells, have
received a great deal of recent attention and share features
with both conventional T lymphocytes and NK cells. Like T
cells, NKT cells have T-cell receptors (TCRs), and some
express CD4. Unlike most T cells, however, the TCRs of NKT
cells are not very diverse and recognize specific lipids and
glycolipids presented by a molecule related to MHC proteins
known as CD1. Like their innate immune relatives, NK cells,
NKT cells have antibody receptors, as well as other receptors
classically associated with NK cells. Activated NKT cells can
release cytotoxic granules that kill target cells, but they can
also release large quantities of cytokines that can both
enhance and suppress the immune response. They appear to
be involved in human asthma, but also may inhibit the development of autoimmunity and cancer. Understanding the
exact role of NKT cells in immunity is one research priority.
Cells, Organs, and Microenvironments of the Immune System
Primary Lymphoid Organs—Where
Immune Cells Develop
The ability of any stem cell to self-renew and differentiate
depends on the structural organization and cellular function
of specialized anatomic microenvironments known as stem
cell niches. These sequestered regions are typically populated by a supportive network of stromal cells. Stem cell
niche stromal cells express soluble and membrane-bound
proteins that regulate cell survival, proliferation, differentiation, and trafficking. The organs that have microenvironments that support the differentiation of hematopoietic stem
cells actually change over the course of embryonic development. However, by mid to late gestation, HSCs take up residence in the bone marrow, which remains the primary site
of hematopoiesis throughout adult life. The bone marrow
supports the maturation of all erythroid and myeloid cells
and, in humans and mice, the maturation of B lymphocytes
(as described in Chapter 10).
HSCs are also found in blood and may naturally recirculate between the bone marrow and other tissues. This observation has simplified the process used to transplant blood
cell progenitors from donors into patients who are deficient
(e.g., patients who have undergone chemotherapy). Whereas
once it was always necessary to aspirate bone marrow from
the donor—a painful process that requires anesthesia—it is
now sometimes possible to use enriched hematopoietic precursors from donor blood, which is much more easily
obtained (see the Clinical Focus Box 2-2 on pages 42–43).
Unlike B lymphocytes, T lymphocytes do not complete
their maturation in the bone marrow. T lymphocyte precursors need to leave the bone marrow and travel to the unique
microenvironments provided by the other primary lymphoid organ, the thymus, in order to develop into functional
cells. The structure and function of the thymus will be discussed below and in more detail in Chapter 9.
The Bone Marrow Provides Niches for
Hematopoietic Stem Cells to Self-Renew
and Differentiate into Myeloid Cells
and B Lymphocytes
The bone marrow is a primary lymphoid organ that supports self-renewal and differentiation of hematopoietic stem
cells (HSCs) into mature blood cells. Although all bones
contain marrow, the long bones (femur, humerus), hip bones
(ileum), and sternum tend to be the most active sites of
hematopoiesis. The bone marrow is not only responsible for
the development and replenishment of blood cells, but it is
also responsible for maintaining the pool of HSCs throughout the life of an adult vertebrate.
The adult bone marrow (Figure 2-5), the paradigmatic
adult stem cell niche, contains several cell types that coordinate HSC development, including (1) osteoblasts, versatile
cells that both generate bone and control the differentiation
|
CHAPTER 2
41
of HSCs, (2) endothelial cells that line the blood vessels and
also regulate HSC differentiation, (3) reticular cells that send
processes connecting cells to bone and blood vessels, and,
unexpectedly, (4) sympathetic neurons, which can control the
release of hematopoietic cells from the bone marrow. A
microscopic cross-section reveals that the bone marrow is
tightly packed with stromal cells and hematopoietic cells at
every stage of differentiation. With age, however, fat cells
gradually replace 50% or more of the bone marrow compartment, and the efficiency of hematopoiesis decreases.
The choices that an HSC makes depend largely on the
environmental cues it receives. The bone marrow is packed
with hematopoietic cells at all stages of development, but it is
likely that the precursors of each myeloid and lymphoid subtype mature in distinct environmental micro-niches within
the bone marrow. Our understanding of the microenvironments within the bone marrow that support specific stages of
hematopoiesis is still developing. Evidence suggests, however,
that the endosteal niche (the area directly surrounding the
bone and in contact with bone-producing osteoblasts) and
the vascular niche (the area directly surrounding the blood
vessels and in contact with endothelial cells) play different
roles (see Figure 2-5c). The endosteal niche appears to be
occupied by quiescent HSCs in close association with osteoblasts that regulate stem cell proliferation. The vascular niche
appears to be occupied by HSCs that have been mobilized to
leave the endosteal niche to either differentiate or circulate. In
addition, the more differentiated a cell is, the farther it
appears to migrate from its supportive osteoblasts and the
closer it moves to the more central regions of the bone. For
example, the most immature B lymphocytes are found closest
to the endosteum and osteoblasts, while the more mature B
cells have moved into the more central sinuses of the bone
marrow that are richly served by blood vessels.
Finally, it is important to recognize that the bone marrow
is not only a site for lymphoid and myeloid development but
is also a site to which fully mature myeloid and lymphoid
cells can return. Mature antibody-secreting B cells (plasma
cells) may even take up long-term residence in the bone
marrow. Whole bone marrow transplants, therefore, do not
simply include stem cells but also include mature, functional
cells that can both help and hurt the transplant effort.
The Thymus Is a Primary Lymphoid Organ
Where T Cells Mature
T cell development is not complete until the cells undergo
selection in the thymus (Figure 2-6). The importance of the
thymus in T-cell development was not recognized until the
early 1960s, when J.F.A.P. Miller, an Australian biologist,
worked against the power of popular assumptions to advance
his idea that the thymus was something other than a graveyard for cells. It was an underappreciated organ, very large in
prepubescent animals, that was thought by some to be detrimental to an organism, and by others to be an evolutionary
dead-end. The cells that populated it—small, thin-rimmed,
42
PA R T I
|
Introduction
CLINICAL FOCUS
Stem Cells—Clinical Uses and Potential
Stem cell transplantation holds great
promise for the regeneration of diseased,
damaged, or defective tissue. Hematopoietic stem cells are already used to restore
hematopoietic function, and their use in
the clinic is described below. However,
rapid advances in stem cell research have
raised the possibility that other stem cell
types may soon be routinely employed
for replacement of a variety of cells and
tissues. Two properties of stem cells
underlie their utility and promise. They
have the capacity to give rise to lineages
of differentiated cells, and they are selfrenewing—each division of a stem cell
creates at least one stem cell. If stem cells
are classified according to their descent
and developmental potential, three levels
of stem cells can be recognized: pluripotent, multipotent, and unipotent.
Pluripotent stem cells can give rise to an
entire organism. A fertilized egg, the
zygote, is an example of such a cell. In
humans, the initial divisions of the zygote
and its descendants produce cells that are
also pluripotent. In fact, identical twins
develop when pluripotent cells separate
and develop into genetically identical
fetuses. Multipotent stem cells arise from
embryonic stem cells and can give rise to
a more limited range of cell types. Further
differentiation of multipotent stem cells
leads to the formation of unipotent stem
cells, which can generate only the same
cell type as themselves. (Note that “pluripotent” is often used to describe the hematopoietic stem cell. Within the context of
blood cell lineages this is arguably true;
however, it is probably strictly accurate to
call the HSC a multipotent stem cell.)
Pluripotent cells, called embryonic
stem cells, or simply ES cells, can be iso-
lated from early embryos, and for many
years it has been possible to grow mouse
ES cells as cell lines in the laboratory. Strikingly, these cells can be induced to generate many different types of cells. Mouse ES
cells have been shown to give rise to
muscle cells, nerve cells, liver cells, pancreatic cells, and hematopoietic cells.
Advances have made it possible to
grow lines of human pluripotent stem cells
and, most recently, to induce differentiated
human cells to become pluripotent stem
cells. These are developments of considerable importance to the understanding of
human development, and they also have
great therapeutic potential. In vitro studies
of the factors that determine or influence
the development of human pluripotent
stem cells along specific developmental
paths are providing considerable insight
into how cells differentiate into specialized
cell types. This research is driven in part by
the great potential for using pluripotent
stem cells to generate cells and tissues that
could replace diseased or damaged tissue.
Success in this endeavor would be a major
advance because transplantation medicine now depends entirely on donated
organs and tissues, yet the need far exceeds
the number of donations, and the need is
increasing. Success in deriving cells, tissues,
and organs from pluripotent stem cells
could provide skin replacement for burn
patients, heart muscle cells for those with
chronic heart disease, pancreatic islet cells
for patients with diabetes, and neurons for
the treatment of Parkinson’s disease or
Alzheimer’s disease.
The transplantation of HSCs is an
important therapy for patients whose
hematopoietic systems must be replaced.
It has three major applications:
featureless cells called thymocytes—looked dull and inactive. However, Miller proved that the thymus was the allimportant site for the maturation of T lymphocytes (see the
Classic Experiment Box 2-3 on pages 46–47).
T-cell precursors, which still retain the ability to give rise
to multiple hematopoietic cell types, travel via the blood
from the bone marrow to the thymus. Immature T cells,
•
Providing a functional immune system
to individuals with a genetically determined immunodeficiency, such as
severe combined immunodeficiency
(SCID).
•
Replacing a defective hematopoietic
system with a functional one to cure
patients with life-threatening nonmalignant genetic disorders in hematopoiesis, such as sickle-cell anemia or
thalassemia.
•
Restoring the hematopoietic system of
cancer patients after treatment with
doses of chemotherapeutic agents and
radiation. This approach is particularly
applicable to leukemias, including acute
myeloid leukemia, which can be cured
only by destroying the patient’s own
hematopoietic system—the source of
the leukemia cells. Clinicians also hope
that this approach can be used to facilitate treatment of solid tumors. Highdose radiation and cytotoxic regimens
can be much more effective at killing
solid tumors than therapies using more
conventional doses of cytotoxic agents;
however, they destroy the immune system, and stem cell transplantation
makes it possible to recover from such
drastic treatment.
Hematopoietic stem cells have extraordinary powers of regeneration. Experiments in mice indicate that as few as one
HSC can completely restore the erythroid
population and the immune system. In
humans, for instance, as little as 10% of a
donor’s total volume of bone marrow can
provide enough HSCs to completely
restore the recipient’s hematopoietic system. Once injected into a vein, HSCs enter
the circulation and find their own way to
known as thymocytes (thymus cells) because of their site of
maturation, pass through defined developmental stages in
specific thymic microenvironments as they mature into
functional T cells. The thymus is a specialized environment
where immature T cells generate unique antigen receptors
(T cell receptors, or TCRs) and are then selected on the basis
of their reactivity to self MHC-peptide complexes expressed
Cells, Organs, and Microenvironments of the Immune System
|
CHAPTER 2
43
BOX 2-2
the bone marrow, where they begin the
process of engraftment. In addition, HSCs
can be preserved by freezing. This means
that hematopoietic cells can be “banked.”
After collection, the cells are treated with a
cryopreservative, frozen, and then stored
for later use. When needed, the frozen
preparation is thawed and infused into the
patient, where it reconstitutes the hematopoietic system. This cell-freezing technology even makes it possible for
individuals to store their own hematopoietic cells for transplantation to themselves
at a later time. Currently, this procedure is
used to allow cancer patients to donate
cells before undergoing chemotherapy
and radiation treatments, then later reconstitute their hematopoietic system using
their own cells.
Transplantation of stem cell populations may be autologous (the recipient is
also the donor), syngeneic (the donor is
genetically identical; i.e., an identical twin
of the recipient), or allogeneic (the donor
and recipient are not genetically identical). In any transplantation procedure,
genetic differences between donor and
recipient can lead to immune-based
rejection reactions. Aside from host rejection of transplanted tissue (host versus
graft), lymphocytes conveyed to the
recipient via the graft can attack the recipient’s tissues, thereby causing graftversus-host disease (GVHD), a
life-threatening affliction. In order to suppress rejection reactions, powerful immunosuppressive drugs must be used.
Unfortunately, these drugs have serious
side effects, and immunosuppression
increases the patient’s risk of infection and
susceptibility to tumors. Consequently,
HSC transplantation has the fewest complications when there is genetic identity
between donor and recipient.
At one time, bone marrow transplantation was the only way to restore the
hematopoietic system. However, both
peripheral blood and umbilical cord
blood are now also common sources of
hematopoietic stem cells. These alternative sources of HSCs are attractive because
the donor does not have to undergo
anesthesia or the highly invasive procedure used to extract bone marrow.
Although peripheral blood may replace
marrow as a major source of hematopoietic stem cells for many applications,
bone marrow transplantation still has
some advantages (e.g., marrow may
include stem cell subsets that are not as
prevalent in blood). To obtain HSCenriched preparations from peripheral
blood, agents are used to induce increased
numbers of circulating HSCs, and then the
HSC-containing fraction is separated from
the plasma and red blood cells in a process called leukapheresis. If necessary,
further purification can be done to
remove T cells and to enrich the CD34
population.
Umbilical cord blood contains an
unusually high frequency of hematopoietic stem cells. Furthermore, it is obtained
from placental tissue (the “afterbirth”),
which is normally discarded. Consequently, umbilical cord blood has become
an attractive source of cells for HSC transplantation. For reasons that remain
incompletely understood, cord blood
stem cell transplants do not engraft as
reliably as peripheral blood stem cell
transplants; however, grafts of cord blood
cells produce GVHD less frequently than
marrow grafts, probably because cord
blood has fewer mature T cells.
Beyond its current applications in
cancer treatment, autologous stem cell
transplantation can also be useful for
on the surface of thymic stromal cells. Those thymocytes
whose T-cell receptors bind self MHC-peptide complexes
with too high affinity are induced to die (negative selection), and those thymocytes that bind self MHC-peptides
with an intermediate affinity undergo positive selection,
resulting in their survival, maturation, and migration to the
thymic medulla. Most thymocytes do not navigate the jour-
gene therapy, the introduction of a
normal gene to correct a disorder
caused by a defective gene. One of the
most highly publicized gene therapy
efforts—the introduction of the adenosine deaminase (ADA) gene to correct a
form of severe combined immunodeficiency (SCID)—was performed successfully on hematopoietic stem cells. The
therapy entails removing a sample of
hematopoietic stem cells from a patient,
inserting a functional gene to compensate for the defective one, and then
reinjecting the engineered stem cells
into the donor.
The advantage of using stem cells in
gene therapy is that they are selfrenewing. Consequently, at least in theory, patients would have to receive only a
single injection of engineered stem cells.
In contrast, gene therapy with engineered mature lymphocytes or other
blood cells would require periodic injections because these cells are not capable
of self-renewal. In the case of the SCID
patients, hematopoietic stem cells were
successfully infected with a retrovirus
engineered to express the ADA gene. The
cells were returned to the patients and
did, indeed, correct the deficiency.
Patients who previously could not generate lymphocytes to protect themselves
from infection were able to generate normal cells and live relatively normal lives.
Unfortunately, in a number of patients,
the retrovirus used to introduce the ADA
gene integrated into parts of the genome
that resulted in leukemia. Investigators
continue to work to improve the safety
and efficiency of gene delivery; more successful gene therapy efforts are clearly in
the future.
ney through the thymus successfully; in fact, it is estimated
that 95% of thymocytes die in transit. The majority of cells
die because they have too low an affinity for the self-antigenMHC combinations that they encounter on the surface of
thymic epithelial cells and fail to undergo positive selection.
These developmental events take place in several distinct
thymic microenvironments (see Figure 2-6). T-cell precursors
PA R T I
44
|
Introduction
(a)
(b)
Sternum
Humerus
Medullary cavity
Endosteal
niche
Ileum
Vascular
niche
Femur
FIGURE 2-5 The bone marrow microenvironment. (a) Multiple bones support hematopoiesis,
including the hip (ileum), femur, sternum, and
humerus. (b) This figure shows a typical cross-section
of a bone with a medullary (marrow) cavity. (c) Blood
vessels (central sinus and medullary artery) run
through the center of the bone and form a network of
capillaries in close association with bone and bone
surface (endosteum). Both the cells that line the blood
vessels (endothelium) and the cells that line the bone
(osteoblasts) generate niches that support hematopoietic stem cell (HSC) self-renewal and differentiation. The most immature cells appear to be associated
with the endosteal (bone) niche; as they mature, they
migrate toward the vascular (blood vessel) niche.
Fully differentiated cells exit the marrow via blood
vessels. [2-5b; Courtesy of Indiana University School of Medicine.]
HSC
(c)
Reticular cells
Hematopoietic
stem cells (HSC)
Osteoblasts
Endosteal
niche
Vascular
niche
Artery
Vein
Medullary
artery
Medullary
cavity
Central sinus
Sympathetic
neuron
|
Cells, Organs, and Microenvironments of the Immune System
CHAPTER 2
(b)
(a)
Thyroid
Thymus
Nerve
Nerve
Thymus
Heart
Diaphragm
Bone marrow
(c)
Cortex
Medulla
FIGURE 2-6 The structure of the thymus. The thymus is
found just above the heart (a, b) and is largest prior to puberty,
when it begins to shrink. Panel (c) depicts a stained thymus tissue
section and (d) a cartoon of the microenvironments: the cortex,
which is densely populated with DP immature thymocytes (blue)
and the medulla, which is sparsely populated with SP mature
thymocytes. These major regions are separated by the corticomedullary junction (CMJ), where cells enter from and exit to the
bloodstream. The area between the cortex and the thymic capsule, the subcapsular cortex, is a site of much proliferation of the
youngest (DN) thymocytes. The route taken by a typical thymocyte during its development from the DN to DP to SP stages is
shown. Thymocytes are positively selected in the cortex. Autoreactive thymocytes are negatively selected in the medulla; some
may also be negatively selected in the cortex. [2-6c: Dr. Gladden
Willis/Getty Images.]
(d)
Thymus
Capsule
DN
Subcapsular
cortex
DP
Cortex
SP
DN
SP
Cortical
medullary
junction
(CMJ)
SP
Medulla
45
46
PA R T I
|
Introduction
CLASSIC EXPERIMENT
The Discovery of a Thymus—and Two
J.F.A.P. MILLER DISCOVERED THE
FUNCTION OF THE THYMUS
In 1961, Miller, who had been investigating the thymus’s role in leukemia, published a set of observations in The Lancet
that challenged notions that, at best, this
organ served as a cemetery for lymphocytes and, at worst, was detrimental to
health (Figure 1). He noted that when this
organ was removed in very young mice
(in a process known as thymectomy), the
subjects became susceptible to a variety
of infections, failed to reject skin grafts,
and died prematurely. On close examination of their circulating blood cells, they
also appeared to be missing a type of cell
that another investigator, James Gowans,
had associated with cellular and humoral
immune responses. Miller concluded that
the thymus produced functional immune
cells.
Several influential investigators could
not repeat the data and questioned
Miller’s conclusions. Some speculated
that the mouse strain he used was peculiar, others that his mice were exposed to
too many pathogens and their troubles
were secondary to infection. Dr. Miller
responded to each of these criticisms
experimentally, assessing the impact of
thymectomy in different mouse strains
and in germ-free facilities. His results were
unequivocal, and his contention that this
organ generated functional lymphocytes
was vindicated. Elegant experiments by
Dr. Miller, James Gowans, and others subsequently showed that the thymus produced a different type of lymphocyte than
the bone marrow. This cell did not produce antibodies directly, but, instead, was
required for optimal antibody production.
It was called a T cell after the thymus, its
organ of origin. Immature T cells are
known as thymocytes. Miller is one of the
few scientists credited with the discovery
of the function of an entire organ.
A SECOND THYMUS
No one expected a new anatomical discovery in immunology in the 21st century.
However, in 2006 Hans-Reimer Rodewald
and his colleagues reported the existence
of a second thymus in mice. The conventional thymus is a bi-lobed organ that
sits in the thorax right above the heart.
Rodewald and his colleagues discovered
thymic tissue that sits in the neck, near the
cervical vertebrae, of mice. This cervical
FIGURE 1
J.F.A.P. Miller (above) in 1961 and the first page
(opposite) of his Lancet article (1961) describing his
discovery of the function of the thymus. [The Walter and
Eliza Hall Institute of Medical Research.]
thymic tissue is smaller in mass than the
conventional thymus, consists of a single
lobe or clusters of single lobes, and is
populated by relatively more mature
thymocytes. However, it contributes to
T-cell development very effectively and
clearly contributes to the mature T-cell
repertoire. Rodewald’s findings raise the
possibility that some of our older observations and assumptions about thymic
function need to be reexamined. In particular, studies based on thymectomy that
indicated T cells could develop outside
the thymus may need to be reassessed.
The cells found may have come from this
more obscure but functional thymic tissue.
The evolutionary implications of this
thymus are also interesting—thymi are
found in the neck in several species,
including the koala and kangaroo.
REFERENCES
Miller, J. F. 1961. Immunological function of the
thymus. Lancet 2:748–749.
Miller, J.F.A.P. 2002. The discovery of thymus
function and of thymus-derived lymphocytes. Immunological Reviews 185:7–14.
Rodewald, H-R. 2006. A second chance for the
thymus. Nature 441:942–943.
Cells, Organs, and Microenvironments of the Immune System
|
CHAPTER 2
47
BOX 2-3
[Immunological Function of the Thymus, J.F.A.P. Miller,
The Lancet, Elsevier 30 September 1961 © 1961, Elsevier]
48
PA R T I
|
Introduction
enter the thymus in blood vessels at the corticomedullary
junction between the thymic cortex, the outer portion of the
organ, and the thymic medulla, the inner portion of the
organ. At this stage thymocytes express neither CD4 nor
CD8, markers associated with mature T cells. They are therefore called double negative (DN) cells. DN cells first travel to
the region under the thymic capsule, a region referred to as
the subcapsular cortex, where they proliferate and begin to
generate their T-cell receptors. Thymocytes that successfully
express TCRs begin to express both CD4 and CD8, becoming
double positive (DP) cells, and populate the cortex, the site
where most (85% or more) immature T cells are found. The
cortex features a distinct set of stromal cells, cortical thymic
epithelial cells (cTECs), whose long processes are perused by
thymocytes testing the ability of their T-cell receptors to bind
MHC-peptide complexes (Video 2-1). Thymocytes that survive selection move to the thymic medulla, where positively
selected thymocytes encounter specialized stromal cells,
medullary thymic epithelial cells (mTECs). Not only do
mTECs support the final steps of thymocyte maturation, but
they also have a unique ability to express proteins that are
otherwise found exclusively in other organs. This allows them
to negatively select a group of potentially very damaging, autoreactive T cells that could not be deleted in the cortex.1
Mature thymocytes, which express only CD4 or CD8 and
are referred to as single positive (SP), leave the thymus as
they entered: via the blood vessels of the corticomedullary
junction. Maturation is finalized in the periphery, where
these new T cells (recent thymic emigrants) explore antigens
presented in secondary lymphoid tissue, including spleen
and lymph nodes.
Secondary Lymphoid Organs—Where
the Immune Response Is Initiated
As just described, lymphocytes and myeloid cells develop to
maturity in the primary lymphoid system: T lymphocytes in
the thymus, and B cells, monocytes, dendritic cells, and
granulocytes in the bone marrow. However, they encounter
antigen and initiate an immune response in the microenvironments of secondary lymphoid organs (SLOs).
Secondary Lymphoid Organs Are Distributed
Throughout the Body and Share Some
Anatomical Features
Lymph nodes and the spleen are the most highly organized
of the secondary lymphoid organs and are compartmentalized from the rest of the body by a fibrous capsule. A some-
1
Note that some investigators describe positive selection as taking place in
the cortex and negative selection solely in the medulla. However, several
lines of evidence suggest that negative selection can also occur in the cortex,
and we have adopted this perspective for this text.
what less organized system of secondary lymphoid tissue,
collectively referred to as mucosa-associated lymphoid tissue (MALT), is found associated with the linings of multiple
organ systems, including the gastrointestinal (GI) and respiratory tracts. MALT includes tonsils, Peyer’s patches (in the
small intestine), and the appendix, as well as numerous lymphoid follicles within the lamina propria of the intestines
and in the mucous membranes lining the upper airways,
bronchi, and genitourinary tract (Figure 2-7).
Although secondary lymphoid organs vary in their location and degree of organization, they share key features. All
SLOs include anatomically distinct regions of T-cell and
B-cell activity, and all develop lymphoid follicles, which are
highly organized microenvironments that are responsible
for the development and selection of B cells that produce
high-affinity antibodies.
Lymphoid Organs Are Connected to Each
Other and to Infected Tissue by Two Different
Circulatory Systems: Blood and Lymphatics
The immune cells are the most mobile cells in a body and use
two different systems to traffic through tissues: the blood
system and the lymphatic system. The blood has access to
virtually every organ and tissue and is lined by endothelial
cells that are very responsive to inflammatory signals. Hematopoietic cells can transit through the blood system—away
from the heart via active pumping networks (arteries) and
back to the heart via passive valve-based systems (veins)
within minutes. Most lymphocytes enter secondary lymphoid organs via specialized blood vessels, and leave via the
lymphatic system.
The lymphatic system is a network of thin walled vessels
that play a major role in immune cell trafficking, including
the travel of antigen and antigen-presenting cells to secondary lymphoid organs and the exit of lymphocytes from
lymph nodes.
Lymph vessels are filled with a protein-rich fluid (lymph)
derived from the fluid component of blood (plasma) that
seeps through the thin walls of capillaries into the surrounding tissue. In an adult, depending on size and activity, seepage can add up to 2.9 liters or more during a 24-hour period.
This fluid, called interstitial fluid, permeates all tissues and
bathes all cells. If this fluid were not returned to the circulation, the tissue would swell, causing edema that would eventually become life threatening. We are not afflicted with such
catastrophic edema because much of the fluid is returned to
the blood through the walls of venules. The remainder of the
interstitial fluid enters the delicate network of primary lymphatic vessels. The walls of the primary vessels consist of a
single layer of loosely apposed endothelial cells. The porous
architecture of the primary vessels allows fluids and even
cells to enter the lymphatic network. Within these vessels,
the fluid, now called lymph, flows into a series of progressively larger collecting vessels called lymphatic vessels (see
Figures 2-7b and 2-7c).
Cells, Organs, and Microenvironments of the Immune System
|
CHAPTER 2
49
(b)
(a)
Adenoids
Right
lymphatic
duct
Right
subclavian
vein
Thymus
Thoracic
duct
Left subclavian
vein
Lymph
nodes
Spleen
Peyer's
patches
Bone marrow
Lymph node
Lymphatic
vessel
(c)
Smooth muscle
Lymphatic
capillary
Arteriole
Tissue
lymphatics
Blood
capillaries
Interstitial
fluid
Venule
Lymph
FIGURE 2-7 The human lymphoid system. The primary
organs (bone marrow and thymus) are shown in red; secondary
organs and tissues, in blue. These structurally and functionally
diverse lymphoid organs and tissues are interconnected by the
blood vessels (not shown) and lymphatic vessels (purple). Most of
the body’s lymphatics eventually drain into the thoracic duct, which
empties into the left subclavian vein. However, the vessels draining
the right arm and right side of the head (shaded blue) converge to
form the right lymphatic duct, which empties into the right subclavian vein. The inset (b) shows the lymphatic vessels in more detail,
and (c) shows the relationship between blood and lymphatic capillaries in tissue. The lymphatic capillaries pick up interstitial fluid,
particulate and soluble proteins, as well as immune cells from the
tissue surrounding the blood capillaries (see arrows). [Part (a):
Adapted from H. Lodish et al., 1995, Molecular Cell Biology, 3rd ed.,
Scientific American Books, New York.]
All cells and fluid circulating in the lymph are ultimately
returned to the blood system. The largest lymphatic vessel,
the thoracic duct, empties into the left subclavian vein. It
collects lymph from all of the body except the right arm and
right side of the head. Lymph from these areas is collected
into the right lymphatic duct, which drains into the right
subclavian vein (see Figure 2-7a). By returning fluid lost
from the blood, the lymphatic system ensures steady-state
levels of fluid within the circulatory system.
The heart does not pump the lymph through the lymphatic system; instead, the slow, low-pressure flow of lymph
is achieved by the movements of the surrounding muscles.
Therefore, activity enhances lymph circulation. Importantly,
a series of one-way valves along the lymphatic vessels
ensures that lymph flows in only one direction.
When a foreign antigen gains entrance to the tissues, it is
picked up by the lymphatic system (which drains all the tissues of the body) and is carried to various organized
50
PA R T I
|
Introduction
lymphoid tissues such as lymph nodes, which trap the foreign
antigen. Antigen-presenting cells that engulf and process the
antigen also can gain access to lymph. In fact, as lymph passes
from the tissues to lymphatic vessels, it becomes progressively enriched in specific leukocytes, including lymphocytes,
dendritic cells, and macrophages. Thus, the lymphatic system
also serves as a means of transporting white blood cells and
antigen from the connective tissues to organized lymphoid
tissues, where the lymphocytes can interact with the trapped
antigen and undergo activation. Most secondary lymphoid
tissues are situated along the vessels of the lymphatic system.
The spleen is an exception and is served only by blood vessels.
All immune cells that traffic through tissues, blood, and
lymph nodes are guided by small molecules known as chemokines. These proteins are secreted by stromal cells, antigenpresenting cells, lymphocytes, and granulocytes, and form
gradients that act as attractants and guides for other immune
cells, which express an equally diverse set of receptors for
these chemokines. The interaction between specific chemokines and cells expressing specific chemokine receptors allows
for a highly refined organization of immune cell movements.
The Lymph Node Is a Highly Specialized
Secondary Lymphoid Organ
Lymph nodes (Figure 2-8) are the most specialized SLOs.
Unlike the spleen, which also regulates red blood cell flow and
fate, lymph nodes are fully committed to regulating an immune
response. They are encapsulated, bean-shaped structures that
include networks of stromal cells packed with lymphocytes,
macrophages, and dendritic cells. Connected to both blood
vessels and lymphatic vessels, lymph nodes are the first organized lymphoid structure to encounter antigens that enter the
tissue spaces. The lymph node provides ideal microenvironments for encounters between antigen and lymphocytes and
productive, organized cellular and humoral immune responses.
Structurally, a lymph node can be divided into three
roughly concentric regions: the cortex, the paracortex, and
the medulla, each of which supports a distinct microenvironment (see Figure 2-8). The outermost layer, the cortex, contains lymphocytes (mostly B cells), macrophages, and
follicular dendritic cells arranged in follicles. Beneath the
cortex is the paracortex, which is populated largely by T
lymphocytes and also contains dendritic cells that migrated
from tissues to the node. The medulla is the innermost layer,
and the site where lymphocytes exit (egress) the lymph node
through the outgoing (efferent) lymphatics. It is more sparsely
populated with lymphoid lineage cells, which include plasma
cells that are actively secreting antibody molecules.
Antigen travels from infected tissue to the cortex of the
lymph node via the incoming (afferent) lymphatic vessels,
which pierce the capsule of a lymph node at numerous sites
and empty lymph into the subcapsular sinus (see Figure 2-8b).
It enters either in particulate form or is processed and
presented as peptides on the surface of migrating antigenpresenting cells. Particulate antigen can be trapped by resident
antigen-presenting cells in the subcapsular sinus or cortex,
and it can be passed to other antigen-presenting cells,
including B lymphocytes. Alternatively, particulate antigen
can be processed and presented as peptide-MHC complexes
on cell surfaces of resident dendritic cells that are already in
the T-cell-rich paracortex.
T Cells in the Lymph Node
It takes every naïve T lymphocyte about 16 to 24 hours to
browse all the MHC-peptide combinations presented by the
antigen-presenting cells in a single lymph node. Naïve lymphocytes enter the cortex of the lymph node by passing
between the specialized endothelial cells of high endothelial
venules (HEV), so-called because they are lined with unusually tall endothelial cells that give them a thickened appearance (Figure 2-9a).
Once naïve T cells enter the lymph node, they browse
MHC-peptide antigen complexes on the surfaces of the dendritic cells present in the paracortex. The paracortex is traversed by a web of processes that arise from stromal cells called
fibroblast reticular cells (FRCs) (Figure 2-9b). This network is
referred to as the fibroblast reticular cell conduit system
(FRCC) and guides T-cell movements via associated adhesion
molecules and chemokines. Antigen-presenting cells also
appear to wrap themselves around the conduits, giving circulating T cells ample opportunity to browse their surfaces as
they are guided down the network. The presence of this specialized network elegantly enhances the probability that T cells
will meet their specific MHC-peptide combination.
T cells that browse the lymph node but do not bind
MHC-peptide combinations exit not via the blood, but via
the efferent lymphatics in the medulla of the lymph node (see
Figure 2-8). T cells whose TCRs do bind to an MHC-peptide
complex on an antigen-presenting cell that they encounter in
the lymph node will stop migrating and take up residence in
the node for several days. Here it will proliferate and,
depending on cues from the antigen-presenting cell itself, its
progeny will differentiate into effector cells with a variety of
functions. CD8 T cells gain the ability to kill target cells.
CD4 T cells can differentiate into several different kinds of
effector cells, including those that can further activate macrophages, CD8 T cells, and B cells.
B Cells in the Lymph Node
The lymph node is also the site where B cells are activated
and differentiate into high-affinity antibody-secreting
plasma cells. B cell activation requires both antigen engagement by the B-cell receptor (BCR) and direct contact with an
activated CD4 TH cell. Both events are facilitated by the
anatomy of the lymph node. Like T cells, B cells circulate
through the blood and lymph and visit the lymph nodes on
a daily basis, entering via the HEV. They respond to specific
signals and chemokines that draw them not to the paracortex but to the lymph node follicle. Although they may initially take advantage of the FRCC for guidance, they
Cells, Organs, and Microenvironments of the Immune System
(a)
CHAPTER 2
51
(b)
Follicle
(B cell zone)
Adenoids
Left subclavian
vein
Right
lymphatic
duct
|
Paracortex
(T cell zone)
Incoming (afferent)
lymphatics
Lymph
nodes
Antigen, APC
Right
subclavian
vein
Medulla
Thymus
Spleen
Outer
cortex
Peyer's
patches
Blood
vessels
Bone marrow
Thoracic
duct
Outgoing (efferent)
lymphatics
Lymphocytes, APC
Naïve lymphocytes
(d)
(c)
B-cell follicle
T-cell zone: paracortex
Follicle
Germinal center
Tissue
lymphatics
FIGURE 2-8 Structure of a lymph node. The microenvironments of the lymph node support distinct cell activities. (a) The lymph
nodes are dispersed throughout the body and are connected by lymphatic vessels as well as blood vessels (not shown). (b) A drawing of the
major features of a lymph node shows the major vessels that serve the
organ: incoming (afferent) and outgoing (efferent) lymphatic vessels,
and the arteries and veins. It also depicts the three major tissue layers:
the outer cortex, the paracortex, and the innermost region, the
medulla. Macrophages and dendritic cells, which trap antigen, are present in the cortex and paracortex. T cells are concentrated in the paracortex; B cells are primarily in the cortex, within follicles and germinal
centers. The medulla is populated largely by antibody-producing
plasma cells and is the site where cells exit via the efferent lymphatics.
Naïve lymphocytes circulating in the blood enter the node via high
endothelial venules (HEV) via a process called extravasation (see
Advances Box 14-2). Antigen and some leukocytes, including antigenpresenting cells, enter via afferent lymphatic vessels. All cells exit via
efferent lymphatic vessels. (c) This stained tissue section shows the
cortex with a number of ovoid follicles, which is surrounded by the
T-cell-rich paracortex. (d) A stained lymph node section showing a follicle that includes a germinal center (otherwise referred to as a secondary follicle). [2-8c: Dr. Gladden Willis/Getty Images; 2-8d: Image Source/Alamy.]
ultimately depend upon follicular dendritic cells (FDCs) for
guidance (Figure 2-9c). FDCs are centrally important in
maintaining follicular and germinal center structure and
“presenting” antigen to differentiating B cells.
B cells differ from T cells in that their receptors can recognize free antigen. A B cell will typically meet its antigen
in the follicle. If its BCR binds to antigen, the B cell
becomes partially activated and engulfs and processes that
antigen. As mentioned above, B cells, in fact, are specialized antigen-presenting cells that present processed peptideMHC complexes on their surface to CD4 TH cells. Recent
data show that B cells that have successfully engaged and
processed antigen change their migration patterns and
move to the T-cell-rich paracortex, where they increase
their chances of encountering an activated CD4 TH cell
that will recognize the MHC-antigen complex they present.
When they successfully engage this TH cell, they maintain
contact for a number of hours, becoming fully activated
and receiving signals that induce B cell proliferation (see
Chapter 14).
Some activated B cells differentiate directly into an antibody-producing cell (plasma cell) but others re-enter the
follicle to establish a germinal center. A follicle that develops
a germinal center is sometimes referred to as a secondary
follicle; a follicle without a germinal center is sometimes
referred to as a primary follicle.
52
PA R T I
|
Introduction
(a) Afferent lymphatic vessel high endothelial venule
Lymphatic endothelial cell
Blood endothelial cell
DC
Perivascular sheath
Basal lamina
Soluble molecule
Particulate antigen
Naïve
B cell
Naïve
T cell
(b) Follicular reticular cell conduit system
FRC
T cell
DC
HEV
(c) Follicular dendritic cell
FDC
B cell
FIGURE 2-9 Features of lymph node microenvironments.
The lymph node microenvironments are maintained and regulated by distinct cell types and structures. (a) The afferent lymphatics are the vessels through which dendritic cells, and particulate
and soluble antigen, enter the lymph node. The high endothelial
venules (HEVs) are the vessels through which naïve T and B cells
enter the lymph node (via extravasation). (b) The paracortex is
crisscrossed by processes and conduits formed by fibroblastic
reticular cells (FRCs), which guide the migration of antigen-
nie Favre and Sanjiv A. Luther, University of Lausanne, Switzerland. 2-9c, left:
Photograph courtesy of Mohey Eldin M. El Shikh.]
Germinal centers are remarkable substructures that
facilitate the generation of B cells with increased receptor
affinities. In the germinal center, an antigen-specific B cell
clone will proliferate and undergo somatic hypermutation
of the genes coding for their antigen receptors. Those receptors that retain the ability to bind antigen with the highest
affinity survive and differentiate into plasma cells that travel
to the medulla of the lymph node. Some will stay and
release antibodies into the bloodstream; others will exit
through the efferent lymphatics and take up residence in the
bone marrow, where they will continue to release antibodies
into circulation.
The initial activation of B cells and establishment of the
germinal center take place within 4 to 7 days of the initial infection, but germinal centers remain active for 3 weeks or more
(Chapter 12). Lymph nodes swell visibly and sometimes painfully, particularly during those first few days after infection. This
swelling is due both to an increase in the number of lymphocytes induced to migrate into the node as well as the proliferation of antigen-specific T and B lymphocytes within the lobe.
presenting cells and T cells, facilitating their interactions. The left
panel shows an immunofluorescence microscopy image with the
FRC shown in red and T cells in green. The right panel shows a
cartoon of the network and cell participants. (c) The B-cell follicle
contains a network of follicular dendritic cells (FDCs), which are
shown as an SEM image as well as a cartoon. FDCs guide the
movements and interactions of B cells. [2-9b, left: Courtesy of Stepha-
Cells, Organs, and Microenvironments of the Immune System
The Generation of Memory T and B Cells
in the Lymph Node
The interactions between TH cells and APCs, and between
activated TH cells and activated B cells, results not only in the
proliferation of antigen-specific lymphocytes and their
functional differentiation, but also in the generation of memory T and B cells. Memory T and B cells can take up residence
in secondary lymphoid tissues or can exit the lymph node and
travel to and among tissues that first encountered the pathogen. Memory T cells that reside in secondary lymphoid organs
are referred to as central memory cells and are distinct in phenotype and functional potential from effector memory T cells
that circulate among tissues. Memory cell phenotype, locale,
and activation requirements are an active area of investigation
and will be discussed in in more detail in Chapters 11 and 12.
The Spleen Organizes the Immune Response
Against Blood-Borne Pathogens
The spleen, situated high in the left side of the abdominal cavity, is a large, ovoid secondary lymphoid organ that plays a
major role in mounting immune responses to antigens in the
bloodstream (Figure 2-10). Whereas lymph nodes are specialized for encounters between lymphocytes and antigen drained
from local tissues, the spleen specializes in filtering blood and
trapping blood-borne antigens; thus, it is particularly important in the response to systemic infections. Unlike the lymph
nodes, the spleen is not supplied by lymphatic vessels. Instead,
blood-borne antigens and lymphocytes are carried into the
spleen through the splenic artery and out via the splenic vein.
Experiments with radioactively labeled lymphocytes show
that more recirculating lymphocytes pass daily through the
spleen than through all the lymph nodes combined.
The spleen is surrounded by a capsule from which a number of projections (trabeculae) extend, providing structural
support. Two main microenvironmental compartments can
be distinguished in splenic tissue: the red pulp and white
pulp, which are separated by a specialized region called the
marginal zone (see Figure 2-10d). The splenic red pulp consists of a network of sinusoids populated by red blood cells,
macrophages, and some lymphocytes. It is the site where old
and defective red blood cells are destroyed and removed;
many of the macrophages within the red pulp contain
engulfed red blood cells or iron-containing pigments from
degraded hemoglobin. It is also the site where pathogens first
gain access to the lymphoid-rich regions of the spleen, known
as the white pulp. The splenic white pulp surrounds the
branches of the splenic artery, and consists of the periarteriolar lymphoid sheath (PALS) populated by T lymphocytes as
well as B-cell follicles. As in lymph nodes, germinal centers are
generated within these follicles during an immune response.
The marginal zone, which borders the white pulp, is populated
by unique and specialized macrophages and B cells, which are
the first line of defense against certain blood-borne pathogens.
Blood-borne antigens and lymphocytes enter the spleen
through the splenic artery, and interact first with cells at the
|
CHAPTER 2
53
marginal zone. In the marginal zone, antigen is trapped and
processed by dendritic cells, which travel to the PALS. Specialized, resident marginal zone B cells also bind antigen via complement receptors and convey it to the follicles. Migrating B
and T lymphocytes in the blood enter sinuses in the marginal
zone and migrate to the follicles and the PALS, respectively.
The events that initiate the adaptive immune response in
the spleen are analogous to those that occur in the lymph
node. Briefly, circulating naïve B cells encounter antigen in
the follicles, and circulating naïve CD8 and CD4 T cells
meet antigen as MHC-peptide complexes on the surface of
dendritic cells in the T-cell zone (PALS). Once activated,
CD4 TH cells then provide help to B cells and CD8 T cells
that have also encountered antigen. Some activated B cells,
together with some TH cells, migrate back into follicles and
generate germinal centers.
It is unclear whether a reticular network operates as
prominently within the spleen as it does in the lymph node.
However, given that T cells, dendritic cells, and B cells find a
way to interact efficiently within the spleen to initiate an
immune response, it would not be surprising if a similar
conduit system were present.
Although animals can lead a relatively healthy life without
a spleen, its loss does have consequences. In children, in particular, splenectomy (the surgical removal of a spleen) can
lead to overwhelming post-splenectomy infection (OPSI)
syndrome characterized by systemic bacterial infections
(sepsis) caused by primarily Streptococcus pneumoniae, Neisseria meningitidis, and Haemophilus influenzae. Although
fewer adverse effects are experienced by adults, splenectomy
can still lead to an increased vulnerability to blood-borne
bacterial infections, underscoring the role the spleen plays in
our immune response to pathogens that enter the circulation.
It is also important to recognize that the spleen has other
functions (e.g., in iron metabolism, thrombocyte storage,
hematopoiesis) that will also be compromised if it is removed.
MALT Organizes the Response to Antigen
That Enters Mucosal Tissues
Lymph nodes and the spleen are not the only organs that
develop secondary lymphoid microenvironments. T- and
B-cell zones and lymphoid follicles are also found in mucosal
membranes that line the digestive, respiratory, and urogenital systems, as well as in the skin.
Mucosal membranes have a combined surface area of about
400 m2 (nearly the size of a basketball court) and are the major
sites of entry for most pathogens. These vulnerable membrane
surfaces are defended by a group of organized lymphoid tissues
known collectively as mucosa-associated lymphoid tissue
(MALT). Lymphoid tissue associated with different mucosal
areas is sometimes given more specific names; for instance, the
respiratory epithelium is referred to as bronchus-associated
lymphoid tissue (BALT) or nasal-associated lymphoid tissue
(NALT), and that associated with the intestinal epithelium is
referred to as gut-associated lymphoid tissue (GALT).
54
PA R T I
|
Introduction
(b)
(a)
Adenoids
Tonsil
Left subclavian
vein
Right
lymphatic
duct
Lymph
nodes
Right
subclavian
vein
Thymus
Thoracic
duct
Spleen
Peyer's
patches
Bone marrow
Large
intestine
(c)
White pulp
Red pulp
Small
intestine
Appendix
Tissue
lymphatics
(d)
Splenic artery
Splenic vein
Follicle
Central
arteriole
Marginal
zone
White
pulp
Sinuses
PALS
Trabeculae
Red pulp
FIGURE 2-10 Structure of the spleen. (a) The spleen, which
is about 5 inches long in human adults, is the largest secondary
lymphoid organ. It is specialized for trapping blood-borne antigens.
Panels (b) and (c) are stained tissue sections of the human spleen,
showing the red pulp, white pulp, and follicles. These microenvironments are diagrammed schematically in (d). The splenic artery
pierces the capsule and divides into progressively smaller arterioles,
ending in vascular sinusoids that drain back into the splenic vein.
The erythrocyte-filled red pulp surrounds the sinusoids. The white
pulp forms a sleeve—the periarteriolar lymphoid sheath (PALS)—
around the arterioles; this sheath contains numerous T cells. Closely
associated with the PALS are the B-cell-rich lymphoid follicles that
can develop into secondary follicles containing germinal centers.
The marginal zone, a site of specialized macrophages and B cells,
surrounds the PALS and separates it from the red pulp. [2-10b: Dr. Keith
The structure of GALT is well described and ranges from
loose, barely organized clusters of lymphoid cells in the lamina
propria of intestinal villi to well-organized structures such as
the tonsils and adenoids (Waldeyer’s tonsil ring), the appendix,
and Peyer’s patches, which are found within the intestinal lining
and contain well-defined follicles and T-cell zones. As shown in
Figure 2-11, lymphoid cells are found in various regions within
the lining of the intestine. The outer mucosal epithelial layer
contains intraepithelial lymphocytes (IELs), many of which
are T cells. The lamina propria, which lies under the epithelial
Wheeler/Photo Researchers; 2-10c: Biophoto Associates/Photo Researchers.]
Cells, Organs, and Microenvironments of the Immune System
|
CHAPTER 2
55
(b)
(a)
MALT
Adenoids
Tonsil
Thoracic duct
Thymus
Right
lymphatic
duct
Right
subclavian
vein
Left subclavian
vein
Lymph
nodes
Spleen
Thoracic
duct
Peyer's
patches
Large
intestine
(c)
Bone marrow
Villi
Small intestine
Appendix
M cells
Lamina propria
Submucosa
High
endothelial
venule
Tissue
lymphatics
Lymphocytes
Lymph
vessel
Lymphocytes
B cells
Germinal center
T cells
Muscle layer
Peyer’s patch
FIGURE 2-11 Mucosa-associated lymphoid tissue (MALT). (a) The Peyer’s patch is a representative of the extensive MALT system that is
found in the intestine. (b) A stained tissue cross-section of Peyer’s patch lymphoid nodules in the intestinal submucosa is schematically diagrammed
in (c). The intestinal lamina propria contains loose clusters of lymphoid cells and diffuse follicles. [2-11b: Dr. Gladden Willis/Visuals Unlimited.]
layer, contains large numbers of B cells, plasma cells, activated
T cells, and macrophages in loose clusters. Microscopy has
revealed more than 15,000 lymphoid follicles within the intestinal lamina propria of a healthy child. Peyer’s patches, nodules
of 30 to 40 lymphoid follicles, extend into the muscle layers that
are just below the lamina propria. Like lymphoid follicles in
other sites, those that compose Peyer’s patches can develop into
secondary follicles with germinal centers. The overall functional importance of MALT in the body’s defense is underscored by its large population of antibody-producing plasma
cells, whose number exceeds that of plasma cells in the spleen,
lymph nodes, and bone marrow combined.
Some cellular structures and activities are unique to MALT.
For instance, the epithelial cells of mucous membranes play an
important role in delivering small samples of foreign antigen
from the respiratory, digestive, and urogenital tracts to the
underlying mucosa-associated lymphoid tissue. In the digestive tract, specialized M cells transport antigen across the epithelium (Figure 2-12). The structure of M cells is striking: they
are flattened epithelial cells lacking the microvilli that characterize the rest of the mucosal epithelium. They have a deep
invagination, or pocket, in the basolateral plasma membrane,
which is filled with a cluster of B cells, T cells, and macrophages. Antigens in the intestinal lumen are endocytosed into
vesicles that are transported from the luminal membrane to
the underlying pocket membrane. The vesicles then fuse with
the pocket membrane, delivering antigens to clusters of lymphocytes and antigen-presenting cells, the most important of
which are dendritic cells, contained within the pocket. Antigen
transported across the mucous membrane by M cells ultimately leads to the activation of B cells that differentiate and
then secrete IgA. This class of antibody is concentrated in
56
PA R T I
|
Introduction
(a)
(b)
M cell
Lumen
Antigen
Antigen
TH cell
Mucosal
epithelium
Intraepithelial
lymphocyte
IgA
M cell
IgA
Pocket
Lamina
propria
B cells
Plasma
cell
Organized
lymphoid
follicle
Macrophage
(c)
M cell
Macrophage
FIGURE 2-12 Structure of M cells and production of IgA
at inductive sites. (a) M cells, situated in mucous membranes,
endocytose antigen from the lumen of the digestive, respiratory,
and urogenital tracts. The antigen is transported across the cell
and released into the large basolateral pocket. (b) Antigen transported across the epithelial layer by M cells at an inductive site
activates B cells in the underlying lymphoid follicles. The activated B cells differentiate into IgA-producing plasma cells, which
migrate along the lamina propria, the layer under the mucosa.
secretions (e.g., milk) and is an important tool used by the
body to combat many types of infection at mucosal sites.
The Skin Is an Innate Immune Barrier
and Also Includes Lymphoid Tissue
The skin is the largest organ in the body and a critical anatomic
barrier against pathogens. It also plays an important role in
nonspecific (innate) defenses (Figure 2-13). The epidermal
(outer) layer of the skin is composed largely of specialized epithelial cells called keratinocytes. These cells secrete a number
of cytokines that may function to induce a local inflammatory
The outer mucosal epithelial layer contains intraepithelial lymphocytes, of which many are T cells. (c) A stained section of
mucosal lymphoid tissue (the Peyer’s patch of the intestine)
shows small, darkly stained intraepithelial lymphocytes encased
by M cells, whose nuclei are labeled. Lymphocytes are also present in the lamina propria. [2-12c: Kucharzik, T., Lugering, N., Schmid,
K.W., Schmidt, M.A., Stoll, R. Domschke, W. Human intestinal M cells
exhibit enterocyte-like intermediate filaments. Gut 1998, 42: 54–52. doi:
10.1136/gut.42.1.54.]
reaction. Scattered among the epithelial-cell matrix of the epidermis are Langerhans cells, skin-resident dendritic cells that
internalize antigen by phagocytosis or endocytosis. These
Langerhans cells undergo maturation and migrate from the
epidermis to regional lymph nodes, where they function as
potent activators of naïve T cells. In addition to Langerhans
cells, the epidermis also contains intraepidermal lymphocytes, which are predominantly T cells; some immunologists
believe that they play a role in combating infections that enter
through the skin, a function for which they are well positioned.
The underlying dermal layer of the skin also contains scattered
lymphocytes, dendritic cells, monocytes, macrophages, and
Cells, Organs, and Microenvironments of the Immune System
Epidermis
Keratinocyte
Langerhans
cell
|
CHAPTER 2
57
may even include hematopoietic stem cells. Most skin lymphocytes appear to be either previously activated cells or memory
cells, many of which traffic to and from local, draining lymph
nodes that coordinate the responses to pathogens that have
breached the skin barrier.
Dermis
Intraepidermal
lymphocyte
Blood vessel
Tertiary Lymphoid Tissues Also Organize
and Maintain an Immune Response
Dendritic cell
Tissues that are the sites of infection are referred to as tertiary lymphoid tissue. Lymphocytes activated by antigen in
secondary lymphoid tissue can return to these organs (e.g.,
lung, liver, brain) as effector cells and can also reside there as
memory cells. It also appears as if tertiary lymphoid tissues
can generate defined microenvironments that organize the
returning lymphoid cells. Investigators have recently found
that the brain, for instance, establishes reticular systems that
guide lymphocytes responding to chronic infection with the
protozoan that causes toxoplasmosis. These observations
together highlight the remarkable adaptability of the immune
system, as well as the intimate relationship between anatomical structure and immune function. The conservation of
structure/function relationships is also illustrated by the
evolutionary relationships among immune systems and
organs (see the Evolution Box 2-4 below on pages 57–59).
Lymphatic
vessel
FIGURE 2-13 The distribution of immune cells in the skin.
Langerhans cells reside in the outer layer, the epidermis. They travel
to lymph nodes via the lymphatic vessels in the dermis, a layer of
connective tissue below the epidermis. Dendritic cells also reside in
the dermis and also can travel via the lymphatic vessels to lymph
nodes when activated. White blood cells, including monocytes and
lymphocytes travel to both layers of the skin via blood vessels,
extravasating in the dermis, as shown.
BOX 2-4
EVOLUTION
Variations on Anatomical Themes
In Chapter 5, we will see that innate
systems of immunity are found in vertebrates, invertebrates, and even in plants.
Adaptive immunity, which depends on
lymphocytes and is mediated by antibodies
and T cells, only evolved in the subphylum
Vertebrata. However, as shown in Figure 1,
the kinds of lymphoid tissues seen in different orders of vertebrates differ dramatically.
All multicellular living creatures, plant
and animal, have cellular and molecular
systems of defense against pathogens.
Vertebrates share an elaborate anatomical
immune system that is compartmentalized in several predictable ways. For example, sites where immune cells develop
(primary lymphoid organs) are separated
from where they generate an immune
response (secondary lymphoid tissue).
Likewise, sites that support the development of B cells and myeloid cells, are separated from sites that generate mature T
cells. The spectrum of vertebrates ranges
from the jawless fishes (Agnatha, the earliest lineages, which are represented by the
lamprey eel and hagfish), to cartilaginous
fish (sharks, rays), which represent the
earliest lineages of jawed vertebrates
(Gnathosomata), to bony fish, reptiles,
amphibians, birds, and mammals, the
most recently evolved vertebrates. If you
view these groups as part of an evolutionary progression, you see that, in general,
immune tissues and organs evolved by
earlier orders have been retained as newer
organs of immunity, such as lymph nodes,
have appeared. For example, all vertebrates have gut-associated lymphoid tissue (GALT), but only jawed vertebrates
have a well-developed thymus and spleen.
The adaptive immune system of jawed
vertebrates emerged about 500 million
years ago. T cells were the first cell population to express a diverse repertoire of anti-
gen receptors, and their appearance is
directly and inextricably linked to the
appearance of the thymus. This dependence is reflected in organisms today: all
jawed vertebrates have a thymus, and the
thymus is absolutely required for the development of T cells. Until recently, it was
thought that jawless vertebrates (e.g., the
lamprey eel) did not have any thymic tissue.
However, recent studies indicate that even
the lamprey supports the development of
two distinct lymphocyte populations reminiscent of T and B cells, and may also harbor
distinct thymic-like tissue in their gill regions
(Figure 2). In contrast, their B-lymphocytelike cells may develop in distinct regions
associated with the kidneys and intestine
(typhlosole). These findings suggest that
jawed and jawless vertebrates share a common ancestor in which lymphocyte lineages and primary lymphoid organs were
already compartmentalized.
(continued)
58
PA R T I
EVOLUTION
|
Introduction
(continued)
Thymus
Lymph
nodes
Thymus
Thymus
Kidney
GALT
Thymus
GALT
GALT
Peyer's
patch
GALT
Spleen
Spleen
Bone
marrow
Lymph
nodes
Bone
marrow
Spleen
Lamprey
GALT
Spleen
Trout
Frog
Bone
marrow
Bursa
Chicken
Mouse
GALT
Thymus
Spleen
Bone marrow
Lymph nodes
Germinal centers
?
Anura
Teleostei
Aves
Mammalia
Reptilia
Amphibia
Osteichthyes
Agnatha
Gnathostomata
Vertebrata
FIGURE 1
Evolutionary distribution of lymphoid tissues. The presence and location of lymphoid tissues in several major orders of vertebrates are shown.
Although they are not shown in the diagram, cartilaginous fish such as sharks and rays have GALT, a thymus, and a spleen. Reptiles also have GALT, a thymus,
and a spleen and may also have lymph nodes that participate in immunological reactions. The sites and nature of primary lymphoid tissues in reptiles are
under investigation. Although jawless animals were thought not to have a thymus, recent data suggest that they might have thymic tissue and two types of
lymphocytes analogous to B cells and T cells. [Adapted from Dupasquier and M. Flajnik, 2004, in Fundamental Immunology, 5th ed., W. E. Paul, ed., Lippincott-Raven, Philadelphia.]
Whereas T-cell development is inextricably linked to the presence of a thymus, B-cell development does not
appear to be bound to one particular
organ. Although the bone marrow is the
site of B-cell development in many mammals, including mouse and human, it is
not the site of B-cell development in all
species. For instance, the anatomical
sites of B-cell development shift over the
course of development in sharks (from
liver to kidney to spleen), in amphibians
and reptiles (from liver and spleen to
bone marrow). In contrast, B-cell development in bony fish takes place primarily
within the kidney.
In birds, B cells complete their development in a lymphoid organ associated
with the gut, the bursa of Fabricius (Figure 3). The gut is also a site of B-cell development in some mammals. In cattle and
sheep, early in gestation, B cells mature in
the fetal spleen. Later in gestation, however, this function is assumed by the
Cells, Organs, and Microenvironments of the Immune System
|
CHAPTER 2
59
BOX 2-4
(c)
(b)
(a)
FIGURE 2
Thymic tissue in the lamprey eel. (a) Jawless vertebrates, including the lamprey eel (shown here in larval form), were thought not to have a
thymus. Recent work suggests, however, that they generate two types of lymphocytes analogous to B and T cells and have thymic tissue at the tip of their
gills (b). The thymic tissue shown in (c) is stained blue for a gene (CDA1) specific for lymphocytes found in lampreys. The thymus of jawed vertebrates arises
from an area analogous to the gill region. [Left: Courtesy of Brian Morland, The Bellflask Ecological Survey Team; middle: Dr. Keith Wheeler/Science Photo Library/
Photo Researchers; right: Courtesy Thomas Boehm.]
patch of tissue embedded in the wall of
the intestine called the ileal Peyer’s patch,
which contains a large number of B cells
as well as T cells. The rabbit, too, uses gutassociated tissues, especially the appendix, as primary lymphoid tissue for
important steps in the proliferation and
diversification of B cells. Recent work suggests that even in animals that depend
largely on the bone marrow to complete
B cell maturation, lymphoid tissue in the
gut can act as a primary lymphoid organ
for generating mature B lymphocytes.
Secondary lymphoid organs are more
variable in their numbers and location
than primary lymphoid organs (e.g.,
rodents do not have tonsils, but do have
well-developed lymphoid tissue at the
base of the nose). The structural and
functional features of these organs are,
however, shared by most vertebrates
with at least one interesting exception:
pig lymph nodes exhibit a striking peculiarity—they are “inverted” anatomically,
so that the medulla of the organ, where
lymphocytes exit the organ, is on the
outside, and the cortex, where lymphocytes meet their antigen and proliferate,
is on the inside. Lymphocyte egress is
also inverted: they exit via blood vessels,
rather than via efferent lymphatics. The
adaptive advantages (if any) of these odd
structural variations are unknown, but
the example reminds us of the remarkable plasticity of structure/function relationships within biological structures and
the creative opportunism of evolutionary
processes.
(b)
(a)
Bursa of Fabricius
(B cell development)
Thymus (T cell
development)
Intestines
FIGURE 3
The avian bursa. (a) Like all vertebrates, birds have a thymus, spleen, and lymph nodes, and hematopoiesis occurs in their bone marrow. However, their
B cells do not develop in the bone marrow. Rather they develop in an outpouching of the intestine, the bursa, which is located close to the cloaca (the
common end of the intestinal and genital tracts in birds). A stained tissue section of the bursa and cloaca is shown in (b). [(b) Courtesy Dr. Thomas Caceci.]
60
PA R T I
|
Introduction
S U M M A R Y
■
■
■
■
■
■
■
■
■
■
The immune response results from the coordinated activities of many types of cells, organs, and microenvironments
found throughout the body.
Many of the body’s cells, tissues, and organs arise from
different stem cell populations. All red and white blood
cells develop from a pluripotent hematopoietic stem cell
during a highly regulated process called hematopoiesis.
In the adult vertebrate, hematopoiesis occurs in the bone
marrow, a stem cell niche that supports both the selfrenewal of stem cells and their differentiation into multiple blood cell types.
Hematopoietic stem cells give rise to two main blood cell
progenitors: common myeloid progenitors and common
lymphoid progenitors.
Four main types of cells develop from common myeloid
progenitors: red blood cells (erythrocytes), monocytes
(which give rise to macrophages and myeloid dendritic
cells), granulocytes (which include the abundant neutrophils and less abundant basophils, eosinophils, and mast
cells), and megakaryocytes.
Macrophages and neutrophils are specialized for the phagocytosis and degradation of antigens. Macrophages also have
the capacity to present antigen with MHC to T cells.
Immature forms of dendritic cells have the capacity to
capture antigen in one location, undergo maturation, and
migrate to another location, where they present antigen to
T cells. Dendritic cells are the most potent antigen-presenting
cell for activating naïve T cells.
There are three types of lymphoid cells: B cells, T cells, and
natural killer (NK) cells. B and T cells are members of
clonal populations distinguished by antigen receptors of
unique specificity. B cells synthesize and display membrane antibody, and T cells synthesize and display T-cell
receptors (TCRs). NK cells do not synthesize antigenspecific receptors; however a small population of TCRexpressing T cells have features of NK cells and are called
NKT cells.
T cells can be further subdivided into helper T cells, which
typically express CD4 and see peptide bound to MHC
class II, and cytotoxic T cells, which typically express CD8
and see peptide bound to MHC class I.
Primary lymphoid organs are the sites where lymphocytes
develop and mature. T cell precursors come from the bone
marrow but develop fully in the thymus; in humans and
mice, B cells arise and develop in bone marrow. In birds,
B cells develop in the bursa.
■
■
■
■
■
■
■
■
Secondary lymphoid organs provide sites where lymphocytes encounter antigen, become activated, and undergo
clonal expansion and differentiation into effector cells.
They include the lymph node, spleen, and more loosely
organized sites distributed throughout the mucosal system
(the mucosa-associated lymphoid tissue, or MALT) and
are typically compartmentalized into T-cell areas and
B-cell areas (follicles).
Lymph nodes and the spleen are the most highly organized secondary lymphoid organs. T-cell and B-cell activity are separated into distinct microenvironments in both.
T cells are found in the paracortex of the lymph nodes and
the periarteriolar sheath of the spleen. B cells are organized into follicles in both organs.
The spleen is the first line of defense against blood-borne
pathogens. It also contains red blood cells and compartmentalizes them in the red pulp from lymphocytes in the
white pulp. A specialized region of macrophages and B cells,
the marginal zone, borders the white pulp.
In secondary lymphoid tissue, T lymphocytes can develop
into killer and helper effector cells. CD4 T cells can help
B cells to differentiate into plasma cells, which secrete
antibody or can also develop into effector cells that further
activate macrophages and CD8 cytotoxic T cells. Both B
and T cells also develop into long-lived memory cells.
B cells undergo further maturation in germinal cells, a
specialized substructure found in follicles after infection;
here they can increase their affinity for antigen and
undergo class switching.
Lymphoid follicles and other organized lymphoid microenvironments are found associated with the mucosa, the
skin, and even tertiary tissues at the site of infection.
Mucosa-associated lymphoid tissue (MALT) is an important defense against infection at epithelial layers and
includes a well-developed network of follicles and lymphoid microenvironments associated with the intestine
(gut-associated lymphoid tissue, GALT).
Unique cells found in the lining of the gut, M cells, are
specialized to deliver antigen from the intestinal spaces to
the lymphoid cells within the gut wall.
R E F E R E N C E S
Banchereau J., et al. 2000. Immunobiology of dendritic cells.
Annual Review of Immunology 18:767.
Directing lymphocyte traffic in the lymph node. Trends in
Immunology 28:346–352.
Bajenoff, M., J. G. Egen, H. Qi, A. Y. C. Huang, F. Castellino, and
R. N. Germain. 2007. Highways, byways and breadcrumbs:
Bajoghil, B., P. Guo, N. Aghaallaei, M. Hirano, C. Strohmeier,
N. McCurley, D. E. Bockman, M. Schorpp, M. D. Cooper, and
Cells, Organs, and Microenvironments of the Immune System
T. Boehm. 2011. A thymus candidate in lampreys. Nature
47:90–95.
Boehm, T., and C. E. Bleul. 2007. The evolutionary history of
lymphoid organs. Nature Immunology 8:131–135.
Borregaard, N. 2011. Neutrophils: From marrow to microbes.
Immunity 33:657–670.
Catron, D.M., A. A. Itano, K. A. Pape, D. L. Mueller, and M. K
Jenkins. 2004. Visualizing the first 50 hr of the primary
immune response to a soluble antigen. Immunity 21:341–347.
Ema, H., et al. 2005. Quantification of self-renewal capacity in
single hematopoietic stem cells from normal and Lnk-deficient
mice. Developmental Cell 6:907.
|
CHAPTER 2
61
Piccirillo, C. A., and E. M. Shevach. 2004. Naturally occurring
CD4CD25 immunoregulatory T cells: Central players in the
arena of peripheral tolerance. Seminars in Immunology 16:81.
Picker, L. J., and M. H. Siegelman. 1999. Lymphoid tissues and
organs. In Fundamental Immunology, 4th ed. W. E. Paul, ed.
Philadelphia: Lippincott-Raven, p. 145.
Rodewald, H-R. 2006. A second chance for the thymus. Nature
441:942–943.
Rothenberg, M., and S. P. Hogan. 2006. The eosinophil. Annual
Review of Immunology 24:147–174.
Scadden, D. T. 2006. The stem-cell niche as an entity of action.
Nature 441:1075–1079.
Falcone, F. H., D. Zillikens, and B. F.Gibbs. 2006. The 21st century renaissance of the basophil? Current insights into its role
in allergic responses and innate immunity. Experimental Dermatology 15:855–864.
Shizuru, J. A., R. S. Negrin, and I. L. Weissman. 2005. Hematopoietic stem and progenitor cells: Clinical and preclinical
regeneration of the hematolymphoid system. Annual Review
of Medicine 56:509.
Godfrey, D. I., H. R. MacDonald, M. Kronenberg, M. J. Smyth,
and L. Van Kaer. 2004. NKT cells: What’s in a name? Nature
Reviews Immunology 4:231.
Weissman, I. L. 2000. Translating stem and progenitor cell biology to the clinic: Barriers and opportunities. Science 287:1442.
Halin, C., J. R. Mora, C. Sumen, and U. H. von Andrian. In vivo
imaging of lymphocyte trafficking. 2005 Annual Review of Cell
and Developmental Biology 21:581–603.
Useful Web Sites
Kiel, M., and S. J. Morrison. 2006. Maintaining hematopoietic
stem cells in the vascular niche. Immunity 25:852–854.
Liu, Y. J. 2001. Dendritic cell subsets and lineages, and their
functions in innate and adaptive immunity. Cell 106:259.
Mebius, R. E., and G. Kraal. 2005. Structure and function of the
spleen. Nature Reviews Immunology. 5:606–616.
Mendez-Ferrer, S., D. Lucas, M. Battista, and P. S. Frenette.
2008. Haematopoietic stem cell release is regulated by circadian oscillations. Nature 452:442–448.
http://bio-alive.com/animations/anatomy.htm A
collection of publicly available animations relevant to biology. Scroll through the list to find videos on blood and
immune cells, immune responses, and links to interactive
sites that reinforce your understanding of immune anatomy.
http://stemcells.nih.gov Links to information on stem
cells, including basic and clinical information, and registries
of available embryonic stem cells.
www.niaid.nih.gov/topics/immuneSystem/Pages/
structureImages.aspx A very accessible site about
Mikkola, H. K. A., and S. A. Orkin. 2006. The journey of developing hematopoietic stem cells. Development 133:3733–3744.
immune system function and structure.
Miller, J. F. 1961. Immunological function of the thymus. Lancet
2:748–749.
www.immunity.com/cgi/content/full/21/3/341/
DC1 A pair of simulations that trace the activities of T cells,
Miller, J. F. A. P. 2002. The discovery of thymus function and of
thymus-derived lymphocytes. Immunological Reviews 185:7–14.
B cells, and dendritic cells in a lymph node. These movies are
discussed in more detail in Chapter 14.
Moon, J. J., H. H. Chu, M. Pepper, S. J. McSorley, S. C. Jameson,
R. M. Kedl, and M. K. Jenkins. 2007. Naïve CD4 T cell frequency varies for different epitopes and predicts repertoire
diversity and response magnitude. Immunity 27:203–213.
www.hhmi.org/research/investigators/cyster.
html Investigator Jason Cyster’s public website that includes
Pabst, R. 2007. Plasticity and heterogeneity of lymphoid organs:
What are the criteria to call a lymphoid organ primary, secondary or tertiary? Immunology Letters 112:1–8.
S T U D Y
many videos of B-cell activity in immune tissue.
www.hematologyatlas.com/principalpage.htm An
interactive atlas of both normal and pathological human
blood cells.
Q U E S T I O N S
RESEARCH FOCUS QUESTION Notch is a surface protein that
regulates cell fate. When bound by its ligand, it releases and
activates its intracellular region, which regulates new gene
transcription. Investigators found that the phenotype of de-
veloping cells in the bone marrow differed dramatically when
they overexpressed the active, intracellular portion of Notch.
In particular, the frequency of BCR cells plummeted, and
the frequency of TCR cells increased markedly. Interestingly,
62
PA R T I
|
Introduction
other investigators found that when you knocked out Notch,
the phenotype of cells in the thymus changed: the frequency
of BCR cells increased and the frequency of TCR cells decreased dramatically.
Propose a molecular model to explain these observations,
and an experimental approach to begin testing your model.
c. TC cells and basophils
d. TH and B cells
3. List two primary and two secondary lymphoid organs and
summarize their functions in the immune response.
4. What two primary characteristics distinguish hematopoi-
etic stem cells from mature blood cells?
CLINICAL FOCUS QUESTION The T and B cells that differenti-
ate from hematopoietic stem cells recognize as self the bodies
in which they differentiate. Suppose a woman donates HSCs
to a genetically unrelated man whose hematopoietic system
was totally destroyed by a combination of radiation and chemotherapy. Further, suppose that, although most of the donor
HSCs differentiate into hematopoietic cells, some differentiate
into cells of the pancreas, liver, and heart. Decide which of the
following outcomes is likely and justify your choice.
a. The T cells that arise from the donor HSCs do not
attack the pancreatic, heart, and liver cells that arose
from donor cells but mount a GVH response against all
of the other host cells.
b. The T cells that arise from the donor HSCs mount a
GVH response against all of the host cells.
c. The T cells that arise from the donor HSCs attack the
pancreatic, heart, and liver cells that arose from donor
cells but fail to mount a GVH response against all of the
other host cells.
d. The T cells that arise from the donor HSCs do not
attack the pancreatic, heart, and liver cells that arose
from donor cells and fail to mount a GVH response
against all of the other host cells.
1. Explain why each of the following statements is false.
a. There are no mature T cells in the bone marrow.
b. The pluripotent stem cell is one of the most abundant
cell types in the bone marrow.
c. There are no stem cells in blood.
d. Activation of macrophages increases their expression of
class I MHC molecules, allowing them to present antigen to TH cells more effectively.
e. Mature B cells are closely associated with osteoblasts in
the bone marrow.
f. Lymphoid follicles are present only in the spleen and
lymph nodes.
g. The FRC guides B cells to follicles.
h. Infection has no influence on the rate of hematopoiesis.
i. Follicular dendritic cells can process and present antigen to T lymphocytes.
j. Dendritic cells arise only from the myeloid lineage.
k. All lymphoid cells have antigen-specific receptors on
their membrane.
l. All vertebrates generate B lymphocytes in bone marrow.
m. All vertebrates have a thymus.
n. Jawless vertebrates do not have lymphocytes.
2. For each of the following sets of cells, state the closest com-
mon progenitor cell that gives rise to both cell types.
a. Dendritic cells and macrophages
b. Monocytes and neutrophils
5. What are the two primary roles of the thymus?
6. At what age does the thymus reach its maximal size?
a. During the first year of life
b. Teenage years (puberty)
c. Between 40 and 50 years of age
d. After 70 years of age
7. Preparations enriched in hematopoietic stem cells are
useful for research and clinical practice. What is the role
of the SCID mouse in demonstrating the success of HSC
enrichment?
8. Explain the difference between a monocyte and a macrophage.
9. What effect would removal of the bursa of Fabricius (bur-
sectomy) have on chickens?
10. Indicate whether each of the following statements about
the lymph node and spleen is true or false. If you think a
statement is false, explain why.
a. The lymph node filters antigens out of the blood.
b. The paracortex is rich in T cells, and the periarteriolar
lymphoid sheath (PALS) is rich in B cells.
c. Only the lymph node contains germinal centers.
d. The FRCC enhances T cell/APC interactions
e. Afferent lymphatic vessels draining the tissue spaces
enter the spleen.
f. Lymph node but not spleen function is affected by a
knockout of the Ikaros gene.
11. For each description below (1–14), select the appropriate
cell type (a–p). Each cell type may be used once, more than
once, or not at all.
Descriptions
1. Major cell type presenting antigen to naïve T cells
2. Phagocytic cell of the central nervous system
3. Granulocytic cells important in the body’s defense
against parasitic organisms
4. Give rise to red blood cells
5. Generally first cells to arrive at site of inflammation
6. Support maintenance of hematopoietic stem cells
7. Give rise to thymocytes
8. Circulating blood cells that differentiate into macro-
phages in the tissues
9. An antigen-presenting cell that arises from the same
precursor as a T cell but not the same as a macrophage
10. Cells that are important in sampling antigens of the
intestinal lumen
Cells, Organs, and Microenvironments of the Immune System
11. Granulocytic cells that release various pharmacologically
active substances
12. White blood cells that play an important role in the devel-
opment of allergies
13. Cells that can use antibodies to recognize their targets
14. Cells that express antigen-specific receptors
Cell Types
a. Common myeloid progenitor cells
b. Monocytes
c. Eosinophils
|
CHAPTER 2
d. Dendritic cells
e. Natural killer (NK) cells
f. Mast cells
g. Neutrophils
h. M cells
i. Osteoblasts
j. Lymphocytes
k. NK1-T cell
l. Microglial cell
m. Myeloid dendritic cell
n. Hematopoietic stem cell
o. Lymphoid dendritic cell
63
This page lelt intentionally blank.
3
Receptors and Signaling:
B and T-Cell Receptors
T
he coordination of physiological functions
throughout the body depends on the ability of
individual cells to sense changes in their
environment and to respond appropriately. One
of the major routes by which a cell interprets its
surroundings is through the binding of signaling
molecules to cell-associated receptor proteins. A
molecule that binds to a receptor is a ligand.
Noncovalent binding of a ligand to its receptor may
induce alterations in the receptor itself, in its
polymerization state, and/or in the environment of that
receptor. These changes act to transmit or transduce the
ligand-binding signal into the interior of the cell, leading
to alterations in cellular functions. In the nervous
system, we would call the signaling molecules
neurotransmitters; in the endocrine system, hormones.
In the immune system, the foreign molecules that signal
the presence of non-self entities are antigens, and the
small molecules that communicate among the various
populations of immune cells are cytokines. Specialized
cytokines that induce chemo-attraction or -repulsion are
termed chemokines.
In this chapter, we provide a general introduction to
receptor-ligand binding and to the broad concepts and
strategies that underlie signal transduction. We then focus
specifically on the antigens and receptors of the adaptive
immune system, introducing the B- and T-cell receptors
and the intracellular signaling events that occur upon
antigen binding.
Because the cells of the immune system are distributed
throughout the body—with some resident in fixed tissues
and others circulating through the various lymphoid
tissues, the blood, and the lymphatics—the ability of these
cells to communicate with and signal to one another via
soluble cytokine and chemokine molecular messengers is
essential to their function. In Chapter 4, we describe the
signaling events that result when cytokines and chemokines
bind to their cognate (matching) receptors. Chapter 5
includes a description of the pattern recognition receptors
(PRRs) of the innate immune system and of the signaling
events initiated by their antigen binding.
All signaling events begin with a
complementary interaction between a ligand
and a receptor. This figure depicts the
molecular interaction between the variable
regions of an antibody molecule (light and
heavy chains are shown in blue and red,
respectively) and the tip of the influenza
hemagglutinin molecule, shown in yellow.
[Illustration based on x-ray crystallography data
collected by P. M. Colman and W. R. Tulip from GJVH
Nossal, 1993, Scientific American 269(3):22.]
■
■
Receptor-Ligand Interactions
Common Strategies Used in Many Signaling
Pathways
■
Frequently Encountered Signaling Pathways
■
The Structure of Antibodies
■
Signal Transduction in B Cells
■
T-Cell Receptors and Signaling
The essential concepts that underlie cell signaling in
the immune system can be summarized as follows:
65
•
A cellular signal is any event that instructs a cell to
change its metabolic or proliferative state.
•
Signals are usually generated by the binding of a ligand
to a complementary cell-bound receptor.
•
A cell can become more or less susceptible to the actions of
a ligand by increasing (up-regulating) or decreasing (downregulating) the expression of the receptor for that ligand.
66
PA R T I
|
Introduction
•
The ligand may be a soluble molecule, or it may be a
peptide, carbohydrate, or lipid presented on the surface
of a cell.
•
The ligand may travel long distances from its entry point
through the body in either the bloodstream or the lymphatics
before it reaches a cell bearing the relevant receptor.
•
Ligand-receptor binding is noncovalent, although it may
be of quite high affinity.
•
Ligand binding to the receptor induces a molecular
change in the receptor. This change may be in the form
of a conformational alteration in the receptor, receptor
dimerization or clustering, a change in the receptor’s
location in the membrane, or a covalent modification.
•
Such receptor alterations set off cascades of intracellular
events that include the activation of enzymes, and changes
in the intracellular locations of important molecules.
•
The end result of cellular signaling is often, but not
always, a change in the transcriptional program of the
target cell.
•
Sometimes a cell must receive more than one signal
through more than one receptor in order to effect a
particular outcome.
•
Integration of all the signals received by a cell occurs at
the molecular level inside the recipient cell.
are individually weak, many such interactions are required
to form a biologically significant receptor-ligand connection.
Furthermore, since each of these noncovalent interactions
operates only over a very short distance—generally about
1 Angstrom, (1 Å 10⫺10 meters)—a high-affinity receptorligand interaction depends on a very close “fit,” or degree of
complementarity, between the receptor and the ligand
(Figure 3-1).
How Do We Quantitate the Strength of
Receptor-ligand Interactions?
Consider a receptor, R, binding to a ligand, L. We can
describe their binding reaction according to the following
equation,
k1
R ⫹ L ∆ RL
k⫺1
(Eq. 3-1)
in which RL represents the bound receptor-ligand complex,
k1 is the forward or association rate constant, and k⫺1 is the
reverse or dissociation rate constant.
The ratio of k1/k⫺1 is equal to Ka, the association constant of the reaction, and is a measure of the affinity of the
Ligand
Receptor
NH2
Receptor-Ligand Interactions
The antigen receptors of the adaptive immune system are
transmembrane proteins localized at the plasma membrane.
Ligand binding to its cognate receptor normally occurs via
specific, noncovalent interactions between the ligand and the
extracellular portion of the membrane receptor. Although a
single lymphocyte expresses only one type of antigen receptor,
it also may express many different receptor molecules for signals such as cytokines and chemokines, and therefore a healthy
cell must integrate the signals from all the receptors that are
occupied at any one time.
CH2
OH ••• O
CH2
CH2 NH3+ –O
C
CH2
C
CH2 Hydrogen bond
CH2
CH2
Ionic bond
O
CH2
CH
CH3
Hydrophobic
interactions
CH3 CH3
CH3 CH
CH
CH2
van der Waals
interactions
CH3 CH3 CH
CH
CH3
O
Receptor-Ligand Binding Occurs via
Multiple Noncovalent Bonds
The surface of a receptor molecule binds to its complementary ligand surface by the same types of noncovalent chemical linkages that enzymes use to bind to their substrates.
These include hydrogen and ionic bonds, and hydrophobic
and van der Waals interactions. The key to a meaningful
receptor-ligand interaction is that the sum total of these
bonding interactions hold the two interacting surfaces
together with sufficient binding energy, and for sufficient
time, to allow a signal to pass from the ligand to the cell
bearing the receptor. Because these noncovalent interactions
CH2
+H
C
O–
3N
CH2
Ionic bond
FIGURE 3-1 Receptor-ligand binding obeys the rules of
chemistry. Receptors bind to ligands using the full-range of noncovalent bonding interactions, including ionic and hydrogen bonds
and van der Waals and hydrophobic interactions. For signaling to
occur, the bonds must be of sufficient strength to hold the ligand
and receptor in close proximity long enough for downstream events
to be initiated. In B- and T-cell signaling, activating interactions also
require receptor clustering. In an aqueous environment, noncovalent
interactions are weak and depend on close complementarity of the
shapes of receptor and ligand.
Receptors and Signaling: B and T-Cell Receptors
receptor-ligand pair. The association constant is defined as
the relationship between the concentration of reaction product, [RL], and the product of the concentrations of reactants,
[R] multiplied by [L]. The units of affinity are therefore M⫺1.
The higher the Ka, the higher the affinity of the interaction.
Ka
3RL4
3R 4 3L4
(Eq. 3-2)
The reciprocal of the association constant, the dissociation
constant, Kd, is often used to describe the interactions
between receptors and ligands. It is defined as:
Kd
3R4 3L 4
3RL4
(Eq. 3-3)
The units of the dissociation constant are in molarity (M).
Inspection of this equation reveals that when half of the
receptor sites are filled with ligand—that is, [R][RL]—the
Kd is equal to the concentration of free ligand [L]. The lower
the Kd, the higher the affinity of the interaction.
For purposes of comparison, it is useful to consider that
the Kd values of many enzyme-substrate interactions lie in
the range of 10⫺3 M to 10⫺5 M. The Kd values of antigenantibody interactions at the beginning of an immune
response are normally on the order of 10⫺4 to 10⫺5 M. However, since the antibodies generated upon antigen stimulation are mutated and selected over the course of an immune
response, antigen-antibody interactions late in an immune
response may achieve a Kd as low as 10⫺12 M. Under these
conditions, if an antigen is present in solution at a concentration as low as 10⫺12 M, half of the available antibody-binding
sites will be filled. This is an extraordinarily strong interaction between receptor and ligand.
Interactions Between Receptors and Ligands
Can Be Multivalent
Many biological receptors, including B-cell receptors, are multivalent—that is, they have more than one ligand binding site
per molecule. When both receptors and ligands are multivalent—as occurs, for example, when a bivalent immunoglobulin
receptor on the surface of a B cell binds to two, identical,
repeated antigens on a bacterial surface—the overall binding
interaction between the bacterial cell and the B-cell receptor is
markedly enhanced compared with a similar, but univalent,
interaction. In this way, multiple concurrent receptor-ligand
interactions increase the strength of binding between two cell
surfaces. Note, however, that binding via two identical receptor
sites to two identical ligands on the same cell may be a little less
than twice as firm as binding through a single receptor site.
This is because the binding of both receptor sites to two ligands
on a single antigen may somewhat strain the geometry of the
binding at one or both of the sites and therefore slightly interfere with the “fit” of the individual interactions.
Much of the benefit of multivalency results from the
fact that noncovalent-binding interactions are inherently
|
CHAPTER 3
67
reversible; the ligand spends some of its time binding to
the receptor, and some of its time in an unbound, or “off ”
state (Figure 3-2a). When more than one binding site is
involved, it is less likely that all of the receptor sites will
simultaneously be in the “off ” state, and therefore that the
receptor will release the ligand (Figures 3-2b and 3-2c).
The term avidity is used to describe the overall strength of
the collective binding interactions that occur during multivalent binding. In the early phases of the adaptive immune
response to multivalent antigens, B cells secrete IgM antibody,
which has 10 available antigen binding sites per molecule. IgM
is therefore capable of binding to antigens at a biologically
significant level, even if the affinity of each individual binding
site for its antigen is low, because of the avidity of the entire
antigen-antibody interaction. Likewise, when a cell-bound
receptor binds to a cell-bound ligand, their interaction may be
functionally multivalent, even if the individual receptor molecules are monovalent, because multiple receptor molecules
can cluster in the cell membrane and participate in the receptor-ligand interaction. In addition, other co-receptor interactions can also contribute to the overall avidity of the
engagement between a cell and its antigen (see below).
The affinity of receptor-ligand interactions can be measured by techniques such as equilibrium dialysis or surface
plasmon resonance (SPR). Both of these methods are
described in Chapter 20.
(a)
“On”
“Off”
FIGURE 3-2 Monovalent and bivalent binding. (a) Monovalent binding. The receptor exists in equilibrium with its ligand, represented here as a circle. Part of the time it is bound (binding is in the
“on” state), and part of the time it is unbound (binding is in the “off ”
state). The ratio of the time spent in the “on” versus the “off ” state
determines the affinity of the receptor-ligand interaction and is
related to the strength of the sum of the noncovalent binding interactions between the receptor and the ligand.
PA R T I
68
|
Introduction
(b)
(c)
FIGURE 3-2 (continued) (b) Bivalent binding. Here, we visualize
a bivalent receptor, binding to two sites on a single multivalent ligand.
This could represent, for example, a bivalent antibody molecule binding to a bacterium with repeated antigens on its surface. In this rendering, both sites are occupied. (c) In this rendering, the bivalent receptor
has momentarily let go from one of its sites, but still holds onto the
ligand with the other. Large proteins such as antibodies continuously
vibrate and “breathe,” resulting in such on-off kinetics. Bivalent or multivalent binding helps to ensure that, when one site momentarily
releases the ligand, the interaction is not entirely lost, as it is in part (a).
Receptor and Ligand Expression Can Vary
During the Course of an Immune Response
cytes and antigen-presenting cells engaged in an immune
response spend significant amounts of time held together by
multiple receptor-ligand interactions. Cytokine signals
released by a T cell, for example, are received by a bound
antigen-presenting cell at the interface between the cells,
before the cytokine has time to diffuse into the tissue fluids.
The secretion of cytokines into exactly the location where
they are needed is facilitated by the ability of antigen receptor signaling to induce redistribution of the microtubule
organizing center (MTOC) within the activated T cell. The
MTOC reorientation, in turn, causes the redistribution of
the secretory organelles (the Golgi body and secretory vesicles) within the T-cell cytoplasm, so that cytokines made by
the T cell are secreted in the direction of the T-cell receptor,
which, in turn, is binding to the antigen-presenting cell. The
technical term for this phenomenon is the vectorial (directional) redistribution of the secretory apparatus. Thus,
cytokines are secreted directly into the space between the
activated T cell and antigen-presenting cell.
Vectorial redistribution of the secretory apparatus in antigen-presenting dendritic cells also occurs upon interaction
with antigen-specific T cells. This ensures that cytokines,
such as IL-12, which are secreted by dendritic cells, are efficiently delivered to antigen-recognizing T cells. Figure 3-4
shows the directional secretion of the cytokine IL-12 by an
antigen-presenting dendritic cell engaged in the activation
of a T cell.
The effector cells of the immune system, such as B cells,
are also capable of presenting antigens on their cell surface
for recognition by helper T cells. A B cell that has specifically
recognized an antigen through its own receptor will process
the antigen and express a particularly high concentration of
those antigenic peptides bound to Major Histocompatibility
Complex (MHC) molecules on its surface. A helper T cell,
binding to an antigen presented on the surface of a B cell,
One of the most striking features of the molecular logic of
immune responses is that receptors for some growth factors
and cytokines are expressed only on an as-needed basis.
Most lymphocytes, for example, express a hetero-dimeric
(two-chain), low-affinity form of the receptor for the cytokine interleukin 2 (IL-2), which initiates a signal that promotes lymphocyte proliferation. (This cytokine and its
receptor are covered in more detail in Chapter 4.) This lowaffinity form of the receptor is unable to bind to IL-2 at
physiological cytokine concentrations. However, when a
lymphocyte is activated by antigen binding, the signal from
the antigen-binding receptor causes an increase in the cell
surface expression of a third chain of the IL-2 receptor (see
Figure 3-3). Addition of this third chain converts the lowaffinity form of the IL-2 receptor to a high-affinity form,
capable of responding to the cytokine levels found in the
lymphoid organs. The functional corollary of this is that only
those lymphocytes that have already bound to antigen and
have been stimulated by that binding have cytokine receptors of sufficiently high affinity to respond to the physiological cytokine concentrations. By waiting until it has interacted
with its specific antigen before expressing the high-affinity
cytokine receptor, the lymphocyte both conserves energy
and prevents the accidental initiation of an immune response
to an irrelevant antigen.
Local Concentrations of Cytokines and Other
Ligands May Be Extremely High
When considering the interactions between receptors and
their ligands, it is important to consider the anatomical environment in which these interactions are occurring. Lympho-
Receptors and Signaling: B and T-Cell Receptors
FIGURE 3-3 The expression of some cytokine receptors is
upregulated following cell stimulation. White blood cells
stained with DAPI, a stain that detects DNA (blue), were either
untreated (left) or stimulated using the human T-cell mitogen
will therefore deliver a high concentration of cytokines
directly to an antigen-specific, activated B cell. The local
concentration of cytokines at the cellular interfaces can
therefore be extremely high, much higher than in the tissue
fluids generally.
Common Strategies Used in Many
Signaling Pathways
A signal transduction pathway is the molecular route by
which a ligand-receptor interaction is translated into a
biochemical change within the affected cell. Communication of the ligand’s message to the cell is initiated by the
complementarity in structure between the ligand and its
receptor, which results in a binding interaction of suffi-
|
CHAPTER 3
69
phytohemagglutinin (right). Upon stimulation, the IL-2 receptor
alpha (IL-2R␣) chain (yellow) is upregulated, increasing affinity of
the IL-2 receptor for IL-2. A similar IL-2R␣ upregulation is seen upon
antigen stimulation. [Courtesy R&D Systems, Inc., Minneapolis, MN, USA.]
cient strength and duration to bring about a biochemical
change in the receptor and/or its associated molecules.
This biochemical change can, in turn, cause a diverse array
of biochemical consequences in the cell (Overview Figure 3-5). The terms upstream and downstream are frequently used to describe elements of signaling pathways.
The upstream components of a signaling pathway are
those closest to the receptor; the downstream components
are those closest to the effector molecules that determine
the outcome of the pathway—for example, the transcription factors or enzymes whose activities are modified
upon receipt of the signal.
As disparate and complex as some of these signaling
pathways may be, many share common features. In this section, in order to provide a framework for analyzing the individual pathways employed by cells of the immune system, we
FIGURE 3-4 Polarized secretion of IL-12 (pink) by dendritic
cells (blue) in the direction of a bound T cell (green). The
higher magnification micrograph shows the secretion of packets of
IL-12 through the dendritic cell membrane. [J. Pulecio et al., 2010, Journal
of Experimental Medicine 207:2,719.]
70
PA R T I
|
Introduction
3-5
OVERVIEW FIGURE
Concepts in lymphocyte signaling
Coreceptor
Ligand
Receptor
Receptor-associated
molecules
Receptor-associated
tyrosine kinase
PIP2
PIP2
PIP3
DAG
PIP3
DAG
Ras-GRP
PLCγ
PI3
kinase
IP3
Ca2+ release from
the endoplasmic
reticulum
P
PKC
SOS
P
P
P
Adapter
P
DAG
P
GDP/GTP
exchange
P
P
Adapter
P
Phosphorylation
Ras
Calmodulin and
calcineurin activation
NF-κB
P
P
P
IκB
Ubiquitination
and destruction
MAP kinase
cascade
NFAT
Cytoplasm
NFAT
Nucleus
Ligand binding to receptors on a cell induces a variety of downstream effects, many of which culminate in transcription factor
activation. Here we illustrate a few of the pathways that are
addressed in this chapter. Binding of receptor to ligand induces
clustering of receptors and signaling molecules into regions of
the membrane referred to as lipid rafts (red). Receptor binding of
ligand may be accompanied by binding of associated coreceptors to their own ligands, and causes the activation of
receptor-associated tyrosine kinases, which phosphorylate receptor-associated proteins. Binding of downstream adapter molecules to the phosphate groups on adapter proteins creates a
scaffold at the membrane that then enables activation of a variety
of enzymes including phospholipase C␥ (PLC␥), PI3 kinase, and
additional tyrosine kinases. PLC␥ cleaves phosphatidyl inositol
bisphosphate (PIP2) to yield inositol trisphosphate (IP3), which
interacts with receptors on endoplasmic reticulum vesicles to
AP-1
NF-κB
Gene activation
cause the release of calcium ions. These in turn activate calcineurin,
which dephosphorylates the transcription factor NFAT, allowing it
to enter the nucleus. Diacylglycerol (DAG), remaining in the membrane after PLC␥ cleavage of PIP2, binds and activates protein
kinase C (PKC), which phosphorylates and activates enzymes
leading to the destruction of the inhibitor of the transcription factor NF-B. With the release of the inhibitor, NF-B enters the
nucleus and activates a series of genes important to the immune
system. Binding of the adapter protein Ras-GRP to the signaling
complex allows for the binding and activation of the Guanine
nucleotide Exchange Factor (GEF) Son of Sevenless (SOS), which
in turn initiates the phosphorylation cascade of the MAP kinase
pathways. This leads to the entry of a third set of transcription factors into the nucleus, and activation of the transcription factor
AP-1. (Many details that are explained in the text have been omitted from this figure for clarity.)
Receptors and Signaling: B and T-Cell Receptors
describe some of the strategies shared by many signaling
pathways. The next section provides three examples of signaling pathways used by the immune system, as well as by
other organ systems.
Ligand Binding Can Induce Conformational
Changes in, and/or Clustering of, the Receptor
The first step necessary to the activation of a signaling pathway is that the binding of the ligand to its receptors in some
way induces a physical or chemical change in the receptor
itself, or in molecules associated with it. In the case of many
growth factor receptors, ligand binding induces a conformational change in the receptor that results in receptor
dimerization (Figure 3-6). Since the cytoplasmic regions of
many growth factor receptors have tyrosine kinase activity,
this results in the reciprocal phosphorylation of the cytoplasmic regions of each of the receptor molecules by its
dimerization partner.
Other receptors undergo conformational changes upon
ligand binding that result in higher orders of receptor
polymerization. The receptors on B cells are membranebound forms of the antibody molecules that the B cell will
eventually secrete. Two different types of antigen receptors
exist on the surface of naïve B cells, which we term immunoglobulins M and D (IgM and IgD). Structural studies of the
IgM form of the receptor have revealed that ligand binding
induces a conformational change in a nonantigen binding
part of the receptor, close to the membrane, that facilitates
aggregation of the receptors into multimeric complexes and
their subsequent movement into specialized membrane
Growth
factor
Cytosol
P
P
P
Inactive receptor
tyrosine kinases
CHAPTER 3
71
regions. T-cell receptors similarly cluster upon antigen binding. Whether or not they undergo a conformational change
upon antigen binding remains controversial.
Some Receptors Require Receptor-Associated
Molecules to Signal Cell Activation
In contrast to growth factor receptors, which have inducible
enzyme activities built into the receptor molecule itself,
B- and T-cell receptors have very short cytoplasmic components and therefore need help from intracellular
receptor-associated molecules to bring about signal transduction. The Ig␣/Ig (CD79␣/) heterodimer in B cells, and
the hexameric CD3 complex in T cells are closely associated
with their respective antigen receptors and are responsible
for transmitting the signals initiated by ligand binding into
the interior of the cell (Figure 3-7). Both of these complexes
have a pair of long cytoplasmic tails that contain multiple
copies of the Immuno-receptor Tyrosine Activation Motif
or ITAM. ITAMs are recurrent sequence motifs found on
many signaling proteins within the immune system, which
contain tyrosines that become phosphorylated following
signal transduction through the associated receptor. Phosphorylation of ITAM-tyrosine residues then allows docking
of adapter molecules, thus facilitating initiation of the signaling cascade.
Other molecules associated with the B- or T-cell antigen
receptors may also interact with the antigen or with other
molecules on the pathogen’s surface. For example, in the case
of B cells, the CD19/CD21 complex binds to complement
molecules covalently attached to the antigen (see Figure
3-7a). Similarly, CD4 and CD8 molecules on T cells bind to
nonpolymorphic regions of the antigen-presenting MHC
molecule and aid in signal transduction (see Figure 3-7b).
Finally, the co-receptor CD28 on naïve T cells must interact
with its ligands CD80 (B7-1) and CD86 (B7-2) for full T-cell
activation to occur.
P
Ligand-Induced Receptor Clustering
Can Alter Receptor Location
P
P
Ligand-induced clustering of B- and T-cell receptors slows
down the rates of their diffusion within the planes of their
respective cell membranes, and facilitates their movement
into specialized regions of the lymphocyte membrane
known as lipid rafts. These rafts are highly ordered, detergent-insoluble, cholesterol- and sphingolipid-rich membrane regions, populated by many molecules critical to
receptor signaling. Moving receptors and co-receptors into
the lipid rafts renders them susceptible to the action of
enzymes associated with those rafts. Figure 3-8 shows how
the raft-associated tyrosine kinase Lyn initiates the B-cell
signaling cascade by phosphorylating the receptor-associated
molecules Ig␣ and Ig. The tyrosine kinase Lck serves a
similar role in the TCR signaling cascade.
P P
Tyrosine
kinase
domain
|
Kinase activity stimulated
by cross-phosphorylation
FIGURE 3-6 Some growth factors induce dimerization of
their receptors, followed by reciprocal tyrosine phosphorylation of the receptor molecules. Many growth factor receptors
possess tyrosine kinase activity in their cytoplasmic regions. Dimerization of the receptor occurs on binding of the relevant ligand and
allows reciprocal phosphorylation of the dimerized receptor chains
at multiple sites, thus initiating the signal cascade. As one example,
stem cell factor binds to its receptor, c-kit (CD117) on the surface of
bone marrow stromal cells.
PA R T I
72
|
Introduction
C3d
Antigen
CD21
mIgM
CD81
(TAPA-1)
Class II
MHC
CD4
CD80 or
CD86
Peptide
CD19
TCR
δ ε
CD28
γ ε
ζ ζ
ss
lgα/lgβ
(CD79α,β)
ITAMs
CD3
FIGURE 3-7 Both B- and T-cell receptors require receptorassociated molecules and co-receptors for signal transduction. (a) B-cell receptors require Ig␣/Ig to transmit their signal. The
CD21 co-receptor, which is associated with CD19, binds to the complement molecule C3d, which binds covalently to the antigen. The
interaction between CD21 on the B cell, and C3d associated with the
antigen, keeps the antigen in contact with the B-cell receptor, even
when the antigen-BCR binding is relatively weak. The yellow bands
on the cytoplasmic regions of the receptor-associated molecules
indicate Immuno-receptor Tyrosine Activation Motifs (ITAMs). Phosphorylation of tyrosine residues in these motifs allows the binding of
downstream adapter molecules and facilitates signal transduction
from the receptors. Both CD19 and Ig␣/Ig bear intracytoplasmic
ITAMs and, along with CD81 (TAPA-1), participate in downstream
signaling events. (b) T-cell receptors use CD3, a complex of ␦, ␥
chains and a pair of chain molecules or a pair. CD4 and CD8 coreceptors bind to the non-polymorphic region of the MHC class II or
class I molecules, respectively. This figure illustrates CD4 binding to
MHC Class II. This binding secures the connection between the T cell
and the antigen presenting cell, and also initiates a signal through
CD4/8. The CD28 co-receptor provides another signal upon binding
to CD80 or CD86. CD28 binding to costimulatory molecules on
antigen-presenting cells is required for stimulation of naïve, but not
memory, T cells.
Antigen
BCR
Raft
lgα/lgβ
Lyn
+ Antigen
Lyn
– Antigen
Lyn
BLNK
P
P
Syk
B cells, the B-cell receptor (BCR) is excluded from the lipid rafts, which
are regions of the membrane high in cholesterol and rich in glycosphingolipids. The raft is populated by tyrosine kinase signaling
molecules, such as Lyn. On binding to antigen, the BCR multimerizes
(clusters), and the changes in the conformation of the BCR brought
BLNK
Syk
Signal
transduction
FIGURE 3-8 Lipid raft regions within membranes. In resting
P
P
Receptor
internalization
about by this multimerization increase the affinity of the BCR for the
raft lipids. Movement of the BCR into the raft brings it into contact
with the tyrosine kinase Lyn, which phosphorylates the receptorassociated proteins Ig␣/Ig, thus initiating the activation cascade. A
similar movement of TCRs into lipid rafts occurs upon T-cell activation. [Adapted from S. K. Pierce, 2002, Nature Reviews Immunology 2:96.]
Receptors and Signaling: B and T-Cell Receptors
Inactive
Active
SH2
SH2
P
pY508
Y397
Kinase
Y508
P
pY397
Kinase
FIGURE 3-9 Activation of Src-family kinaes. Src-family
kinases are maintained in an inactive closed configuration by the
binding of a phosphorylated inhibitory tyrosine residue (pY508 in this
example) with an SH2 domain in the same protein. Dephosphorylation of this tyrosine opens up the molecule, allowing substrate access
to the enzymatic site. Opening up the kinase also allows phosphorylation of a different internal tyrosine (pY397), which further stimulates
the Src kinase activity.
Tyrosine Phosphorylation is an Early Step in
Many Signaling Pathways
Many signaling pathways, particularly those that signal cell
growth or proliferation, are initiated with a tyrosine phosphorylation event. Many of the tyrosine kinases that initiate
BCR and TCR activation belong to the family of enzymes
known as the Src-family kinases. Src (pronounced “sark”)
family kinases have homology to a gene, c-src, that was first
identified in birds. A constitutively active Rous sarcoma virus
TABLE 3-1
|
CHAPTER 3
73
homolog of these genes, v-src, was shown to induce a form of
cancer, fibrosarcoma, in birds. The fact that a simple mutation in this viral form of a tyrosine kinase gene could result
in the development of a tumor was the first hint that tyrosine
kinase genes may be important in the regulation of cell proliferation and, indeed, in many aspects of cell signaling.
Src-family kinases are important in the earliest stages of
activation of both T and B cells: Lck and Fyn are critical to
T-cell activation, and Lyn, Fyn, and Blk play the corresponding
initiating roles in B cells. Since inadvertent activation of these
enzymes can lead to uncontrolled proliferation—a precursor
to tumor formation—it is not surprising that their activity is
tightly regulated in two different but interconnected ways.
Inactive Src-family tyrosine kinase enzymes exist in a
closed conformation, in which a phosphorylated tyrosine is
tightly bound to an internal SH2 domain (Src homology 2
domain) (Figure 3-9, Table 3-1). (SH2 domains in proteins
bind to phosphorylated tyrosine residues.) For as long as the
inhibitory tyrosine is phosphorylated, the Src-family kinase
remains folded in on itself and inactive. In lymphocytes, the
tyrosine kinase enzyme Csk is responsible for maintaining
phosphorylation of the inhibitory tyrosine. However, upon
cell activation, a tyrosine phosphatase removes the inhibitory
phosphate and the Src-family kinase opens up into a partially
active conformation. Thus, the initiating event in signal transduction is often the movement of the tyrosine kinase into a
region of the cell or the membrane populated by an appropriate, activating phosphatase and distant from the inhibitory
kinase Csk. Full activity is then achieved when the Src-kinase
phosphorylates itself on a second, activating tyrosine residue.
Selected domains of adapter proteins and their binding target motifs
Domain of adapter protein
Binding specificity of adapter domain
Src-homology 2 (SH2)
Specific phosphotyrosine (pY)–containing motifs in the context of 3–6 amino acids located carboxyterminal to the pY (An invariant arginine in the SH2 domain is required for pY binding.)
Src-homology 3 (SH3)
Proline-rich sequences in a left-handed polyproline helix (Proline residues are usually preceded by an
aliphatic residue.)
Phosphotyrosine-binding (PTB)
pY-containing peptide motifs (DPXpY, where X is any amino acid) in the context of amino-terminal
sequences
Pleckstrin homology (PH)
Specific phosphoinositides, which allow PH-containing proteins to respond to the generation of lipid
second messengers generated by enzymes such as PI3 kinase
WW
Bind proline-rich sequences (Derive their name from two conserved tryptophan [W] residues 20–22
amino acids apart)
Cysteine-rich sequences (C1)
Diacylglycerol (DAG) (On association with DAG, the C1 domain exhibits increased affinity for the lipid
membrane, promoting membrane recruitment of C1-containing proteins.)
Tyrosine kinase-binding (TKB)
Phosphotyrosine-binding domain divergent from typical SH2 and PTB domains (Consists of three
structural motifs [a four-helix bundle, an EF hand, and a divergent SH2 domain], that together form an
integrated phosphoprotein-recognition domain)
Proline-rich
Amino-acid sequence stretches rich in proline residues, able to bind modular domains including SH3
and WW domains
14-3-3- binding motifs
Phosphorylated serine residues in the context of one of the two motifs RSXpSXP and RXXXpSXp
[After G. A. Koretsky and P. S. Myung, 2001, Nature Reviews Immunology 1:95]
74
PA R T I
|
Introduction
Tyrosine phosphorylation can bring about changes in a
signaling pathway in more than one way. Sometimes, phosphorylation of a tyrosine residue can induce a conformational change in the phosphorylated protein itself, which can
turn its enzymatic activity on or off. Alternatively, tyrosine
phosphorylation of components of a receptor complex can
permit other proteins to bind to it via their SH2 or phosphotyrosine binding domains (see Table 3-1), thereby altering
their locations within the cell. Note that a phosphorylated
tyrosine is often referred to as a “pY” residue.
specificity for a particular molecular structure. In many signaling pathways, multiple adapter proteins may participate in
the formation of a protein scaffold that provides a structural
framework for the interaction among members of a signaling
cascade. Particularly common domains in adapter proteins
functioning in immune system signaling are the SH2 domain
that binds to phosphorylated tyrosine (pY) residues, the SH3
domain that binds to clusters of proline residues, and the
pleckstrin homology (PH) domain that binds to phosphatidyl inositol trisphosphate in the plasma membrane.
Binding of adapter proteins may simply bring molecules
into contact with one another, so that, for example, an enzyme
may act on its substrate. Alternatively, the binding of an
adapter protein may induce a conformational change, which in
turn can stabilize, destabilize, or activate the binding partner.
Signal transduction may induce conformational changes
in proteins that in turn result in the uncovering of one or
more protein domains with specific affinity for other proteins. Such interactions may be homotypic (interactions
between identical domains) or heterotypic (interactions
between different domains).
Adapter Proteins Gather Members
of Signaling Pathways
It is easy to imagine the cytoplasm as an ocean of macromolecules, with little structure or organization, in which proteins bump into one another at a frequency dependent only
on their overall cytoplasmic concentrations. However, nothing could be further from the truth. The cytoplasm is, in fact,
an intricately organized environment, in which threedimensional arrays of proteins form and disperse in a manner directed by cellular signaling events. Many of these
reversible interactions between proteins are mediated by
adapter proteins as well as by interactions between members of signaling pathways with cytoskeletal components.
Adapter proteins have no intrinsic enzymatic or receptor
function, nor do they act as transcription factors. Their function is to bind to specific motifs or domains on proteins or
lipids, linking one to the other, bringing substrates within the
range of enzymes, and generally mediating the redistribution
of molecules within the cell. Table 3-1 lists a number of representative adapter domains along with their binding specificities. Adapter proteins are characterized by having multiple
surface domains, each of which possesses a precise binding
⫹ NH3
CH2
CH
COO ⫺
Phosphorylation on Serine and Threonine
Residues is also a Common Step in
Signaling Pathways
Whereas tyrosine phosphorylation is frequently seen at the
initiation of a signaling cascade, the phosphorylation of proteins at serine and threonine residues (Figure 3-10) tends to
occur at later steps in cell activation. Serine or threonine
phosphorylation may serve to activate a phosphorylated
enzyme, to induce the phosphorylated protein to interact
with a different set of proteins, to alter the protein’s location
within the cell, to protect the protein from destruction, or, in
COO ⫺
COO ⫺
⫹
H3N
⫹
CH
H3N
CH
CH2OH
CH
H3C
OH
Tyrosine
Threonine
Serine
Tyrosine
kinase
OH
Serine/threonine
kinases
⫹ NH3
CH2
CH
COO ⫺
COO ⫺
COO
⫺
⫹
H3N
OPO32⫺
Phosphotyrosine
⫹
H3N
CH
CH2OPO32⫺
3-Phosphoserine
FIGURE 3-10 Phosphorylated tyrosine, serine, and threonine.
CH
CH
H3C
OPO32⫺
Phosphothreonine
Receptors and Signaling: B and T-Cell Receptors
some important instances, to convert the phosphorylated
protein into a target for proteasomal destruction.
Phosphorylation of Membrane Phospholipids
Recruits PH Domain-Containing Proteins to
the Cell Membrane
The phospholipid Phosphatidyl Inositol bis-Phosphate
(PIP2) is a component of the inner face of eukaryotic plasma
membranes. However, PIP2 is much more than a structural
phospholipid; it also actively participates in cell signaling. The
enzyme Phosphatidyl Inositol-3-kinase (PI3 kinase), activated during the course of signaling through many immune
receptors, phosphorylates PIP2 to form Phosphatidyl Inositol
tris-Phosphate (PIP3) (Figure 3-11). PIP3 remains in the
membrane and serves as a binding site for proteins bearing
pleckstrin homology (PH) domains (see Table 3-1). Thus, this
lipid phosphorylation event serves to move proteins from the
cytosol to the membrane by providing them with a binding site
O
O
O
O
O
O⫺
OH
P
O
O
OH
O
O⫺
O
P
O⫺
O
O
O
O
O⫺
P
P
O⫺
O⫺
⫺
O
FIGURE 3-11 PIP2 can be phosphorylated (red) by PI3
kinase to create PIP3 during cellular activation. PIP3 then creates binding sites for proteins with PH domains at the internal face of
the plasma membrane.
|
CHAPTER 3
75
at the inner membrane surface. Localization of proteins at the
membrane brings them into contact with other members of a
signaling cascade, allowing enzymes access to new substrates
and enabling the modification of adapter proteins with the
subsequent assembly of signaling complexes.
Signal-Induced PIP2 Breakdown by PLC
Causes an Increase in Cytoplasmic Calcium
Ion Concentration
In addition to being phosphorylated by PI3 kinase, PIP2 is also
susceptible to a second type of signal-induced biochemical
modification. A second enzyme (more correctly, a family of
enzymes), Phospholipase C (PLC), is also activated upon antigen signaling of lymphocytes (Figure 3-12). PLC hydrolyzes
PIP2, cleaving the sugar inositol trisphosphate (IP3) from the
diacylglyerol (DAG) backbone. The DAG remains in the
membrane, where it binds and activates a number of important
signaling enzymes. IP3 is released into the cytoplasm, where it
interacts with specific IP3 receptors on the surface of endoplasmic reticulum vesicles, inducing the release of stored calcium
ions (Ca2⫹) into the cytoplasm. These calcium ions then bind
several cellular proteins, changing their conformation and
altering their activity.
The calcium ion’s double positive charge and the size of its
ionic radius render it particularly suitable for binding to many
proteins; however, in the field of cell signaling, calmodulin
(CaM) is unquestionably the most important calcium-binding
protein. As shown in Figure 3-13a, CaM is a dumbbell-shaped
protein, with two globular subunits separated by an ␣-helical
rod. Each of the two subunits has two calcium binding sites,
such that calmodulin has the capacity to bind four Ca2⫹ ions.
On binding calcium, CaM undergoes a dramatic conformational change (Figure 3-13b), which allows it to bind to and
activate a number of different cellular proteins.
As cytoplasmic Ca2⫹ ions are used up in binding to cytoplasmic proteins, and the free cytoplasmic Ca2⫹ ion concentration drops, Ca2⫹ channel proteins begin to assemble at the
membrane. Eventually the channels open to allow more calcium to flood in from the extracellular fluid and complete
the activation of the calcium-regulated proteins. The assembly of these so-called store-operated calcium channel proteins
is prohibited in the presence of high intracellular calcium.
Many different types of ions are present in a cell, and it is
reasonable to ask why eukaryotes should have evolved proteins with activities that are so responsive to intracellular
Ca2⫹ ion concentrations. The answer to this question in part
lies in the fact that it is relatively easy to alter the intracytoplasmic concentration of Ca2⫹ ions. The concentration of
Ca2⫹ in the blood and tissue fluids is on the order of 1 mM,
whereas the cytosolic concentration in a resting cell is closer
to 100 nM—10,000 times lower than the extracellular concentration. This difference is maintained by an efficient
(although energetically expensive) system of membrane
pumps. Furthermore, higher-concentration Ca2⫹ stores exist
PA R T I
76
|
Introduction
O
O
O
O
P HO
O
2⫺
OH OPO3
OPO32⫺
OH
`
O`
⫺
H
O
O
Phosphatidylinositol 4,5-bisphosphate (PIP2)
Phospholipase C
O
OH
2⫺
OH OPO3
`
O`
H
+
O
HO
2⫺O PO
3
OPO32⫺
OH
O
Diacylglycerol (DAG)
(Remains associated with the membrane)
Inositol 1,4,5-trisphosphate (IP3)
Enters the cytoplasm and induces
Ca2+ ion release from ER vesicles
FIGURE 3-12 PIP2 can be hydrolyzed by PLC to create DAG and IP3.
within intracellular vesicles associated with the endoplasmic
reticulum and the mitochondria. Thus, any signaling event
that opens up channels in the membrane of the endoplasmic
reticulum, or in the plasma membrane, and allows for free
flow of Ca2⫹ ions into the cytoplasm, facilitates a rapid rise
in intracytoplasmic Ca2⫹ ion concentration.
Ubiquitination May Inhibit or Enhance
Signal Transduction
The protein ubiquitin is a small, highly conserved, 76-residue,
monomeric protein. Binding of the carboxy-terminal of
ubiquitin to lysine residues of target proteins is often followed
Ca2+
(b)
(a)
Ca2+
Calmodulin
1
Ca2+
+NH
2
COOH
3
Ca2+
Calciumcalmodulin
complex
Target
protein
Calmodulinbinding site
3
Calmodulin binds
Ca2+ ions.
Calmodulin changes
conformation,
becomes active.
Calmodulin binds
a target protein.
FIGURE 3-13 The calcium-binding regulatory protein calmodulin undergoes a conformational shift on binding to calcium,
which enables it to bind and activate other proteins. [Part (a) PDB ID 3CLN.]
Receptors and Signaling: B and T-Cell Receptors
by polymerization of a chain of ubiquitins onto selected
lysine residues of the conjugated ubiquitin. This results in
poly-ubiquitination of the target protein. In most instances,
mono- or poly-ubiquitination serves as a mechanism by
which the tagged proteins are targeted for destruction by the
proteasome of the cell. However, we now know that ubiquitination on certain residues of proteins can also serve as an
activating signal, and we will come across this alternative role
of ubiquitin below as we learn about the activation of the
NF-B transcription factor, as well as in Chapters 4 and 5.
Frequently Encountered
Signaling Pathways
Analysis of the biochemical routes by which molecular signals pass from cell surface receptors to the interior of the cell
has revealed that a few signal transduction pathways are
used repeatedly in the cellular responses to different ligands.
These pathways are utilized, in slightly different forms, in
multiple cellular and organ systems, and in many species.
The end result of most of these pathways is an alteration in
the transcriptional program of the cell. Here we briefly
describe three such pathways of particular relevance to the
adaptive immune system (see Overview Figure 3-5). Each of
these pathways is triggered by antigen binding to the receptor and leads to the activation of a different family of transcription factors. The generation of these active transcription
factors in turn initiates the up-regulation of a cascade of
genes important to the immune response, including those
encoding cytokines, antibodies, survival factors, and proliferative signals.
Although the conceptual framework of these signaling
pathways is shared among many cell types, small modifications in the molecular intermediates and transcription factors involved allows for variations in the identities of the
genes controlled by particular signals and for cell-type specific transcriptional control mechanisms. For example, the
enzyme PLC exists in different forms; PLC␥1 is used in T
cells and PLC␥2 in B cells. Similarly, the Nuclear Factor of
Activated T cells (NFAT) family of transcription factors
includes five members, some of which are expressed as differentially spliced variants, and each of which is affected by
different arrays of molecular signals.
Furthermore, activation of transcription from some promoters requires the binding of multiple transcription factors. For example, full expression of the gene encoding
interleukin 2 (IL-2) requires the binding of AP-1, NF-B,
and NFAT, in addition to several other transcription factors,
to the interleukin promoter. In this way, the promoter can be
partially activated in the presence of some of these transcription factors, but not fully active until all of them are in place.
In the next section, we focus on the general principles that
govern the control of these three pathways, from the receipt
of an antigen signal, to the activation of a particular type of
transcription factor.
|
CHAPTER 3
77
The PLC Pathway Induces Calcium
Release and PKC Activation
We have already met the essential elements of this pathway in
our discussion of the enzymes that modulate PIP2 upon cell
activation (see Figure 3-12), and we now know that PLC breaks
down PIP2 to IP3 and DAG. But how does PLC become activated in the first place? We use T cells here as our example.
As mentioned, T cells use the PLC␥1 form of the enzyme.
On antigen stimulation, tyrosine phosphorylation of the ITAM
residues of the CD3 receptor-associated complex results in the
localization of the adapter protein LAT to the membrane (Figure 3-14). LAT in turn is phosphorylated, and PLC␥1 then
binds to phosphorylated LAT. This binding ensures that PLC␥1
localizes to the cell membrane, the site of its substrate, PIP2.
CD4
TCR/CD3
PIP3
PIP2 DAG
Itk
PLCγ1
P
Lck
P
P
LAT
IP3
Endoplasmic
reticulum
Ca2+
Calcineurin
Calmodulin
(inactive)
Calmodulin
(active)
P
NFAT
P
Cytoplasm
NFAT
Nucleus
Transcription
factor activation
FIGURE 3-14 Activation of calcineurin by binding of the
calcium-calmodulin complex induces NFAT dephosphorylation and nuclear entry.
78
PA R T I
|
Introduction
Once located at the plasma membrane, PLC␥1 is further activated by tyrosine phosphorylation mediated by the receptoractivated kinase, Lck, as well as by a second tyrosine kinase, Itk.
The phosphorylated and activated PLC␥1 then mediates cleavage of PIP2 to IP3 and DAG as described above.
How do these secondary mediators then function to
bring about lymphocyte activation? Calcium ions, released
from intracellular stores by IP3, bind to the signaling intermediate, calmodulin. Each molecule of calmodulin binds to
four calcium ions, and this binding results in a profound
alternation in its conformation (see Figure 3-13). The
calcium-calmodulin complex then binds and activates the
phosphatase calcineurin, which dephosphorylates the transcription factor NFAT (see Figure 3-14). This dephosphorylation induces a conformational change in NFAT, revealing a
nuclear localization sequence, which directs NFAT to enter
the nucleus and activate the transcription of a number of
important T-cell target genes. Genes activated by NFAT
binding include that encoding interleukin 2, a cytokine pivotally important in the control of T cell proliferation.
The importance of the PLC␥1 pathway to T cell activation
is illustrated by the profound immunosuppressant effects of the
drug cyclosporine, which is used to treat both T-cell–mediated
autoimmune disease and organ transplant rejection. In T cells,
cyclosporine binds to the protein cyclophilin (immunophilin),
and the cyclosporine/cyclophilin complex binds and inhibits
calcineurin, effectively shutting down T-cell proliferation.
PIP2 cleavage by PLC is a particularly efficient signaling strategy as both products of PIP2 breakdown are active in subsequent
cellular events. DAG, the second product of PIP2 breakdown
(see Figure 3-12) remains in the membrane, where it binds and
activates enzymes of the protein kinase C (PKC) family. These
kinases are serine/threonine kinases active in a variety of signaling pathways. Protein kinase C, important in T-cell signaling,
only requires DAG binding, whereas its relative, protein kinase
C, must also bind Ca2⫹ ions for full activation. DAG is also
implicated in the Ras pathway (described in the next section).
The Ras/Map Kinase Cascade Activates
Transcription Through AP-1
The Ras signaling pathway was discovered initially after a
mutated, viral form of the Ras protein was found to induce
cancer in a rat model. Ras is a monomeric, GTP-binding
protein (G protein, for short). When Ras binds to GTP, its
conformation changes into an active state, in which it is
capable of binding, and activating a number of serine/threonine kinases. However, the Ras protein possesses an intrinsic
GTPase activity that hydrolyzes GTP to GDP, and the Ras–
GDP form of the protein is incapable of transmitting a positive signal to downstream kinases (Figure 3-15). Activation
of the Ras pathway is therefore dependent on the ability to
maintain Ras in its GTP-bound state. Modulation between
the active, GTP-bound form and the inactive, GDP-bound
form is brought about by two families of enzymes. Guaninenucleotide Exchange Factors (GEFs) activate Ras by induc-
Guanine nucleotide
Exchange Factors (GEFs)
GTP
GDP
GDP
Inactive
GTP
Pi
Active
GTPase Activating Proteins
(GAPs)
FIGURE 3-15 Small monomeric G proteins, such as Ras,
alter conformation depending on whether they are bound to
GTP or GDP. The GDP-bound form of small G proteins such as Ras
(pale green) is inactive, and the GTP-bound form (dark green) is
active. Small G proteins have intrinsic GTPase activity, which is further enhanced by GTPase Activating Proteins (GAPs). Guanine
nucleotide Exchange Factors (GEFs) act in an opposite manner to
release GDP and promote GTP binding.
ing it to release GDP and accept GTP; GTPase Activating
Proteins (GAPs) inhibit the protein’s activity by stimulating
Ras’s intrinsic ability to hydrolyze bound GTP.
Like the PLC␥ pathway, the Ras pathway is initiated during
both B and T lymphocyte activation and, again, we will use T
cells as our example. DAG, released after PLC␥1-mediated
PIP2 cleavage, binds and activates Ras-GRP, an adapter protein that then recruits the GEF, Son of Sevenless (SOS). SOS
binds to Ras, inducing it to bind GTP, at which point Ras gains
the ability to bind and activate the first member of a cascade
of serine/threonine kinase enzymes that phosphorylate and
activate one another. Because the members of this cascade
were first identified in experiments that studied the activation
of cells by mitogens (agents that induce proliferation), it is
referred to as the Mitogen Activated Protein Kinase or MAP
kinase cascade. The members of the cascade are held in close
proximity to one another by the adapter protein KSR.
In the T-cell activation form of this cascade (illustrated in
Figure 3-16), RasGTP binds to the MAP kinase kinase
kinase (MAPKKK) Raf. Binding of RasGTP alters the conformation of Raf and stimulates its serine/threonine kinase
activity. Raf then phosphorylates and activates the next
enzyme in the relay, a MAP kinase kinase (MAPKK), in this
case MEK. Activated MEK then phosphorylates its substrate,
Extracellular signal-Related Kinase, or Erk, a MAP kinase,
which consequently gains the ability to pass through the
nuclear membrane.
Once inside the nucleus, Erk phosphorylates and activates a transcription factor, Elk-1, which cooperates with a
second protein, serum response factor (SRF), to activate the
transcription of the fos gene. The Fos protein is also phosphorylated by Erk, and along with its partner, Jun, forms the
master transcription factor, AP-1. Jun is phosphorylated and
activated via a slightly different form of the MAP kinase
pathway. AP-1 is another of the transcription factors that
facilitates the transcription of the IL-2 gene.
|
Receptors and Signaling: B and T-Cell Receptors
CHAPTER 3
79
TCR/CD3
TCR-mediated signals
DAG
Pi
GDP
GTP
Ras-GRP
SOS
(inactive)
DAG
PKCθ
GTP
GDP
(active)
Raf (MAPKKK)
P
MAP
kinase
cascade
MEK (MAPKK)
P
TAK1
complex
Bcl10
MALT1
TRAF6
ERK (MAPK)
NEMO Ub
P
IKKα IKKβ P
Cytoplasm
Nucleus
Elk-1 P
NF-κB
SRF
Carma-1
Fos
Ub
NF-κB
pathway
Gene activation
Fos
P
IκB
P
Cytoplasm
Jun
P
NF-κB
P Jun Fos
P
5´
Gene activation
Nucleus
3´
AP-1
FIGURE 3-17 Phosphorylation and ubiquitination acti-
Gene activation
vates IKK, which phosphorylates and inactivates IB, allowing translocation of NF-B into the nucleus.
FIGURE 3-16 The Ras pathway involves a cascade of serine/threonine phosphorylations and culminates in the entry
of the MAP kinase, Erk, into the nucleus where it phosphorylates the transcription factors Elk-1 and Fos.
The Ras pathway is an important component in many
developmental and cell activation programs. Each pathway
uses slightly different combinations of downstream protein
kinases, but all adhere to the same general mode of passing
the signal from the cell surface through to the nucleus via a
cascade of phosphorylation reactions, with the resultant
activation of a new transcriptional program.
PKC Activates the NF-B Transcription Factor
NF-B belongs to a family of heterodimeric transcription
factors, and each dimer activates its own repertoire of promoters. In resting cells, NF-B heterodimers are held in the
cytoplasm by binding to the Inhibitor of NF-B (IB) protein. Cell activation induces the phosphorylation of these
inhibitory proteins by an IB kinase (IKK) complex. The
phosphorylated IB protein is then targeted for proteasomal
degradation, releasing the freed NF-B to enter the nucleus
and bind to the promoters of a whole array of immunologically important genes.
NF-B is important in the transcription control of proteins needed for the proper functioning of many innate and
adaptive immune system cell types; in general, NF-Bmediated transcription is associated with proinflammatory
and activation events, rather than with regulatory processes. Different types of cells and receptors use various
forms of protein kinase activation as well as diverse combinations of adapter molecules in the pathway leading to
NF-B activation. However, all of these pathways culminate in the phosphorylation and subsequent destruction of
the inhibitor of NF-B, IB. Below, we describe the pathway to NF-B activation that is triggered by TCR antigen
recognition.
80
PA R T I
|
Introduction
In T cells, activation of NF-B begins when DAG, generated by PLC␥1, recruits the serine/threonine kinase PKC to
the membrane (Figure 3-17). As mentioned above, PKC is
the form of the serine/threonine kinase that is used in the
T-cell receptor signaling pathway. Once bound to DAG,
PKC is activated and phosphorylates adapter proteins
including Carma1, initiating a cascade that ultimately recruits
the ubiquitin ligase TRAF6. TRAF6 ubiquitinates and partially activates the IKK complex, which is made up of the
regulatory protein NEMO, and two catalytic IKK subunits.
The IKK complex is fully activated only when it is both phosphorylated and ubiquitinated. (Note that this is one of those
instances in which ubiquitination does not result in subsequent protein degradation, but rather in its activation.) Other
T cell mediated signals activate the TAK1 complex, which
phosphorylates IKK, thus completing its activation and
allowing it to phosphorylate IB. In this step, phosphorylation of IB signals its degradation, rather than its activation.
NF-B is then free to move into the nucleus and activate the
transcription of its target genes, including the IL2 gene.
In addition to this T-cell receptor-mediated pathway of
NF-B activation, signaling through CD28, the T-cell coreceptor, also positively controls NF-B activation in T cells.
These three pathways together illustrate how the broad
concepts introduced earlier in this chapter can be applied to
understand the mechanisms underlying T-cell activation.
These themes recur in numerous forms throughout the
immune system.
The Structure of Antibodies
Having outlined some of the general concepts that underlie
many different receptor-ligand interactions and signaling
pathways, we now turn our attention more specifically to the
antigen receptors of the adaptive immune system.
On stimulation with antigen, B cells secrete antibodies
with antigen-binding sites identical to those on the B-cell
membrane antigen receptor. The identity between the binding sites of the secreted antibody and the membrane-bound
B-cell receptor (BCR) was first demonstrated by making
reagents that bound to antibodies secreted by a particular
clone of B cells and showing that those reagents also bound
to the receptors on the cells that had secreted the antibodies.
Since working with soluble proteins is significantly easier
than manipulating membrane receptor proteins, the presence of a soluble form of the receptor greatly facilitated the
characterization of B-cell receptor structure. Consequently,
the basic biochemistry of the B-cell receptor was established
long before that of the T-cell receptor (TCR), which, unlike
the BCR, is not released in a secreted form. Box 3-1: Classic
Experiment details the Nobel Prize–winning experiments
that established the four-chain structure of the antibody
molecule.
We begin this section with a discussion of antibody
structure, and then go on to describe the B-cell membrane
receptor and its associated signaling pathways. Secreted
antibodies and their membrane-bound receptor forms
belong to the immunoglobulin family of proteins. This large
family of proteins, which includes both B- and T-cell receptors, adhesion molecules, some tyrosine kinases, and other
immune receptors, is characterized by the presence of one
or more immunoglobulin domains.
Antibodies Are Made Up of Multiple
Immunoglobulin Domains
The immunoglobulin domain (Figure 3-18) is generated when
a polypeptide chain folds into an organized series of antiparallel -pleated strands. Within each domain, the  strands
are arranged into a pair of  sheets that form a tertiary, compact domain. The number of strands per sheet varies among
individual proteins. In antibody molecules, most immunoglobulin domains contain approximately 110 amino acids,
and each  sheet contains three to five strands. The pair of 
sheets within each domain are stabilized with respect to one
another by an intrachain disulfide bond. Neighboring domains
are connected to one another by a stretch of relatively unstructured polypeptide chain. Within the  strands, hydrophobic
and hydrophilic amino acids alternate and their side chains
are oriented perpendicular to the plane of the sheet. The
hydrophobic amino acids on one sheet are oriented toward
the opposite sheet, and the two sheets within each domain are
therefore stabilized by hydrophobic interactions between the
two sheets as well as by the covalent disulfide bond.
The immunoglobulin fold provides a perfect example of
how structure determines and/or facilitates function. At the
ends of each of the  sheets, more loosely folded polypeptide
regions link one  strand to the next, and these loosely folded
regions can accommodate a variety of amino acid side-chain
lengths and structures without causing any disruption to the
overall structure of the molecule. Hence, in the antibody molecule, the immunoglobulin fold is superbly adapted to provide
a single scaffold onto which multiple different binding sites can be
built, as the antigen-binding sites can simply be built into these
loosely folded regions of the antigen-binding domains. These
properties explain why the immunoglobulin domain has been
used in so many proteins with recognition or adhesive functions. The essential domain structure provides a molecular
backbone, while the loosely folded regions can be adapted to
bind specifically to many adhesive or antigenic structures.
Indeed, the immunoglobulin domain structure is used by
many proteins besides the BCR chains. As we will see later in
this chapter, the T-cell receptor is also made up of repeating
units of the immunoglobulin domain. Other proteins that
use immunoglobulin domains include Fc receptors; the
T-cell receptor accessory proteins CD2, CD4, CD8, and
CD28; the receptor-associated proteins of both the TCR and
the BCR; adhesion molecules; and others. Some of these
immunoglobulin domain-containing proteins are illustrated
in Figure 3-19. Each of these proteins is classified as a member
Receptors and Signaling: B and T-Cell Receptors
(a)
CL domain
|
CHAPTER 3
81
VL domain
Loops
β strands
COOH
NH2
Disulfide bond
CDRs
β-strand arrangement
(b)
COOH
NH2
COOH
NH2
CDRs
FIGURE 3-18 Diagram of the immunoglobulin fold structure of the antibody light chain variable (VL) and constant
(CL) region domains. (a) The two  pleated sheets in each domain
are held together by hydrophobic interactions and the conserved
disulfide bond. The  strands that compose each sheet are shown
in different colors. The three loops of each variable domain show
considerable variation in length and amino acid sequence; these
hypervariable regions (blue) make up the antigen-binding site.
of the immunoglobulin superfamily, a term that is used to
denote proteins derived from a common primordial gene
encoding the basic domain structure.
Antibodies Share a Common Structure of Two
Light Chains and Two Heavy Chains
All antibodies share a common structure of four polypeptide
chains (Figure 3-20), consisting of two identical light (L)
chains and two identical heavy (H) chains. Each light chain
is bound to its partner heavy chain by a disulfide bond
between corresponding cysteine residues, as well as by noncovalent interactions between the VH and VL domains and
the CH1 and CL domains. These bonds enable the formation
Hypervariable regions are usually called complementarity-determining
regions (CDRs). Heavy-chain domains have the same characteristic
structure. (b) The -pleated sheets are opened out to reveal the relationship of the individual  strands and joining loops. Note that the
variable domain contains two more  strands than the constant
domain. [Part (a) adapted from M. Schiffer et al., 1973, Biochemistry 12: 4620;
reprinted with permission; part (b) adapted from A. F. Williams and A. N. Barclay,
1988, Annual Review of Immunology 6:381.]
of a closely associated heterodimer (H-L). Multiple disulfide
bridges link the two heavy chains together about halfway
down their length, and the C-terminal parts of the two heavy
chains also participate in noncovalent bonding interactions
between corresponding domains.
As shown in Figure 3-21, the antibody molecule forms a
Y shape with two identical antigen-binding regions at the
tips of the Y. Each antigen-binding region is made up of
amino acids derived from both the heavy- and the lightchain amino-terminal domains. The heavy and light chains
both contribute two domains to each arm of the Y, with the
non–antigen-binding domain of each chain serving to
extend the antigen-binding arm. The base of the Y consists
of the C-terminal domains of the antibody heavy chain.
CLASSIC EXPERIMENT
The Elucidation of Antibody Structure
FIRST SET OF EXPERIMENTS
Arne Tiselius, Pers Pederson, Michael
Heidelberger, and Elvin Kabat
Since the late nineteenth century, it has
been known that antibodies reside in the
blood serum—that is, in that component of
the blood which remains once cells and
clotting proteins have been removed. However, the chemical nature of those antibodies remained a mystery until the experiments
of Tiselius and Pederson of Sweden and
Heidelberger and Kabat, in the United
States, published in 1939. They made use of
the fact that, when antibodies react with a
multivalent protein antigen, they form a
multimolecular cross-linked complex that
falls out of solution. This process is known as
immunoprecipitation (see Chapter 20 for
modern uses of this technique). They immunized rabbits with the protein ovalbumin
(the major component of egg whites), bled
the rabbits to obtain an anti-ovalbumin
antiserum, and then divided their antiserum
into two aliquots. They subjected the first
aliquot to electrophoresis, measuring the
amount of protein that moved different
distances from the origin in an electric field.
The blue plot in Figure 1 depicts the four
major protein subpopulations resolved by
their technique. The first, and largest, is the
albumin peak, the most abundant protein in
serum, with responsibility for transporting
lipids through the blood. They named the
other peaks globulins. Two smaller peaks are
denoted the ␣ and  globulin peaks, and
then a third globulin peak, ␥ globulin, clearly
represents a set of proteins in high concentration in the serum.
However, the most notable part of the
experiment occurred when the investigators mixed their second serum aliquot with
ovalbumin, the antigen. The antibodies in
the serum bound to ovalbumin in a multivalent complex, which fell out of solution
into a precipitate. The precipitate was then
82
removed by centrifugation. Now that they
had succeeded in removing the antibodies
from an antiserum, the question was,
which protein peak would be affected?
The black plot in Figure 1 illustrates
their results. Very little protein was lost from
the albumin, or the ␣ or  globulin peaks.
However, immunoprecipitation resulted in
a dramatic decrease in the size of the ␥
globulin peak, demonstrating that the
majority of their anti-ovalbumin antibodies
could be classified as ␥ globulins.
We now know that most antibodies of
the IgG class are indeed found in the ␥
globulin class. However, antibodies of
other classes are found in the ␣ and 
globulin peaks, which may account for
the slight decrease in protein concentration found after immunoprecipitation in
these other protein peaks.
SECOND SET OF EXPERIMENTS
Sir Rodney Porter, Gerald Edelman,
and Alfred Nisonoff
Knowing the class of serum protein into
which antibodies fall was a start, but immunochemists next needed to figure out
what antibodies looked like. The fact that
they could form precipitable multivalent
complexes suggested that each antibody
was capable of binding to more than one
site on a multivalent antigen. But the scientists still did not know how many polypeptide chains made up an antibody molecule
and how many antigen-binding sites were
present in each molecule.
Two lines of experimentation conducted in a similar time frame on both
sides of the Atlantic combined to provide
the answers to these two questions. Ultracentrifugation experiments had placed the
molecular weight of IgG antibody molecules at approximately 150,000 Daltons.
Digestion of IgG with the enzyme papain
produced three fragments, two of which
were identical and a third which was clearly
different (Figure 2). The third fragment, of
approximately 50,000 Daltons, spontaneously formed crystals and was therefore
named Fragment crystallizable, or Fc. By
demonstrating that they could competitively inhibit the binding of antibodies to
their antigen, the other two fragments
were shown to retain the antigen-binding
+
−
Albumin
Globulins
γ
Absorbance
The experiments that first identified antibodies as serum immunoglobulins and
then characterized their familiar four-chain
structure represent some of the most elegant adaptations of protein chemistry ever
used to solve a biological problem.
α
β
Migration distance
FIGURE 1
Experimental demonstration that most antibodies are in the ␥-globulin fraction
of serum proteins. After rabbits were immunized with ovalbumin (OVA), their antisera were
pooled and electrophoresed, which separated the serum proteins according to their electric charge
and mass. The blue line shows the electrophoretic pattern of untreated antiserum. The black line shows
the pattern of antiserum that was first incubated with OVA to remove anti-OVA antibody and then
subjected to electrophoresis. [Adapted from A. Tiselius and E. A. Kabat, 1939, Journal of Experimental Medicine
69:119, with copyright permission of the Rockefeller University Press.]
BOX 3-1
capacity of the original antibody. These
fragments were therefore named Fab, or
Fragment antigen binding. This experiment indicated that a single antibody molecule contained two antigen binding sites
and a third part of the molecule that did
not participate in the binding reaction.
Use of another proteolytic enzyme, pepsin, resulted in the formation of a single
fragment of 100,000 Daltons, which contained two antigen-binding sites that were
still held together in a bivalent molecule.
Because the molecule acted as though it
contained two Fab fragments, but clearly
had an additional component that facilitated the combination of the two fragments
into one molecule, it was named F(ab’)2.
Pepsin digestion does not result in a recoverable Fc fragment, as it apparently digests it
into small fragments. However, F(ab’)2 derivatives of antibodies are often used in experiments in which scientists wish to avoid
artifacts resulting from binding of antibodies to Fc receptors on cell surfaces.
In the second set of experiments, investigators reduced the whole IgG molecule
using -mercaptoethanol, in order to break
disulfide bonds, and alkylated the reduced
product so that the disulfide bonds could
not spontaneously reform. They then used
a technique called gel filtration to separate
and measure the size of the protein fragments generated by this reduction and
alkylation. (Nowadays, we would use SDS
PAGE gels to do this experiment.) In this
way, it was shown that each IgG molecule
contained two heavy chains, MW 50,000
and two light chains, MW 22,000.
Now the challenge was to combine
the results of these experiments to create a consistent model of the antibody
molecule. To do this, the scientists had to
determine which of the chains was implicated in antigen binding, and which
chains contributed to the crystallizable
fragments. Immunologists often use
immunological means to answer their
questions, and this was no exception.
They elected to use Fab and Fc fragments purified from rabbit IgG antibodies to immunize two separate goats.
From these goats, they generated antiFab and anti-Fc antibodies, which they
reacted, in separate experiments, with
the heavy and light chains from the
reduction and alkylation experiments.
The answer was immediately clear.
Anti-Fab antibodies bound to both
heavy and light chains, and therefore the
antigen binding site of the original rabbit
IgG was made up of both heavy and light
chain components. However, anti-Fc antibodies bound only to the heavy chains,
but not to the light chains of the IgG molecule, demonstrating that the Fc part of
the molecule was made up of heavy
chains only. Finally, careful protein chemistry demonstrated that the amino-termini
of the two chains resided in the Fab portion of the molecule. In this way, the familiar four-chain structure, with the binding
sites at the amino-termini of the heavy
and light chain pairs, was deduced from
some classically elegant experiments.
In 1972, Sir Rodney Porter and Gerald
Edelman were awarded the Nobel Prize
in Physiology and Medicine for their
work in discovering the structure of
immunoglobulins.
Disulfide
bonds
SS
SS
SS
Edelman, G. M., B. A. Cunningham, W. E. Gall,
P. D. Gottlieb, U. Rutishauser, and M. J.
Waxdal. 1969. The covalent structure of an
entire gammaG immunoglobulin molecule. Proceedings of the National Academy of
Sciences of the United States of America
63:78–85.
Fleishman, J, B., R. H. Pain, and R. R. Porter.
1962 Sep. Reduction of gamma-globulins.
Archives of Biochemistry and Biophysics
Suppl 1:174–80.
Heidelberger, M., andK. O. Pedersen. 1937. The
molecular weight of antibodies. The Journal
of Experimental Medicine 65:393–414.
Nisonoff, A., F. C. Wissler, and L. N. Lipman. 1960.
Properties of the major component of a
peptic digest of rabbit antibody. Science
132:1770–1771.
Porter, R. R. (1972). Lecture for the Nobel Prize
for physiology or medicine 1972: Structural
studies of immunoglobulins. Scandinavian
Journal of Immunology 34(4):381–389.
Tiselius, A., and E. A. Kabat. 1939. An electrophoretic study of immune sera and purified
antibody preparations. The Journal of Experimental Medicine 69:119–131.
L chain
SS
F(ab')2
Pepsin
digestion
H chain
SS
SS
SS
SS
+
Fc fragments
Mercaptoethanol
reduction
Papain digestion
Fab
Fab
+
SS
SS
SS
SS
HS
HS
SH
SH
+
+
+
SH
SH
SH L chains
SH
Fc
H chains
FIGURE 2
Prototype structure of IgG, showing chain structure and interchain disulfide
bonds. The fragments produced by enzymatic digestion with pepsin or papain or by cleavage of the
disulfide bonds with mercaptoethanol are indicated. Light (L) chains are in light blue, and heavy (H)
chains are in dark blue.
83
PA R T I
84
|
Introduction
Immunoglobulin (IgM)
V
S
S
V
S
S
S
S
S
S
S S
S S
S
C
S
S
S
S
C
S
S
Ig-α/ Ig-β
heterodimer
V
S
C
S
S
C
S
C
S
V
C
MHC molecules
T-cell receptor
C
S
S
S
S
C
C
S
S
S
S
C
V
C
S
S
S
S
C
C
CHO
S
S
S
S
Class I
Class II
α
β
V
S
S
S
S
S
S
C
S
S
C
S S
S
S
S
S
C
SS
β2
microglobulin
C
S
S
S
S
C
Adhesion molecules
VCAM-1
S
S
C
Poly- Ig
receptor
S
S
C
T-cell accessory proteins
ICAM-1
V
S
S
V
C
S
S
C
S
S
γ
V
S
S
V
S
S
CD3
CD2
δ
ε
S
S
C
S
S
S
S
C
C
V
CD8
V
V
C
S
S
S
S
C
CD4
S
S
S
S
V
C
S
S
S
S
S
S
S
S
C
S
S
C
S
S
C
CD64
Fcγ RI
S
S
C
S
S
ICAM-2
C
S
S
C
S
S
C
S
S
C
C
S
S
C
S
S
C
S
S
C
C
S
S
LFA–3
S
S
C
S
S
C
S
S
C
S
S
γ γ
S S
FIGURE 3-19 Some examples of proteins bearing immunoglobulin domains. Each immunoglobulin domain is depicted by a blue
loop and, where relevant, is labeled as variable (V) or constant (C).
Figure 3-21 further shows that the overall structure of the
antibody molecule consists of three relatively compact regions,
joined by a more flexible hinge region. The hinge region is particularly susceptible to proteolytic cleavage by the enzyme
papain. Papain cleavage resolves the antibody molecule into
two identical fragments that retain the antigen-binding specificity of the original antibody (shown as Fab regions in Figure
3-21), and the remaining region of the molecule, which consists
of the non antigen-binding portion. This latter region, which is
identical for all antibodies of a given class, crystallizes easily and
was thus called the Fc region (fragment crystallizable).
Each Fab region and Fc region of antibodies mediates its
own particular functions during an antibody response to
an antigen. The Fab regions bind to the antigen, and the Fc
region of the antigen-coupled antibody binds to Fc receptors on phagocytic or cytolytic cells, or to immune effector
molecules. In this way, antibodies serve as physiological
bridges between an antigen present on a pathogen, and
the cells or molecules that will ultimately destroy it. A family of Fc receptors exists; each Fc receptor is expressed on
a different array of cells and binds to a different class of
antibodies.
Receptors and Signaling: B and T-Cell Receptors
S
H
3+
L
V
H1
CH2
CH2
S
S
O–
CHO
S S
S S
CH3
S S
S
CO
Antigen
binding
Effector
activity
CH3
S
CHO
S S
S
L
S
S S
S S
C
S
C
S
1
CH
CL
S
–
O
Fab
S
S
S
S
CO
85
Light
chains
NH
V
S
S
VL
S
4
21
Fab
3+
Hinge
S
CHAPTER 3
S
+
3
VH
NH
NH
+
3
S
NH
|
Antigenbinding site
Antigenbinding site
Fc
446
COO–
Heavy chains
COO–
FIGURE 3-20 Schematic diagram of the structure of immunoglobulins derived from amino acid sequence analysis.
Each heavy (dark blue) and light (light blue) chain in an immunoglobulin molecule contains an amino-terminal variable (V) region
that consists of 100 to 110 amino acids and differs from one antibody to the next. The remainder of each chain in the molecule—
the constant (C) regions—exhibits limited variation that defines the
two light-chain subtypes and the five heavy-chain subclasses.
Some heavy chains (␥, ␦, and ␣) also contain a proline-rich hinge
region. The amino-terminal portions, corresponding to the V
regions, bind to antigen; effector functions are mediated by the carboxy-terminal domains. The and ⑀ heavy chains, which lack a hinge
region, contain an additional domain in the middle of the molecule.
CHO denotes a carbohydrate group linked to the heavy chain.
There are Two Major Classes of
Antibody Light Chains
Amino acid sequencing of antibody light chains revealed
that the amino-terminal half (approximately 110 amino
acids) of the light chain was extremely variable, whereas the
sequence of the carboxyl-terminal half could be classified
into one of two major sequence types. The N-terminal half
of light chains is thus referred to as the variable, or VL,
region of the light chain, and the less variable part of the
sequence is termed the constant, or CL, region. The two
major light chain constant region sequences are referred to
as (kappa) or (lambda) chains. As more light-chain
sequences were generated, it became apparent that the
chain constant region sequences could be further subdivided
into four subtypes—1, 2, 3, and 4—based on amino
acid substitutions at a few positions. In humans, the light
chains are fairly evenly divided between the two light-chain
classes; 60% of human light chains are whereas only 40%
are . In mice, the situation is quite different: Only 5% of
mouse light chains are of the light-chain type. All light
chains have a molecular weight of approximately 22 kDa.
Further analysis of light-chain sequences demonstrated
that, even within the variable regions of the light chain, there
Hinge region
= open region;
site of segmental
flexibility
FIGURE 3-21 The three-dimensional structure of IgG.
Clearly visible in this representation are the 110 amino acid immunoglobulin domains of which the molecule is constructed, along with
the open hinge structure in the center of the molecule, which affords
flexibility in binding to multivalent antigens. In IgG, the light chain
contains two immunoglobulin domains, the heavy chain four. Fab
Antigen-binding portion of the antibody; contains paired VL/VH and
CL/CH1 domains. Fc Non–antigen-binding region of the antibody,
with paired CH2/CH2 and CH3/CH3 domains. [Source: D. Sadava, H. C.
Heller, G. H. Orians, W. K. Purves, and D. M. Hillis. Life: The Science of Biology,
7e (© 2004 Sinauer Associates, Inc., and W. H. Freeman & Co.), Figure 18-10
with modifications.]
were regions of hypervariability. Since these hypervariable
regions could be shown to interact with the bound antigen,
they were renamed the complementarity-determining
regions, or CDRs.
Those readers with expertise in genetics will immediately
identify a problem: How is it possible to encode a single
protein chain with some regions that are extremely diverse
and other regions that are relatively constant, while maintaining that distinction across millions of years of evolutionary drift? The solution to this puzzle involves the encoding
of a single antibody variable region in multiple genetic segments that are then joined together in different combinations in each individual antibody-producing cell. This
unique mechanism will be fully described in Chapter 7.
There are Five Major Classes of
Antibody Heavy Chains
Using antibodies directed toward the constant region of immunoglobulins and amino acid sequencing of immunoglobulins
derived from plasmacytoma tumor cells, investigators discovered that the sequences of the heavy-chain constant regions fall
into five basic patterns. These five basic sequences have been
named with Greek letters: (mu), ␦ (delta), ␥ (gamma), ⑀
(epsilon), and ␣ (alpha). Each different heavy-chain constant
region is referred to as an isotype, and the isotype of the heavy
86
PA R T I
TABLE 3-2
|
Introduction
Chain composition of the five immunoglobulin classes
Heavy
chain
Number of CH
Ig domains
IgG
␥
3
␥1, ␥2, ␥3, ␥4 (human)
␥1, ␥2a, ␥2b, ␥3 (mouse)
or
None
␥22
␥22
IgM
4
None
or
Yes
(22)n
(22)n
n 1 or 5
IgA
␣
3
␣1, ␣2
or
Yes
(␣22)n
(␣22)n
n 1, 2, 3, or 4
IgE
4
None
or
None
⑀22
⑀22
IgD
␦
3
None
or
None
␦22
␦22
Class*
Subclasses
Light chain
J chain
Molecular
formula
*
See Figure 3-22 for general structures of the five antibody classes.
chains of a given antibody molecule determines its class. Thus,
antibodies with a heavy chain of the isotype are of the IgM
class; those with a heavy chain are IgD; those with ␥, IgG;
those with ⑀, IgE; and those with ␣, IgA. The length of the
constant region of the heavy chains is either 330 amino acid
residues (for ␥, ␦, and ␣ chains) or 440 amino acids (for and
chains). Correspondingly, the molecular weights of the heavy
chains vary according to their class. IgA, IgD, and IgG heavy
chains weigh approximately 55 kDa, whereas IgM and IgE
antibodies are approximately 20% heavier.
Minor differences in the amino acid sequences of groups
of ␣ and ␥ heavy chains led to further subclassification of
these heavy chains into sub-isotypes and their corresponding antibodies therefore into subclasses (Table 3-2). There
are two sub-isotypes of the ␣ heavy chain, ␣1 and ␣2, and
thus two IgA subclasses, IgA1 and IgA2. Similarly, there are
four sub-isotypes of ␥ heavy chains, ␥1, ␥2, ␥3, and ␥4, with
the corresponding formation of the four subclasses of IgG:
IgG1, IgG2, IgG3, and IgG4. In mice, the four ␥ chain subisotypes are ␥1, ␥2␣, ␥2, and ␥3, and the corresponding
subclasses of mouse IgG antibodies are IgG1, IgG2a, IgG2b,
and IgG3, respectively.
Further analysis revealed that the exact number, and precise positions of the disulfide bonds between the heavy
chains of antibodies, vary among antibodies of different
classes and subclasses (Figure 3-22).
Antibodies and Antibody Fragments
Can Serve as Antigens
The essential principles of antibody structure were established prior to the development of the technology required
to artificially generate monoclonal antibodies, and indeed
much of the basic work of structure determination was
completed before techniques were available for rapid DNA
sequencing. As a source of homogenous antibodies, immunologists therefore turned to the protein products of antibody-secreting tumors. Plasmacytomas are tumors of
plasma cells, the end-stage cell of B-cell differentiation, and
the cells from that tumor are normally located in the bone
marrow. When a single clone of plasma cells becomes cancerous, it is called a plasmacytoma for as long as it remains
in a single bone. However, once it metastasizes into multiple bone marrow sites, the tumor is referred to as multiple
myeloma. Plasmacytoma or myeloma tumors secrete large
amounts of monoclonal antibodies into the serum and tissue fluids of the patients, and these antibodies can be purified in large quantities. Rather than secreting the whole
antibody, some of these tumors will secrete only the light
chains, or sometimes both the light chains and the whole
antibodies, into the serum. The homogenous light chains
secreted by these myeloma tumors are referred to as
Bence-Jones proteins.
In the middle to late twentieth century, tumor-derived
antibodies were used to generate a great deal of information
regarding antibody structure and sequence. When a means
was developed by which to generate these tumors artificially
in mice, data from human tumors were supplemented with
information derived from laboratory-generated murine cell
lines, many of which are still in use to this day.
Both tumor-derived and purified serum antibodies
were also used as antigens, and the anti-immunoglobulin
antibodies so derived proved to be extraordinarily useful
in the elucidation of antibody structure. Antibodies, or
antibody fragments derived from one animal species (for
example, a rabbit), can be injected into a second species
|
Receptors and Signaling: B and T-Cell Receptors
(a) IgG
(b) IgD
VH
VL
Cγ1
CL
(c) IgE
VH
VL
CL
CHAPTER 3
VH
VL
Cδ1
87
CL
Cε1
Cε2
Hinge
region
Cγ2
Cε3
Cδ2
Cγ3
Cε4
Cδ3
(d) IgA (dimer)
(e) IgM (pentamer)
VH
Cα1
VH
VL
VL
Cµ1
CL
CL
Cα2
Cµ2
Disulfide
bond
Hinge
region
Cα3
J chain
Cµ3
Cµ4
J chain
FIGURE 3-22 General structures of the five major classes
of antibodies. Light chains are shown in lighter shades, disulfide
bonds are indicated by thick black lines. Note that the IgG, IgA, and
IgD heavy chains contain four domains and a hinge region, whereas
the IgM and IgE heavy chains contain five domains but no hinge
region. The polymeric forms of IgM and IgA contain a polypeptide,
called the J chain, that is linked by two disulfide bonds to the Fc
region in two different monomers. Serum IgM is always a pentamer;
most serum IgA exists as a monomer, although dimers, trimers, and
even tetramers are sometimes present. Not shown in these figures
are intrachain disulfide bonds and disulfide bonds linking light and
heavy chains (see Figure 3-23).
(for example, a goat), or even into another animal of the
same species, in which the antibody or antibody fragment
will serve as an antigen. Box 3-1 describes the series of
experiments in which antibodies made against isolated
proteolytic fragments of immunoglobulin molecules were
used to help determine immunoglobulin structure. To
understand these experiments, and many others that also
use antibodies directed against whole immunoglobulins or
against immunoglobulin fragments, it is important to have
a good grasp of the nature of the antigenic determinants
on immunoglobulin molecules. An antigenic determinant is defined as a region of an antigen that makes con-
tact with the antigen-combining region on an antibody.
The different types of antibodies that recognize immunoglobulin determinants are described in Table 3-3.
Anti-isotype antibodies are directed against antigenic
determinants present on the constant region of one particular heavy- or light-chain class of antibody, but not on
any of the others. For example, an anti-isotype antibody
may bind only to human heavy chains, but not to human
, ␥, ␣, or ⑀ constant regions. Alternatively, it may bind to
but not to light chains. Thus, an anti-isotype antibody
binds only to a single antibody constant region class or
subclass. Anti-isotype antibodies are usually specific for
88
TABLE 3-3
PA R T I
|
Introduction
Antibodies that recognize other antibodies
Type of Antibody
What it recognizes
Anti-Fab (Fragment, antigen-binding)
Papain-generated fragment of antibodies consisting of the VHCH1–VLCL domains
Anti-Fc (Fragment, crystallizable)
Papain-generated fragment of antibodies consisting of the paired CH2CH3 (and, for IgM
and IgE, CH4) domains
Anti-isotype
Antigenic determinants specific to each heavy chain class
Anti-allotype
Antigenic determinants that are allele specific—small differences in the constant region
of light and heavy chains that vary among individuals
Anti-idiotype
Antigenic determinants characteristic of a particular antigen combining site. Each antibody will have its own characteristic idiotypic determinants made up of residues from
the heavy and light chains that contribute to the antigen-binding regions.
determinants present on the constant region of antibodies from a particular species of animal, although some
cross-reactivity among related species—for example, mice
and rats—may occur.
Some heavy- or light-chain genes occur in multiple allelic
forms, and alternative allelic forms of the same isotype are
referred to as allotypes. Generally, two antibodies with allotypic differences will vary in just a few residues of one of the
immunoglobulin chains, and these residues constitute the
allotypic determinants. Anti-allotype antibodies are
generated by immunizing an individual of one species with
antibodies derived from a second animal of the same species
bearing an alternative form (allele) of the particular immunoglobulin gene. After an immune response has occurred,
the investigator purifies or selects those antibodies that recognize the allotypic determinants from the immunized individual. Anti-allotype antibodies are used, for example, to
trace particular cell populations in experiments in which B
cells from a strain of animals bearing a particular allotype
are transferred into a second strain that differs in immunoglobulin allotype.
Finally, antibodies directed against the antigen-binding
site of a particular antibody are referred to as anti-idiotypic
antibodies. They can be generated by immunizing an animal
with large quantities of a purified monoclonal antibody,
which, by definition, bears a single antigen-binding site.
Those antibodies from the immunized animal that recognize
the antigen-binding site of the original antibody can then be
purified. Anti-idiotypic antibodies were used in the initial
experiments that proved that the B-cell receptor had the
same antigen-binding site as the antibody secreted by that B
cell, and they are now used to follow the fate of B cells bearing a single receptor specificity in immuno-localization
experiments.
As described above, treatment of whole antibody molecules with the enzyme papain cleaves the antibody molecule at the hinge region and releases two antigen-binding
Fab fragments and one Fc fragment per antibody molecule.
Anti-Fab antibodies and anti-Fc antibodies are made by
immunizing a different species of animal from that which
provided the antibody fragments with Fab or Fc fragments,
respectively.
Each of the Domains of the Antibody Heavy
and Light Chains Mediate Specific Functions
Antibodies protect the host against infection, by binding to
pathogens and facilitating their elimination. Antibodies of
different heavy-chain classes are specialized to mediate particular protective functions, such as complement activation
(Chapter 6), pathogen agglutination, or phagocytosis (Chapter 13), and each different domain of the antibody molecule
plays its own part in these host defense mechanisms. Here we
briefly examine the structure of each of the antibody heavychain domains in turn, beginning with the CH1 and CL
domains, which are those proximal to the VH and VL domains
respectively (see Figure 3-20).
CH1 and CL Domains
As discussed above, the strength (avidity) of receptor binding to antigen is greatly enhanced by receptor multivalency.
Antibodies have evolved to take advantage of this property
by employing two antigen-binding sites, each of which can
bind to individual determinants on multivalent antigens,
such as are found on bacterial surfaces. The CH1 and CL
domains serve to extend the antigen-binding arms of the
antibody molecule, facilitating interactions with multivalent antigens and maximizing the ability of the antibody to
bind to more than one site on a multivalent antigen. An
interchain disulfide bond between these two domains
holds them together and may facilitate the ability of some
VH-VL pairs to hold on to one another and form a stable
combining site.
Receptors and Signaling: B and T-Cell Receptors
IgG1
IgG2
IgG3
|
CHAPTER 3
89
IgG4
Disulfide
bond
FIGURE 3-23 General structure of the four subclasses of human IgG, which differ in the number and arrangement of the interchain disulfide bonds (thick black lines) linking the heavy chains. A notable feature of human IgG3 is its 11 interchain disulfide bonds.
The Hinge Regions
The ␥, ␦, and ␣ heavy chains contain an extended peptide
sequence between the CH1 and CH2 domains that has no
homology with the other domains (see Figures 3-20 and
3-22). This so-called hinge region is rich in proline residues,
rendering it particularly flexible, and as a consequence, the
two antigen-binding arms of IgG, IgD, and IgA antibodies
can assume a wide variety of angles with respect to one
another, which facilitates efficient antigen binding. The
extended nature of the amino acid chain in the hinge region
contributes to the vulnerability of this part of the molecule
to protease cleavage, a vulnerability that was ingeniously
exploited in the early experiments that characterized antibody structure (see Box 3-1).
In addition to these proline residues, the hinge region
also contains a number of cysteines, which participate in
heavy-chain dimerization. The actual number of interchain
disulfide bonds in the hinge region varies considerably
among different heavy-chain classes and subclasses of antibodies (Figure 3-23) as well as between species. Lacking a
hinge region, the heavy chains of IgE make their inter-heavy
chain disulfide bonds between the CH1 and CH3 domains. In
IgM, disulfide bonds bridge the pairs of heavy chains at the
level of CH3 and CH4. Although and chains have no
hinge regions, they do have an additional immunoglobulin
domain that retains some hingelike qualities.
Carbohydrate Chains
The two CH2 domains of ␣, ␦, and ␥ chains and the two CH3
domains of and chains are separated from their partner
heavy-chain domains by oligosaccharide side chains that prevent the two heavy chains from nestling close to one another
(Figure 3-24). As a result, the paired domains are significantly
more accessible to the aqueous environment than other constant region domains. This accessibility is thought to contribute to the ability of IgM and IgG antibodies to bind to
complement components. Immunoglobulins are in general
quite extensively glycosylated, and some antibodies even have
carbohydrates attached to their light chains.
FIGURE 3-24 Carbohydrate residues, shown in pink,
prevent close contact between the CH2 domains of human
IgG1. [PDB ID 1IGT.]
The Carboxy-Terminal Domains
The five classes of antibodies can be expressed as either
membrane or secreted immunoglobulin. Secreted antibodies
90
PA R T I
|
Introduction
Primary transcript
AAA
Alternative
splicing
Processed transcripts
AAA
Secreted form
Hydrophilic
segment
AAA
Membrane-bound form
lgα/lgβ
Membrane-bound segments
FIGURE 3-25 Membrane vs. secreted forms of immunoglobulin are created by alternative mRNA splicing. The dark
blue boxes correspond to the mRNA transcript sections of the rearranged variable region sequences, and the light-blue boxes correspond to the parts of the mRNA transcript corresponding to the
individual constant region domains of the heavy-chain genes. The
IgM gene, shown here, contains one variable region and four constant region exons. The pink segment represents the transcript
encoding the C-terminal portion of the secreted immunoglobulin.
The green segments represent the transcript encoding the C-terminal portions of the membrane-bound immunoglobulin receptor,
including the transmembrane and cytoplasmic regions. Alternative
splicing creates the two different types of immunoglobulin molecules. Membrane-bound and secreted immunoglobulins on any
one B cell share the identical antigen-binding regions and most of
the heavy-chain sequences.
have a hydrophilic amino acid sequence of various lengths at
the carboxyl terminus of the final CH domain. In membranebound immunoglobulin receptors, this hydrophilic region is
replaced by three regions (Figure 3-25):
X-ray Crystallography Has Been Used to
Define the Structural Basis of AntigenAntibody Binding
• An extracellular, hydrophilic “spacer” sequence of
approximately 26 amino acids
• A hydrophobic transmembrane segment of about 25
amino acids
• A very short, approximately three amino acid,
cytoplasmic tail
Crystallography has been used to explore the nature of
antigen-antibody binding for a large number of antibodies
and has demonstrated that either or both chains might provide the majority of contact residues with any one antigen.
Thus, some antibodies bind to the antigen mainly via contacts with heavy-chain variable region residues, with the
light chain merely providing structural support to the binding site; for other antigen-antibody interactions, the opposite is true. Still other antibodies use residues from both
chains to directly contact the antigen. In most cases, though,
contacts between antigen and antibody occur over a broad
face and protrusions or depressions in the antibody surface
are likely to be matched by complementary depressions or
protrusions in the antigen. Figure 3-26 illustrates the binding of an antibody to the tip of the influenza hemagglutinin
molecule. In another well-studied case of an antibody binding to the protein lysozyme, the surface area of interaction
was shown to be quite large, ranging in different antibodylysozyme pairs from 650 to 900 square Angstroms. Given the
tight binding between an antibody and its complementary
antigen, it should not be surprising that, at least in some
cases, binding of antigen to antibody induces a conformational change in the antibody, which can be visualized by
x-ray crystallography (Figure 3-27).
B cells express different classes of membrane immunoglobulin at particular developmental stages and under different stimulatory conditions. Immature, pre-B cells express
only membrane IgM. Membrane IgD co-expression along
with IgM is one of the markers of differentiation to a fully
mature B cell that has yet to encounter antigen. Following
antigen stimulation, IgD is lost from the cell surface, and the
constant region of the membrane and secreted immunoglobulin can switch to any one of the other isotypes. The
antibody class secreted by antigen-stimulated B cells is
determined by cytokines released by T cells and antigen
presenting cells in the vicinity of the activated B cell. Antibodies of different heavy-chain classes have selective affinities for particular cell surface Fc receptors, as well as for
components of the complement system. The effector functions of particular antibody classes are further discussed in
Chapters 6 and 13.
Receptors and Signaling: B and T-Cell Receptors
(a)
|
CHAPTER 3
91
(b)
Antigen
Antibody
FIGURE 3-26 Computer simulation of an interaction between antibody and influenza virus antigen. (a) The antigen (yellow) is
shown interacting with the antibody molecule; the variable region of the heavy chain is red, and the variable region of the light chain is blue.
(b) The complementarity of the two molecules is revealed by separating the antigen from the antibody by 8 Å. [Based on x-ray crystallography data
collected by P. M. Colman and W. R. Tulip. From G.J.V.H. Nossal, 1993. Scientific American 269(3):22.]
H3
L1
H2
H1
L2
L3
FIGURE 3-27 Conformational change can occur on binding of antigen to antibody. This figure shows a complex between
a peptide derived from HIV protease and an Fab fragment from an
anti-protease antibody. The peptide is shown in black. The red line
shows the structure of the Fab fragment before it binds the peptide,
and the blue line shows its structure after binding. There are significant conformational changes in the CDRs of the Fab on binding to
the antigen. These are especially pronounced in the light chain CDR1
(L1) and the heavy chain CDR3 (H3). [From J. Lescar et al., 1997, Journal of
Molecular Biology 267:1207; courtesy of G. Bentley, Institute Pasteur.]
Signal Transduction in B Cells
Having described the structure of antibody molecules and their
membrane-bound form, the B-cell receptor (BCR), we now
turn our attention to BCR function. Recall that the structure of
the BCR is identical to that of the antibodies that the cell will
secrete on antigen stimulation, with the exception of the
C-terminal portion of the heavy chain; this portion is modified
so as to anchor the receptor into the B-cell plasma membrane.
Prior to antigen recognition, mature B cells residing in the
secondary lymphoid tissues, such as the spleen or lymph
nodes, express membrane-bound forms of both IgM and IgD.
As described above, the cytoplasmic tail of the BCR heavy
chain is extremely short—only three amino acids—and so one
of the puzzles that had to be solved by those exploring BCRmediated antigen activation was how such a short cytoplasmic
tail could efficiently pass a signal into the cytoplasm. This
problem was solved when co-immunoprecipitation experiments revealed that each BCR molecule was noncovalently
associated with a heterodimer, Ig␣/Ig (see Figure 3-7a), that
is responsible for transducing the antigen signal into the interior of the cell. Recall that Ig␣/Ig chains contain ITAMs,
which include tyrosine residues that become phosphorylated
on activation through the receptor, and serve as docking residues for downstream signaling components. The BCR is
therefore structurally and functionally divided into two components: a recognition component (the immunoglobulin
receptor) and a signal transduction component (Ig␣/Ig).
In previous sections of this chapter, we described some
general principles of signal transduction and outlined three
commonly employed signal transduction pathways (see Overview Figure 3-5). In this section, we show how those principles
apply to the intracellular events that follow upon antigenreceptor interactions in B cells.
Antigen Binding Results in Docking of Adapter
Molecules and Enzymes into the BCR-Ig␣/Ig
Membrane Complex
Antigen binding induces conformational alterations in the
BCR, which expose regions in the C4 domains of the
92
PA R T I
|
Introduction
CD21
CD19
PIP3
Lyn
PIP3
P
P
P
P
P
P
P
P
P Syk
P
Akt PDK1
BCAP
PI3K
lgα/lgβ
P
PIP3
P
Lyn
PLCγ2
Btk
P
P
BLNK
VAV
P
P
Grb-2
Phosphorylation
and inactivation
Rac/Rho/
cdc42
SOS
Ca2+
P PI3K
Cytoskeletal
reorganization
AntiBax Bad apoptotic
Ras
PKC
Calmodulin
Cell survival
MAP kinase
cascade
Calcineurin
Cytoplasm
NFAT
NF-κB
AP-1
Nucleus
Gene activation
FIGURE 3-28 Signal transduction pathways emanating
from the BCR. Antigen-mediated receptor clustering into the
lipid raft regions of the membrane leads to src-family kinase
phosphorylation of the co-receptors Ig␣/Ig and CD19, the
adapter proteins BLNK and BCAP, and the tyrosine kinase, Syk.
BCAP and CD19 recruit PI3 kinase to the membrane with generation of PIP3 and subsequent localization of PDK1 and Akt to the
membrane. Phosphorylation by Akt enhances cell survival and
leads to activation of the transcription factors NF-B and CREB as
described. Activation of the B cell isoform of PLC, PLC␥2 occurs
on binding to the membrane-localized adapter, BLNK and phosphorylation by Syk, resulting in the generation of DAG and IP3,
with activation of the NFAT and NF-B pathways as described in
the text. MAP kinase pathways are activated through the binding
of Grb2 to BLNK, with subsequent activation of Ras and through
the activation of the GEF protein VAV which activates Rac.
receptor heavy chains. Interactions between neighboring
receptor molecules through these domains seed receptor
oligomerization (formation of small clusters of antigenreceptor complexes) and subsequent movement of these
receptor clusters into specialized lipid raft regions of the B-cell
membrane. There, the ITAM residues of Ig␣/Ig are brought
into close proximity with the Src-family kinases Lyn (Figure 3-28), Fyn, and Blk. Tyrosine phosphorylation of the
Ig␣/Ig ITAM residues by these Src-family kinases, particularly Lyn, then provides attachment sites for the adapter protein BLNK and an additional tyrosine kinase, Syk, which is
phosphorylated and activated by the Src-family kinases. Syk
then phosphorylates BLNK, providing docking sites for multiple downstream components of the signaling pathway. The
adapter protein BCAP and CD19, the B-cell co-receptor, are
also phosphorylated by these tyrosine kinases and serve to
recruit the enzyme PI3 kinase to the plasma membrane.
[Adapted from M. E. Conley et al. 2009. Annual Reviews of Immunology
27:199–227.]
B Cells Use Many of the Downstream Signaling
Pathways Described Above
The downstream signaling pathways used by the BCR will now
be familiar. The tyrosine kinases Syk and Btk together phosphorylate and activate PLC␥2, which hydrolyzes PIP2, as
Receptors and Signaling: B and T-Cell Receptors
described above. The resultant increase in intracytoplasmic
Ca2⫹ concentrations induces the activation of calcineurin and
the movement of NFAT into the nucleus (see Figure 3-14). The
other product of PIP2 hydrolysis, DAG (see Figure 3-12),
remains in the membrane and binds the B-cell isoform of protein kinase C, leading to the phosphorylation and release of the
NF-B inhibitor as described above for T cells (see Figure 3-17).
This results in the nuclear localization and activation of NFB.
Additional downstream effector pathways elicit the many
other changes that take place upon B-cell activation. For
example, PI3 kinase, now localized at the membrane, phosphorylates PIP2 to PIP3 (see Figure 3-11), allowing the
recruitment of the PH domain–containing proteins PDK1
and Akt. On phosphorylation by the serine-threonine kinase
PDK1, Akt promotes cell survival by phosphorylating and
inactivating pro-apoptotic molecules such as Bax and Bad. It
also phosphorylates and further activates the transcription
factors NF-B and CREB, which both support proliferation,
differentiation, and survival functions of the activated B cells.
The MAP kinase pathway (see Figure 3-16) is also activated during B-cell activation. Grb2 attachment to BLNK
|
CHAPTER 3
93
brings about the binding of the GEF SOS, followed by the
binding and activation of Ras, as described above. Similarly,
Rac, another small monomeric G protein of the Ras family,
is activated by binding to the GEF protein Vav. Ras and Rac
are both small G proteins that act to initiate MAP kinase
signaling pathways. As described above, activation of the
MAP kinase pathway culminates in the expression of the
phosphorylated transcription factor Elk. In B cells, Elk promotes the synthesis of the transcription factor Egr-1, which
acts to induce alterations in the cell-surface expression of
important adhesion molecules, and ultimately serves to aid
B lymphocyte migration into and within the secondary lymphoid tissues. Downstream effectors from Rac promote actin
polymerization, further facilitating B-cell motility.
In conclusion, antigen binding at the BCR leads to multiple changes in transcriptional activity, as well as in the
localization and motility of B cells, which together result in
their enhanced survival, proliferation, differentiation, and
eventual antibody secretion. As exemplified in Box 3-2,
defects in proteins involved in B-cell signaling can lead to
immunodeficiencies.
Box 3-2
CLINICAL FOCUS
Defects in the B-Cell Signaling Protein Btk Lead to X-Linked
Agammaglobulinemia
The characterization of the proteins necessary for B- and T-cell signaling
opened up new avenues of exploration
for clinicians working with patients suffering from immunodeficiency disorders. Clinicians and immunogeneticists now work
closely together to diagnose and treat
patients with immunodeficiencies, to the
benefit of both the clinical and the basic
sciences.
Characterization of the genes responsible for disorders of the immune system is
complicated by the fact that antibody
deficiencies may result from defective
genes encoding either T- or B-cell proteins
(since T cells provide helper factors necessary for B-cell antibody production), or even
from mutations in genes encoding proteins
in stromal cells important for healthy B-cell
development in the bone marrow. However, no matter what the cause, all antibody
deficiencies manifest clinically in increased
susceptibility to bacterial infections, particularly those of the lung, intestines, and (in
younger children) the ear.
In 1952, a pediatrician, Ogden Bruton,
reported in the journal Pediatrics the case
of an eight-year-old boy who suffered from
multiple episodes of pneumonia. When
the serum of the boy was subjected to
electrophoresis, it was shown to be completely lacking in serum globulins, and his
disease was therefore named agammaglobulinemia. This was the first immunodeficiency disease for which a laboratory
finding explained the clinical symptoms,
and the treatment that Bruton applied—
administering subcutaneous injections of
gamma globulin—is still used today. As
similar cases were subsequently reported,
it was noted that most of the pediatric
cases of agammaglobulinemia occurred in
boys, whereas when the disease was
reported in adults, both men and women
appeared to be similarly affected. Careful
mapping of the disease susceptibility to
the X chromosome resulted in the pediatric form of the disease being named XLA,
for X-linked agammaglobulinemia.
With the characterization of the BCR
signal transduction pathway components
in the 1980s and 1990s came the opportunity to define which proteins are damaged
or lacking in particular immunodeficiency
syndromes. In 1993, 41 years after the initial
description of the disease, two groups
independently reported that many cases
of XLA resulted from mutations in a cytoplasmic tyrosine kinase called Bruton’s tyrosine kinase, or Btk; at this point, we now
know that fully 85% of patients affected
with XLA have mutations in the Btk gene.
Btk is a member of the Tec family of
cytoplasmic tyrosine kinases, which are
predominantly expressed in hematopoietic
cells. Tec family kinases share a C-terminal
kinase domain, preceded by SH2 and SH3
domains, a proline-rich domain, and an
amino-terminal PH domain, capable of
binding to PIP3 phospholipids generated by
PI3 kinase activity. Btk is expressed in both B
cells and platelets and is activated following
signaling through the BCR, the pre-BCR
(which is expressed in developing B cells),
the IL-5 and IL-6 receptors, and also the
CXCR4 chemokine receptor. Its involvement in pre-BCR signaling explains why
(continued)
94
PA R T I
CLINICAL FOCUS
|
Introduction
Box 3-2
(continued)
children with XLA suffer from defective
B-cell development.
Following activation, Btk moves to the
inner side of the plasma membrane, where
it is phosphorylated and partially activated
(see Figure 3-28). Activation is completed
when it autophosphorylates itself at a second phosphorylation site. Btk binds to the
adapter protein BLNK, along with PLC␥2.
Btk then phosphorylates and activates
PLC␥2, leading, as described, to calcium
flux and activation of the NF-B and NFAT
pathways. Btk therefore occupies a central
position in B-cell activation, and it is no
surprise that mutations in its gene result in
such devastating consequences.
Over 600 different mutations have
been identified in the btk gene, with the
vast majority of these resulting from single base pair substitutions, or the insertion or deletion of less than five base pairs.
As for other X-linked mutations that are
lethal without medical intervention, XLA
disease is maintained in the population by
the generation of new mutations.
Patients with XLA are usually healthy in
the neonatal (immediately after birth)
period, when they still benefit from maternal antibodies. However, recurrent bacterial infections begin between the ages of
3 months and 18 months, and currently
the mean age at diagnosis in North
America is 3 years. XLA is a so-called “leaky”
defect; almost all children with mutations
in btk have some serum immunoglobulin,
and a few B cells in the peripheral circulation. The prognosis for patients who are
treated with regular doses of gamma
globulin has improved dramatically over
the last 25 years, with the use of prophylactic injections of gamma globulin.
The B cells in patients with XLA have a
distinctive phenotype that can be used for
diagnostic purposes. CD19 expression is
low and variable in patients with XLA,
whereas membrane IgM expression, normally variable in mature B cells, is relatively
high and consistent in patients with XLA.
This phenotype can be observed in Figure
1, which shows the flow cytometric profiles of a normal control individual (left two
graphs) and a patient with a defective btk
gene (right two graphs). In the top two
plots, we note that the XLA patient has
very few CD19⫹ B cells compared with
the control and that the levels of CD19 on
the surface of those B cells that do exist are
lower than those of the control cells. In the
lower two plots, we note that, although
there are fewer CD19⫹ B cells overall in the
btk-compromised patient, all of those
CD19⫹ B cells have relatively high levels of
surface IgM, whereas the levels of membrane IgM are much more variable in the
healthy controls.
Bruton, O. C. 1952. Agammaglobulinemia. Pediatrics 9:722–728.
Conley, M. E., et al. 2009. Primary B cell immunodeficiencies: Comparisons and contrasts.
Annual Review of Immunology 27:199–227.
Tsukuda, S., et al. 1993. Deficient expression of
a B cell cytoplasmic tyrosine kinase in
human X-linked agammaglobulinemia. Cell
72:279–290
Vetrie, D., et al. 1993. The gene involved in
X-linked agammaglobulinemia is a member
of the src family of protein tyrosine kinases.
Nature 361:226–233.
Control
CD19
Btk
CD19
Control antibody
CD19
Control antibody
CD19
sIgM
sIgM
FIGURE 1
FACS profiles of a normal individual and an XLA patient. [Adapted from Conley et al. 2009.
Annual Review of Immunology 27:199.]
B Cells Also Receive Signals Through
Co-Receptors
The immunoglobulin receptor on the B cell membrane is
noncovalently associated with three transmembrane molecules: CD19, CD21, and CD81 (TAPA-1) (see Figure 3-7a).
Antigens are sometimes presented to the BCR already covalently bound to complement proteins, in particular to the
complement component C3d. (The complement cascade is
discussed in Chapter 6.) The B-cell co-receptor CD21 specifically binds to C3d, on C3d-coated antigens. This coengagement of the BCR and CD21 brings the co-receptor
and the BCR into close apposition with one another. When
this happens, tyrosine residues on the cytoplasmic face of
the co-receptor become phosphorylated by the same
enzymes that phosphorylate the ITAMs on Ig␣/Ig, providing sites of attachment for PI3 kinase. As illustrated in Figure
3-28, localization of PI3 kinase to the co-receptor enhances
Receptors and Signaling: B and T-Cell Receptors
both cell survival and the alterations in the transcriptional
program that accompany cell activation. CD19 also serves as
an additional site of recruitment of PLC␥.
T-Cell Receptors and Signaling
T cells bind complex antigens made up of peptides located
in the groove of membrane-bound MHC proteins. When
the T-cell receptor makes contact with its MHC-peptide
antigen on the surface of an antigen-presenting cell, the
two cell membranes are brought into close apposition with
one another. This adds an additional layer of complexity to
the process of T-cell activation that finds no parallel in
B-cell signaling. Notwithstanding this additional complexity, the events of T-cell activation still unfold according to
a mix of the strategies described above, and bear many
similarities to B-cell receptor signaling. Here, we briefly
describe T-cell receptor structure and then turn to a characterization of the signaling routes through this receptor.
In Box 3-3, we provide a description of the experiments
that resulted in the isolation and characterization of the
␣TCR.
The T-Cell Receptor is a Heterodimer with
Variable and Constant Regions
There are two types of T-cell receptors, both of which are
heterodimers (dimers made up of two different polypeptides). The majority of recirculating T cells bear ␣ heterodimers, which bind to ligands made up of an antigenic
peptide presented in a molecular groove on the surface of
a type I or type II MHC molecule. A second subset of T
cells instead expresses a heterodimeric T-cell receptor
composed of a different pair of protein chains, termed ␥
and ␦. T cells bearing ␥␦ receptors have particular localization patterns (often in mucosal tissues) and some ␥␦ T cells
recognize different types of antigens from those bound by
␣ T cells. Although some ␥␦ T cells recognize conventional MHC-presented peptide antigens, other ␥␦ T cells
bind lipid or glycolipid moieties presented by noncanonical
MHC molecules. Yet other ␥␦ T-cell clones appear to recognize self-generated heat shock proteins or phosphoantigens derived from microbes. A unified theory of the
precise nature of antigens recognized by ␥␦ T cells remains
elusive. However, this ability of ␥␦ T cells to break the rules
of MHC restriction may account for the evolution of a
slight difference in the angle between the antigen-binding
and constant regions of the T-cell receptor, which is apparent in an x-ray crystallographic analysis of the two types of
receptor (Figure 3-29). Notwithstanding these functional
differences in the ␣ versus the ␥␦ receptors, their essential
biochemistry is quite similar.
For the remainder of this chapter, we focus on the structure and signaling of the dominant ␣ TCR type, recognizing
|
CHAPTER 3
95
that minor differences may exist between the two types of
receptors and their downstream components.
Although the TCR is not an immunoglobulin per se, the
TCR proteins are members of the immunoglobulin superfamily of proteins and therefore the domain structures of
␣ and ␥␦ TCR heterodimers are strikingly similar to
those of the immunoglobulins (see Figure 3-19). The ␣
chain has a molecular weight of 40–50 kDa, and the 
chain’s is 40–45 kDa. Like the antibody light chains, the
TCR chains have two immunoglobulin-like domains, each
of which contains an intrachain disulfide bond spanning 60
to 75 amino acids. The C␣ domain of the TCR differs from
most immunoglobulin domains in that it possesses only a
single  sheet, rather than a pair, and the remainder of the
sequence is more variably folded. The amino-terminal
(variable) domain in both chains exhibits marked sequence
variation, but the sequences of the remainder of each chain
are conserved (constant). Each of the TCR variable domains
has three hypervariable regions, which appear to be equivalent to the complementarity-determining regions (CDRs)
in immunoglobulin light and heavy chains. A fourth
hypervariable region on the TCR chain does not appear
to contact antigen, and its functional significance is therefore uncertain.
At the C-terminal end of the constant domain, each
TCR chain contains a short connecting sequence, in which
a cysteine residue forms a disulfide link with the other
chain of the heterodimer. C-terminal to this disulfide is a
transmembrane region of 21 or 22 amino acids, which
anchors each chain in the plasma membrane. The transmembrane domains of the TCR ␣ and  chains are unusual
in that they each contain positively charged amino acid
residues that promote interaction with corresponding
negatively charged residues on the chains of the signaltransducing CD3 complex. Finally, like BCRs, each TCR
chain contains only a very short cytoplasmic tail at the
carboxyl-terminal end.
V domains
C domains
␥␦ TCR
111⬚
␣ TCR
147⬚
FIGURE 3-29 Comparison of the crystal structures of ␥␦
and ␣ TCRs. The difference (highlighted with black lines) in the
elbow angle between the ␥␦ and ␣ forms of the TCR.
CLASSIC EXPERIMENT
The Discovery of the ␣ T-Cell Receptor
Once scientists had established
that the BCR was simply a membranebound form of the secreted antibody,
the elucidation of BCR structure became
a significantly more tractable problem.
1
However, investigators engaged in characterizing the T-cell receptor (TCR) did
not enjoy the same advantage, as the
TCR is not secreted in soluble form.
Understanding of TCR biochemistry
Generation of a T cell hybridoma
with known antigen specificity
4
Production of antibodies that bind
to TCR on the T cell hybridoma
Immunize a new mouse with
cells from an OVA-specific
T cell hybridoma
Immunize mouse
with ovalbumin (OVA)
Wait several days
Remove lymph nodes
Culture lymph node cell T cells with OVA
Wait several days
Isolate spleen
5
2
therefore lagged behind that of the BCR
until the 1980s, when an important scientific breakthrough—the ability to
make monoclonal antibodies from artificially constructed B-cell tumors, or
Fuse B cells from spleen
with long-term B cell line
Add polyethylene glycol
to induce fusion of
antigen-specific T cells
with long-lived T cell line
6
3
Dilute out fused cells so that each well
contains a single T cell hybridoma.
Allow cells to divide to form clones and
test each clone for its ability to secrete IL-2
when stimulated with ovalbumin
peptides. Grow up individual clones
of T cell hybridomas that recognize OVA.
Selection and expansion of those long-term B cell
clones that secrete monoclonal antibodies that bind
to the T cell hybridoma
Collect monoclonal antibodies that bind to T cell lines
FIGURE 1
The generation of antibodies specific for the TCR
96
BOX 3-3
hybridomas—made the analysis of the
TCR more technically feasible.
A hybridoma is a fusion product of two
cells. B-cell hybridomas are generated by
artificially fusing antibody-producing,
short-lived lymphocytes with long-lived
tumor cells in order to generate long-lived
daughter cells secreting large amounts of
monoclonal antibodies. (The term monoclonal refers to the fact that all of the cells
in a given hybridoma culture are derived
from the single clone of cells, and therefore carry the same DNA; details of the
technology are described in Chapter 20.)
Although this technique was first developed for the generation of long-lived B
cells, scientists working in the laboratory
of John Kappler and Philippa Marrack also
applied it to T lymphocytes.
The researchers began by immunizing
a mouse with the protein ovalbumin
(OVA), allowing OVA-specific T cells to
divide and differentiate for a few days, and
then harvesting the lymph nodes from
the immunized animal. To enrich their
starting population with as many OVAspecific T cells as possible, they cultured
the harvested lymph node cells in vitro
with OVA for several hours (Figure 1, step 1).
7
After some time in culture, they fused
these activated, OVA-specific T cells with
cells derived from a T-cell tumor (Figure 1,
step 2), thus generating a number of
long-lived T-cell hybridoma cultures
that recognized OVA peptides in the context of the MHC of the original mouse, an
MHC allele called H-2d. They then diluted
out the fused cells in each culture, generating several T-cell hybridoma lines in
which all the cells in an individual hybridoma line derived from the product of a
single fusion event (Figure 1, step 3). This
is referred to as cloning by limiting dilution. In this way, they isolated a T-cell
hybridoma that expressed a TCR capable
of recognizing a peptide from OVA, in the
context of MHC Class 2 proteins from
mice of the H-2d strain.
These T cells could now be used as
antigens and injected into a mouse (Figure
1, step 4). The spleen of this mouse was
removed a few days later, and the mouse B
cells were fused with B lineage tumor cells
(Figure 1, step 5). After culturing to stabilize
the hybrids, the investigators cloned the
B-cell hybridomas and selected a B-cell
hybridoma line that produced monoclonal
antibodies that bound specifically to the
T-cell hybridoma (Figure 1, step 6). Most
important, these antibodies interfered with
the T cell’s ability to recognize its cognate
antigen (Figure 1, step 7). The fact that this
monoclonal antibody inhibited TCR antigen binding suggested that the antibody
was binding directly to the receptor, and
competing with the antigen for TCR binding. They then used these antibodies to
immunoprecipitate the TCR from detergent-solubilized membrane preparations
and purify the TCR protein (Figure 1, step 8).
Concurrent with these experiments,
the laboratories of Stephen Hedrick and
Mark Davis at NIH and Tak Mak in Toronto
had been making headway searching for
the genes encoding the T-cell receptor.
These experiments, as well as subsequent
work by Susumu Tonegawa, which completed the identification of the TCR genes,
are described in detail in Chapter 7.
Haskins et al. 1983. The major histocompatibility complex-restricted antigen receptor on T
cells. I. Isolation with a monoclonal antibody. The Journal of Experimental Medicine
157:1149–1169.
Haskins et al. 1984. The major histocompatibility complex-restricted antigen receptor
on T cells. Annual Review of Immunology
2:51–66.
Identification of monoclonal antibody that interferes
with antigen recognition by T cell hybridoma cells
Mixing T cell hybridoma
with OVA-presenting
cells in absence of
antibody results in T
hybridoma proliferation
OVA
Mixing T cell hybridoma
with OVA-presenting
cells in presence of
antibody results in no T
hybridoma proliferation
OVA
8
The two chains of the αβ receptor
can be seen in lane 1 of the gel,
running on top of one another at
approximately 40-45 kDa. The two
higher molecular weight bands seen
in this lane were shown to represent
non-specific bands.
45 kDa αβ TCR bands
97
PA R T I
98
|
Introduction
The T-Cell Signal Transduction Complex
Includes CD3
Just as B-cell signaling requires the participation of the
Ig␣/Ig signal transduction complex, signaling through the
TCR depends on a complex of proteins referred to collectively
as CD3 (Figure 3-30). The CD3 complex is made up of three
dimers: a ␦⑀ (delta epsilon) pair, a ␥⑀ (gamma epsilon) pair,
and a third pair that is made up either of two CD3 (zeta)
(a)
molecules or a (zeta, eta) heterodimer. (Note that the CD3
␥ and ␦ chains are different from the chains that make up the
␥␦ TCR.) Like Ig␣ and Ig, the cytoplasmic tails of the CD3
molecules are studded with ITAM sequences that serve as
docking sites for adapter proteins following activationinduced tyrosine phosphorylation. Each of the CD3 dimers
contains negatively charged amino acids in its transmembrane
domain that form ionic bonds with the positively charged
residues on the intramembrane regions of the T-cell receptor.
TCR
NH2
23
222 S S 255
9
+
+
−
−
30
NH2
NH2
90
+
COOH COOH
(248) (282)
NH2
105
105
NH2
S S
201
εδ
S S
184
140
S S
S S
134
S S
γε
91
S S
90
ζ ζ
β
S S
22
S S
NH2
S S
α
80
−
−
−
−
116
130
130
106
160
COOH
ITAM
185
COOH
185
COOH
150
COOH
143
COOH
COOH
TCR–CD3
(b)
TCR
␥
␦
⫺
⫹
⫹
⫹
⫺
⫺
ITAM
FIGURE 3-30 Schematic diagram of the TCR-CD3 complex, which constitutes the T-cell antigen-binding receptor.
(a) Components of the CD3 complex include the homodimers
(alternately, a heterodimer) plus ␥⑀ and ␦⑀ heterodimers. The
external domains of the ␥, ␦, and ⑀ chains of CD3 consist of immunoglobulin folds, which facilitates their interaction with the T-cell receptor and with each other. The long cytoplasmic tails of the CD3 chains
contain a common sequence, the Immunoreceptor Tyrosine-based
Activation Motif (ITAM), which functions in signal transduction; these
sequences are shown as blue boxes. (b) Ionic interactions also may
occur between the oppositely charged transmembrane regions in
the TCR and CD3 chains. Proposed interactions among the CD3 components and the ␣ TCR are shown. [Adapted from M. E. Call and K. W.
Wucherpfenning, 2004. Molecular mechanisms for the assembly of the T cell
receptor-CD3 complex. Molecular Immunology 40:1295.]
|
Receptors and Signaling: B and T-Cell Receptors
TABLE 3-4
CHAPTER 3
99
Selected T cell accessory molecules participating in T-cell signal transduction
FUNCTION
Name
Ligand
Adhesion
Signal
transduction
Member of
Ig superfamily
CD4
Class II MHC
⫹
⫹
⫹
CD8
Class I MHC
⫹
⫹
⫹
CD2 (LFA-2)
CD58 (LFA-3)
⫹
⫹
⫹
CD28
CD80, CD86
?
⫹
⫹
CTLA-4
CD80, CD86
?
⫹
⫺
CD45R
CD22
⫹
⫹
⫹
CD5
CD72
?
⫹
⫺
β
S S
D2
CD8
α
S S
D1
S S
CD4
S S
The T-cell receptor is noncovalently associated with a
number of accessory molecules on the cell surface (Table
3-4). However, the only two such molecules that also recognize the MHC-peptide antigen are CD4 and CD8. Recall
that mature T cells can be subdivided into two populations
according to their expression of CD4 or CD8 on the
plasma membrane. CD4⫹ T cells recognize peptides that
are combined with class II MHC molecules, and function
primarily as helper or regulatory T cells, whereas CD8⫹ T
cells recognize antigen that is expressed on the surface of
class I MHC molecules, and function mainly as cytotoxic
T cells.
CD4 is a 55 kDa monomeric membrane glycoprotein
that contains four extracellular immunoglobulin-like
domains (D1–D4), a hydrophobic transmembrane region,
and a long cytoplasmic tail containing three serine residues that can be phosphorylated (Figure 3-31). CD8 takes
the form of a disulfide-linked ␣ heterodimer or ␣␣
homodimer. (These are not the same as the ␣ and 
chains that constitute the TCR heterodimer.) Both the ␣
and  chains of CD8 are small glycoproteins of approximately 30 to 38 kDa. Each chain consists of a single,
extracellular, immunoglobulin-like domain, a stalk
region, a hydrophobic transmembrane region, and a cytoplasmic tail containing 25 to 27 residues, several of which
can be phosphorylated.
The extracellular domains of CD4 and CD8 bind to conserved regions of MHC class II and MHC class I molecules
respectively (see Figure 3-7b). The co-engagement of a single
MHC molecule by both the TCR and its CD4 or CD8 coreceptor enhances the avidity of T-cell binding to its target.
This co-engagement also brings the cytoplasmic domains of
the TCR/CD3 and the respective co-receptor into close
proximity, and it helps to initiate the cascade of intracellular
events that activate a T cell.
Signaling through the antigen receptor, even when combined with that through CD4 or CD8, is insufficient to activate a T cell that has had no prior contact with antigen (a
naïve T cell). A naïve T cell needs to be simultaneously signaled through the TCR and its co-receptor, CD28, in order
to be activated. The TCR and CD28 molecules on a naïve T
cell must co-engage the MHC-presented peptide and the
CD28 ligand, CD80 (or CD86), respectively, on the antigenpresenting cell for full activation to occur. The signaling
events mediated through CD28, which include the stimulation of interleukin 2 synthesis by the T cell, were alluded to
above and are discussed fully in Chapter 11.
D3
D4
S S
The T Cell Co-receptors CD4 and CD8
Also Bind the MHC
S S
FIGURE 3-31 General structure of the CD4 and CD8 coreceptors; the Ig-like domains are shown as circles. CD8 takes
the form of an ␣ heterodimer or an ␣␣ homodimer. The monomeric CD4 molecule contains four Ig-fold domains; each chain in the
CD8 molecule contains one.
PA R T I
100
|
Introduction
CD45
TCR/CD3
CD4
CD28
ss
PIP3
PIP3
Itk
GADS
P
LAT SLP76
P P
P
VAV
P
P
PLCγ1
P
P
P
ZAP-70 P
P
P
P
Lck
P
Grb-2
SOS
Ras
DAG
Ca2+
Akt PDK1
Grb-2 PI3K
PKCθ
Bax Bad
Rac/Rho/
cdc42
Cytoskeletal
reorganization
Survival
MAP kinase
cascade
Calmodulin
Calcineurin
Cytoplasm
Nucleus
NFAT
NF-κB
AP-1
Gene activation
Lck is the First Tyrosine Kinase Activated
in T Cell Signaling
When the T-cell receptor interacts with its cell-bound antigen, receptors, co-receptors, and signaling molecules cluster
into the cholesterol-rich lipid rafts of the plasma membrane
(Figure 3-32). The Src-family tyrosine kinase Lck is normally
found associated with CD4 and CD8, and the association
between Lck and CD4 is particularly close. Antigen-induced
clustering of the receptor–co-receptor complex brings Lck
into the vicinity of the membrane-associated tyrosine phosphatase, CD45, which removes the inhibitory phosphate
group on Lck. Reciprocal phosphorylation by nearby Lck
molecules at their activating tyrosine sites (see Figure 3-9)
then induces Lck to phosphorylate CD3 ITAM residues.
(Note that these early events parallel those induced in B cells
by the Src-family kinase Lyn, which is also regulated by the
phosphatase activity of CD45.)
Once the CD3 ITAMs are phosphorylated, a second tyrosine kinase, ZAP-70, docks at the phosphorylated tyrosine
FIGURE 3-32 Signal transduction pathways emanating from the TCR. T cell antigen binding activates
the src-family kinase Lck, which phosphorylates the
kinase ZAP-70. ZAP-70 in turn phosphorylates the adapter
molecules LAT, SLP76, and GADS which form a scaffold
enabling the phosphorylation and activation of PLC␥1
and PKC with the consequent effects on transcription
factor activation described in the text. The GEF proteins
Vav and SOS are also activated on binding to LAT, leading
to activation of the Ras/MAP kinase transcription factor
pathway and the Rac/Rho/cdc42 pathway, leading to
changes in cell shape and motility. PI3 kinase, translocated to the cytoplasmic side of CD28, forms PIP3, inducing localization of the enzymes PDK1 and Akt to the
membrane. This leads to further NF-B activation and
increased cell survival as described.
residues of the CD3 chains. ZAP-70 is activated by Lckmediated phosphorylation and goes on to phosphorylate
many adapter molecules including SLP-76 and LAT, as well
as enzymes important in T-cell activation, such as PLC␥1.
T Cells Use Downstream Signaling Strategies
Similar to Those of B Cells
Just as in B cells, signals initiating at the antigen receptor of
T cells with tyrosine phosphorylation events are then fanned
out to intracellular enzymes and transcription factors using
a network of adapter molecules and enzymes.
In T cells, one of the earliest adapter molecules to be incorporated into the signaling complex is LAT (Linker protein of
Activated T cells), a transmembrane protein associated with
lipid rafts in the plasma membrane. Following TCR ligation,
LAT is phosphorylated on multiple residues by ZAP-70, and
these phosphorylated residues now provide docking sites for
several important enzymes bearing SH2 domains, including
PLC␥1 (see Figure 3-32). Phosphorylated LAT also binds to
Receptors and Signaling: B and T-Cell Receptors
the adapter protein GADS, which is constitutively associated
with the adapter SLP-76. This combination of adapter proteins
is critical to T-cell receptor signaling, providing the structural
framework for most downstream signaling events.
Many of those downstream events will now be familiar.
PLC␥1, localized to the plasma membrane by binding to LAT,
is further activated by tyrosine phosphorylation, mediated by
the kinase Itk (which belongs to a family of kinases referred to
as Tec kinases). As described earlier, PLC␥1 breaks down PIP2,
releasing IP3, which induces the release of calcium and the
activation of NFAT via calcineurin activation. The DAG created by PIP2 hydrolysis binds, in T cells, to a specialized form
of PKC called PKC (theta). As described above, this part of
the signaling cascade similarly culminates in the degradation
of the inhibitors of NF-B and the translocation of the active
transcription factor into the nucleus (see Figure 3-17).
Phosphorylated LAT also associates with the SH2 domain
of Grb2, the now-familiar adapter molecule that brings in
components of the Ras pathway to the signaling complex.
Recall that Grb2 binds constitutively to SOS, the GEF that
|
CHAPTER 3
101
facilitates activation of the Ras pathway. In T cells, the Ras
pathway is important both to the activation of the transcription
factor AP-1, which functions to signal cytokine secretion, and
to the passage of the signals that reorganize the actin cytoskeleton for directed cytokine release.
Thus, as for B cells, TCR-antigen binding leads to a multitude of consequences including transcription factor upregulation, reorganization of the cytoskeleton, and cytokine secretion.
Again, as for B cells, T-cell signaling also affects the expression
of adhesion molecules such as integrins on the cell surface, and
chemokines, which has subsequent effects on cell localization.
Clearly the description of adaptive immune signaling
offered in this chapter represents just the tip of the iceberg,
and these signaling cascades contain more components and
outcomes that can be alluded to in this brief outline. However, just as the adapter proteins provide a scaffold for the
immune system to organize its signaling proteins, so we
hope that this chapter has provided a similar scaffold for the
organization of the reader’s thoughts regarding these fascinating and complex processes.
S U M M A R Y
■
■
■
■
■
■
■
Antigens bind to receptors via noncovalent bonding interactions.
The interactions between antigens and receptors of the
immune system are enhanced by simultaneous interactions
between lymphocyte-expressed co-receptors and molecules on antigen-presenting cells or on complex antigens.
Most receptor-antigen interactions are multivalent, and
this multivalency significantly increases the avidity of the
receptor-antigen binding interaction.
Binding of antigen to receptor induces a signaling cascade
in the receptor-bearing cell, which leads to alterations in
the motility, adhesive properties, and transcriptional program of the activated cell.
Antigen signaling is initiated in both T and B cells by
antigen-mediated receptor clustering. The clustered receptors are located in specialized regions of the membranes
called lipid rafts.
The CD3 and Ig␣/Ig proteins, which are T- and B-cell
receptor-associated signal transduction elements, are
phosphorylated on Immunoreceptor Tyrosine Activation
Motifs (ITAMs), and these serve as docking points for
adapter molecules.
Downstream signaling enzymes and GEFs dock onto the
adapter molecules and make contact with their substrates.
These enzymes include phospholipase C␥, which breaks
down PIP2 into DAG and IP3. IP3 interaction with ERlocated receptors leads to release of intracellular calcium
and activation of calcium-regulated proteins such as calcineurin phosphatase. Calcineurin dephosphorylates the
transcription factor NFAT, allowing it access to the
■
■
■
■
■
■
■
nucleus. DAG binds and activates protein kinase C, leading eventually to NF-B activation. Docking of the adapter
molecule Grb2 onto adapter proteins facilitates its binding
to the GEF protein SOS and activation of the MAP kinase
pathway, resulting in activation of the transcription factor
AP-1.
The antigen receptor on B cells is a membrane-bound
form of the four-chain immunoglobulin molecule that the
B cell secretes upon stimulation. An immunoglobulin
molecule is commonly known as an antibody. Antibodies
have two heavy and two light chains.
The antibodies secreted by B cells upon stimulation are
classified according to the amino acid sequence of the
heavy chain, and antibodies of different classes perform
different functions during an immune response.
Antigen signaling in B cells proceeds according to signaling strategies shared among many cell types.
The antigen receptor on T cells is not an immunoglobulin
molecule, although its protein domains are classified as
belonging to the immunoglobulin superfamily of proteins.
Most T cells bear receptors made up of an ␣ heterodimer
that recognizes a complex antigen, made up of a short
peptide inserted into a groove on the surface of a protein
encoded by the major histocompatibility complex (MHC)
of genes.
Some T cells bear receptors made up of ␥␦ receptor chains
that recognize a different array of antigens.
Antigen signaling in T cells shares many characteristic
strategies with B cell signaling.
102
PA R T I
|
Introduction
R E F E R E N C E S
Andreotti, A. H., P. L. Schwartzberg, R. E. Joseph, and L. J. Berg.
2010. T-cell signaling regulated by the Tec family kinase, Itk.
Cold Spring Harbor Perspectives in Biology 2:a002287.
Porter, R. R. (1972). Lecture for the Nobel Prize for physiology
or medicine 1972: Structural studies of immunoglobulins.
Scandinavian Journal of Immunology 34(4):381–389.
Chaturvedi, A., Z. Siddiqui, F. Bayiroglu, and K. V. Rao. 2002. A
GPI-linked isoform of the IgD receptor regulates resting B cell
activation. Nature Immunology 3:951.
Tiselius, A. 1937. Electrophoresis of serum globulin. I. The Biochemical Journal 31:313.
Choudhuri, K., and M. L. Dustin. 2010. Signaling microdomains in T cells. FEBS Letters 584:4823.
Edelman G. M., B. A. Cunningham, W. E. Gall, P. D. Gottlieb, U.
Rutishauser, and M. J. Waxdal. 1969. The covalent structure of
an entire gammaG immunoglobulin molecule. Proceedings of
the National Academy of Sciences U.S.A. 63:78.
Finlay, D., and D. Cantrell. The coordination of T-cell function
by serine/threonine kinases. 2011. Cold Spring Harbor Perspectives in Biology 3:a002261.
Fitzgerald, K. A., and Z. J. Chen. 2006. Sorting out Toll signals.
Cell 125:834.
Heidelberger, M., and F. E. Kendall. 1936. Quantitative studies on
antibody purification : I. The dissociation of precipitates formed
by Pneumococcus specific polysaccharides and homologous
antibodies. Journal of Experimental Medicine 64:161.
Heidelberger, M., and K. O. Pedersen. 1937. The Molecular
Weight of Antibodies. Journal of Experimental Medicine 65:393.
Kanneganti, T. D., M. Lamkanfi, and G. Nunez. 2007. Intracellular NOD-like receptors in host defense and disease. Immunity 27:549.
Koretzky, G. A., and P. S. Myung. 2001. Positive and negative
regulation of T-cell activation by adaptor proteins. Nature
Reviews Immunology 1:95.
Kurosaki, T. 2011. Regulation of BCR signaling. Molecular
Immunology 48:1287.
Love, P. E., and S. M. Hayes. 2010. ITAM-mediated signaling by
the T-cell antigen receptor. Cold Spring Harbor Perspectives in
Biology 2:a002485.
Macian, F. 2005. NFAT proteins: Key regulators of T-cell development and function. Nature Reviews Immunology 5:472.
Nisonoff, A., F. C. Wissler, and L. N. Lipman. 1960. Properties
of the major component of a peptic digest of rabbit antibody.
Science 132:1770.
Palacios, E. H., and A. Weiss. 2004. Function of the Src-family
kinases, Lck and Fyn, in T-cell development and activation.
Oncogene 23:7990.
Park, S. G., J. Schulze-Luehrman, M. S. Hayden, N. Hashimoto,
W. Ogawa, M. Kasuga, and S. Ghos. 2009. The kinase PDK1
integrates T cell antigen receptor and CD28 coreceptor signaling to induce NF-kappaB and activate T cells. Nature Immunology 10:158.
Tiselius, A., and E. A. Kabat. 1938. Electrophoresis of immune
serum. Science 87:416.
Tiselius, A., and E. A. Kabat. 1939. An electrophoretic study of
immune sera and purified antibody preparations. Journal of
Experimental Medicine 69:119.
Wang, H., T. A. Kadlecek, B. B. Au-Yeung, H. E. Goodfellow, L.
Y. Hsu, T. S. Freedman, and A. Weiss. 2010. ZAP-70: An essential kinase in T-cell signaling. Cold Spring Harbor Perspectives
in Biology 2:a002279.
Yamamoto, M., and K. Takeda. 2010. Current views of toll-like
receptor signaling pathways. Gastroenterology Research and
Practice 2010:240365.
Useful Web Sites
www.genego.com This site offers a searchable database
of metabolic and regulatory pathways.
www.nature.com/subject/cellsignaling This is a collection of original research articles, reviews and commentaries pertaining to cell signaling.
http://stke.sciencemag.org This is the Signal Transduction Knowledge Environment Web site, maintained by
Science magazine. It is an excellent site, but parts of it are
closed to those lacking a subscription.
www.signaling-gateway.org This gateway is powered
by University of California at San Diego and is supported by
Genentech and Nature. An excellent resource, complete with
featured articles.
www.qiagen.com/geneglobe/pathwayview.
aspx Qiagen. A useful commercial Web site.
www.biosignaling.com Part of Springer Science⫹Business Media.
www.youtube.com Many excellent videos and animations of signaling Web sites are available on YouTube, too
numerous to mention here. Just type your pathways into a
Web browser and go. But do be cognizant of the derivation
of your video. Not all videos are accurate, so check your facts
with the published literature.
www.hhmi.org/biointeractive/immunology/tcell.
html Howard Hughes Medical Institute (HHMI) movie:
Cloning an Army of T Cells for Immune Defense.
Receptors and Signaling: B and T-Cell Receptors
S T U D Y
|
CHAPTER 3
103
Q U E S T I O N S
1. The NFAT family is a ubiquitous family of transcription
factors.
a. Under resting conditions, where is NFAT localized in a
cell?
b. Under activated conditions, where is NFAT localized in
a cell?
c. How is it released from its resting condition and per-
mitted to relocalize?
d. Immunosuppressant drugs such as cyclosporin act via
inhibition of the calcineurin phosphatase. If NFAT is ubiquitous, how do you think these drugs might act with so few
side effects on other signaling processes within the body?
2. In the early days of experiments designed to detect the
T-cell receptor, several different research groups found that
antibodies directed against immunoglobulin proteins
appeared to bind to the T-cell receptor. Given what you
know about the structure of immunoglobulins and the
T-cell receptor, why is this not completely surprising?
3. True or false? Explain your answers.
Interactions between receptors and ligands at the cell surface:
a. are mediated by covalent interactions.
b. can result in the creation of new covalent interactions
within the cell.
4. Describe how the following experimental manipulations
were used to determine antibody structure.
a. Reduction and alkylation of the antibody molecule
b. Enzymatic digestion of the antibody molecule
c. Antibody detection of immunoglobulin fragments
5. What is an ITAM, and what proteins modify the ITAMs in
Ig␣ and Ig?
6. Define an adapter protein. Describe how an interaction
between proteins bearing SH2 and phosphorylated tyrosine (pY) groups helps to transduce a signal from the T-cell
receptor to the ZAP-70 protein kinase.
7. IgM has ten antigen-binding sites per molecule, whereas
IgG only has two. Would you expect IgM to be able to bind
five times as many antigenic sites on a multivalent antigen
as IgG? Why/why not?
8. You and another student are studying a cytokine receptor
on a B cell that has a Kd of 10⫺6 M. You know that the cytokine receptor sites on the cell surface must be at least 50%
occupied for the B cell to receive a cytokine signal from a
helper T cell. Your lab partner measures the cytokine concentration in the blood of the experimental animal and
detects a concentration of 10⫺7 M. She tells you that the
effect you have been measuring could not possibly result
from the cytokine you’re studying. You disagree. Why?
9. Activation of Src-family kinases is the first step in several
different types of signaling pathways. It therefore makes
biological sense that the activity of this family of tyrosine
kinases is regulated extremely tightly. Describe how phosphorylation of Src-family kinases can deliver both activating and inhibitory signals to Src kinases.
10. You have generated a T-cell clone in which the Src-family
tyrosine kinase Lck is inactive. You stimulate that clone
with its cognate antigenic peptide, presented on the appropriate MHC platform and test for interleukin 2 secretion,
as a measure of T cell activation. Do you expect to see IL-2
secretion or not? Explain.
11. Name one protein shown to be defective in many cases of
X-linked agammaglobulinemia, and describe how a reduction in the activity of this protein could lead to immunodeficiency.
12. The B- and T-cell receptor proteins have remarkably short
intracytoplasmic regions of just a few amino acids. How can
you reconcile this structural feature with the need to signal
the presence of bound antigen to the interior of the cell?
13. Describe one way in which the structure of antibodies is
superbly adapted to their function.
14. As a graduate student, your adviser has handed you a T-cell
clone that appears to be constitutively (always) activated,
although at a low level, even in the absence of antigenic
stimulation, and he has asked you to figure out why. Your
benchmate suggests you start by checking out the sequence
of its lck gene, or the status of the Csk activity in the cell. You
agree that those are good ideas. What is your reasoning?
This page lelt intentionally blank.
Receptors and Signaling:
Cytokines and
Chemokines
Immunoglobulin
family receptors
Hematopoietin-type
receptors (class I)
Interferon-type
receptors (class II)
S
S
α
β
JAK
JAK
S
S
T
he hundreds of millions of cells that comprise
the vertebrate immune system are distributed
throughout the body of the host (see Chapter 2).
Some cells circulate through the blood and
lymph systems, whereas others are sessile (remain in
place) in the primary and secondary lymphoid tissues,
the skin, and the mucosa of the respiratory, alimentary,
and genito-urinary tracts. The key to success for such a
widely dispersed organ system is the ability of its various
components to communicate quickly and efficiently with
one another, so that the right cells can home to the
appropriate locations and take the necessary measures to
destroy invading pathogens.
Molecules that communicate among cells of the
immune system are referred to as cytokines. In general,
cytokines are soluble molecules, although some also exist
in membrane-bound forms. The interaction of a cytokine
with its receptor on a target cell can cause changes in the
expression of adhesion molecules and chemokine
receptors on the target membrane, thus allowing it to
move from one location to another. Cytokines can also
signal an immune cell to increase or decrease the activity
of particular enzymes or to change its transcriptional
program, thereby altering and enhancing its effector
functions. Finally, they can instruct a cell when to survive
and when to die.
In an early attempt to classify cytokines,
immunologists began numbering them in the order of
their discovery, and naming them interleukins. This
name reflects the fact that interleukins communicate
between (Latin, inter) white blood cells (leukocytes).
Examples include interleukin 1 (IL-1), secreted by
macrophages, and interleukin 2 (IL-2), secreted by
activated T cells. However, many cytokines that were
named prior to this attempt at rationalizing nomenclature
have resisted reclassification, and so students will come
across cytokines such as Tumor Necrosis Factor or
Interferons, that are also “interleukins” in all but name.
Although the term cytokine refers to all molecules that
communicate among immune cells, the name chemokine
is used specifically to describe that subpopulation of
4
S
S
α
γ
β
TNF receptors
IL-17 receptors
Chemokine receptors
G protein
Cytokine and Chemokine Receptor Families
■
■
General Properties of Cytokines and Chemokines
Six Families of Cytokines and Associated Receptor
Molecules
■
Cytokine Antagonists
■
Cytokine-Related Diseases
■
Cytokine-Based Therapies
cytokines that share the specific purpose of mobilizing
immune cells from one organ, or indeed, from one part of
an organ, to another. Chemokines belong to the class of
molecules called chemoattractants, molecules that attract
cells by influencing the assembly, disassembly, and
contractility of cytoskeleton proteins and the expression
of cell-surface adhesion molecules. Chemokines attract
cells with the appropriate chemokine receptors to regions
where the chemokine concentration is highest. For
example, chemokines are important in attracting cells of
105
106
PA R T I
|
Introduction
the innate immune system to the site of infection and
inducing T cells to move toward antigen-presenting cells
in the secondary lymphoid tissues. Leukocytes change
their pattern of expression of chemokine receptors over
the course of an immune response, first migrating to the
secondary immune organs, in which they undergo
differentiation to mature effector cells, and then moving
out into the affected tissues to fight the infection,
responding to different chemokine gradients with each
movement. As we will learn in a later section, chemokines
are also capable of instructing cells to alter their
transcriptional programs.
The classification and nomenclature of chemokines is
more logical than that of interleukins, and is based on
their biochemical structures. Although chemokines
technically fall under the umbrella classification of
“cytokines,” normal usage is evolving such that the term
chemokine is used when referring to molecules that move
immune cells from place to place, and the term cytokine is
employed when referring to any other messenger
molecule of the immune system.
Like all signaling molecules, cytokines can be further
classified on the basis of the distance between the cell
secreting the signaling ligand and the cell receiving that
chemical signal. Cytokines that act on cells some distance
away from the secreting cell, such that they must pass
through the bloodstream before reaching their target, are
referred to as endocrine (Figure 4-1). Those that act on
cells near the secreting cell, such that the cytokine merely
has to diffuse a few Ångstroms through tissue fluids or
Circulation
Endocrine action
Paracrine action
Distant cell
Nearby cell
Autocrine action
FIGURE 4-1 Most immune system cytokines exhibit
autocrine and/or paracrine action; fewer exhibit endocrine
action.
across an immunological synapse, are referred to as
paracrine. Sometimes, a cell needs to receive a signal
through its own membrane receptors from a cytokine that
it, itself, has secreted. This type of signaling is referred to as
autocrine. Of note, the T-cell interleukin IL-2 acts
effectively in all three modes. Unlike the classical
hormones, such as insulin and glucagon, that generally act
at long range in an endocrine fashion, many cytokines act
over a short distance in an autocrine or paracrine fashion.
We begin this chapter with an introduction to the
general properties of cytokines and chemokines followed
by a discussion of the specific receptors and signaling
pathways used by the six families of immune system
cytokines and chemokines. Next, we describe the ways in
which cytokine signaling can be regulated by antagonists.
Finally, we turn to the role of cytokines and chemokines
in disease and medicine.
General Properties of Cytokines
and Chemokines
The activity of cytokines was first recognized in the mid1960s, when supernatants derived from in vitro cultures of
lymphocytes were found to contain soluble factors, usually
proteins or glycoproteins, that could regulate proliferation,
differentiation, and maturation of immune system cells. Production of these factors by cultured lymphocytes was
induced by activation with antigens or with nonspecific
mitogens (molecules inducing cell division, or mitosis).
However, biochemical isolation and purification of cytokines was initially hampered because of their low concentrations in the culture supernatants and the absence of
well-defined assay systems for individual cytokines.
The advent of hybridoma technology (see Chapter 20)
allowed the production of artificially generated T-cell tumors
that constitutively produced IL-2, allowing for its purification and characterization. Gene cloning techniques developed during the 1970s and 1980s then made it possible to
generate pure cytokines by expressing the proteins from
cloned genes derived from hybridomas or from normal leukocytes, after transfection into bacterial or yeast cells. Using
these pure cytokine preparations, researchers were able to
identify cell lines whose growth depended on the presence of
a particular cytokine, thus providing them with biological
cytokine assay systems. Since then, monoclonal antibodies
specific for many cytokines have made it possible to develop
rapid, quantitative, cytokine-specific immunoassays. ELISA
assays measure the concentrations of cytokines in solution,
Elispot assays quantitate the cytokines secreted by individual
cells, and cytokine-specific antibodies can be used to identify cytokine-secreting cells using intracellular cytokine
staining followed by flow cytometry or immuno-fluorescence
microscopy (see Chapter 20).
Receptors and Signaling: Cytokines and Chemokines
Cytokines Mediate the Activation, Proliferation,
and Differentiation of Target Cells
Cytokines bind to specific receptors on the membranes of
target cells, triggering signal transduction pathways that
ultimately alter enzyme activity and gene expression (Figure 4-2). The susceptibility of a target cell to a particular
cytokine is determined by the presence of specific membrane receptors. In general, cytokines and their fully assembled receptors exhibit very high affinity for one another, with
dissociation constants for cytokines and their receptors
ranging from 108 to 1012 M1. Because their receptor
affinities are so high and because cytokines are often secreted
in close proximity to their receptors, such that the cytokine
concentration is not diluted by diffusion (as mentioned in
Chapter 3), the secretion of very few cytokine molecules can
mediate powerful biological effects.
Cytokines regulate the intensity and duration of the
immune response by stimulating or inhibiting the activation, proliferation, and/or differentiation of various cells, by
regulating the secretion of other cytokines or of antibodies,
or in some cases by actually inducing programmed cell death
in the target cell. In addition, cytokines can modulate the
Inducing stimulus
Cytokine gene
activation
Cytokine-producing cell
Cytokine
secretion
Cytokine receptor
Signal
Gene or enzyme
activation
Target cell
Biological effect
(e.g., proliferation,
differentiation, cell death)
FIGURE 4-2 Overview of the induction and function of
cytokines. An inducing stimulus, which may be an antigen or
another cytokine, interacts with a receptor on one cell, inducing it to
secrete cytokines that in turn act on receptors of a second cell, bringing about a biological consequence. In the case of IL-2, both cells
may be antigen-activated T cells that secrete IL-2, which acts both on
the secreting cell and on neighboring, activated T cells.
|
CHAPTER 4
107
expression of various cell-surface receptors for chemokines,
other cytokines, or even for themselves. Thus, the cytokines
secreted by even a small number of antigen-activated lymphocytes can influence the activity of many different types of
cells involved in the immune response.
Cytokines exhibit the attributes of pleiotropy, redundancy, synergism, antagonism, and cascade induction (Figure 4-3), which permit them to regulate cellular activity in a
coordinated, interactive way. A cytokine that induces different biological effects depending on the nature of the target
cells is said to have a pleiotropic action, whereas two or
more cytokines that mediate similar functions are said to be
redundant. Cytokine synergy occurs when the combined
effect of two cytokines on cellular activity is greater than the
additive effects of the individual cytokines. In some cases,
the effects of one cytokine inhibit or antagonize the effects
of another. Cascade induction occurs when the action of
one cytokine on a target cell induces that cell to produce one
or more additional cytokines.
Cytokines Have Numerous Biological Functions
Although a variety of cells can secrete cytokines that instruct
the immune system, the principal producers are TH cells,
dendritic cells, and macrophages. Cytokines released from
these cell types are capable of activating entire networks of
interacting cells (Figure 4-4). Among the numerous physiological responses that require cytokine involvement are the
generation of cellular and humoral immune responses, the
induction of the inflammatory response, the regulation of
hematopoiesis, and wound healing.
The total number of proteins with cytokine activity grows
daily as research continues to uncover new ones. Table 4-1
summarizes the activities of some commonly encountered
cytokines. An expanded list of cytokines can be found in
Appendix II. Note, however, that many of the listed functions have been identified from analyses of the effects of
recombinant cytokines, sometimes added alone to in vitro
systems at nonphysiologic concentrations. In vivo, cytokines
rarely, if ever, act alone. Instead, a target cell is exposed to a
milieu containing a mixture of cytokines whose combined
synergistic or antagonistic effects can have a wide variety of
consequences. In addition, as we have learned, cytokines
often induce the synthesis of other cytokines, resulting in
cascades of activity.
Cytokines Can Elicit and Support the
Activation of Specific T-Cell Subpopulations
As described in Chapters 2 and 11, helper T cells can be
classified into subpopulations, each of which is responsible
for the support of a different set of immune functions. For
example, TH1 cells secrete cytokines that promote the differentiation and activity of macrophages and cytotoxic T cells,
thus leading to a primarily cytotoxic immune response, in
which cells that have been infected with viruses and
108
PA R T I
|
Introduction
Target Cell
Effect
CASCADE INDUCTION
Activation
Proliferation
Differentiation
Activated TH cells
PLEIOTROPY
B cell
IL-4
Proliferation
and differentiation
IFN-γ
Cytotoxic T cell precursor
Activated TH cells
Proliferation
Mast cell
Macrophage
REDUNDANCY
IL-2
IL-4
IL-5
Proliferation
IL-12
Activated TH cells
B cell
SYNERGY
IL-4
+
IL-5
Activated TH cells
Induces class switch to IgE
B cell
Activated TH cells
ANTAGONISM
Blocks class switch to IgE
induced by IL-4
IL-4
IFN-γ
Activated TH cells
B cell
IFN-γ, TNF, IL-2, and
other cytokines
FIGURE 4-3 Cytokine attributes of (a) pleiotropy, redundancy, synergism, antagonism, and (b) cascade induction.
intracellular bacteria are recognized and destroyed. The
cytokines IL-12 and interferon (IFN) induce TH1 differentiation. In contrast, TH2 cells activate B cells to make
antibodies, which neutralize and bind extracellular pathogens, rendering them susceptible to phagocytosis and
complement-mediated lysis. IL-4 and IL-5 support the generation of TH2 cells. TH17 cells promote the differentiation
of activated macrophages and neutrophils, and support the
inflammatory state; their generation is induced by IL-17
and IL-23. The differentiation and activity of each distinctive T-cell subpopulation is therefore supported by the
binding of different combinations of cytokines to T-cell
surface receptors, with each cytokine combination inducing its own characteristic array of intracellular signals, and
sending the helper T cell down a particular differentiation
pathway.
|
Receptors and Signaling: Cytokines and Chemokines
CHAPTER 4
109
IFN-α
IFN-β
IFN-γ
IL-6
IL-10
IL-15
Macrophage
IL-1 IFN-α
IL-6 IFN-γ
IL-8 IFN-α
IL-10
IL-12
IL-1
IL-15
IFN-α
IL-18
IFN-β
IL-1
TNF-α
IL-3
TNF-β
IL-4
IL-10
IL-13
IL-6 IFN-α
IL-10 TNF-β
IL-10
TNF-β
TNF-α
IL-1
IL-8
TNF-α
IL-1
IL-8
IFN-α
IFN-β
IFN-γ
TNF-α
TNF-β
Neutrophil
TNF-α
IL-1 IFN-γ
IL-3 TNF-α
IL-4 TNF-β
IL-8
B-cell IL-1 IFN-α
IL-2 IFN-β
IL-4 IFN-γ
IL-6
IL-10
IL-13
IL-14
IL-1
IL-3
IL-4
IL-8
Basophil
IL-6
IL-10
IL-2
IL-4
IL-6
IL-8
IL-9
IL-16
IL-17
IFN-α
IFN-β
IFN-γ
T-cell
IL-3
IL-4
IL-5
Eosinophil
IL-3
IL-9
IL-10
IL-4
IL-4
IL-5
IL-4
IL-10
IL-4
IL-4
Mast cell
FIGURE 4-4 The cells of the immune system are subject to control by a network of cytokine actions.
Cell Activation May Alter the Expression
of Receptors and Adhesion Molecules
The ability of cytokines to activate most, if not all, members
of particular immune cell subpopulations appears to conflict
with the established specificity of the immune system. What
keeps cytokines from activating all T cells, for example, in a
nonspecific fashion during the immune response?
In order for a cell to respond to a signaling molecule, it
must express receptors for that molecule, and responsiveness to a molecular signal can thus be controlled by signal
receptor expression. For example, antigen stimulation of a
T cell induces alterations in the T-cell surface expression of
chemokine receptors. Reception of chemokine signals
through these receptors therefore instructs only those cells
that have previously been activated by antigen to migrate to
nearby lymph nodes or to the spleen. Furthermore, activationinduced changes in the adhesion molecules that are
expressed on the cell membrane ensures that stimulated
cells migrate to, and then remain in, the location best suited
to their function. T-cell activation by antigen also up-regulates
the expression of the receptors for cytokines that provide
proliferative signals, such as IL-2 (as described in Chapter 3),
and also for differentiative cytokines such as IL-4. In this
way, following antigen encounter, only those T cells that
have been activated by antigen are primed to relocate and
to receive the proliferative and differentiative signals they
need to function as a mature immune effector cell. This
110
PA R T I
TABLE 4-1
|
Introduction
Functional groups of selected cytokines*
Cytokine
Secreted by†
Targets and effects
SOME CY TOKINES OF INNATE IMMUNITY
Interleukin 1 (IL-1)
Monocytes, macrophages,
endothelial cells, epithelial cells
Vasculature (inflammation); hypothalamus (fever); liver (induction of
acute phase proteins)
Tumor necrosis
factor- (TNF-)
Macrophages, monocytes,
neutrophils, activated T cells and
NK cells
Vasculature (inflammation); liver (induction of acute phase proteins);
loss of muscle, body fat (cachexia); induction of death in many cell
types; neutrophil activation
Interleukin 12 (IL-12)
Macrophages, dendritic cells
NK cells; influences adaptive immunity (promotes TH1 subset)
Interleukin 6 (IL-6)
Macrophages, endothelial cells,
and TH2 cells
Liver (induces acute phase proteins); influences adaptive immunity
(proliferation and antibody secretion of B-cell lineage)
Interferon- (IFN-) (this
is a family of molecules)
Macrophages dendritic cells,
virus-infected cells
Induces an antiviral state in most nucleated cells; increases MHC Class I
expression; activates NK cells
Interferon (IFN-)
Macrophages, dendritic cells,
virus-infected cells
Induces an antiviral state in most nucleated cells; increases MHC Class I
expression; activates NK cells
SOME CY TOKINES OF ADAPTIVE IMMUNITY
Interleukin 2 (IL-2)
T cells
T-cell proliferation; can promote AICD. NK cell activation and proliferation; B-cell proliferation
Interleukin 4 (IL-4)
TH2 cells, mast cells
Promotes TH2 differentiation; isotype switch to IgE
Interleukin 5 (IL-5)
TH2 cells
Eosinophil activation and generation
Transforming growth
factor (TGF-)
T cells, macrophages,
other cell types
Inhibits T-cell proliferation and effector functions; inhibits B-cell proliferation; promotes isotype switch to IgA; inhibits macrophages
Interferon (IFN-)
TH1 cells, CD8 cells, NK cells
Activates macrophages; increases expression MHC Class I and Class II
molecules; increases antigen presentation
*Many cytokines play roles in more than one functional category.
†
Only the major cell types providing cytokines for the indicated activity are listed; other cell types may also have the capacity to synthesize the given cytokine. Activated
cells generally secrete greater amounts of cytokine than unactivated cells.
pattern of activation-induced alteration in the cell surface
expression of adhesion molecules, chemokine receptors,
and cytokine receptors is a common strategy employed by
the immune system.
Cytokines Are Concentrated Between
Secreting and Target Cells
During the process of T-cell activation by an antigen-presenting dendritic cell, or of B-cell activation by a cognate T cell,
the respective pairs of cells are held in close juxtaposition for
many hours (see Chapter 14). Over that time period, the cells
release cytokines that bind to relevant receptors on the partner cell surface, without ever entering the general circulation.
Furthermore, during this period of close cell-cell contact, the
secretory apparatus of the stimulating cell is oriented so that
the cytokines are released right at the region of the cell membrane that is in closest contact with the recipient cell (see
Figure 3-4). The close nature of the cell-cell interaction and
the directional release of cytokines by the secretory apparatus
means that the effective concentration of cytokines in the
region of the membrane receptors may be orders of magnitude higher than that experienced outside the contact region
of the two cells. Thus, discussions of membrane receptor
affinity and cytokine concentrations within tissue fluids must
always take into account the biology of the responding system and the geography of the cell interactions involved. In
addition, the half-life of cytokines in the bloodstream or
other extracellular fluids into which they are secreted is usually very short, ensuring that cytokines usually act for only a
limited time and over a short distance.
Signaling Through Multiple Receptors
Can Fine Tune a Cellular Response
Cytokine and chemokine signaling in the immune response
can be a strikingly complex and occasionally redundant
affair. Effector molecules such as cytokines can bind to more
Receptors and Signaling: Cytokines and Chemokines
than one receptor, and receptors can bind to more than one
signaling molecule. Nowhere is the latter concept more
clearly illustrated than in the chemokine system, in which
approximately 20 receptors bind to close to 50 distinct chemokines (see Appendix III). Effector molecule signaling can
also cooperate with signaling through antigen-specific
receptors. Signals received through more than one receptor
must then be integrated at the level of the biological
response, with multiple pathways acting to tune up or tune
down the expression of particular transcription factors or
the activity of particular enzymes. Thus, the actual biological
response mounted by a cell to a particular chemical signal
depends not only on the nature of the individual receptor for
that signal, but also on all of the downstream adapters and
enzymes present in the recipient cell.
|
CHAPTER 4
111
Six Families of Cytokines and
Associated Receptor Molecules
In recent years, immunologists have enjoyed an explosion of
information about new cytokines and cytokine receptors as a
result of advances in genomic and proteomic analysis.
Advances Box 4-1 describes a recently developed proteomic
approach to the search for new, secreted cytokines and illustrates the manner in which a sophisticated appreciation of the
molecular and cell biology of secretory pathways aids in the
identification of new cytokines. The purpose of this chapter is
not to provide an exhaustive list of cytokines and their receptors (see Appendices II and III for a comprehensive and current list of cytokines and chemokines), but rather to outline
BOX 4-1
ADVANCES
Methods Used to Map the Secretome
The related approaches of genomics
and proteomics provide scientists with
tools they can use to assess the complex
changes that occur in gene and protein
expression induced by stimuli, such as
antigen or cytokine stimulation. Vast
arrays of information regarding the derivation and readout of genes in different
cells and organisms, and the expression of
particular proteins, can be analyzed and
presented in ways not available to scientists just a few years ago.
Recently, the science of proteomics has
been extended to address the mapping of
proteins that are secreted by various cell
types. The array of proteins secreted by a
cell is referred to as its secretome, and the
secretome can be more formally defined
as the “proteins released by a cell, tissue, or
organisms through classical and nonclassical secretion mechanisms.”
Scientists first became interested in the
concept of the secretome as a way to diagnose and identify various types of cancer.
They reasoned that they could use the set
of proteins secreted into the serum or other
tissue fluids as a biological marker for spe-
cific tumor types. If particular proteins can
be shown to be secreted at high concentration only under conditions of malignancy,
then rapid and inexpensive tests can be
developed that have the potential to screen
for tumors at an early stage, when they are
still amenable to treatment. Although such
tumor-specific profiles of secreted proteins
are surprisingly difficult to develop, given
the range of mutations associated with the
generation of a cancer, the ability to diagnose a tumor at an early stage using only a
serum sample provides intense motivation,
and many such attempts are ongoing.
The approaches used to define a set of
cancer secretomes have since been
applied to studies of many other, nonmalignant cell populations for which the
description of a secretome would be a
useful analytical tool. These populations
include stem cells, cells of the immune
system, and adipose cells. Given the diversity of cytokines that can be secreted by a
single cell, and the manner in which the
activities of cytokines can interact at the
level of the target cell, cytokine biology is
a superb target for such a global approach.
A recent secretome analysis (Botto
et al., 2011) addressed the question of how
the human cytomegalovirus induces the
formation of new blood vessels (angiogenesis). Virus-free supernatant from
virus-infected endothelial cells was found
to induce angiogenesis. Secretome analysis of the infected endothelial cell supernatant revealed the presence of multiple
cytokines, including IL-8, GM-CSF, and IL-6.
The addition of a blocking anti-IL-6 antibody at the same time as the virus-free
supernatant was then shown to inhibit its
angiogenic activity, thus demonstrating
that it was the IL-6 activity in the supernatant that was primarily responsible for
inducing the new blood vessel growth.
One difficulty that is frequently
encountered in trying to analyze the secretome of a particular type of cell is the
need of many cells to grow in a tissue
culture fluid supplemented with serum,
which is itself a complex mixture of proteins. In this case it is important to distinguish between proteins released by the
cells under study and those which were
originally present in the serum. Several
(continued)
PA R T I
112
ADVANCES
|
Introduction
BOX 4-1
(continued)
techniques are available to discriminate
between secreted proteins and those
from the tissue culture media, including
adding inhibitors of secretion to some
cultures and then comparing those proteins present in the culture supernatant in
the presence and absence of inhibitors.
Alternatively, culturing the cells in the
presence of radioisotopes that only label
newly synthesized proteins, such as 35S
methionine, can be used to distinguish
these proteins from preexisting proteins
in the culture medium.
In the case, such as that described
above, that a cell line is being tested to
determine whether it secretes a set of
cytokines for which antibody assays
already exist, two different types of multiplex measurements may be used (see Figure 1). Both of these approaches utilize
antibodies to the array of cytokines to be
analyzed, attached to some sort of solid
phase support. This support may be glass,
a membrane, or a set of beads, with each
antibody attached to a bead of a different
color. The sample of tissue culture fluid is
added to the solid phase antibody, excess
fluid is washed away, and then biotinylated antibodies are added. (Biotin, a small
molecule, is used because it has an
extremely high affinity for a protein, streptavidin, and is therefore used to couple
two molecules together in assays such as
these. For more details, see Chapter 20.)
After antibody binding, the excess biotinylated antibodies are removed by washing and the cytokine concentrations are
assessed by the addition of fluorescent
streptavidin, which will bind to the biotin.
A fluorescent signal indicates the presence
of the cytokine in the sample, and the level
of the signal reveals its concentration
Since each bead fluoresces at a different
wavelength, the fluorescence associated
with each cytokine can be distinguished.
Various bioinformatics tools have been
developed that have particular application to secretome analysis. These include
SignalP, which identifies the presence of
signal peptides and also shows the location of signal peptide cleavage sites in
bacterial and eukaryotic proteins. In addition, SecretomeP can be used for the pre-
Planar arrays
a-IL-8 a-TGF-β
a-bFGF a-TNF-α
Bead assays
Panel of antibodies
recognizing cytokines,
chemokines and
growth factors
a-IL-8
a-bFGF
a-TGF-β
a-TNF-α
Antibodies immobilized
on beads that fluoresce
at different wavelengths
Antibodies printed
on glass or membrane
Incubation with sample
Incubation with
biotinylated antibodies
a-IL-8 a-TGF-β a-bFGF a-TNF-α
Incubation with
labelled streptavidin
Detection of signal,
data analysis
Fluorescent or
chemiluminescent detection
Fluorescent detection
FIGURE 1
Principle of planar and bead-based multiplex detection and quantitation of
cytokines, chemokines, growth factors, and other proteins. Assays use antibodies
against (a-) various cytokines, and biotin (yellow)-streptavidin (green) conjugation. See text for details.
[Adapted from H. Skalnikova et al., Mapping of the secretome of primary isolates of mammalian cells, stem cells and
derived cell lines, 2011, Proteomics 11:691.]
diction of nonclassically secreted proteins.
Several bioinformatics tools, including
TargetP and Protein Prowler, use the protein sequence to predict its subcellular
localization. Finally, Ingenuity Pathway
Analysis allows the investigator to search
for protein interaction partners and to
predict the involvement of the protein of
interest in functional networks.
Botto, S., D. N. Streblow, V. DeFilippis, L. White, C.
N. Kreklywich, P. P. Smith, and P. Caposio.
(2011). IL-6 in human cytomegalovirus secretome promotes angiogenesis and survival
of endothelial cells through the stimulation
of survivin. Blood 117:352–361.
Receptors and Signaling: Cytokines and Chemokines
TABLE 4-2
|
CHAPTER 4
113
Six Cytokine Families
Family name
Representative members of family
Comments
Interleukin 1 family
IL-1, IL-1, IL-1Ra, IL-18, IL-33
IL-1 was the first noninterferon cytokine to be identified.
Members of this family include important inflammatory
mediators.
Hematopoietin (Class I
cytokine) family
IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-12, IL-13, IL15,
IL-21, IL-23, GM–CSF, G-CSF, Growth hormone,
Prolactin, Erythropoietin/hematopoietin
This large family of small cytokine molecules exhibits
striking sequence and functional diversity.
Interferon (Class II
cytokine) family
IFN-, IFN-, IFN-, IL-10, IL-19, IL-20,
IL-22, IL-24
While the IFNs have important roles in anti-viral responses,
all are important modulators of immune responses.
Tumor Necrosis Factor
family
TNF-, TNF-, CD40L, Fas (CD95), BAFF,
APRIL, LT
Members of this family may be either soluble or membrane bound; they are involved in immune system development, effector functions, and homeostasis.
Interleukin 17 family
IL-17 (IL17-A), IL17B, C, D, and F
This is the most recently discovered family; members
function to promote neutrophil accumulation and activation, and are proinflammatory.
Chemokines (see
Appendix III)
IL-8, CCL19, CCL21, RANTES, CCL2 (MCP-1),
CCL3 (MIP-1)
All serve chemoattractant function.
some general principles of cytokine and receptor architecture
and function that should then enable the reader to place any
cytokine into its unique biological context.
Detailed studies of cytokine structure and function have
revealed common features among families of cytokines.
Cytokines are relatively small proteins and generally have a
molecular mass of less than 30 kDa. Many are glycosylated,
and glycosylation appears to contribute to cytokine stability,
although not necessarily to cytokine activity. Cytokines
characterized so far belong to one of six groups: the Interleukin 1 (IL-1) family, the Hematopoietin (Class I cytokine)
family, the Interferon (Class II cytokine) family, the Tumor
Necrosis Factor (TNF) family, the Interleukin 17 (IL-17)
family, and the Chemokine family (Table 4-2). Each of these
six families of cytokines, the receptors that engage them, and
the signaling pathways that transduce the message received
upon cytokine binding into the appropriate biological outcome are described in the following pages.
Cytokines of the IL-1 Family Promote
Proinflammatory Signals
Cytokines of the interleukin 1 (IL-1) family are typically
secreted very early in the immune response by dendritic cells
and monocytes or macrophages. IL-1 secretion is stimulated
by recognition of viral, parasitic, or bacterial antigens by
innate immune receptors. IL-1 family members are generally
proinflammatory, meaning that they induce an increase in
the capillary permeability at the site of cytokine secretion,
along with an amplification of the level of leukocyte migration into the infected tissues. In addition, IL-1 has systemic
(whole body) effects and signals the liver to produce acute
phase proteins such as the Type I interferons (IFNs and ),
IL-6, and the chemokine CXCL8. These proteins further
induce multiple protective effects, including the destruction
of viral RNA and the generation of a systemic fever response
(which helps to eliminate many temperature-sensitive bacterial strains). IL-1 also activates both T and B cells at the
induction of the adaptive immune response.
Cytokines of the IL-1 Family
Members of the IL-1 cytokine and receptor family are
shown in Figure 4-5. The canonical (most representative)
members of the IL-1 family, IL-1 and IL-1, are both synthesized as 31 kDa precursors, pro-IL-1 and pro-IL-1. Pro
IL-1 is biologically active, and often occurs in a membranebound form, whereas pro-IL-1 requires processing to the
fully mature soluble molecule before it can function. Pro-IL-1
and are both trimmed to their 17 kDa active forms by the
proteolytic enzyme caspase-1 inside the secreting cell. Active
caspase-1 is located in a complex set of proteins referred to
as the inflammasome (see Chapter 5).
Other IL-1 family members, IL-18 and IL-33, have also
been shown to be processed by caspase-1 in vitro (although
there is ambiguity as to whether IL-33 requires this processing for full activity in vivo). IL-18 is related to IL-1, uses the
same receptor family, and has a similar function; like IL-1,
IL-18 is expressed by monocytes, macrophages, and dendritic
cells and is secreted early in the immune response. In contrast, IL-33 is constitutively expressed in smooth muscle and
in bronchial epithelia, and its expression can be induced by
IL-1 and TNF- in lung and skin fibroblasts. IL-33 has been
shown to induce TH2 cytokines that promote T-lymphocyte
interactions with B cells, mast cells, and eosinophils. IL-33
PA R T I
114
|
Introduction
has also been implicated in the pathology of diseases such as
asthma and inflammatory airway and bowel diseases.
Two additional members of this cytokine family act as
natural inhibitors of IL-1 family function. The soluble protein
IL-1Ra (IL-1 Receptor antagonist) binds to the IL-1RI receptor, but prevents its interaction with its partner receptor
chain, IL-1RAcP, thus rendering it incapable of transducing
a signal to the interior of the cell. IL-1Ra therefore functions
as an antagonist ligand of IL-1. IL-18BP adopts a different
strategy of inhibition, binding to IL-18 in solution and preventing IL-18 from interacting productively with its receptor. The inhibitory effect of IL-18BP is enhanced by the
further binding of IL-1F7 (see Figure 4-5b).
The Interleukin 1 family of receptors includes the receptors
for IL-1, IL-18, and IL-33. Both forms of IL-1—IL-1 and
IL-1—bind to the same receptors and mediate the same
responses. Two different receptors for IL-1 are known, and
both are members of the immunoglobulin superfamily of
proteins (see Chapter 3). Only the type I IL-1R (IL-1RI),
which is expressed on many cell types, is able to transduce a
cellular signal; the type II IL-1R (IL-1RII) is limited to B cells
and is inactive. For full functioning, the Type 1 IL-1R also
requires the presence of an interacting accessory protein, IL1RAcP (IL-1 Receptor Accessory Protein (see Figure 4-5a).
(a)
(b)
The IL-1 Family of Cytokine Receptors
IL-18
Inhibitory ligand and receptors
IL-1Ra
IL-1
Inhibitory ligands
S
S
IL-1RI
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
IL-1RAcP
S
S
S
S
S
S
IL-18Rα
S
S
sIL-1RII
S
S
S
S
S
S
S
S
S
S
S
S
S
S
IL-18Rβ
IL-18
sIL-1RAcP
Blocked binding
S
S
Cytoplasm
Cytoplasm
TIR
domains
IL-18BP
IL-1F7
TIR
domains
IL-1RII
Signal
Signal
(c)
IL-33
Inhibitory receptors
T1/ST2
S
S
S
S
S
S
S
S
S
S
S
S
Cytoplasm
IL-1RAcP
TIR
domains
Signal
S
S
S
S
S
S
S
S
S
S
S
S
sST2
sIL-1RAcP
FIGURE 4-5 Ligands and receptors of the IL-1 family. (a) The
two agonist ligands, IL-1 and IL-1, are represented by IL-1 and the
antagonist ligand by IL-1Ra. The IL-1 receptor, IL-1RI, has a long cytoplasmic domain and, along with IL-1RAcP, activates signal transduction pathways. IL-1Ra functions as an IL-1 inhibitor by binding to
IL-1RI while not allowing interaction with IL-1RAcP. IL-1RII does not
activate cells but functions as an IL-1 inhibitor both on the plasma
membrane and in the cell microenvironment as a soluble receptor
(sIL-1RII). IL-1RAcP can also inhibit IL-1 signals by cooperating with
IL-1RII in binding IL-1 either on the plasma membrane or as a soluble
molecule (sIL-1RAcP). (b) IL-18 binds to the IL-18R chain, and this
complex then engages the IL-18R chain to initiate intracellular signals. The soluble protein IL-18BP functions as an inhibitor of IL-18 by
binding this ligand in the fluid phase, preventing interaction with the
IL-18R chain. IL-1F7 appears to enhance the inhibitory effect of
IL-18BP. (c) IL-33 binds to the T1/ST2 receptor, and this complex
engages the IL-1RAcP as a co-receptor. A soluble form of ST2 (sST2)
may function as an inhibitor of IL-33 by binding IL-33 in the cell microenvironment and sIL-1RAcP may enhance the inhibitory effects of sST2.
|
Receptors and Signaling: Cytokines and Chemokines
Note that both the IL-1RI and the IL-1RII receptor chains
as well as the receptor accessory protein exist in both soluble
and membrane-bound forms. However, a full signal is transmitted only from the dimer of the membrane-bound forms of
IL-1RI and IL-1RAcP. The alternative, membrane-bound
and soluble forms of IL-1 binding proteins, can “soak up”
excess cytokine, but they are unable to transduce the interleukin signal. Thus, by secreting more or fewer of these inactive receptors, at different times during an immune response,
the organism has the opportunity to fine-tune the cytokine
signal by allowing the inactive and soluble receptors to compete with the signal-transducing receptor for available cytokine. This theme finds echoes in other immune system
receptor families, and appears to be a frequently evolved
strategy for controlling the strength of signals that give rise
to important outcomes. In the case of IL-1, the ultimate
result of successful IL-1 signaling is a global, proinflammatory state, and so the penalty paid by the host for an inappropriately strong IL-1 response would be physiologically
significant and even potentially fatal.
The receptor for IL-18 is also a heterodimer, made up of
IL-18R and IL-18R. IL-33 is recognized by the IL-1RAcP
in combination with a novel receptor protein, variously
termed T1/ST-2 or IL-1RL1. As for IL-1, inhibitory receptors
exist for IL-33 (see Figure 4-5c).
Signaling from IL-1 Receptors
Productive ligand binding to the extracellular portion of the
IL-1 receptor leads to a conformational alteration in its cytoplasmic domain. This structural alteration in the receptor
leads to a series of downstream signaling events (Figure 4-6).
Most of the themes of these events will be familiar to the
reader from Chapter 3, and we will encounter them again in
the discussion of innate immune receptors in Chapter 5.
First, binding of the adapter protein MyD88 to the occupied receptor allows recruitment to the receptor complex of
one or more members of the IL-1 Receptor Activated
Kinase (IRAK) protein family. One of these, IRAK-4, is
activated by autophosphorylation and phosphorylates its fellow IRAKs, resulting in the generation of binding sites for
TNF Receptor Associated Factor 6 (TRAF6), which is associated with a ubiquitin-ligase complex capable of generating
polyubiquitin chains. The IRAK-TRAF6 complex now dissociates from the receptor and interacts with a preformed
cytosolic complex made up of the kinase TGFAssociated
Kinase 1 (TAK1) and two TAK1-Binding proteins, TAB1
and TAB2. Binding of polyubiquitin chains to the TAB proteins in the TAK1 complex activates it.
The TAK1 complex now performs two functions with
which the reader should be familiar. It phosphorylates and
activates the IKK complex, leading to the destruction of IB
and the resultant activation of the transcription factor NF-B
(see Figure 3-17). In addition, TRAF6 also plays a role in
IKK activation by providing ubiquitination sites to which
the NEMO component of IKK can bind, resulting in its further activation. TAK1 also activates downstream members
CHAPTER 4
115
IL-1
IL-1RI
S
S
S
S
S
S
S
S
S
S
S
S
IL-1RAcP
Cytoplasm
TIR domains
MyD88
IRAKs
TRAF6
TAB1
Ub
TAB2
TAK1
MAP kinase
cascade
NF-κB
pathway
AP-1
NF-κB
Cytokine expression
FIGURE 4-6 Signaling from members of the IL-1 receptor
family. IL-1 binding to its receptor induces a conformational alteration in the receptor’s Toll-IL-1R (TIR) domain that allows binding of the
adapter protein MyD88 via its TIR domain. MyD88 recruits one or
more IL-1 receptor activated kinases (IRAKs) to the receptor complex,
which phosphorylate one another providing binding sites for TRAF6.
The IRAK-TRAF6 complex dissociates from the receptor complex and
interacts with the cytoplasmic protein TAK1 and its two binding proteins, TABs 1 and 2. TRAF6, together with a ubiquitin-ligase complex,
catalyzes the generation of polyubiquitin chains that activate the
TAK1 complex. TAK1 activates downstream events leading to the
activation and nuclear localization of the transcription factor NF-B.
TAK1 also activates downstream members of the MAP kinase cascade, leading to activation of the AP-1 transcription factor.
of the MAP kinase cascade, which in turn activate the AP-1
transcription factor (see Figure 3-16). Binding of IL-1 family
cytokines to their receptors thereby leads to a global alteration in the transcription patterns of the affected cells, which
in turn results in the up-regulation of proinflammatory cytokines and adhesion molecules.
116
PA R T I
|
Introduction
Hematopoietin (Class I) Family Cytokines Share
Three-Dimensional Structural Motifs, but
Induce a Diversity of Functions in Target Cells
Members of the hematopoietin (Class I) cytokine family
are small, soluble cytokines that communicate between and
among cells of the immune system. Their name is somewhat
misleading in that not all members of this family are implicated in hematopoietic (blood-cell forming) functions per
se. However, some of the earliest members of this family to
be characterized indeed have hematopoietic functions, and
the cytokine family was then defined on the basis of structural similarities among all the participants. Because the
hematopoietin family contains some of the earliest cytokines
to be structurally characterized, it is sometimes also referred
to as the Class I cytokine family.
Cytokines of the Hematopoietin (Class I) Family
As more hematopoietin family members have been defined,
it has become clear that their cellular origins and target cells
are as diverse as their ultimate functions, which range from
signaling the onset of T- and B-cell proliferation (e.g., IL-2),
to signaling the onset of B-cell differentiation to plasma cells
and antibody secretion (e.g., IL-6), to signaling the differentiation of a T helper cell along one particular differentiation
pathway versus another (e.g., IL-4 vs. IL-12) and, finally, to
initiating the differentiation of particular leukocyte lineages
(e.g., GM-CSF, G-CSF). Appendix II lists the cytokines
described in this book, along with their cells of origin, their
target cells, and the functions they induce.
Significant homology in the three-dimensional structure
of hematopoietin family cytokines defines them as members
of a single protein family, despite a relatively high degree of
amino acid sequence diversity. The defining structural feature of this class of cytokines is a four-helix bundle motif,
organized into four anti-parallel helices (Figure 4-7). Members of this family can then be further subclassified on the
basis of helical length. Cytokines such as IL-2, IL-4, and IL-3
typically have short helices of 8 to 10 residues in length. In
contrast, the so-called long-chain cytokines, which include
IL-6 and IL-12, typically have helical lengths of 10 to 20
residues.
The Hematopoietin or Class I Receptor Family
Most hematopoietin cytokine receptors include two types of
protein domains: an immunoglobulin-like domain, made
up of sheets, as described in Chapter 3, and domains that
bear structural homology to the FNIII domain of the extracellular matrix protein fibronectin. Binding sites for most
cytokines are to be found in a structure made up of two,
tandem (side-by-side) FNIII domains referred to as
Cytokine-binding Homology Regions (CHRs). As we will
see, the CHR motif is common to cytokine receptors from
several families.
A feature common to most of the Hematopoietin and Interferon cytokine receptor families is the presence of multiple
subunits. Table 4-3 lists the three subfamilies of hematopoietin
(a)
36
114
30
α
42
73
80
58
D
133
C
A
6
C
96
B
S S
105
52
N
(b)
FIGURE 4-7 The four-helix bundle is the defining structural feature of the Hematopoietin family of cytokines. Structure of interleukin 2—the defining member of the Hematopoietin
family—showing the four -helices of the hematopoietin cytokines
point in alternating directions. (a) Topographical representation of
the primary structure of IL-2 showing -helical regions ( and A-D)
and connecting chains of the molecule. (b) Ribbon representation of
the crystallographic structure of human Il-2. [Part (b) PDB ID 1M47.]
receptors, each subfamily being defined by a receptor subunit
that is shared among all members of that family.
The -Chain Bearing, or IL-2 Receptor, Subfamily
Expression of a common chain defines the IL-2 receptor
subfamily, which includes receptors for IL-2, IL-4, IL-7, IL-9,
IL-15, and IL-21. The IL-2 and the IL-15 receptors are heterotrimers, consisting of a cytokine-specific chain and two
chains— and —responsible for both signal transduction
and cytokine recognition. The IL-2 receptor chain also
functions as the signal-transducing subunit for the other
receptors in this subfamily, which are all dimers. Congenital
X-linked severe combined immunodeficiency (XSCID)
results from a defect in the -chain gene, which maps to the
X chromosome. The immunodeficiency observed in this
disorder, which includes deficiencies in both T-cell and
|
Receptors and Signaling: Cytokines and Chemokines
TABLE 4-3
Subfamilies of hematopoietin
family cytokine receptors share
common subunits
Common cytokine
receptor subunit
Cytokines recognized by receptors
bearing that common subunit
IL-2, IL-4, IL-7, IL-9, IL-15, IL-12
IL-3, IL-5, GM-CSF
gp130
IL-6, IL-11, LIF, OSM, CNTF, IL-27
(a)
Intermediate
affinity, IL-2R
High affinity,
IL-2R
Dissociation
constant (Kd):
Cells
expressed by:
(b)
α
β
IL-2Rβ
IL-2Rγ
IL-2Rα
IL-2Rβ
IL-2Rγ
IL-2Rα
10−9 M
10−11 M
10−8 M
NK cells
Resting T cells
(low numbers)
Activated CD4+ and CD8+ T cells
Activated B cells
(low numbers)
IL-2Rα
IL-2
Low affinity,
IL-2R
γ
β
Subunit
composition:
117
NK-cell activity, results from the loss of all the cytokine
functions mediated by the IL-2 subfamily receptors.
The IL-2 receptor occurs in three forms, each exhibiting a
different affinity for IL-2: the low-affinity monomeric IL-2R
(CD25) (which can bind to IL-2, but is incapable of transducing
a signal from it), the intermediate-affinity dimeric IL-2R
(which is capable of signal transduction), and the high-affinity
trimeric IL-2R (which is responsible for most physiologically relevant IL-2 signaling) (Figure 4-8a). A recent x-ray
crystallographic structure of the high-affinity trimeric form of
the IL-2 receptor with an IL-2 molecule in its binding site
reveals that IL-2 binds in a pocket formed by the and chains
α
γ
CHAPTER 4
IL-2Rα
γc
IL-2
IL-2Rβ
IL-2Rβ
FIGURE 4-8 Comparison of the three forms of the IL-2
receptor. (a) Schematic of the three forms of the receptor and listing
of dissociation constants and properties for each. Signal transduction
is mediated by the and chains, but all three chains are required
for high-affinity binding of IL-2. (b) Three-dimensional structure of the
γc
three-chain form of the IL-2 receptor with bound IL-2 (views rotated
by 90). Note that the chain completes the pocket to which IL-2
binds, accounting for the higher affinity of the trimeric form. [From X.
Wang, M. Rickert, and K. C. Garcia, 2005, Structure of the quaternary complex of
interleukin–2 with its , , and c receptors. Science 310:1159.]
118
PA R T I
|
Introduction
(Figure 4-8b). Important additional contacts with the IL-2
ligand are contributed when the chain is present, accounting
for the higher affinity of binding by the trimer.
The expression of the three chains of the IL-2 receptor
varies among cell types and in different activation states. The
intermediate affinity () IL-2 receptors are expressed on
resting T cells and on NK cells, whereas activated T and B
cells express both the low-affinity () and the high-affinity
() receptor forms (see Figure 4-8a). Since there are
approximately ten times as many low-affinity as high-affinity
receptors on activated T cells (50,000 vs. 5000), one must ask
what the function of the low-affinity receptor might be, and
two possible ideas have been advanced. It may serve to concentrate IL-2 onto the recipient cell surface for passage to the
high-affinity receptor. Conversely, it may reduce the local
concentration of available IL-2, ensuring that only cells bearing the high-affinity receptor are capable of being activated.
Whatever the answer to this question may be, the restriction
of the high-affinity IL-2 receptor expression to activated T
cells ensures that only antigen-activated CD4 and CD8 T
cells will proliferate in response to physiologic levels of IL-2.
The Chain Bearing, or GM-CSF, Receptor Subfamily
Members of the GM-CSF receptor subfamily, which includes
the receptors for IL-3, IL-5, and GM-CSF, share the signaling subunit. Each of these cytokines binds with relatively low
affinity to a unique, cytokine-specific receptor protein, the
subunit of a dimeric receptor. All three low-affinity subunits
associate noncovalently with the common signal-transducing
subunit. The resulting dimeric receptor has a higher
affinity for the cytokine than the specific chain alone, and
is also capable of transducing a signal across the membrane
upon cytokine binding (Figure 4-9a).
IL-3, IL-5, and GM-CSF exhibit redundant activities. IL-3
and GM-CSF both act on hematopoietic stem cells and progenitor cells, activate monocytes, and induce megakaryocyte
differentiation, and all three of these cytokines induce
eosinophil proliferation and basophil degranulation with
release of histamine.
Since the receptors for IL-3, IL-5, and GM-CSF share a common signal-transducing subunit, each of these cytokines
would be expected to transduce a similar activation signal,
accounting for the redundancy seen among their biological
effects, and indeed, all three cytokines induce the same patterns
of protein phosphorylation upon cell activation. However,
when introduced simultaneously to a cell culture, IL-3 and GMCSF appear to antagonize one another; the binding of IL-3 is
inhibited by GM-CSF, and binding of GM-CSF is inhibited by
IL-3. This antagonism is caused by competition for a limited
number of subunits available to associate with the cytokinespecific subunits of the dimeric receptors (Figure 4-9b).
The gp130 Receptor Subfamily
The importance of the gp130 cytokine receptor family to the
development and health of the individual is underscored by
the results of deletion studies which have demonstrated that
(a)
IL-3
IL-5
GM-CSF
Low-affinity
receptors
Exterior
Membrane
α
α
α
β subunit
α
α
α
Interior
High-affinity
receptors
β
(b)
β
β
IL-3
α α
GM-CSF
α α α α
β
α
α
β
α
β
β β
α α α
β
FIGURE 4-9 Interactions between cytokine-specific subunits
and a common signal-transducing subunit of the -chain family
of cytokine receptors. (a) Schematic diagram of the low-affinity and
high-affinity receptors for IL-3, IL-5, and GM-CSF. The cytokine-specific
subunits exhibit low-affinity binding and cannot transduce an activation signal. Noncovalent association of each subunit with a common
subunit yields a high-affinity dimeric receptor that can transduce a
signal across the membrane. (b) Competition of ligand-binding chains
of different receptors for a common subunit can produce antagonistic
effects between cytokines. Here binding of IL-3 by subunits of the IL-3
receptor allows them to outcompete chains of the GM-CSF receptor
for subunits. [Part a adapted from T. Kishimoto et al., 1992, Interleukin-6 and its
receptor: A paradigm for cytokines, Science 258:593.]
the targeted disruption of gp130 in mice during embryonic
development is lethal. Receptors in this family include those
for IL-6, important in the initiation of the immune response,
and IL-12, critical for signaling differentiation of helper T
cells along the TH1 pathway. Targeted disruption of individual cytokine receptors, as well as of the cytokines themselves,
has provided much functional information about signaling
via these cytokine family members.
Cytokine specificity of the gp130 family of receptors is
determined by the regulated expression of ligand-specific
chains in dimers, or trimers, with the gp130 component. The
Receptors and Signaling: Cytokines and Chemokines
gp130 subunit family of cytokine receptors is further subdivided into receptors specific for monomeric cytokines, such as
IL-6, and those which bind the dimeric cytokines, such as IL-12.
Because the Hematopoietin (Class I) and Interferon
(Class II) cytokine receptor families utilize similar signaling
pathways, we will first describe the Interferons and their
receptors and then consider the signaling pathways used by
the two families together.
The Interferon (Class II) Cytokine Family Was
the First to Be Discovered
In the late 1950s, investigators studying two different viral
systems in two laboratories half a world apart almost simultaneously discovered interferons. Yasu-Ichi Nagano and Yashuhiko Kojima, Japanese virologists, were using a rabbit skin
and testes tissue culture model to develop a vaccine against
smallpox. They noted that immunization with a UV-inactivated
form of the cowpox virus resulted in the localized inhibition
of viral growth, following a subsequent injection of the same
virus. Viral growth inhibition was restricted to a small area of
skin close to the site of the original immunization, and the
scientists postulated that the initial injection had resulted in
the production of a “viral inhibitory factor.” After showing
that their “inhibitory factor” was not simply antibody, they
published a series of papers about it. With hindsight, scientists
now believe that their protective effect was mediated by interferons. However, the technical complexity of their system, and
the fact that their papers were published in French, rather
than in English, delayed the dissemination of their findings to
the broader scientific community.
Meanwhile, in London, Alick Isaacs and Jean Lindenman
were growing live influenza virus on chick egg chorioallantoic
membranes (a method that is still used today), and noticed
that exposure of their membranes to a heat-inactivated form
of influenza interfered with subsequent growth of a live virus
preparation on that surface preparation. They proved that the
growth inhibition resulted from the production of an inhibiting molecule by the chick membrane. They named it “interferon” because of its ability to “interfere” with the growth of
the live virus. Their more straightforward in vitro assay system
enabled them to rapidly characterize the biological effects of
the molecule involved, and they wrote a series of papers
describing the biological effects of interferon(s) in the late
1950s. However, since interferons are active at very low concentrations, it was not until 1978 that they were produced in
quantities sufficient for biochemical and crystallographic
analysis. Since that time, investigators have shown that there
are two major types of interferons, Types 1 and 2, and that
Type 1 interferons can be subdivided into two subgroups.
Interferons
Type I interferons are composed of Interferons , a family of
about 20 related proteins, and interferon-, which are secreted
by activated macrophages and dendritic cells, as well as by
virus-infected cells. Interferons and are also secreted by
|
CHAPTER 4
119
virally infected cells after recognition of viral components by
pattern recognition receptors (PRRs) located either at the
cell surface, or inside the cell (see Chapter 5). Intracellular
PRRs may interact with virally derived nucleic acids or with
endocytosed viral particles. The secreted Type I interferons
then interact in turn with membrane-bound interferon
receptors on the surfaces of many different cell types. The
results of their interaction with these receptors are discussed
in detail in Chapter 5, but they include the induction of ribonucleases that destroy viral (and cellular) RNA, and the cessation of cellular protein synthesis. Thus, interferons prevent
virally infected cells from replicating and from making new
viral particles. However, they simultaneously inhibit normal
cellular functions and destroy virally infected cells so that the
infection cannot spread.
Type I interferons are dimers of 18 to 20 kDa polypeptides,
predominantly helical in structure, and some members of this
family are naturally glycosylated. Type I interferons are used
in the treatment of a variety of human diseases, most notably
hepatitis infections.
Type II interferon, otherwise known as interferon-, is
produced by activated T and NK cells and is released as a
dimer (Figure 4-10). Interferon- is a powerful modulator of
FIGURE 4-10 The complex between IFN- and the ligandbinding chains of its receptor. This model is based on the x-ray
crystallographic analysis of a crystalline complex of interferon- (dark
and light purple) bound to ligand-binding chains of the receptors
(green and yellow). Note that IFN- is shown in its native dimeric
form; each member of the dimer engages the chain of an IFN-
receptor, thereby bringing about receptor dimerization and signal
transduction. [From M. R. Walter et al., 1995, Crystal structure of a complex
between interferon- and its soluble high-affinity receptor. Nature 376:230, courtesy M. Walter, University of Alabama.]
120
PA R T I
|
Introduction
CLINICAL FOCUS
Therapy with Interferons
Interferons are an extraordinary
group of proteins with important effects
on the immune system. Their actions
affect both the adaptive and the innate
arms of the immune system and include
the induction of increases in the expression of both Class I and Class II MHC molecules and the augmentation of NK-cell
activity. Cloning of the genes that encode
IFN-, IFN-, and IFN- has made it possible for the biotechnology industry to
produce large amounts of each of these
interferons at costs that make their clinical
use practical (Table 1).
IFN- (also known by its trade names
Roferon and Intron A) has been used for
the treatment of hepatitis C and hepatitis
B. It has also been found useful in a number of different applications in cancer
therapy. A type of B-cell leukemia known
as hairy-cell leukemia (because the cells
are covered with fine, hairlike cytoplasmic
projections) responds well to IFN-.
Chronic myelogenous leukemia, a disease
characterized by increased numbers of
granulocytes, begins with a slowly developing chronic phase that changes to an
accelerated phase and terminates in a
blast phase, which is usually resistant to
treatment. IFN- is an effective treatment
for this leukemia in the chronic phase
(70% response rates have been reported),
and some patients (as many as 20% in
some studies) undergo complete remission. Kaposi’s sarcoma, the cancer most
often seen in AIDS patients in the United
States, also responds to treatment with
IFN-, and there are reports of a trend
toward longer survival and fewer oppor-
tunistic infections in patients treated with
this agent. Most of the effects mentioned
above have been obtained in clinical
studies that used IFN- alone, but certain
applications such as hepatitis C therapy
commonly use it with an antiviral drug
such as ribavirin. The clearance time of
IFN- is lengthened by using it in a form
complexed with polyethylene glycol
(PEG) called pegylated interferon.
IFN- has emerged as the first drug
capable of producing clinical improvement in multiple sclerosis (MS). Young
adults are the primary target of this autoimmune neurologic disease, in which
nerves in the central nervous system
(CNS) undergo demyelination. This results
in progressive neurologic dysfunction,
leading to significant and, in many cases,
severe disability. This disease is often characterized by periods of nonprogression
and remission alternating with periods of
relapse. Treatment with IFN- provides
longer periods of remission and reduces
the severity of relapses. Furthermore,
magnetic resonance imaging (MRI) studies of CNS damage in treated and
untreated patients revealed that MSinduced damage was less severe in a
group of IFN-treated patients than in
untreated ones.
IFN- has been used, with varying
degrees of success, to treat a variety of
malignancies that include non-Hodgkin’s
lymphoma, cutaneous T-cell lymphoma,
and multiple myeloma. A more successful
clinical application of IFN- in the clinic is
in the treatment of the hereditary immunodeficiency chronic granulomatous dis-
the adaptive immune response, biasing T cell help toward
the TH1 type and inducing the activation of macrophages,
with subsequent destruction of any intracellular pathogens
and the differentiation of cytotoxic T cells. All three interferons increase the expression of MHC complex proteins on the
surface of cells, thus enhancing their antigen-presentation
capabilities.
ease (CGD; see Chapter 18). CGD features
a serious impairment of the ability of
phagocytic cells to kill ingested microbes,
and patients with CGD suffer recurring
infections by a number of bacteria (Staphylococcus aureus, Klebsiella, Pseudomonas,
and others) and fungi such as Aspergillus
and Candida. Before interferon therapy,
standard treatment for the disease
included attempts to avoid infection,
aggressive administration of antibiotics,
and surgical drainage of abscesses. A failure to generate microbicidal oxidants
(H2O2, superoxide, and others) is the basis
of CGD, and the administration of IFN-
significantly reverses this defect. Therapy
of CGD patients with IFN- significantly
reduces the incidence of infections. Also,
the infections that are contracted are less
severe, and the average number of days
spent by patients in the hospital is
reduced.
IFN- has also been shown to be effective in the treatment of osteopetrosis (not
osteoporosis), a life-threatening congenital
disorder characterized by overgrowth of
bone that results in blindness and deafness. Another problem presented by this
disease is that the buildup of bone
reduces the amount of space available for
bone marrow, and the decrease in hematopoiesis results in fewer red blood cells
and anemia. The decreased generation of
white blood cells causes an increased
susceptibility to infection.
The use of interferons in clinical
practice is likely to expand as more is
learned about their effects in combination with other therapeutic agents.
Interferon- is used medically to bias the adaptive immune
system toward a cytotoxic response in diseases such as leprosy and toxoplasmosis, in which antibody responses are less
effective than those that destroy infected cells. Clinical Focus
Box 4-2 describes additional roles of interferons in the clinic.
Minority members of the Interferon family of cytokines
include IL-10, secreted by monocytes and by T, B, and
Receptors and Signaling: Cytokines and Chemokines
|
CHAPTER 4
121
BOX 4-2
TABLE 1
Cytokine-based therapies in clinical use
Agent
Nature of agent
Clinical application
Enbrel
Chimeric TNF-receptor/IgG
constant region
Rheumatoid arthritis
Remicade or Humira
Monoclonal antibody against
TNF- receptor
Rheumatoid arthritis, Crohn’s disease
Roferon
Interferon--2a*
Hepatitis B, Hairy-cell leukemia, Kaposi’s sarcoma, Hepatitis C†
Intron A
Interferon-–2b
Melanoma
Betaseron
Interferon-–1b
Multiple sclerosis
Avonex
Interferon-–1a
Multiple sclerosis
Actimmune
Interferon-–1b
Chronic granulomatous disease (CGD), Osteopetrosis
Neupogen
G-CSF (hematopoietic
cytokine)
Stimulates production of neutrophils; reduction of infection in
cancer patients treated with chemotherapy, AIDS patients
Leukine
GM-CSF (hematopoietic
cytokine)
Stimulates production of myeloid cells after bone marrow
transplantation
Neumega or Neulasta
Interleukin 11 (IL-11), a
hematopoietic cytokine
Stimulates production of platelets
Epogen
Erythropoietin (hematopoietic
cytokine)
Stimulates red-blood-cell production
Ankinra (kineret)
Recombinant IL-1Ra
Rheumatoid arthritis
Daclizumab (Zenapax)
Humanized monoclonal
antibody against IL-2R
Prevents rejection after transplantation
Basiliximab (Simulect)
Human/mouse chimeric
monoclonal antibody against
IL-2R
Prevents transplant rejection
*Interferon-–2a is also licensed for veterinary use to combat feline leukemia.
†
Normally used in combination with an antiviral drug (ribavirin) for hepatitis C treatment.
Although interferons, in common with
other cytokines, are powerful modifiers
of biological responses, the side effects
accompanying their use are fortunately
relatively mild. Typical side effects include
flu-like symptoms, such as headache,
fever, chills, and fatigue. These symptoms
can largely be managed with acetaminophen (Tylenol) and diminish in intensity
during continued treatment. Although
dendritic cells that regulates immune responses. IL-10 shares
structural similarities with interferon-, and these similarities enable it to bind to the same class of receptors. In addition, a third class of interferons, the so-called interferon- ,
or type III Interferon family, was discovered in 2003. There
are currently three members of this family: interferon- 1
(IL-29), interferon- 2 (IL-28A), and interferon- 3 (IL-28B).
interferon toxicity is usually not severe,
treatment is sometimes associated with
serious manifestations such as anemia
and depressed platelet and white-bloodcell counts.
Like Type I interferons, the Type III interferons up-regulate
the expression of genes controlling viral replication and host
cell proliferation.
Interferon Receptors
Members of the Interferon receptor family are heterodimers
that share similarly located, conserved cysteine residues with
122
PA R T I
|
Introduction
members of the Hematopoietin receptor family. Initially,
only interferon-, -, and - were thought to be ligands for
these receptors. However, recent work has shown that the
receptor family consists of 12 receptor chains that, in their
various assortments, bind no fewer than 27 different cytokines, including six members of the IL-10 family, 17 Type I
interferons, one Type II interferon, and three members of
the recently described interferon– family, including IL-28A,
IL-28B, and IL-29.
The JAK-STAT Signaling Pathway
Early experiments in cytokine signaling demonstrated that a
series of protein tyrosine phosphorylations rapidly followed
the interaction of a cytokine with a receptor from the Class I
or Class II cytokine receptor families. These results were
initially puzzling, since the cytokine receptors lack the
immunotyrosine activation motifs (ITAMs) characteristic of
B- and T-cell receptors. However, studies of the molecular
events triggered by binding of interferon gamma (IFN-) to
its receptor shed light on the mode of signal transduction
used by members of both the Hematopoietin and Interferon
cytokine families.
In the absence of cytokine, the receptor subunits are
associated only loosely with one another in the plane of the
membrane, and the cytoplasmic region of each of the
receptor subunits is associated noncovalently with inactive
tyrosine kinases named Janus Activated Kinases (JAKs).
(Some members of this family of kinases retain their earlier
name of Tyk, but share structural and functional properties
with the JAK family of kinases.) The process of signal
transduction from Class I and Class II cytokine receptors
has been shown to proceed according to the following steps
(Figure 4-11):
•
Cytokine binding induces the association of the two separate cytokine receptor subunits and activation of the
receptor-associated JAKs.
•
The receptor-associated JAKs phosphorylate specific
tyrosines in the receptor subunits.
•
These phosphorylated tyrosine residues serve as docking
sites for inactive transcription factors known as Signal
Transducers and Activators of Transcription (STATs).
•
The inactive STATs are phosphorylated by JAK and Tyk
kinases.
•
Phosphorylated STAT transcription factors dimerize,
binding to one another via SH2/phosphotyrosine
interactions.
•
Phosphorylation also results in a conformational change in
the STAT dimer that reveals a nuclear localization signal.
•
The STAT dimer translocates into the nucleus, where it
initiates the transcription of specific genes.
Currently, we know of seven STAT proteins (STAT 1-4,
5A, 5B, and 6) and four JAK proteins (JAK 1–3 and Tyk2) in
Cytokine
α
β
β
α
Dimerization
of receptor
P
JAK
P
P
SH2
P
P
Activation of JAK
family tyrosine kinases,
phosphorylation of receptor
P
Tyrosine phosphorylation of
STAT by JAK kinase
STAT
P
P
Dimerization
of STAT
P
P
DNA
Specific gene transcription
FIGURE 4-11 General model of signal transduction mediated by most Class I and Class II cytokine receptors. Binding of
a cytokine induces dimerization of the receptor subunits, which leads
to the activation of receptor subunit-associated JAK tyrosine kinases
by reciprocal phosphorylation. Subsequently, the activated JAKs
phosphorylate various tyrosine residues, resulting in the creation of
docking sites for STATs on the receptor and the activation of one or
more STAT transcription factors. The phosphorylated STATs dimerize
and translocate to the nucleus, where they activate transcription of
specific genes.
mammals. Specific STATs play essential roles in the signaling
pathways of a wide variety of cytokines (Table 4-4).
Given the generality of this pathway among Class I and
Class II cytokines, how does the immune system induce a
specific response to each cytokine? First, there is exquisite
specificity in the binding of cytokines to their receptors. Secondly, particular cytokine receptors are bound to specific
partner JAK enzymes that in turn activate unique STAT transcription factors. Third, the transcriptional activity of activated STATs is specific because a particular STAT homodimer
or heterodimer will only recognize certain sequence motifs
and thus can interact only with the promoters of certain
genes. Finally, only those target genes whose expression is
permitted by a particular cell type can be activated within
that variety of cell. For example, promoter regions in some
cell types may be caught up in heterochromatin and
Receptors and Signaling: Cytokines and Chemokines
TABLE 4-4
|
CHAPTER 4
123
STAT and JAK interaction with selected cytokine receptors during signal transduction
Each cytokine receptor must signal through a pair of Janus kinases. The JAKs may operate as either homo- or heterodimers.
Cytokine receptor
Janus kinase
STAT
IFN-/-
JAK 1, Tyk 2*
STATs 1 and 2
IFN-
JAK 1, JAK 2
STAT 1
IL-2
JAK 1, JAK 3
Mainly STATs 3 and 5. Also STAT 1.
IL-4
JAK 1, JAK 3
Mainly STAT 6. Also STAT 5.
IL-6
JAK 1, JAK 2
STAT 3
IL-7
JAK 1, JAK 3
STATs 5 and 3
IL-12
JAK 2, Tyk2
STATS 2, 3, 4, and 5
IL-15
JAK 1, JAK 3
STAT 5
IL-21
JAK 1, JAK 3
Mainly STATs 1 and 3; also STAT 5
* Despite its name, Tyk2 is also a Janus kinase.
inaccessible to transcription factors. In this way, the Class I
cytokine IL-4 can induce one set of genes in T cells, another
in B cells, and yet a third in eosinophils.
JAK-STAT pathways are not unique to the immune system. Among the many genes known to be regulated by
mammalian STAT proteins are those encoding cell survival
factors such as the Bcl-2 family members, those involved in
cell proliferation such as cyclin D1 and myc, and those implicated in angiogenesis or metastasis such as vascular endothelial growth factor, or VEGF.
At the close of cytokine signaling, negative regulators of
the STAT pathway, such as protein inhibitor of activated
STAT (PIAS), suppressor of cytokine signaling (SOCS), and
protein tyrosine phosphatases are believed to be responsible
for turning off JAK-STAT signaling and returning the cell to
a quiescent state.
Members of the TNF Cytokine Family Can
Signal Development, Activation, or Death
The Tumor Necrosis Family (TNF) family of cytokines
regulates the development, effector function, and homeostasis of cells participating in the skeletal, neuronal, and
immune systems, among others.
Cytokines of the TNF Family Can Be Soluble
or Membrane Bound
TNF-related cytokines are unusual in that they are often
firmly anchored into the cell membrane. Generally they are
Type 2 transmembrane proteins with a short, intracytoplasmic N-terminal region, and a longer, extracellular C-terminal
region. The extracellular region typically contains a canonical TNF-homology domain responsible for interaction with
the cytokine receptors. Members of the TNF family can also act
as soluble mediators, following cleavage of their extracellular
regions, and in some cases, the same cytokine exists in both
soluble and membrane-bound forms.
There are two eponymous (having the same name as)
members of the TNF family: TNF- and TNF-, though
TNF- is more commonly known as Lymphotoxin-, or LT-.
Both of these are secreted as soluble proteins. TNF- (frequently referred to simply as TNF) is a proinflammatory
cytokine, produced primarily by activated macrophages, but
also by other cell types including lymphocytes, fibroblasts,
and keratinocytes (skin cells), in response to infection,
inflammation, and environmental stressors. TNF elicits its
biological effects by binding to its receptors, TNF-R1 or
TNF-R2, which are described below. Lymphotoxin- is produced by activated lymphocytes and can deliver a variety of
signals. On binding to neutrophils, endothelial cells, and
osteoclasts (bone cells), Lymphotoxin- delivers activation
signals; in other cells, binding of Lymphotoxin- can lead to
increased expression of MHC glycoproteins and of adhesion
molecules.
We will also encounter five physiologically significant,
membrane-bound members of the TNF cytokine family
throughout this book. Lymphotoxin-, a membrane-bound
cytokine, is important in lymphocyte differentiation. We will
learn about BAFF and APRIL in the context of B-cell development and homeostasis (Chapter 10). CD40L is a cytokine
expressed on the surface of T cells that is required to signal
for B-cell differentiation (Chapter 12). Fas ligand (FasL), or
CD95L, induces apoptosis on binding to its cognate receptor,
Fas, or CD95.
Whether membrane-bound or in soluble form, active
cytokines of the TNF family assemble into trimers. Although
in most cases they are homotrimeric, heterotrimeric cytokines
do form between the TNF family members Lymphotoxin-
and Lymphotoxin- and between APRIL and BAFF. Crystallographic analysis of TNF family members has revealed that
124
PA R T I
|
Introduction
tant questions still await resolution. One reason that these
pathways have been so difficult to define is that the same
receptor, TNF-R1, can transduce both activating and deathpromoting signals, depending on the local cellular and
molecular environment in which the signal is received, and
investigators have yet to determine the trigger that shifts the
signaling program from life to death. However, much is
known about how each of these signaling pathways work,
once that all-important decision has been made.
We will start by describing the proapoptotic (deathinducing) pathway that is initiated when the membranebound TNF family member FasL on one cell binds to a Fas
receptor on a second cell, leading to death in the cell bearing
the Fas receptor. With this as our foundation, we will then
illustrate how the TNF-R1 receptor mediates both life- and
death- promoting signals. Signaling through other TNF-R
family members, such as CD40, BAFF, and April, will be
described in later chapters in the context of the various
immune responses in which they are involved.
FIGURE 4-12 The TNF-family members act as trimers in
vivo. [PDB ID 1TNF]
they have a conserved tertiary structure and fold into a
-sheet sandwich. The conserved residues direct the folding
in the internal strands that, in turn, promote the trimer
formation (Figure 4-12).
TNF Receptors
Members of the TNF receptor superfamily are defined by the
presence of Cysteine-Rich Domains (CRDs) in the extracellular, ligand-binding domain. Each CRD typically contains
six cysteine residues, which form three disulfide bonded
loops, and individual members of the superfamily can contain from one to six CRDs.
Although most TNF receptors are Type 1 membrane proteins (their N-terminals are outside the cell), a few family
members are cleaved from the membrane to form soluble
receptor variants. Alternatively, some lack a membrane
anchoring domain at all, or are linked to the membrane only
by covalently bound, glycolipid anchors. These soluble forms
of TNF family receptors are known as “decoy receptors,” as
they are capable of intercepting the signal from the ligand
before it can reach a cell, effectively blocking the signal. This
is a theme that we have encountered before in our consideration of the IL-1 receptor family.
Signaling Through TNF Superfamily Receptors
The work of delineating the precise pathways of signaling
through TNF family receptors is ongoing, and some impor-
Signaling Through the Fas Receptor
At the close of an immune response, when the pathogen is
safely demolished and the immune system needs to eliminate the extra lymphocytes it has generated to deal with the
invader, responding lymphocytes begin to express the TNF
family receptor Fas on their cell surfaces. Fas, and its ligand
FasL, are specialized members of the TNF receptor and the
TNF cytokine families, respectively, and they work together
to promote lymphocyte homeostasis. Mice with mutations
in either the fas (mrl/lpr mice) or the fasL (gld mice) genes
consequently suffer from severe lympho-proliferative disorders, indicative of their inability to eliminate lymphocytes
that are no longer serving a useful purpose.
On interaction with other immune cells bearing FasL,
the Fas receptor trimerizes and transduces a signal to the
interior of the Fas-bearing cell that results in its elimination by apoptosis. Apoptosis, or programmed cell death,
is a mechanism of cell death in which the cell dies from
within and is fragmented into membrane-bound vesicles
that can be rapidly phagocytosed by neighboring macrophages (Figure 4-13a). By using such well-controlled apoptotic pathways, the organism ensures that minimal
inflammation is associated with the natural end of an
immune response. Activation of the apoptotic pathway
invokes the activation of caspases; these are proteases,
bearing Cysteine residues at their active sites, which cleave
after ASPartic acid residues.
Binding of Fas to FasL results in the clustering of the Fas
receptors (Figure 4-13b). This, in turn, promotes interaction between their cytoplasmic regions, which include
domains common to a number of proapoptotic signaling
molecules called death domains. This type of interaction,
between homologous protein domains expressing affinity
for one another, is referred to as a homotypic interaction. As
they bind to one another, the clustered Fas protein death
domains incorporate death domains from the adapter
Receptors and Signaling: Cytokines and Chemokines
(a)
Hoechst
Transmission
(b)
|
CHAPTER 4
125
(a)
T cell
FasL
Control
Death
domains
Fas
Death
effector
domains
Procaspase-8
(inactive)
FADD
Early
Caspase-8
(active)
Late
Procaspase-3
and Procaspase-7
(inactive)
Caspase-3 and Caspase-7
(active)
Inactive
pro-apoptotic
enzymes
Active
apoptotic
enzymes
Apoptosis
FIGURE 4-13 Apoptotic signaling through Fas receptors.
domains also located on the FADD adapter proteins incorporate
the DED domains of procaspase-8 into the membrane complex.
Clustering of procaspase-8 induces cleavage of the pro domains
of procaspase-8, leading to the release of the active caspase-8
protease. Caspase-8 cleaves the pro domains from the executioner caspases, caspase-3 and caspase-7, which in turn cleave
and activate nucleases leading to the degradation of nuclear
DNA. Caspase-8 also cleaves and activates the proapoptotic Bcl-2
family member protein, BID. [Taimen, Pekka and Kallajoki, Markku.
(a) Human HeLa cells were activated to undergo apoptosis
through the Fas receptors. Hoechst-stained cells show the gradual
condensation of nuclear DNA into membrane-bounded blebs, as
the cell breaks up into vesicular packages that are recognized and
phagocytosed by macrophages in the absence of inflammation.
The same cells are also shown under transmission microscopy.
Arrows show staining of the Nuclear Mitotic Apparatus protein, an
early nuclear caspase target in apoptosis. (b) Signaling from Fas
leads to apoptosis. Binding of FasL to Fas induces clustering of the
Fas receptors and corresponding clustering of the Fas Death
Domains (DDs). The DDs of the adapter protein FADD bind to the
clustered Fas DDs via a homotypic interaction. Death effector
NuMA and nuclear lamins behave differently in Fasmediated apoptosis
J Cell Sci 2003 116:(3):571-583; Advance Online Publication December 11,
2002, doi:10.1242/jcs.00227. Reproduced with permission of Journal of
Cell S]
protein FADD (Fas-Associated Death Domain-containing
protein). FADD contains not only death domains, but also a
related type of domain called a Death Effector Domain
(DED). This, in turn, binds homotypically to the DED
domains of procaspase-8, resulting in the clustering of
procaspase-8 molecules. Procaspase-8 molecules contain
the active caspase-8 enzyme, held in an inactive state by
binding to prodomains.
The multimerization of procaspase-8 molecules results in
mutual cleavage of their prodomains and induces capsase-8
activation. Caspase-8 then cleaves many target proteins
critical to the generation of apoptosis. The target proteins of
caspase-8 include the executioner caspases, -3 and -7 (which
cleave and activate nucleases leading to the degradation of
nuclear DNA), and the proapoptotic Bcl-2 family member
BID. The complex of Fas, FADD, and procaspase-8 is
referred to as the Death-Inducing Signaling Complex
(DISC). The ultimate result of activation of this cascade is
the condensation of nuclear material (see Figure 4-13a), the
degradation of nuclear DNA into 240 base pair, nucleic acid
fragments, and the subsequent breakdown of the cell into
“easily digestible” membrane-bound fragments.
126
PA R T I
|
Introduction
Signaling Through the TNF-R1 Receptor
The TNF-R1 receptor is present on the surface of all vertebrate cells and, like Fas, has an intracytoplasmic death
domain (DD). Although this receptor is capable of binding
to both TNF- and Lymphotoxin-, we will focus on the
signaling that is elicited by TNF- (TNF). TNF binding to
the TNF-R1 receptor can lead to two very different outcomes: apoptosis (death) or survival (life). How it does so is
still the focus of intensive investigation, but the story as it is
unfolding is already a fascinating one.
The mechanism by which TNF binding leads to apoptosis
is slightly different from that which follows Fas-FasL binding. Like FasL, binding of TNF to the TNF-R1 receptor
induces trimerization of the receptor as well as an alteration
in its conformation, and these together result in the binding
of a DD-containing adapter molecule, in this case TRADD,
to the internal face of the receptor molecule (Figure 4-14).
The TRADD adapter molecule provides additional binding
sites for the components RIP1 (a serine/threonine kinase,
rather evocatively named Rest In Peace 1), which binds via
its own DDs and TRAF2, the TNF Receptor Associated Factor 2. This is known as complex I. Intracellular localization
experiments have shown that this complex can dissociate
from the TNF-R at the membrane, and migrate to the cytoplasm where it binds to the now familiar DD-containing protein FADD. FADD recruits procaspase-8, as described above,
resulting in the generation of an apoptotic signal. The proapoptotic cytoplasmic complex generated upon TNF-R1
receptor binding is shown in Figure 4-14a as complex II.
Counterintuitively, binding of this same molecule, TNF,
to the same receptor, TNF-R1, can result in the delivery of
survival as well as of proapoptotic signals. How can the same
cytokine, acting through the same receptor, bring about two
apparently opposing actions?
In the TNF-mediated survival pathway (Figure 4-14b),
the generation of the original membrane complex appears to
initiate in the same general manner as for the proapoptotic
pathway. However, in the case of the prosurvival pathway,
the TRADD-containing complex does not dissociate from
the membrane, but rather remains at the cell surface and
recruits a number of other components, including the ubiquitin ligases cIAP1 and cIAP2.
Once cIAP1 and cIAP2 join the TNF-R1 complex at the
cell membrane, they recruit the LUBAC proteins, which
attach linear ubiquitin chains to RIP1. Polyubiquitinated
RIP1 then binds to the NEMO component of IKK as well as
to TAK1, which is already complexed with its associated
TAB proteins as described above. RIP1 and TAK1 activate
the IKK complex, leading to IB phosphorylation and
destruction, and subsequent activation of NF-B. Once
NF-B is fully activated, it turns on the expression of the
cFLIP protein that then inhibits the activity of caspase-8.
This effectively shuts down the antagonistic, proapoptotic
pathway (Figure 4-14a). As previously described, the TAK1
complex also acts to activate the MAP kinase pathway, which
further enhances survival signaling.
(a) Apoptosis
(b) Survival
TNF
TNF
TNF-R1
DD
DD
RIP1
TRADD
TRAF2
Complex I
RIP1
RIP1
cIAP1 cIAP2
Dissociation
LUBACs
TRADD
TRAF2
FADD
Complex II
TAB1
Procaspase-8
(inactive)
Caspase-8
(active)
Apoptosis
TRADD
TRAF2
TAB2
TAK1
NF-κB MAP kinase
activation cascade
cFLIP
Survival
FIGURE 4-14 Signaling through TNF-R family receptors.
Signaling through TNF-R family receptors can lead to pro- or antiapoptotic outcomes depending on the nature of the signal, the
receptor, and the cellular context. (a) Apoptosis. Binding of TNF to
TNF-R1 induces trimerization of the receptor and conformational
alteration in its cytoplasmic domain, resulting in the recruitment of
the DD-containing adapter molecule TRADD to the cytoplasmic face
of the receptor. TRADD binds to the serine-threonine kinase RIP1 and
the TNF receptor associated factor TRAF2. This complex of TRADD,
RIP1, and TRAF2, known as complex I, dissociates from the receptor
and migrates to the cytoplasm where it binds to the adapter protein
FADD. FADD recruits procaspase-8, leading to apoptosis as described
in Figure 4-13. The proapoptotic complex generated upon TNF-R1
receptor binding is shown as Complex II. (b). Survival. As for the apoptotic pathway, TNF ligation results in receptor trimerization, TRADD
binding, and RIP1 recruitment. In this case, however, TRADD also
recruits the ubiquitin ligases cIAP1 and cIAP2, which in turn bind to
the proteins of the linear ubiquitin assembly complex (LUBAC) proteins. Polyubiquitination of RIP1 allows it to bind to the NEMO component of the IKK complex as well as to TAK1. TAK1 and RIP1 together
activate the IKK complex, leading to IB phosphorylation and
destruction, and release of NF-B to enter the nucleus. Among other
prosurvival effects, NF-B activates the transcription of the cFLIP
protein, which inhibits caspase-8 action, thus tipping the scales in
favor of survival. The TAK1 complex also activates MAP kinase signaling, which enhances cell survival.
Receptors and Signaling: Cytokines and Chemokines
The survival versus death decisions that are made at the
level of the TNF-R1 receptor depend upon the outcome of
the race between the generation of active caspase-8 on the
one hand and the generation of the caspase-8 inhibitor cFLIP
on the other. Although we now understand the molecular
mechanisms that bring about the consequences of these
decisions, we still have much to learn regarding how the cell
integrates the signals received through TNF-R1 with other
signals delivered to the cell in order to determine which of
the two competing pathways will prevail. Since the generation of the membrane-bound complex that is capable of
activating NF-B is entirely dependent on interactions
between various ubiquitinated proteins, it now appears that
the life-death decision for a cell may be executed by a small
protein previously thought to have only destructive intent.
The IL-17 Family Is a Recently Discovered,
Proinflammatory Cytokine Cluster
The most recently described family of cytokines, the IL-17
family, includes interleukins 17A, 17B, 17C, 17D, and 17F.
Signaling through most members of this family culminates
in the generation of inflammation. IL-17 receptors are found
on neutrophils, keratinocytes, and other nonlymphoid cells.
Members of the IL-17 family therefore appear to occupy a
TABLE 4-5
|
CHAPTER 4
127
location at the interface of innate and adaptive immunity.
IL-17 cytokines do not share sequence similarity with other
cytokines, but intriguingly the amino acid sequence of
IL-17A is 58% identical to an open reading frame (ORF13)
found in a T-cell–tropic herpesvirus. The significance of this
sequence relationship is so far unknown; did the virus hijack
the cytokine sequence for its own needs, or did the pilfering
occur in the opposite direction?
IL-17 Cytokines
IL-17A, the first member of this family to be identified, is
released by activated T cells and stimulates the production of
factors that signal a proinflammatory state, including IL-6,
CXCL8, and granulocyte colony-stimulating factor (G-CSF).
As characterization of IL-17A and the T cells that secreted it
progressed, it became clear that the T cells secreting this
cytokine represent a new lineage, the TH17 cell subset, which
is currently the focus of intense investigation (see Chapter
11). Genomic sequencing has since led to the identification
of a number of homologs of IL-17A (see Table 4-5). Most of
the interleukins in the IL-17 family share the property of
operating at the interface of innate and adaptive immunity,
serving to coordinate the release of proinflammatory and
neutrophil-mobilizing cytokines. However, IL-17E provides an exception to this general rule, instead promoting
Expression and known functions of members of the extended IL-17 receptor family
Family
member
Other common
names
IL-17A
Receptors
Expression by which cells
Main functions
IL-17 and CTL-8
IL-17RA and
IL-17RC
TH17 cells, CD8 T cells, T cells, NK cells,
and NKT cells
Autoimmune pathology, neutrophil
recruitment, and immunity to
extracellular pathogens
IL-17B
NA
IL-17RB
Cells of the GI tract, pancreas, and
neurons
Proinflammatory activities?
IL-17C
NA
IL-17RE
Cells of the prostate and fetal kidney
Proinflammatory activities?
IL-17D
NA
Unknown
Cells of the muscles, brain, heart, lung,
pancreas, and adipose tissue
Proinflammatory activities?
IL-17E
IL-25
IL-17RA and
IL-17RB
Intraepithelial lymphocytes, lung
epithelial cells, alveolar macrophages,
eosinophils, basophils, NKT cells, TH2 cells,
mast cells, and cells of the gastrointestinal
tract and uterus
Induces TH2 responses and
suppresses TH17 responses
IL-17F
NA
IL-17RA and
IL-17RC
TH17 cells, CD8 T cells, T cells, NK cells,
and NKT cells
Neutrophil recruitment and immunity to extracellular pathogens
IL-17A/IL-17F
heterodimer
NA
IL-17RA and
IL-17RC
TH17 cells, CD8 T cells, T cells, NK cells,
and NKT cells
Neutrophil recruitment and immunity to extracellular pathogens
vIL-17
ORF13
IL-17RA (and
IL-17RC?)
Herpesvirus saimiri
Unknown
Adapted from Gaffen, S. L. 2009. Structure and signalling in the IL-17 receptor family. Nature Reviews Immunology 9:556–567.
PA R T I
128
IL-17F
IL-17A
FN1
|
IL-17A–
IL-17F
Introduction
IL-17E
IL-17B
IL-17C
IL-17RB
IL-17RE
IL-17D
Unknown
ligand
Unknown
ligand
IL-17A
FN2
SEFIR
TILL
CBAD
IL-17RB
IL-17RC
IL-17RA
IL-17RA
FIGURE 4-15 The IL-17 family of cytokines and their associated receptors. The cytokines that form the IL-17 family share
a highly conserved structure, with four conserved cysteines. Only
one of the proteins has so far been subject to x-ray analysis, which
demonstrates that the structure is that of a “cysteine knot,” a
tightly folded protein that exists naturally as a dimer. The five proteins that make up the IL-17 receptor family are IL-17RA, IL-17RB,
IL-17RC, IL-17RD, and IL-17RE. These are arranged into homo- and
hetero- dimers and trimers to create the complete receptor mol-
the differentiation of the anti-inflammatory TH2 subclass,
while suppressing further TH17 cell responses, in what
amounts to a negative feedback loop.
In general, members of the IL-17 family exist as homodimers, but heterodimers of IL-17A and IL-17F have been
described. Monomeric units of IL-17 family members range in
molecular weight from 17.3 to 22.8 kDa, and crystallographic
analysis has revealed that they share a structure that is primarily sheet in nature, stabilized by intrachain disulfide bonds.
The IL-17 Family Receptors
The IL-17 receptor family is composed of five protein chains—
IL-17RA, IL-17RB, IL-17RC, IL-17RD, and IL-17RE—which
are variously arranged into homo- and hetero- dimeric and
trimeric units to form the complete receptor molecules (Figure 4-15). Members of the IL-17 receptor family share fibronectin domains with the Hematopoietin and Interferon family
cytokine receptors and are single transmembrane proteins.
They all contain cytoplasmic SEF/IL-17R (Similar Expression
to Fibroblast growth factor interleukin 17 Receptor, or SEFIR)
domains, responsible for mediating the protein-protein interactions of the IL-17R signal transduction pathway. The IL17RA chain also contains a TIR-like loop (TILL) domain,
analogous to structures found in the Toll and IL-1 receptor
molecules, as well as a C/EBP activation domain (CBAD),
capable of activating the C/EBP transcription factor.
Signaling Through IL-17 Receptors
Analogous to signaling through IL-1 receptors, signaling
through most IL-17 receptors results in an inflammatory
response, and so it should not come as a surprise to learn
that signaling through the IL-17 receptor results in activa-
Unknown
receptor
IL-17RD
IL-17RD
IL-17RA
ecules shown. Each receptor protein includes one or more fibronectin (FN) domains, as well as a cytoplasmic SEF/IL-17R (SEFIR)
domain that is important in mediating downstream signaling
events. The IL-17RA protein also includes a TIR-like loop domain
(TILL), similar to that found in Toll-like receptors and IL-1 receptors,
as well as a C/EBP activation domain, capable of interacting with
the downstream transcription factor C/EBP. [Adapted from S. Gaffen,
2009, Structure and signalling in the IL-17 receptor family, Nature Reviews
Immunology, 9:556.]
tion of NF-B, a hallmark transcription factor of inflammation. Details of the signaling pathways that emanate from the
IL-17R are still being worked out, but Figure 4-16 illustrates
the major features of our current knowledge.
1.
NF-B activation via IL-17RA and IL-17RC. Binding of
IL-17A to the receptor molecules IL-17RA and IL-17RC
results in the recruitment of the adapter protein ACT1 to
the SEFIR domain. ACT1 binds other proteins, including
TRAF3 and TRAF6, which then engage with the TAK1
complex. TAK1 activation results in the phosphorylation
and inactivation of the inhibitor of NF-B (IB), allowing
NF-B activation and nuclear migration.
2.
Activation of MAP kinase pathway and cytokine mRNA
stabilization. Adapter proteins bound to the receptor also
recruit components of the MAP kinase pathway, resulting
in the activation of MAP kinases, including the extracellular signal-regulated kinase Erk1. Though unusual, it
appears that the most important role of Erk1 in IL-17
signaling is not in the generation of phosphorylated
transcription factors (as is the case for its involvement in
TCR- and BCR-mediated cell signaling), but rather in
controlling the stability of cytokine mRNA transcripts.
Many of the target genes of IL-17 signaling are cytokines
and chemokines whose transcription is up-regulated on
receipt of an IL-17 signal. The levels of cytokine mRNA
are controlled in part by binding of the cytoplasmic protein tristetraprolin to AU-rich elements (AREs) in the
3-untranslated regions of mRNA transcripts. Tristetraprolin then delivers the cytokine mRNAs to the exosome
complexes of the cells, where they are degraded. However,
phosphorylation of tristetraprolin by MAP kinases inhibits
Receptors and Signaling: Cytokines and Chemokines
1
SEFIR
ACT1
TRAFs
2
TAB2
MAP kinase
cascade
3
TAK1
ERK1
NF-κB
activation
mRNA
stabilization
NF-κB
C/EBPs
Cytokine expression
IL-6
expression
FIGURE 4-16 Signaling from the IL-17 receptor. Binding of
IL-17 to its receptor initiates three signaling pathways. (1) Binding of
IL-17 to its receptor results in the recruitment of the adapter protein
ACT1 to the cytoplasmic region of the receptor. ACT1 then serves as
a docking point for TRAF proteins 3 and 6, which in turn recruit members of the TAK1 complex, consisting of the TAK1 kinase and TAK1
binding proteins. TAK1 activation results in the phosphorylation and
activation of the IKK complex and resultant NF-B activation, as
described previously. (2) Adapter proteins bound to the SEFIR (Similar
Expression to Fibroblast growth factor interleukin 17 Receptor)
domain also recruit components of the MAP kinase pathway. The
MAP kinase Erk1 phosphorylates the cytoplasmic protein tristetraprolin, and inhibits its ability to bind to AU-rich elements on mRNA
encoding cytokines. Since tristetraprolin binding results in mRNA
degradation, activation of this arm of the pathway results in enhancing the stability of cytokine mRNA. (3) IL-17 binding to its receptor
also results in the activation of transcription factors of the C/EBP family, which promote the expression of the inflammatory cytokine IL-6.
its ability to recruit the degradative machinery and hence
results in increased stability of the cytokine and chemokine mRNAs.
3.
CHAPTER 4
129
Chemokines Direct the Migration
of Leukocytes Through the Body
IL-17R
TAB1
|
Activation of the transcription factors C/EBP and
C/EBP In addition to up-regulation of NF-B and
members of the MAP kinase pathway, signaling through
IL-17RA and IL-17RC proteins has also been shown to
activate the transcription factors C/EBP and C/EBP ,
which promote expression of IL-6, one of the quintessential inflammatory cytokines.
Chemokines are a structurally related family of small cytokines that bind to cell-surface receptors and induce the
movement of leukocytes up a concentration gradient and
toward the chemokine source. This soluble factor-directed
cell movement is known as chemotaxis, and molecules that
can elicit such movement are referred to as chemoattractants
(Box 4-3). Some chemokines display innate affinity for the
carbohydrates named glycosaminoglycans, located on the
surfaces of endothelial cells, a property that enables them to
bind to the inner surfaces of blood vessels and set up a cellbound chemoattractant gradient along blood vessel walls,
directing leukocyte movement.
Chemokine Structure
Chemokines are relatively low in molecular weight (7.5–
12.5kDa) and structurally homologous. The tertiary structure of chemokines is constrained by a set of highly conserved
disulfide bonds; the positions of the cysteine residues determine the classification of the chemokines into six different
structural categories (Figure 4-17). Within any one category,
chemokines may share 30% to 99% sequence identity.
The grouping of chemokines into the subclasses shown in
Figure 4-17 has functional, as well as structural, significance.
For example, the seven human CXC chemokines within the
ELR subclass share the same receptor (CXCR2), attract neutrophils, are angiogenic, and have greater than 40% sequence
identity. (A substance is angiogenic if it promotes the formation of new blood vessels; it is angiostatic if it prevents the
formation of new blood vessels.) The non-ELR, CXCL chemokines CXCL9, CXCL10, and CXCL11, are also more than
40% identical to one another; however, this group is angiostatic, not angiogenic, and utilizes the CXCR4 receptor.
Members of the two, structurally distinct CC groups are
chemoattractants that attract monocytes and macrophages
(although not neutrophils) to the site of infection. See
Appendix III for a more comprehensive tabulation of chemokines and their immunologic roles.
Chemokine Receptors
In the 1950s, investigations of the mechanisms by which
glucagon and adrenaline signaling led to an increase in the
rate of glycogen metabolism revealed the existence of a class
of receptors that threads through the membrane seven times
and transduces the ligand signal via interactions with a polymeric GTP/GDP-binding “G protein.” This class of G-Protein–
Coupled Receptors (GPCRs) is used in the recognition of
many types of signals, including those mediated by chemokines. Certain essential features of this pathway are conserved in all GPCR-type responses. (These larger, polymeric,
seven membrane pass receptor-associated G proteins are
different from the small, monomeric G proteins such as ras,
which participate farther downstream in intracellular signaling pathways. Although both types of G proteins are
130
PA R T I
|
Introduction
ADVANCES
How Does Chemokine Binding to a Cell-Surface Receptor Result
in Cellular Movement Along the Chemokine Gradient?
Chemotaxis is
the mechanism by
which the speed and direction of cell
movement are controlled by a concentration gradient of signaling molecules.
Receptors on the surface of responsive
cells bind to chemotactic factors and
direct the cell to move toward the source
of factor secretion. Chemotaxis was recognized as a biological phenomenon as
early as the 1880s, and is found in organisms as simple as bacteria, which use
chemotaxis to locate sources of nutrition.
However, only as scientists have developed the ability to analyze signaling pathways and to observe the movement of
individual cells under in vivo conditions
have we been able to approach a mechanistic understanding of the complex
series of intracellular events that culminate in chemically directed cell movement in the immune system. Although
we still do not understand all the details of
this process, some general principles have
begun to emerge.
We first address the means by which
cells sense chemoattractant signals. We
next develop an appreciation for how
cells move, and finally we can describe a
little of what we know regarding how
they become polarized in order to direct
their movement as specified by the chemotactic signal.
SIGNAL SENSING
Leukocytes recognize chemoattractant signals, or chemokines, using the G-protein–
coupled receptors described in this chapter.
Such receptors are located at the plasma
membrane, and the cytoplasmic face of
these receptors is associated with a polymeric G protein, so called because of its
affinity for guanosine phosphates. On binding to its chemokine ligand, the receptor
alters its conformation, passing on that
conformational change to its associated G
protein. The bound G protein then loses
affinity for guanosine diphosphate (GDP)
and instead binds guanosine triphosphate
(GTP), thus achieving its active conformation. Next, the G protein dissociates into
two subunits, termed G-GTP and G.
The signaling pathways emanating from
each of these subunits are described in
the text. It is difficult to overstate the
effectiveness of these receptor signals in
determining the direction and speed of
cellular movement. For example, neutrophils can recognize and move in response
to extremely shallow chemoattractant
gradients, in which the concentration of
the chemoattractant at the front of the
gradient is as little as 2% higher than that
at the back.
and almost like wires (filopodia), or protrude from the surface of the cell like a
sheet (lamellipodia). The nucleus of the
migrating cell lies behind these protrusions in the cell body, and at the posterior
part of the cell lies a near-cylindrical tail,
the uropod, which may be as long as 10 m.
Formation of protrusions
Protrusions
Direction of
movement
Formation of adhesions
Uropod
Adhesions
Contraction
HOW CELLS MOVE
Figure 1 illustrates a model of amoeboid
cellular movement, the mechanism used
by leukocytes. Ameboid migration is a
rapid type of cell movement; leukocytes
and stem cells, which use this mode of
migration, may move as quickly as 30 m/
minute. Cells routinely move over other
cells and tissues (the substrata) by first
forming cellular protrusions at their leading edge—the edge closest to the direction of movement. Depending on the cell
type and the nature of the chemoattractant signal, these protrusions may be thin,
FIGURE 1
Schematic description of the classical
migratory cycle. [Adapted from P. Mrass et al.,
2010, Cell-autonomous and environmental contributions to the interstitial migration of T cells, Seminars in
Immunopathology 32:257.]
Receptors and Signaling: Cytokines and Chemokines
|
CHAPTER 4
131
BOX 4-3
The protrusions at the leading edge of
a moving leukocyte form noncovalent
attachments to the substratum using
adhesive interactions between proteins
on the cell protrusion and other proteins
on the substratum. The substratum may,
for example, be a capillary endothelial cell,
when a white blood cell rolls along the
inside of a blood capillary. Alternatively,
during leukocyte movement within a
lymph node, the adhesive interaction may
occur between B or T lymphocyte surface
integrins and proteins located on a cellular extension from a follicular reticular cell.
Once the migrating cell is temporarily
attached to the substratum, contractions
of the cell body mediated by actin and
myosin, and resulting cytoskeletal rearrangements, bring the rest of the cell forward toward the leading edge and then
the cycle of reaching forward with cellular
protrusions and contraction of the rest of
the cell in the direction of the leading
edge can recur.
GENERATION OF POLARITY SIGNALS:
HOW CELLS DECIDE ON THE
DIRECTION OF MOVEMENT
FIGURE 2
Polarization of neutrophils after
receipt of a chemoattractant signal. A
neutrophil exposed to a chemoattractant organizes filamentous actin at the protrusions on the
leading edge of the cell (shown here in red) and
an actino-myosin contractile complex at the
uropod and along the sides of the cell (shown
here stained green). [Fei Wang, The signaling mechanisms underlying cell polarity and chemotaxis. ColdSpring Harbor Perspectives in Biology, October 2009,
1(4):a002980. doi: 10.1101/cshperspect.a002980. ? 2009
Cold Spring Harbor Laboratory Press, all rights reserved.]
Upon receipt of a chemotactic signal,
leukocytes polarize very rapidly, and leukocyte locomotion is clearly visible within
30 seconds after the receipt of the chemokine by the cell.
Scientists initially hypothesized that
movement of a leukocyte toward a particular chemoattractant may result from
an asymmetric distribution of chemoattractant receptors on the cell surface,
with a higher concentration of receptors
localized to the leading edge of the cell.
However, microscopic observations suggest that this initial hypothesis was incorrect. Instead, investigators found an
asymmetrical distribution of signaling
molecules and cytoskeletal components
(Figure 2). Thus, although there is a uniform distribution of receptors on the cell
surface, the distribution of occupied
receptors is asymmetrical, which means
that the signal received from chemokines
on the side of the cell facing the chemokine source is stronger than that received
elsewhere on the cell surface. The ultimate result of this asymmetrical signal
from the occupied receptors is an internal
redistribution of signaling pathway components and cytoskeletal elements; this
in turn causes the cell to move toward
the source of the signal.
Many of the signaling pathways that
mediate this directional movement are
already familiar to you from Chapter 3
and from this chapter. The dimer of
the activated G protein recruits a particular subclass of PI3 kinase enzymes to
the inner leaflet of the leading edge of
the membrane, where it phosphorylates
PIP2 and other phosphatidyl inositol
phosphates. These phosphorylated lipids then serve as docking sites for PHdomain–containing proteins, including
Akt, which in turn phosphorylates downstream effectors leading to actin polymerization. Depending on the cell type
and the nature of the stimulation, the
small GTPases Rac, Cdc42, and Rho are
also implicated in the chemokinemediated cytoskeletal modifi cations
that are necessary for cell movement,
and these are activated via tyrosine
kinases in lymphocytes and by pathways
that are as yet not fully characterized in
other leukocyte cell types.
A cell can only move in response to a
gradient if the trailing edge of the cell
moves at the same rate as the leading
edge, and the uropod of moving leukocytes is susceptible to its own set of
G-protein–mediated signals. Following
redistribution of the internal signaling
components, the chemokine receptors at
the trailing edge of the cell are coupled to
a set of G-protein trimers different from
those at the leading edge. Interaction of
the uropod-localized G-Protein–Coupled
Receptors (GPCRs) leads to the formation
and contraction of actin-myosin complexes and subsequent retraction of the
uropod.
132
PA R T I
|
Introduction
Class
Structural signature
Names
Number (n) in
class
CX3C
.............CXXXC.............C.........C.........
CX3CL1
1
Non-ELR CXC
.............CX__C.............C.........C.........
CXCL#
9
ELR CXC
...ELR....CX__C.............C.........C.........
CXCL#
7
4C CC
.............C___C.............C.........C.........
CCL#
19
6C CC
.............C___C......C.....C.........C....C...
CCL#
5
C
.....................C........................C
XCL#
2
FIGURE 4-17 Disulfide bridges in chemokine structures.
A schematic of the locations of cysteine residues in chemokines that
shows how the locations of cysteines determine chemokine class.
Chemokines are proteins of small molecular weight which share two,
four, or six conserved cysteine residues at particular points in their
sequence that form intrachain disulfide bonds. The number of cyste-
activated by GTP binding, their structures and functions
are quite different).
The GPCRs are classified according to the type of chemokine they bind. For example, the CC receptors (CCRs) recognize CC chemokines, the CXCRs recognize CXCL
chemokines, and so on. Chemokine receptors bind to their
respective ligands quite tightly (Kd ≅ 109 M). Interestingly,
the intrinsic specificity of the receptors is balanced by the
capacity of many receptors to bind more than one chemokine from a particular family and of several chemokines to
bind to more than one receptor. For example, the receptor
CXCR2 recognizes seven different chemokines, and CCL5
can bind to both CCR3 and CCR5.
Signaling Through Chemokine Receptors
The cytoplasmic faces of seven membrane pass GPCRs associate with intracellular, trimeric GTP-binding proteins
consisting of G, G, and G subunits (Figure 4-18a).
When the receptor site is unoccupied, the G subunit of the
trimeric G protein binds to GDP. Chemokine ligation to the
receptor results in a conformational change that is transmitted to the G protein and in turn induces an exchange of GDP
for GTP at the G binding site that is analogous to what
occurs upon GTP binding to the small GTP-binding proteins such as Ras. This results in the dissociation of the G
protein into a G-GTP monomer and a G dimer. In the
case of the chemokine receptor associated G proteins,
chemokine signaling is mediated by both the dissociated
G dimer (which has no nucleotide binding site) and the
G-GTP subunit.
Just as for the small G-protein–coupled pathways, the
duration of signaling through the chemokine receptor is
limited by the intrinsic GTPase activity of the G subunit,
which in turn can be increased by GTPase Activating
ines as well as the positions of the disulfide bonds determine the
subclass of these cytokines as shown. The overscores indicate the
cysteines between which disulfide bonds are made. The naming of
chemokines in part reflects the cysteine-determined class (see
Appendix III). [Adapted from W. E. Paul, 2008, Fundamental Immunology,
6th ed. Lippincott Williams & Wilkins, Philadelphia, Figure 26.1]
Proteins (GAPs), also known as Regulators of G-protein
Signaling (RGSs). Because the pathway is active only when
the protein binds to GTP, GAPs down-regulate the activity
mediated by the receptor (see Figure 3-15). Once the GTP in
the G binding site is hydrolyzed to GDP, G re-associates
with the G dimers, effectively terminating signaling.
There are multiple subtypes of G and G subunits that vary
in representation between different cell types. Signaling
through different subunits can give rise to different consequences, depending on the downstream pathways that are
elicited. Just as for the small G protein, there are several different large polymeric G proteins, which vary in their cellular distribution and receptor partners.
Once released from its G partner, the G subunit of
the trimeric G protein activates a variety of downstream
effector molecules, including those of the Ras/MAP kinase
pathway (path 1 in Figure 4-18). Full activation of MAP
kinase is further facilitated by tyrosine phosphorylation
mediated by G-GTP-activated tyrosine kinases (not
shown). Activation of the Ras pathway culminates in the
initiation of transcription as well as in up-regulation of integrin adhesion molecules on the cell membrane. G signaling also cooperates with signaling mediated by G-GTP to
activate one isoform of phospholipase C, PLC, resulting in
an increase in the activity of the transcription factor NF-B
(path 2).
The GGTP complex also activates a signaling pathway
that is initiated by the small G protein, Rho (path 3). This
pathway leads to actin polymerization and the promotion of
cell migration, so it is this third pathway that is responsible
for the most commonly described aspect of chemokine signaling: cell movement (see also Box 4-3). Rho signaling is
also instrumental in bringing about changes in the transcriptional program of the cell. Finally, a JAK associated with
Receptors and Signaling: Cytokines and Chemokines
Chemokine
Chemokine receptor
C
JAK
4
β
γ
α
2
PLCβ
G protein
3
1
PKC
Ras
Rho
Cell
movement
Akt
MAP kinase
cascade
Survival
|
CHAPTER 4
133
the GPCR initiates signaling through PKC-mediated pathways that culminate in the activation of Akt and increased
cell survival (path 4) as well as in further transcriptional
alterations.
The quality of the response elicited by particular chemokines in different types of cells is dependent on the nature of
the chemokine ligand, as well as on the signaling microenvironment, which in turn is determined by the range of G
protein subtypes and regulatory molecules present in that
cell. But from the complexity of signaling options available
to the receptor, it is not difficult to see how binding of a chemokine molecule to its receptor can simultaneously bring
about alterations in the location, the adhesion molecule
binding capacity, and the transcriptional program of the
chemokine-activated cell.
Cytokine Antagonists
AP-1
NF-κB
Gene expression
FIGURE 4-18 G-protein–coupled receptors interact with
G proteins that transduce chemokine signals into the interior of the cell. Different chemokines can induce different signaling pathways. This figure therefore represents a composite of some
of the most common pathways elicited by chemokine binding,
which lead collectively to alterations in the transcriptional program,
the enabling of cell movement, and changes in the adhesive properties of the signaled cell. (1) The G subunit binds to the adapter
molecule Grb2, activating it and initiating the Ras signaling pathway
that leads eventually to activation of MAP kinase and an alteration in
the cell’s transcriptional program, as shown in Figure 3-16. Ras pathway activation also leads eventually to activation of integrin adhesion molecules on the cell surface. G GTP simultaneously binds and
activates a protein tyrosine kinase that phosphorylates and further
activates MAP kinase (not shown). (2) Both GGTP and G cooperate to activate PLC, which activates the NF-B pathway. (3) GGTP
activates the small cytoplasmic G protein, Rho, initiating actin
polymerization and cell movement. Other pathways emanate from
Rho that lead to the activation of the transcription factor Serum
Response Factor (SRF). (4) A JAK is stimulated by chemokine binding
to the receptor, and it turns on the activity of PKC, leading eventually
to the activation of the enzyme Akt. Akt affects cell survival by phosphorylating the proapoptotic genes Bax and Bad (not shown) and
marking them for destruction and enhancing cell survival. It also
phosphorylates and further activates the transcription factor NF-B.
JAK-mediated PKC activation can also lead to phosphorylation of
the transcription factor Jun, and its dimerization with Fos to create
the complete transcription factor AP-1.
A number of proteins that inhibit the biological activity of
cytokines have been reported. These proteins act in one of
two ways: either they bind directly to a cytokine receptor but
fail to activate the cell, thus blocking the active cytokine
from binding, or they bind directly to the cytokine itself,
inhibiting its ability to bind to the cognate receptor. In this
section, we describe some naturally occurring cytokine
antagonists that modulate and refine the power of particular
cytokine responses, as well as the ways in which various
pathogens have hijacked cytokine responses to their own
ends.
The IL-1 Receptor Antagonist Blocks the IL-1
Cytokine Receptor
The best-characterized cytokine inhibitor is the IL-1 receptor antagonist (IL-1Ra), which binds to the IL-1 receptor but
does not elicit activation of the signaling pathway (see
above). As previously described, ligation of IL-1Ra to the
IL-1 receptor blocks the binding of both IL-1 and IL-1,
thus accounting for its inhibitory properties. IL-1Ra is synthesized by the same cells that secrete IL-1 and IL-1, and
its synthesis in the liver is up-regulated under inflammatory
conditions, along with that of IL-1. Several animal and
human models exist in which the levels of IL-1Ra are naturally reduced, and humans carrying an allele that decreases
the expression of IL-1Ra suffer from arthritis and a variety of
other autoimmune diseases. This observation suggests that
the normal function of IL-1Ra is to provide for the host a
means by which to modulate the numbers of receptors that
are capable of mounting a physiological response to IL-1.
Given the fiercely proinflammatory effects of IL-1, it makes
sense that responses to this powerful cytokine should be
carefully controlled. Indeed, we note quite often in biological
systems that a process—if it has the potential to lead to deleterious consequences to the organism—is subject to several
different means of regulation.
134
PA R T I
|
Introduction
Recombinant IL-Ra has been used clinically, under the
name of anakrina, for the treatment of rheumatoid arthritis.
Investigations into how cells control the balance of IL-1 and
IL-1Ra secretion are still ongoing, but preliminary studies
suggest that the activation of different isoforms of PI3 kinase
may play an important role in determining the relative
amounts of IL-1 and IL-1Ra that are secreted by a stimulated
monocyte.
Cytokine Antagonists Can Be Derived
from Cleavage of the Cytokine Receptor
Some naturally occurring soluble antagonists arise from
enzymatic cleavage of the extracellular domains of cytokine
receptors. These soluble receptor components can compete
with the membrane-bound receptor for cytokine binding
and thus down-modulate the potential cytokine response.
The best characterized of the soluble cytokine receptors
consists of a segment containing the amino-terminal 192
amino acids of the IL-2R (CD25) subunit, which is
released by proteolytic cleavage, forming a 45-kDa soluble
IL-2 receptor (sIL-2R or sCD25). The shed receptor retains
its ability to bind IL-2 and can therefore prevent the cytokine’s productive interaction with the membrane-bound
IL-2 receptor.
The origin of sIL-2R is still a matter for debate. Recently,
regulatory T cells, which express high levels of CD25 on
their membrane surfaces, have been shown to release sIL-2R
upon activation. Since these T cells serve the function of
down-regulating ongoing immune responses, it has been
suggested that the soluble IL-2 receptors may serve the
physiological function of soaking up excess IL-2 and thus
reducing the amount of the cytokine that is available to irrelevant or even to competing effector T cells. The soluble IL-2
receptor is also found in the serum and bodily fluids of
patients suffering from a number of hematologic malignancies (blood cell cancers), and high levels of sIL-2R in the
blood correlate with a poor disease prognosis. However, the
issue of whether the sIL-2R is released from the tumor cells
per se, or whether it is released from regulatory T cells that
may be acting to dampen the host anti-tumor response, has
yet to be resolved.
Some Viruses Have Developed Strategies
to Exploit Cytokine Activity
The cytokine antagonists described above derive from an
organism’s own immune system. However, as is so often the
case in immunology, some pathogens have evolved ways in
which to circumvent cytokine responses, by mimicking molecules and pathways used by the host. The evolution of anticytokine strategies by microbial pathogens provides
biological evidence of the importance of cytokines in organizing and promoting effective antimicrobial immune
responses. The following are among the various anti-cytokine
strategies used by viruses:
• The generation of viral products that interfere with
cytokine secretion
• The generation of cytokine homologs that compete with
natural cytokines or inhibit anti-viral responses
• The production of soluble cytokine-binding proteins
• The expression of homologs of cytokine receptors
• The generation of viral products that interfere with
intracellular signaling
• The induction of cytokine inhibitors in the host cell
Epstein-Barr virus (EBV), for example, produces an
IL-10–like molecule (viral IL-10 or vIL-10) that binds to the
IL-10 receptor. Just like host-derived IL-10, this viral homologue suppresses TH1-type cell-mediated responses that
would otherwise be effective in fighting a viral infection.
Other cytokine mimics produced by viruses allow them to
manipulate the immune response in alternative ways that aid
the survival of the pathogen. For example, EBV produces an
inducer of IL-1Ra, the host antagonist of IL-1. Poxviruses
have also been shown to encode a soluble TNF-binding protein and a soluble IL-1-binding protein that block the ability
of the bound cytokines to elicit a response. Since both
TNF- and IL-1 are critical to the early phases of an inflammatory, antiviral response, these soluble cytokine-binding
proteins may allow the viruses an increased time window in
which to replicate.
Yet other viruses produce molecules that inhibit the production of cytokines. One such example is the cytokine
response modifier (Crm) protein of the cowpox virus, which
inhibits the production of caspase-1 and hence prevents the
processing of IL-1 precursor proteins. Finally, some viruses
produce soluble chemokines and chemokine-binding proteins that interfere with normal immune cell trafficking, and
allow the producing viruses and virally infected cells to
evade an immune response. Table 4-6 lists a number of viral
products that inhibit cytokines, chemokines, and their
activities.
Cytokine-Related Diseases
Defects in the complex regulatory networks governing the
expression of cytokines and cytokine receptors have been
implicated in a number of diseases. Genetic defects in cytokines, their receptors, or the molecules involved in cytokinedirected signal transduction lead to immunodeficiencies
such as those described in Chapter 18. Other defects in the
cytokine network can cause an inability to defend against
specific families of pathogens. For example, people with a
defective receptor for IFN- are susceptible to mycobacterial
Receptors and Signaling: Cytokines and Chemokines
TABLE 4-6
|
CHAPTER 4
135
Viruses use many different strategies to evade cytokine-mediated immune
mechanisms
Virus
Virally encoded proteins
Epstein-Barr Virus (EBV), Cytomegalovirus
IL-10 homolog
Vaccinia virus, Variola virus
Soluble IL-1 receptors
Myxoma virus
Soluble IFN- receptor
Variola virus
Soluble TNF receptors
Adenovirus
RID complex proteins induce internalization of Fas receptor
Measles virus
Viral hemagglutinin binds to complement receptor, CD46, signaling disruption of
IL-12 production and therefore inhibition of TH1 pathway differentiation
Herpes simplex virus
Reverses translation block induced by Type 1 interferons
Adenovirus
Blocks interferon-induced JAK/STAT signaling
infections that rarely occur in the general population. In
addition to the diseases rooted in genetic defects in cytokine
activity, a number of other pathologic states result from
overexpression or underexpression of cytokines or cytokine
receptors. Several examples of these diseases are given below,
followed by an account of therapies aimed at preventing the
potential harm caused by cytokine activity.
Septic Shock Is Relatively Common
and Potentially Lethal
Despite the widespread use of antibiotics, bacterial infections remain a major cause of septic shock, which may
develop a few hours after infection by certain bacteria,
including Staphyloccocus aureus, E. coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, Enterobacter aerogenes,
and Neisseria meningitidis.
Bacterial septic shock is one of the conditions that falls
under the general heading of sepsis. Sepsis, in turn, may be
caused not only by bacterial infection but also by trauma,
injury, ischemia (decrease in blood supply to an organ or a
tissue), and certain cancers. Sepsis is the most common
cause of death in U.S. hospital intensive-care units and the
13th leading cause of death in the United States. A common
feature of sepsis, whatever the underlying cause, is an overwhelming production of proinflammatory and fever-inducing
cytokines such as TNF- and IL-1. The cytokine imbalance
induces abnormal body temperature, alterations in the respiratory rate, and high white blood cell counts, followed by
capillary leakage, tissue injury, widespread blood clotting,
and lethal organ failure.
Bacterial septic shock often develops because bacterial cell
wall endotoxins bind to innate immune system pathogen
receptors, such as Toll-like receptors (see Chapter 5), on dendritic cells and macrophages, causing them to produce IL-1
and TNF- at levels that lead to pathological capillary perme-
ability and loss of blood pressure. A condition resembling
bacterial septic shock can be produced in the absence of any
bacterial infection simply by injecting mice with recombinant TNF-. Several studies offer hope that neutralizing
TNF- or IL-1 activity with monoclonal antibodies or antagonists will prevent fatal shock from developing. In one such
investigation, monoclonal antibodies to TNF- protected
animals from endotoxin-induced shock. In another, injection
of a recombinant IL-1 receptor antagonist (IL-1Ra), which
prevents binding of IL-1 to the IL-1 receptor as described
above, resulted in a threefold reduction in mortality.
However, neutralization of TNF- does not reverse the
progression of septic shock in all cases, and antibodies
against TNF- give little benefit to patients with advanced
disease. Recent studies in which the cytokine profiles of
patients with septic shock were followed over time shed
some light on this apparent paradox.
The increases in TNF- and IL-1 occur rapidly in early
sepsis, so neutralizing these cytokines is most beneficial
early in the process. Indeed, in animal experiments, early
intervention can prevent sepsis altogether. However, approximately 24 hours following the onset of sepsis, the levels of
TNF- and IL-1 fall dramatically, and other factors become
more important. Cytokines critical in the later stages of sepsis include IL-6, MIF, and CCL-8. Sepsis remains an area of
intense investigation, and clarification of the process involved
in bacterial septic shock and other forms of sepsis can be
expected to lead to advances in therapies for this major killer
in the near future.
Bacterial Toxic Shock Is Caused by
Superantigen Induction of T-Cell
Cytokine Secretion
A variety of microorganisms produce toxins that act as superantigens. As described in Chapters 9 and 11, superantigens
136
PA R T I
|
Introduction
BOX 4-4
CLINICAL FOCUS
Cytokines and Obesity
Even as those in the Third World
suffer from malnutrition, among the leading current causes of disease and death in
developed countries are obesity and its
corollary, Type 2 diabetes.
Type 1, or juvenile onset, diabetes has
long been known to have an autoimmune
etiology (cause). Cells of the adaptive
immune system kill the cells of the Islets
of Langerhans in the pancreas, leading to
the complete absence of insulin in the
diabetic patient, who must rely on exogenously delivered hormone to survive.
In Type 2 diabetes, the body fails to
respond to insulin in an appropriate manner, and the cells fail to take in glucose
from the blood and into the tissues. Just as
in Type 1 diabetes, the presence of high
levels of glucose in the blood and tissue
fluids facilitates nonenzymatic glucose
conversions to reactive carbohydrate
derivatives such as glyoxals. These derivatives cross-link proteins and carbohydrates
in the membranes and walls of blood vessels and neurons and within the extracellular matrix, leading to the familiar array of
diabetic symptoms: poor peripheral circulation, plaque buildup in the arteries, and
heart disease. However, we are now beginning to understand that Type 2, just like
Type 1, diabetes is a disease closely related
to the workings of the immune system.
In 1993, a seminal publication by
Hotamisligil and colleagues made the
link between inflammation and metabolic conditions such as Type 2 diabetes
and obesity. These authors demonstrated
that adipocytes (fat cells) constitutively
express the proinflammatory cytokine
TNF-, and that TNF- expression in adipocytes of obese animals is markedly
increased. Later research demonstrated
that this finding could be extended to
humans. The adipose tissue of human
subjects was found to constitutively
express TNF-, and blood levels of TNF-
fall after weight loss.
But is this increase in TNF- expression
affecting insulin sensitivity? On binding of
insulin to its receptor, tyrosine kinase
activity in the cytoplasmic region of the
receptor is activated. This Receptor Tyrosine Kinase (RTK) then phosphorylates
both itself (autophosphorylation) and
some nearby proteins. The signaling cascade is initiated from the insulin receptor
by the binding of adapter proteins such as
Grb2 and IRS-1 via their SH2 domains to
the phosphorylated tyrosine sites on the
receptor molecule. We now know that
TNF- signaling in adipocytes inhibits the
autophosphorylation of tyrosine residues
of the insulin receptor and instead induces
phosphorylation of serine residues of
both the insulin receptor and the IRS-1
adapter. The serine phosphorylation
inhibits any subsequent phosphorylation
at tyrosine residues, and thus the passage
bind to MHC Class II molecules at a location in the MHC
molecule that is outside the groove normally occupied by
antigenic peptides (see Figure 11-6). They then bind to a part
of the V chain of the T-cell receptor that is outside the normal antigen-binding site, and this binding is sufficient to
trigger T-cell activation. This means that a given superantigen can simultaneously activate all T cells bearing a particular V domain. Because of their unique binding ability,
superantigens can activate large numbers of T cells irrespective of the antigenic specificity of their canonical antigenbinding site.
of signals from insulin to the interior of the
fat cell is prevented. Recently, the interleukin IL-6 has also been shown to inhibit
insulin signal transduction in hepatocytes
(liver cells) through a similar mechanism.
The decrease in effectiveness of insulin
signaling then becomes a self-reinforcing
problem, as insulin signaling itself is antiinflammatory, and so any decrease in the
insulin signal can give rise to inflammatory side effects.
However, TNF- and IL-6 are not the
only cytokines implicated in the etiology
of Type 2 diabetes. With the discovery
and characterization of the proinflammatory cytokine family represented by IL-17,
interest has arisen in the relationship
between members of this family, obesity,
and the control of fat cell metabolism.
Since IL-6 is implicated in the differentiation of T lymphocytes to secrete IL-17,
obesity and its associated inflammation
tend to predispose an individual to
secrete IL-17. However, again we find ourselves in a positive feedback loop, as IL-17
acts on monocytes to induce the further
secretion of IL-6, thus ensuring the maintenance of an inflammatory state.
It is therefore clear that cytokine signaling plays a profound role in a disease
that is emblematic of our time, and which
is predicted to afflict close to 40% of the
U.S. population by the middle of the next
decade.
Although less than 0.01% of T cells respond to a given
conventional antigen (see Chapter 11), 5% or more of the
T-cell population can respond to a given superantigen.
Bacterial superantigens have been implicated as the causative agent of several diseases, such as bacterial toxic
shock and food poisoning. Included among these bacterial
superantigens are several enterotoxins, exfoliating toxins,
and toxic shock syndrome toxin (TSST1) from Staphylococcus aureus; pyrogenic exotoxins from Streptococcus
pyrogenes; and Mycoplasma arthritidis supernatant (MAS).
The large number of T cells activated by these superantigens
Receptors and Signaling: Cytokines and Chemokines
results in excessive production of cytokines. The TSST1,
for example, has been shown to induce extremely high
levels of TNF- and IL-1. As in bacterial septic shock,
these elevated concentrations of cytokines can induce systemic reactions that include fever, widespread blood clotting, and shock.
In addition to those diseases described above, in which
cytokines or their receptors are directly implicated, recent
information has indicated the importance of cytokine
involvement in the most important public health crisis currently afflicting the developed world: the increasing incidence of Type 2 diabetes. The roles of TNF- and IL-6 in the
induction and maintenance of this disease are described in
Box 4-4.
Cytokine Activity Is Implicated in Lymphoid
and Myeloid Cancers
Abnormalities in the production of cytokines or their
receptors have been associated with some types of cancer.
For example, abnormally high levels of IL-6 are secreted
by cardiac myxoma (a benign heart tumor), myeloma and
plasmacytoma cells, as well as cervical and bladder cancer
cells. In myeloma and plasmacytoma cells, IL-6 appears to
operate in an autocrine manner to stimulate cell proliferation. When monoclonal antibodies to IL-6 are added to in
vitro cultures of myeloma cells, their growth is inhibited.
In contrast, transgenic mice that express high levels of
IL-6 have been found to exhibit a massive, fatal, plasmacell proliferation, called plasmacytosis. In addition, as
described above, high serum concentrations of the sIL-2R
are found in patients suffering from various blood cell
cancers, which may impede a vigorous anti-tumor
response.
Cytokine Storms May Have Caused Many
Deaths in the 1918 Spanish Influenza
Occasionally, a particularly virulent infection may induce
the secretion of extremely high levels of cytokines, that
then feed back on the immune cells to elicit yet more cytokines. Normally, these positive feedback loops represent
effective modes of immune amplification; they are themselves usually kept in check by self-regulating immune
mechanisms, such as the activation of regulatory T cells
(see Chapter 11). However, some viruses cause a localized,
exaggerated response, resulting in the secretion of extraordinarily high levels of cytokines. If this occurs in the lungs,
for example, the localized swelling, inflammation, and
increase in capillary permeability can lead to the accumulation of fluids and leukocytes that block the airways,
thereby causing exacerbation of symptoms, or even death,
before the cytokine levels can be controlled. It is unclear
why some viruses induce these cytokine storms and others
do not.
|
CHAPTER 4
137
Historical documents detailing the symptoms of the 1918
Spanish influenza suggest that the massive fatalities associated with that pandemic most likely resulted from cytokine
storms, and there is some evidence that the severe acute
respiratory syndrome (SARS) epidemic of 1993 may have
caused a similar, unregulated, immune cell cytokine secretion. Transplant surgeons also observe this phenomenon on
occasions when leukocytes associated with a graft—often a
bone marrow transplant—mount an immune response
against the host. Effective treatments for patients undergoing
cytokine storms are still being developed. Currently, patients
are offered steroidal and nonsteroidal anti-inflammatory
medications, but other drugs that are more specifically
directed at the reduction of cytokine secretion and/or activity are being tested.
Cytokine-Based Therapies
The availability of purified cloned cytokines, monoclonal
antibodies directed against cytokines, and soluble cytokine
receptors offers the prospect of specific clinical therapies to
modulate the immune response. Cytokines such as interferons (see Clinical Focus Box 4-2), colony-stimulating factors
such as G-CSF, and IL-2 have all been used clinically. In
addition, several reagents that specifically block the proinflammatory effects of TNF- have proven to be therapeutically useful in certain diseases. Specifically, soluble TNF-
receptor (Enbrel) and monoclonal antibodies against TNF-
(Remicade and Humira) have been used to treat rheumatoid
arthritis and ankylosing spondylitis in more than a million
patients. These anti–TNF- drugs reduce proinflammatory
cytokine cascades; help to alleviate pain, stiffness, and joint
swelling; and promote healing and tissue repair. In addition,
as described above, the recombinant form of IL-1Ra—
anakinra (Kineret)—has been shown to be relatively effective in the treatment of rheumatoid arthritis. Monoclonal
antibodies directed against the chain of the IL-2R—
basiliximab (Simulect) and daclizumab (Zenapax)—are also
in clinical use for the prevention of transplantation rejection
reactions.
As powerful as these reagents may be, interfering with
the normal course of the immune response is not without
its own intrinsic hazards. Reduced cytokine activity brings
with it an increased risk of infection and malignancy, and
the frequency of lymphoma is higher in patients who are
long-term users of the first generation of TNF- blocking
drugs.
In addition, the technical problems encountered in
adapting cytokines for safe, routine medical use are far
from trivial. As described above, during an immune
response, interacting cells may produce extremely high
local concentrations of cytokines in the vicinity of target
cells, but achieving such high concentrations over a clinically significant time period, when cytokines must be
138
PA R T I
|
Introduction
administered systemically, is difficult. Furthermore, many
cytokines have a very short half-life—recombinant human
IL-2 has a half-life of only 7 to 10 minutes when administered intravenously—so frequent administration may be
required. Finally, cytokines are extremely potent biological
response modifiers, and they can cause unpredictable and
undesirable side effects. The side effects from administration of recombinant IL-2, for example, range from mild
(e.g., fever, chills, diarrhea, and weight gain) to serious
(e.g., anemia, thrombocytopenia, shock, respiratory distress, and coma).
The use of cytokines and anti-cytokine therapies in
clinical medicine holds great promise, and efforts to
develop safe and effective cytokine-related strategies continue, particularly in those areas of medicine that have so
far been resistant to more conventional approaches, such as
inflammation, cancer, organ transplantation, and chronic
allergic disease.
S U M M A R Y
■
■
■
■
■
■
■
■
■
■
Cytokines are proteins that mediate the effector functions
of the immune system.
Most cytokines are soluble proteins, but some—for example, members of the TNF family, may be expressed in a
membrane-bound form.
Some cytokines are secreted following stimulation of the
innate immune system (e.g., IL-1, TNF-, CXCL8), whereas
others are secreted by the T and B lymphocytes of the
adaptive immune system (IL-2, IL-4, IL-17).
Cytokines bind to receptors on the plasma membrane and
elicit their effects through the activation of an intracellular
signaling cascade.
Cytokines can effect alterations in the differentiative, proliferative, and survival capacities of their target cells.
Cytokines exhibit the properties of redundancy, pleiotropy, synergy, antagonism, and cascade induction.
The levels of expression of cytokine receptors on the cell surface may change according to the activation status of a cell.
There are six families of cytokines with associated receptors, distinguished on the basis of the structures of the
cytokines and the receptor molecules, and on the nature of
their signaling pathways.
IL-1 family members interact with dimeric receptors to
induce responses that are primarily proinflammatory. The
physiological responses to some IL-1 family members are
modulated by the presence of soluble forms of the receptors and soluble cytokine-binding proteins.
The Hematopoietin (Class I cytokine) family is the largest
family of cytokines, and members mediate diverse effects,
including proliferation, differentiation, and antibody
secretion. The Hematopoietin family members share a
common, four-helix bundle structure.
■
■
■
■
■
■
■
■
■
■
Receptors for cytokines from the Hematopoietin family
are classified into three subgroups—the , , or gp130
receptors—each of which shares a common signaling
chain.
The Interferon (Class II cytokine) family includes the Type
I interferons (interferon and interferon ), which were
the first cytokines to be discovered and mediate early antiviral responses.
Type II interferons (interferon ) activate macrophages,
interact with cells of the adaptive immune system and support the generation of TH1 cells.
The TNF family of cytokines act as trimers and may occur
in either soluble or membrane-bound forms.
FasL, a TNF family member, interacts with its receptor,
Fas, to stimulate apoptosis in the recipient cell. This interaction is important at the close of the immune response.
TNF interacts with the TNF-R1 receptor on the surface of
the cell to induce either apoptosis or survival, depending
on the physiological environment.
The IL-17 family of cytokines has been defined quite
recently, and its members are primarily proinflammatory
in action.
Chemokines act on GPCR-coupled receptors to promote
chemoattraction, the movement of immune system cells
into, within, and out of lymphoid organs.
Naturally occurring and pathogen-derived inhibitors of
cytokine function may modulate their activity in vivo.
Levels of inflammatory cytokines such as IL-1, IL-6, and
TNF may be increased in certain disease states such as
rheumatoid arthritis, and such diseases are susceptible to
treatment with drugs that inhibit cytokine activities.
R E F E R E N C E S
Abram, C. L., and C. A. Lowell. 2009. The ins and outs of leukocyte
integrin signaling. Annual Review of Immunology 27:339–362.
Alcami, A. 2003. Structural basis of the herpesvirus M3-chemokine interaction. Trends in Microbiology 11:191–192.
Ahmed, M., and S. L. Gaffen. 2010. IL-17 in obesity and adipogenesis. Cytokine & Growth Factor Reviews 21:449–453.
Arend, W.P., G. Palmer, and C. Gabay. 2008. IL-1, IL-18, and
IL-33 families of cytokines. Immunological Reviews 223:20–38.
Receptors and Signaling: Cytokines and Chemokines
|
CHAPTER 4
139
Botto, S., et al. 2011. IL-6 in human cytomegalovirus secretome
promotes angiogenesis and survival of endothelial cells
through the stimulation of survivin. Blood 117:352–361.
Micheau, M., and J. Tschopp. 2003. Induction of TNF Receptor
1-mediated apoptosis via two sequential signaling complexes.
Cell 114:181–190.
Boulanger, M. J., and K. C. Garcia. 2004. Shared cytokine signaling receptors: Structural insights from the gp130 system.
Advances in Protein Chemistry 68:107–146.
Raab, M., et al. 2010. T cell receptor “inside-out” pathway via
signaling module SKAP1-RapL regulates T cell motility and
interactions in lymph nodes. Immunity 32:541–556.
Crabtree, G. R., S. Gillis, K. A. Smith, and A. Munck. 1980.
Mechanisms of glucocorticoid-induced immunosuppression:
Inhibitory effects on expression of Fc receptors and production of T-cell growth factor. Journal of Steroid Biochemistry
12:445–449.
Raman, D., T. Sobolik-Delmaire, and A. Richmond. 2011. Chemokines in health and disease. Experimental Cell Research
317:5755–89.
Crispin, J. C., and G. C. Tsokos. 2009. Transcriptional regulation of IL-2 in health and autoimmunity. Autoimmunity
Reviews 8:190–195.
Dandona, P., A. Aljada, and A. Bandyopadhyay. 2004. Inflammation: The link between insulin resistance, obesity and diabetes. Trends in Immunology 25:4–7.
Dandona, P., A. Aljada, A. Chaudhuri, P. Mohanty, and G.
Rajesh. 2004. A novel view of metabolic syndrome. Metabolic
Syndrome and Related Disorders 2:2–8.
Eisenbarth, S. C., and R. A. Flavell. 2009. Innate instruction of
adaptive immunity revisited: The inflammasome. EMBO
Molecular Medicine 1:92–98.
Fickenscher, H., et al. 2002. The interleukin-10 family of cytokines. Trends in Immunology 23:89–96.
Gaffen, S. L. (2009). Structure and signalling in the IL-17 receptor family. Nature Reviews. Immunology 9:556–567.
Gee, K., C. Guzzo, N. F. Che Mat, W. Ma, and A. Kumar. 2009.
The IL-12 family of cytokines in infection, inflammation and
autoimmune disorders. Inflammation & Allergy Drug Targets
8:40–52.
Gillis, S., R. Mertelsmann, and M. A. Moore. 1981. T-cell
growth factor (interleukin 2) control of T-lymphocyte proliferation: Possible involvement in leukemogenesis. Transplantation Proceedings 13:1884–1890.
Guilherme, A., J. V. Virbasius, V. Puri, and M. P. Czech. 2008.
Adipocyte dysfunctions linking obesity to insulin resistance
and type 2 diabetes. Nature Reviews. Molecular Cell Biology
9:367–377.
Horng, T., and G. S. Hotamisligil. 2011. Linking the inflammasome to obesity-related disease. Nature Medicine 17:164–165.
Hotamisligil, G. S. 2003. Inflammatory pathways and insulin
action. International Journal of Obesity and Related Metabolic
Disorders 27:Suppl 3, S53–55.
Hotamisligil, G. S., N. S. Shargill, and B. M. Spiegelman. 1993.
Adipose expression of tumor necrosis factor-alpha: Direct role
in obesity-linked insulin resistance. Science 259:87–91.
Rochman, Y., R. Spolski, and W. J. Leonard. 2009. New insights
into the regulation of T cells by gamma(c) family cytokines.
Nature Reviews. Immunology 9:480–490.
Rot, A., and U. H. von Andrian. 2004. Chemokines in innate
and adaptive host defense: Basic chemokinese grammar for
immune cells. Annual Review of Immunollogy 22:891–928.
Sadler, A. J., and B. R. Williams. 2008. Interferon-inducible
antiviral effectors. Nature Reviews. Immunology 8:559–568.
Skalnikova, H., J. Motlik, S. J. Gadher, and H. Kovarova. Mapping of the secretome of primary isolates of mammalian cells,
stem cells and derived cell lines. Proteomics 11:691–708.
Skiniotis, G., P. J. Lupardus, M. Martick, T. Walz, T., and K. C.
Garcia. 2008. Structural organization of a full-length gp130/
LIF-R cytokine receptor transmembrane complex. Molecular
Cell 31:737–748.
Smith, K. A., P. E. Baker, S. Gillis, and F. W. Ruscetti. (1980).
Functional and molecular characteristics of T-cell growth factor. Molecular Immunology 17:579–589.
Stull, D., and S. Gillis. 1981. Constitutive production of interleukin 2 activity by a T cell hybridoma. Journal of Immunology
126:1680–1683.
Walczak, H. 2011. TNF and ubiquitin at the crossroads of gene
activation, cell death, inflammation and cancer. Immunological Reviews 244:9–28.
Wellen, K. E., and G. S. Hotamisligil. 2003. Obesity-induced
inflammatory changes in adipose tissue. The Journal of Clinical Investigation 112:1785–1788.
Useful Web Sites
www.invitrogen.com
www.miltenyibiotec.com/cytokines
www.prospecbio.com/Cytokines
www.peprotech.com
www.rndsystems.com Many companies that sell recom-
Jones, L. L., and D. A. Vignali. 2011. Molecular interactions
within the IL-6/IL-12 cytokine/receptor superfamily. Immunologic Research 51:5–14.
binant cytokines or cytokine-related products provide useful
information on their websites, or in print copy. The preceding are a few that are particularly helpful.
Li, W. X. 2008. Canonical and non-canonical JAK-STAT signaling. Trends in Cell Biology 18:545–551.
www.jakpathways.com/understandingjakpathways A useful animation of JAK-STAT signaling.
140
PA R T I
|
Introduction
www.youtube.com/watch?vZUUfdP87Ssg&featu
rerelated A rather wonderful movie of chemotaxis.
You Tube. This one shows a macrophage recognizing a
pathogen and releasing cytokines in response.
www.youtube.com/watch?vEpC6G_DGqkI&
featurerelated A striking movie of a neutrophil chas-
www.netpath.org A curated set of pathways, with infor-
ing a bacterium.
mation on interacting proteins. Many interleukin pathways
are included.
www.youtube.com/watch?vKiLJl3NwmpU An
increasing number of medical animations are available on
S T U D Y
Q U E S T I O N S
1. Distinguish between a hormone, a cytokine, a chemokine,
and a growth factor. What functional attributes do they
share, and what properties can be used to discriminate
among them?
2. Measurement of the blood concentration of a particular
cytokine reveals that it is rarely present above 1010 M, even
under the conditions of an ongoing immune response.
However, when you measure the affinity of the cognate
receptor, you discover that its dissociation constant is close
to 108 M, implying that the receptor occupancy must
rarely exceed 1%. How do you account for this discrepancy?
3. Describe how dimerization and phosphorylation of intra-
cellular signaling molecules contribute to activation of cells
by Type 1 cytokines.
4. Define the terms pleiotropy, synergy, redundancy, antago-
nism, and cascade induction as they apply to cytokine
action.
system ensure that only T cells that have been stimulated
by antigen are susceptible to IL-2 signaling?
10b. The following diagram represents the results of a flow
cytometry experiment in which mouse spleen cells were
stained with antibodies directed against different components of the IL-2R. The more antibody that binds to the
cells, the further they move along the relevant axis. The
number of cells stained with fluorescein-conjugated anti IL-2R antibodies are shown along the x-axis of the flow
cytometry plot, and cells that stain with phycoerythrinlabeled antibodies to the subunit of the IL-2 receptor
move along the Y-axis. We have drawn for your reference a
circle that represents cells that stain with neither antibody.
On this plot, draw, as circles, and label where you
would expect to find the populations representing unstimulated T cells and T cells after antigen activation, after
treatment with the two fluorescent labels described above.
tion of the location of a lymphocyte?
6. Cytokines signaling through the Class I cytokine receptors
can compete with one another, even though the recognition units of the receptors are different. Explain.
7. Describe one mechanism by which Type I interferons
“interfere” with the production of new viral particles.
8. Signaling by tumor necrosis factor can paradoxically lead
to cell activation or cell death. Explain how, by drawing
diagrams of the relevant signaling pathways.
9. Describe two examples of ways in which vertebrates tune
Populations staining with anti-IL-2Rα
antibodies labeled with phycoerythrin
5. How might receipt of a cytokine signal result in the altera-
down the intensity of their own cytokine signaling.
10a. The cytokine IL-2 is capable of activating all T cells to
proliferation and differentiation. How does the immune
Populations staining with anti-IL-2Rβ and
anti-IL-2Rγ antibodies labeled with fluorescein
5
Innate Immunity
V
ertebrates are protected by both innate
immunity and adaptive immunity. In contrast
to adaptive immune responses, which take
days to arise following exposure to antigens,
innate immunity consists of the defenses against
infection that are ready for immediate action when a
host is attacked by a pathogen (viruses, bacteria, fungi,
or parasites). The innate immune system includes
anatomical barriers against infection—both physical
and chemical—as well as cellular responses (Overview
Figure 5-1). The main physical barriers—the body’s
first line of defense—are the epithelial layers of the skin
and of the mucosal and glandular tissue surfaces
connected to the body’s openings; these epithelial
barriers prevent infection by blocking pathogens from
entering the body. Chemical barriers at these surfaces
include specialized soluble substances that possess
antimicrobial activity as well as acid pH. Pathogens that
breach the physical and chemical barriers due to
damage to or direct infection of the epithelial cell layer
can survive in the extracellular spaces (some bacteria,
fungi, and parasites) or they can infect cells (viruses
and some bacteria and parasites), eventually replicating
and possibly spreading to other parts of the body.
The cellular innate immune responses to invasion by
an infectious agent that overcomes the initial epithelial
barriers are rapid, typically beginning within minutes of
invasion. These responses are triggered by cell surface or
intracellular receptors that recognize conserved molecular
components of pathogens. Some white blood cell types
(macrophages and neutrophils) are activated to rapidly
engulf and destroy extracellular microbes through the
process of phagocytosis. Other receptors induce the
production of proteins and other substances that have a
variety of beneficial effects, including direct antimicrobial
activity and the recruitment of fluid, cells, and molecules
to sites of infection. This influx causes swelling and other
physiological changes that collectively are called
inflammation. Such local innate and inflammatory
responses usually are beneficial for eliminating pathogens
and damaged or dead cells and promoting healing. For
example, increased levels of antimicrobial substances and
A macrophage (yellow) on the surface of a
blood vessel (red) binds and phagocytoses
bacteria (orange). [Dennis Kunkel Microscopy,
Inc./Visuals Unlimited, Inc.]
■
Anatomical Barriers to Infection
■
Phagocytosis
■
Induced Cellular Innate Responses
■
Inflammatory Responses
■
Natural Killer Cells
■
■
■
Regulation and Evasion of Innate and Inflammatory
Responses
Interactions Between the Innate and Adaptive
Immune Systems
Ubiquity of Innate Immunity
phagocytic cells help to eliminate the pathogens, and
dendritic cells take up pathogens for presentation to
lymphocytes, activating adaptive immune responses.
Natural killer cells recruited to the site can recognize and kill
virus-infected, altered, or stressed cells. However, in some
situations these innate and inflammatory responses can be
harmful, leading to local or systemic consequences that can
cause tissue damage and occasionally death. To prevent
these potentially harmful responses, regulatory mechanisms
have evolved that usually limit such adverse effects.
Despite the multiple layers of the innate immune
system, some pathogens may evade the innate defenses.
On call in vertebrates is the adaptive immune system,
which counters infection with a specific tailor-made
141
142
PA R T I I
|
Innate Immunity
5-1
OVERVIEW FIGURE
Innate Immunity
Physical barriers
to infection
Epithelial layers
of skin and
mucosal/glandular tissues
H+
H+
Chemical barriers
to infection
Pathogens
H+
H+
Acidic pH and
anti-microbial
proteins and peptides
H+
+
H
+
H
H+
Damage/infection
Cellular responses to infection
Pathogens
Macrophage
Infected cell
Phagocytosis
and
degradation
Dendritic cell
Killing by
NK cell
Binding
to cells
Activation
of adaptive
immune
responses
NK cell
Anti-microbial
substances
(e.g., peptides,
interferons)
Pathogen
elimination
Cell activation
Inflammation:
recruitment and
activation of
protective cells and
molecules (e.g.,
complement) to the
infection site
Key elements of innate immunity include the physical and chemical barriers that prevent infection, provided by the epithelial cell
layers of the skin, mucosal tissues (e.g., gastrointestinal, respiratory,
and urogenital tracts), and glandular tissues (e.g., salivary, lacrimal,
and mammary glands). Once pathogens enter the body, such as
through a breach in an epithelial layer, they are confronted by an
array of cells with cell surface and intracellular receptors that recognize pathogen components and trigger a variety of cellular
responses. Pathogen recognition by these receptors activates some
cells to phagocytose and degrade the pathogen, and many cells
response to the attacking pathogen. This attack occurs in
the form of B and T lymphocytes, which generate
antibodies, and effector T cells that specifically recognize
and neutralize or eliminate the invaders. Table 5-1
compares innate and adaptive immunity. While innate
Antibodies
T-cell
responses
Cytokines
and
chemokines
Systemic
effects
(e.g., fever)
are activated through their receptors to produce a variety of antimicrobial substances that kill pathogens, as well as cytokine and
chemokine proteins that recruit cells, molecules, and fluid to the
site of infection, leading to swelling and other symptoms collectively known as inflammation. The innate natural killer (NK) cells
recognize and kill some virus-infected cells. Cytokines and chemokines can cause systemic effects that help to eliminate an infection,
and also contribute—along with dendritic cells that carry and present pathogens to lymphocytes—to the activation of adaptive
immune responses.
immunity is the most ancient form of defense, found in
all multicellular plants and animals, adaptive immunity is
a much more recent evolutionary invention, having arisen
in vertebrates. In these animals, adaptive immunity
complements a well-developed system of innate immune
Innate Immunity
TABLE 5-1
|
CHAPTER 5
143
Innate and adaptive immunity
Attribute
Innate immunity
Adaptive immunity
Response time
Minutes/hours
Days
Specificity
Specific for molecules and molecular
patterns associated with pathogens
and molecules produced by
dead/damaged cells
Highly specific; discriminates between even
minor differences in molecular structure of
microbial or nonmicrobial molecules
Diversity
A limited number of conserved, germ line–
encoded receptors
Highly diverse; a very large number of
receptors arising from genetic recombination
of receptor genes in each individual
Memory responses
Some (observed in invertebrate innate
responses and mouse/human NK cells)
Persistent memory, with faster response of
greater magnitude on subsequent exposure
Self/nonself discrimination
Perfect; no microbe-specific self/nonself
patterns in host
Very good; occasional failures of
discrimination result in autoimmune disease
Soluble components of blood
Many antimicrobial peptides, proteins, and
other mediators
Antibodies and cytokines
Major cell types
Phagocytes (monocytes, macrophages,
neutrophils), natural killer (NK) cells, other
leukocytes, epithelial and endothelial cells
T cells, B cells, antigen-presenting cells
mechanisms that share important features with those of
our invertebrate ancestors. A large and growing body of
research has revealed that as innate and adaptive
immunity have co-evolved in vertebrates, a high degree of
interaction and interdependence has arisen between the
two systems. Recognition by the innate immune system
not only kicks off the adaptive immune response but also
helps to ensure that the type of adaptive response
generated will be effective for the invading pathogen.
This chapter describes the components of the innate
immune system—physical barriers, chemical agents, and
a battery of protective cellular responses carried out by
numerous cell types—and illustrates how they act
together to defend against infection. We conclude with an
overview of innate immunity in plants and invertebrates.
Anatomical Barriers to Infection
The most obvious components of innate immunity are the
external barriers to microbial invasion: the epithelial layers
that insulate the body’s interior from the pathogens of the
exterior world. These epithelial barriers include the skin and
the tissue surfaces connected to the body’s openings: the
mucous epithelial layers that line the respiratory, gastrointestinal, and urogenital tracts and the ducts of secretory glands
such as the salivary, lacrimal, and mammary glands (which
produce saliva, tears, and milk, respectively) (Figure 5-2).
Skin and other epithelia provide a kind of living “plastic
wrap” that encases and protects the inner domains of the
body from infection. But these anatomical barriers are more
than just passive wrappers. They contribute to physical and
mechanical processes that help the body shed pathogens and
also generate active chemical and biochemical defenses by
synthesizing and deploying molecules, including peptides
and proteins, that have or induce antimicrobial activity.
Epithelial Barriers Prevent Pathogen Entry
into the Body’s Interior
The skin, the outermost physical barrier, consists of two distinct layers: a thin outer layer, the epidermis, and a thicker
layer, the dermis. The epidermis contains several tiers of tightly
packed epithelial cells; its outer layer consists of mostly dead
cells filled with a waterproofing protein called keratin. The
dermis is composed of connective tissue and contains blood
vessels, hair follicles, sebaceous glands, sweat glands, and scattered myeloid leukocytes such as dendritic cells, macrophages,
and mast cells. In place of skin, the respiratory, gastrointestinal,
and urogenital tracts and the ducts of the salivary, lacrimal, and
mammary glands are lined by strong barrier layers of epithelial
cells stitched together by tight junctions that prevent pathogens
from squeezing between them to enter the body.
A number of nonspecific physical and chemical defense
mechanisms also contribute to preventing the entry of pathogens through the epithelia in these secretory tissues. For
example, the secretions of these tissues (mucus, urine, saliva,
tears, and milk) wash away potential invaders and also contain antibacterial and antiviral substances. Mucus, the viscous fluid secreted by specialized cells of the mucosal
epithelial layers, entraps foreign microorganisms. In the
144
PA R T I I
|
Innate Immunity
Organ or tissue
Innate mechanisms protecting skin/epithelium
Skin
Antimicrobial peptides, fatty acids in sebum
Mouth and upper
alimentary canal
Enzymes, antimicrobial peptides, and sweeping of
surface by directional flow of fluid toward stomach
Stomach
Low pH, digestive enzymes, antimicrobial peptides,
fluid flow toward intestine
Small intestine
Digestive enzymes, antimicrobial peptides,
fluid flow to large intestine
Large intestine
Normal intestinal flora compete with invading
microbes, fluid/feces expelled from rectum
Airway and lungs
Cilia sweep mucus outward, coughing, sneezing
expel mucus, macrophages in alveoli of lungs
Urogenital tract
Flushing by urine, aggregation by urinary mucins;
low pH, anti-microbial peptides, proteins in vaginal
secretions
Salivary, lacrimal,
and mammary
glands
Flushing by secretions; anti-microbial peptides and
proteins in vaginal secretions
Skin
Lacrimal
glands
Mouth
Salivary
glands
Airway
Lung
Epithelial lining of
airway and lung
Epithelial lining of
alimentary canal
Mammary glands
Stomach
Large
intestine
Small
intestine
Rectum
Urogenital tract
In addition to serving as physical barriers, the skin and mucosal and
glandular epithelial layers are defended against microbial coloniza-
tion by a variety of mechanisms: mechanical (cilia, fluid flow, smooth
muscle contraction), chemical (pH, enzymes, antimicrobial peptides),
and cellular (resident macrophages and dendritic cells).
lower respiratory tract, cilia, hairlike protrusions of the cell
membrane, cover the epithelial cells. The synchronous movement of cilia propels mucus-entrapped microorganisms from
these tracts. Coughing is a mechanical response that helps us
get rid of excess mucus, with trapped microorganisms, that
occurs in many respiratory infections. The flow of urine
sweeps many bacteria from the urinary tract. With every
meal, we ingest huge numbers of microorganisms, but they
must run a gauntlet of defenses in the gastrointestinal tract
that begins with the antimicrobial compounds in saliva and
in the epithelia of the mouth and includes the hostile mix of
digestive enzymes and acid found in the stomach. Similarly,
the acidic pH of vaginal secretions is important in providing
protection against bacterial and fungal pathogens.
Some organisms have evolved ways to evade these
defenses of the epithelial barriers. For example, influenza
FIGURE 5-2 Skin and other epithelial barriers to infection.
Innate Immunity
FIGURE 5-3 Electron micrograph of E. coli bacteria adhering to the surface of epithelial cells of the urinary tract. E. coli is
an intestinal bacterial species that causes urinary tract infections
affecting the bladder and kidneys. [From N. Sharon and H. Lis, 1993,
Scientific American 268(1):85; courtesy of K. Fujita; Matthew A. Mulvey, Joel D.
Schilling, Juan J. Martinez, and Scott J. Hultgren, Bad bugs and beleaguered
bladders: Interplay between uropathogenic Escherichia coli and innate host
defenses. PNAS August 1, 2000 vol. 97 no. 16 8829-8835. Copyright 2000 National
Academy of Sciences, U.S.A.]
virus has a surface molecule that enables it to attach
firmly to cells in mucous membranes of the respiratory
tract, preventing the virus from being swept out by the
ciliated epithelial cells. Neisseria gonorrhoeae, the bacteria
that causes gonorrhea, binds to epithelial cells in the
mucous membrane of the urogenital tract. Adherence of
these and other bacteria to mucous membranes is generally mediated by hairlike protrusions on the bacteria
called fimbriae or pili that have evolved the ability to bind
to certain glycoproteins or glycolipids only expressed by
epithelial cells of the mucous membrane of particular tissues (Figure 5-3).
|
CHAPTER 5
145
(which also incudes calprotectin) with potent antibacterial
activity against Escherichia coli, an enteric (intestinal) bacterial
species. This finding answered a long-standing question: Why
is human skin resistant to colonization by E. coli despite exposure to it from fecal matter resulting from lack of cleanliness or
poor sanitation? As shown in Figure 5-4, incubation of E. coli
on human skin for as little as 30 minutes kills the bacteria.
Antimicrobial proteins can show some specificity toward particular pathogens. For example, psoriasin is a key antimicrobial
for E. coli on the skin (and also on the tongue), but, as shown
in Figure 5-4, it does not kill Staphylococcus aureus (a major
cause of food poisoning and skin infections), whereas the
related protein calprotectin kills S. aureus but not E. coli.
S. aureus
E. coli
Inoculation
30 minutes
Fresh culture plates
Incubate
Antimicrobial Proteins and Peptides Kill
Would-be Invaders
To provide strong defense at these barrier layers, epithelial
cells secrete a broad spectrum of proteins and peptides that
provide protection against pathogens. The capacity of skin
and other epithelia to produce a wide variety of antimicrobial
agents on an ongoing basis is important for controlling the
microbial populations on these surfaces, as breaks in these
physical barriers from wounds provide routes of infection
that would be readily exploited by pathogenic microbes if
not defended by biochemical means.
Among the antimicrobial proteins produced by the skin
and other epithelia in humans (Table 5-2), several are enzymes
and binding proteins that kill or inhibit growth of bacterial and
fungal cells. Lysozyme is an enzyme found in saliva, tears, and
fluids of the respiratory tract that cleaves the peptidoglycan
components of bacterial cell walls. Lactoferrin and calprotectin
are two proteins that bind and sequester metal ions needed by
bacteria and fungi, limiting their growth. Among the many
antibacterial agents produced by human skin, recent research
has identified psoriasin, a small protein of the S-100 family
FIGURE 5-4 Psoriasin prevents colonization of the skin by
E. coli. Skin secretes psoriasin, an antimicrobial protein that kills
E. coli. Fingertips of a healthy human were inoculated with Staphylococcus aureus (S. aureus) and E. coli. After 30 minutes, the fingertips
were pressed on a nutrient agar plate and the number of colonies of
S. aureus and E. coli determined. Almost all of the inoculated E. coli were
killed; most of the S. aureus survived. [Photograph courtesy of Nature
Immunology; from R. Gläser et al., 2005, Nature Immunology 6:57–64.]
146
TABLE 5-2
PA R T I I
|
Innate Immunity
Some human antimicrobial proteins and peptides at epithelial surfaces
Proteins and peptides*
Location
Antimicrobial activities
Lysozyme
Mucosal/glandular secretions (e.g., tears,
saliva, respiratory tract)
Cleaves glycosidic bonds of peptidoglycans in cell
walls of bacteria, leading to lysis
Lactoferrin
Mucosal/glandular secretions (e.g., milk,
intestine mucus, nasal/respiratory and
urogenital tracts)
Binds and sequesters iron, limiting growth of
bacteria and fungi; disrupts microbial membranes;
limits infectivity of some viruses
Secretory leukocyte protease
inhibitor
Skin, mucosal/glandular secretions
(e.g., intestines, respiratory, and urogenital
tracts, milk)
Blocks epithelial infection by bacteria, fungi,
viruses; antimicrobial
S100 proteins, e.g.:
- psoriasin
- calprotectin
Skin, mucosal/glandular secretions
(e.g., tears, saliva/tongue, intestine, nasal/
respiratory and urogenital tracts)
Defensins ( and )
Skin, mucosal epithelia (e.g., mouth,
intestine, nasal/respiratory tract,
urogenital tract)
Disrupt membranes of bacteria, fungi, protozoan
parasites, and viruses; additional toxic effects
intracellularly; kill cells and disable viruses
Cathelicidin (LL37)**
Mucosal epithelia (e.g., respiratory tract,
urogenital tract)
Disrupts membranes of bacteria; additional toxic
effects intracellularly; kills cells.
Surfactant proteins SP-A, SP-D
Secretions of respiratory tract, other
mucosal epithelia
Block bacterial surface components; promotes
phagocytosis
- Disrupts membranes, killing cells
- Binds and sequesters divalent cations
(e.g., manganese and zinc), limiting growth of
bacteria and fungi
*Examples listed in this table are all produced by cells in the epithelia of mucosal and glandular tissues; examples of prominent epithelial sites are listed. Most proteins
and peptides are produced constitutively at these sites, but their production can also be increased by microbial or inflammatory stimuli. Many are also produced constitutively in neutrophils and stored in granules. In addition, synthesis and secretion of many of these molecules may be induced by microbial components during
innate immune responses by various myeloid leukocyte populations (monocytes, macrophages, dendritic cells, and mast cells).
**While some mammals have multiple cathelicidins, humans have only one.
Another major class of antimicrobial components
secreted by skin and other epithelial layers is comprised of
antimicrobial peptides, generally less than 100 amino acids
long, which are an ancient form of innate immunity present
in vertebrates, invertebrates, plants, and even some fungi.
The discovery that vertebrate skin produces antimicrobial
proteins came from studies in frogs, where glands in the skin
were shown to secrete peptides called magainins that have
potent antimicrobial activity against bacteria, yeast, and protozoans. Antimicrobial peptides generally are cysteine-rich,
cationic, and amphipathic (containing both hydrophilic and
hydrophobic regions). Because of their positive charge and
amphipathic nature, they interact with acidic phospholipids
in lipid bilayers, disrupting membranes of bacteria, fungi,
parasites, and viruses. They then can enter the microbes,
where they have other toxic effects, such as inhibiting the
synthesis of DNA, RNA, or proteins, and activating antimicrobial enzymes, resulting in cell death.
The main types of antimicrobial peptides found in humans
are - and -defensins and cathelicidin. Human defensins
kill a wide variety of bacteria, including E. coli, S. aureus,
Streptococcus pneumoniae, Pseudomonas aeruginosa, and
Hemophilus influenzae. Antimicrobial peptides also attack
the lipoprotein envelope of enveloped viruses such as influ-
enza virus and some herpesviruses. Defensins and cathelicidin LL-37 (the only cathelicidin expressed in humans) are
secreted constitutively by epithelial cells in many tissues, as
well as stored in neutrophil granules where they contribute to
killing phagocytosed microbes. Recent studies have shown
that human -defensin antimicrobial peptides secreted into
the gut by intestinal epithelial Paneth cells, located in the
deep valleys (crypts) between the villi, are important for
maintaining beneficial bacterial flora that are necessary for
normal intestinal immune system functions. As we will see
later, production of these antimicrobial peptides also can be
induced in many epithelial and other cell types by the binding
of microbial components to cellular receptors.
The epithelium of the respiratory tract secretes a variety of
lubricating lipids and proteins called surfactants. Two surfactant proteins, SP-A and SP-D, which are present in the lungs
as well as in the secretions of some other mucosal epithelia,
are members of a class of microbe-binding proteins called collectins. SP-A and SP-D bind differentially to sets of carbohydrate, lipid, and protein components of microbial surfaces and
help to prevent infection by blocking and modifying surface
components and promoting pathogen clearance. For example,
they differentially bind two alternating states of the lung
pathogen Klebsiella pneumonia that differ in whether or not
Innate Immunity
they are coated with a thick polysaccharide capsule: SP-A
binds the complex polysaccharides coating many of the capsulated forms, while SP-D only binds the exposed cell wall lipopolysaccharide of the nonencapsulated form.
Phagocytosis
Despite the strong defenses of our protective epithelial layers,
some pathogens have evolved strategies to penetrate these
defenses, and epithelia may be disrupted by wounds, abrasions, and insect bites that may transmit pathogens. Once
pathogens penetrate through the epithelial barrier layers into
the tissue spaces of the body, an array of cellular membrane
receptors and soluble proteins that recognize microbial components play the essential roles of detecting the pathogen and
triggering effective defenses against it. Phagocytic cells make
up the next line of defense against pathogens that have penetrated the epithelial cell barriers. Macrophages, neutrophils,
and dendritic cells in tissues and monocytes in the blood are
the main cell types that carry out phagocytosis—the cellular
uptake (eating) of particulate materials such as bacteria—a
key mechanism for eliminating pathogens. This major role of
the cells attracted to the site of invading organisms is evolutionarily ancient, present in invertebrates as well as vertebrates. Elie Metchnikoff initially described the process of
phagocytosis in the 1880s using cells from starfish (echinoderm invertebrates) similar to vertebrate white blood cells
and ascribed to phagocytosis a major role in immunity. He
was correct in this conclusion; we now know that defects in
phagocytosis lead to severe immunodeficiency.
As described in Chapter 2, most tissues contain resident
populations of macrophages that function as sentinels for
|
CHAPTER 5
147
the innate immune system. Through various cell surface
receptors they recognize microbes such as bacteria, extend
their plasma membrane to engulf them, and internalize
them in phagosomes (endosomes resulting from phagocytosis, Figure 5-5). Lysosomes then fuse with the phagosomes, delivering agents that kill and degrade the microbes.
Neutrophils are a second major type of phagocyte, usually
recruited to sites of infection. Finally, dendritic cells also
can bind and phagocytose microbes. As will be described
later in this chapter and more extensively in Chapters 8
and 11, uptake and degradation of microbes by dendritic
cells play key roles in the initiation of adaptive immune
responses. In addition to triggering phagocytosis, various
receptors on phagocytes recognize microbes and activate
the production of a variety of molecules that contribute in
other ways to eliminating infection, as will be described
later in this chapter. A phagocyte’s recognition of microbes
and the responses that result are shown in Overview
Figure 5-6.
Microbes are Recognized by Receptors on
Phagocytic Cells
How does a phagocytic cell recognize microbes, triggering
their phagocytosis? Phagocytes express on their surfaces a
variety of receptors, some of which directly recognize specific conserved molecular components on the surfaces of
microbes, such as cell wall components of bacteria and
fungi. These conserved motifs, usually present in many copies on the surface of a bacterium, fungal cell, parasite, or
virus particle, are called pathogen-associated molecular
patterns (PAMPs). Note that they can be expressed by
microbes whether or not the microbes are pathogenic (cause
(b)
(a)
1
2
Bacterium becomes
attached to membrane
evaginations called
pseudopodia.
Bacterium is ingested,
forming phagosome.
3
Phagosome fuses with
lysosome.
4
Bacterium is killed and
then digested by
lysosomal enzymes.
5
Digestion products are
released from cell.
FIGURE 5-5 Phagocytosis. (a) Scanning electron micrograph of alveolar macrophage phagocytosis of E. coli bacteria on the outer surface
of a blood vessel in the lung pleural cavity. (b) Steps in the phagocytosis of a bacterium. [Part (a) from Dennis Kunkel Microscopy, Inc./Visuals Unlimited, Inc.]
5-6
OVERVIEW FIGURE
Effectors of Innate Immune Responses to Infection
Phagocytes
(neutrophils,
dendritic cells,
macrophages)
Pathogen-associated
molecular patterns
(PAMPs)
PAMPs recognized
by pattern recognition
receptors (PRRs)
Pathogen killed and
degraded in lysosomes
Phagocyte
activated to
produce
anti-microbial
components,
followed by
secretion of
inflammationpromoting
cytokines and
chemokines
Phagocytosis
Pathogen
C-reactive
protein (CRP)
Complement
proteins
SP-A, SP-D,
Mannosebinding
lectin (MBL)
Opsonins are recognized
by opsonin receptors,
enhancing phagocytosis
Innate initiation of
adaptive response
Dendritic cell PRRs
recognize PAMPs,
activating phagocytosis
and signaling pathways.
Dendritic cells migrate
to lymph nodes, carrying
intact or degraded
pathogens.
Antigen fragments bound
to cell surface MHC
proteins are recognized
by T cells.
T cell
Activated T cells initiate
adaptive responses.
Microbial invasion brings many effectors of innate immunity into
play. Entry of microbial invaders through lesions in epithelial barriers generates inflammatory signals and exposes the invaders to
attack by different effector molecules and cells. Microbes with surface components recognized by C-reactive protein (CRP), mannosebinding lectin (MBL), or surfactant proteins A or D (SP-A and SP-D)
are bound by these opsonizing molecules, marking the microbes
for phagocytosis by neutrophils and macrophages. Some bacteria
and fungi can activate complement directly, or via bound CRP or
MBL, leading to further opsonization or direct lysis. Inflammatory
signals cause phagocytes such as monocytes and neutrophils to
bind to the walls of blood vessels, extravasate, and move to the
site of infection, where they phagocytose and kill infecting
Antimicrobial
peptides
Membrane
damage kills
pathogen
Opsonized
pathogen
CRP, MBL, complement
proteins activate
complement pathway
Complement destroys
membrane, stimulates
inflammation, attracts
neutrophils and other
cells
microorganisms. Binding of microbes to receptors on phagocytes
activates phagocytosis and production of additional antimicrobial
and proinflammatory molecules that intensify the response, in part
by recruiting more phagocytes and soluble mediators (CRP, MBL,
and complement) from the bloodstream to the site of infection.
Inset: Dendritic cells bind microbes via receptors and are activated
to mature; they also internalize and degrade microbes. These dendritic cells migrate through lymphatic vessels to nearby lymph
nodes, where they present antigen-derived peptides on their MHC
proteins to T cells. Antigen-activated T cells then initiate adaptive
immune responses against the pathogen. Cytokines produced during innate immune responses also support and direct the adaptive
immune responses to infection.
Innate Immunity
TABLE 5-3
|
CHAPTER 5
149
Human receptors that trigger phagocytosis
Receptor type on phagocytes
Examples
Ligands
Pattern recognition receptors
C-type lectin receptors (CLRs)
Scavenger receptors
Microbial ligands (found on microbes)
Mannose receptor
Mannans (bacteria, fungi, parasites)
Dectin 1
-glucans (fungi, some bacteria)
DC-SIGN
Mannans (bacteria, fungi, parasites)
SR-A
Lipopolysaccharide (LPS), lipoteichoic acid
(LTA) (bacteria)
SR-B
LTA, lipopeptides, diacylglycerides (bacteria),
-glucans (fungi)
Microbe-binding opsonins
(soluble; bind to microbes)
Opsonin receptors
Collagen-domain receptor
CD91/calreticulin
Collectins SP-A, SP-D, MBL; L-ficolin; C1q
Complement receptors
CR1, CR3, CR4, CRIg, C1qRp
Complement components and fragments*
Immunoglobulin Fc receptors
FcR
Specific IgA antibodies bound to antigen#
FcRs
Specific IgG antibodies bound to antigen;#
C-reactive protein
* See Table 6-3 for specific complement components or fragments that are bound by individual receptors
#
Opsonization of antibody-bound antigens is an adaptive immune response clearance mechanism
disease); hence some researchers have started to use the
more general term microbe-associated molecular patterns
(MAMPs). The receptors that recognize PAMPs are called
pattern recognition receptors (PRRs). Some PRRs that
bind microbes and trigger phagocytosis of the bound
microbes are listed at the top of Table 5-3, along with the
PAMPs they recognize. As we will see later, there are other
PRRs that, after PAMP binding, do not activate phagocytosis
but trigger other types of responses. Most PAMPs that
induce phagocytosis are cell wall components, including
complex carbohydrates such as mannans and -glucans,
lipopolysaccharides (LPS), other lipid-containing molecules,
peptidoglycans, and surface proteins.
As shown in Figure 5-6, activation of phagocytosis can
also occur indirectly, by phagocyte recognition of soluble
proteins that have bound to microbial surfaces, thus enhancing phagocytosis, a process called opsonization (from the
Greek word for “to make tasty”). Many of these soluble
phagocytosis-enhancing proteins (called opsonins) also
bind to conserved, repeating components on the surfaces of
microbes such as carbohydrate structures, lipopolysaccharides, and viral proteins; hence they are sometimes referred
to as soluble pattern-recognition proteins. Once bound to
microbe surfaces, opsonins are recognized by membrane
opsonin receptors on phagocytes, activating phagocytosis
(see Table 5-3, bottom).
A variety of soluble proteins function as opsonins; many
play other roles as well in innate immunity. For example, the
two surfactant collectin proteins mentioned earlier, SP-A and
SP-D, are found in the blood as well as in mucosal secretions
in the lungs and elsewhere, where they function as opsonins.
After binding to microbes they are recognized by the CD91
opsonin receptor (see Table 5-3) and promote phagocytosis
by alveolar and other macrophage populations. This function
of SP-A and SP-D contributes to clearance of the fungal respiratory pathogen Pneumocystis carinii, a major cause of pneumonia in individuals with AIDS. Mannose-binding lectin
(MBL), a third collectin with opsonizing activity, is found in
the blood and respiratory fluids. L-ficolin, a member of the
ficolin family that is related to MBL and other collectins, is
found in the blood, where it binds to acetylated sugars on
microbes, including some streptococcal bacteria. The complement component C1q also functions as an opsonin, binding bacterial cell wall components such as lipopolysaccharides
and some viral proteins.
MBL (and other collectins), ficolins, and C1q share structural features, including similar polymeric structures with
collagen-like shafts, but have recognition regions with different binding specificities (Figure 5-7). As a result of their
structural similarities, all are bound by the CD91 opsonin
receptor (see Table 5-3) and activate pathogen phagocytosis.
Another opsonin, C-reactive protein (CRP), recognizes
phosphorylcholine and carbohydrates on bacteria, fungi,
and parasites, and is then bound by Fc receptors (FcRs) for
IgG found on most phagocytes (see Chapter 3). Fc receptors
also are important for the opsonizing activity of IgA antibodies and some IgG antibody subclasses. After binding
specifically to antigens on microbe surfaces, the Fc regions of
these antibodies can be recognized by specific FcRs, triggering phagocytosis. As an important mechanism by which the
PA R T I I
150
|
Innate Immunity
(a) Mannose-binding lectin
(c) C1 bound to LPS
C1q
Cysteine-rich region
Collagen-like region
MASP-2
C1r
MASP-3
C1s
Recognition domain
Carbohydrate
(b) H-ficolin
LPS
(d) C1 bound to lgG
C1q
MASP-2
MASP-3
Carbohydrate
FIGURE 5-7 Structures of opsonins. (a) Mannose-binding
lectin (MBL), a collectin, and (b) H-ficolin are polymers of multiple
polypeptide chains, each containing an N-terminal cysteine-rich
region followed by a collagen-like region, an -helical neck region,
and a recognition domain. Both MBL and H-ficolin have bound
MBL-associated serine proteases (MASPs), which become active after
the recognition domains bind to specific carbohydrate residues
on pathogen surfaces. The MASPs then activate the complement
adaptive immune response clears antigens, opsonization by
antibodies (not shown in Figure 5-6) will be discussed in
Chapter 13.
As mentioned above, among the most effective opsonins
are several components of the complement system, which is
described detail in Chapter 6. Present in both invertebrates
and vertebrates, complement straddles both the innate and
adaptive immune systems, indicating that it is ancient and
important. In vertebrates, complement consists of more than
30 binding proteins and enzymes that function in a cascade
of sequential activation steps. It can be triggered by several
innate soluble pattern-recognition proteins (including the
C1r
C1s
lgG
pathway. C1, the first component of complement, has a multimeric
structure similar to that of MBL and H-ficolin. C1q can bind LPS on
bacterial cell walls (c) or antibodies bound to antigens (d). After C1
binds to the pathogen-associated LPS or antibodies, the associated
C1r and C1s subunit proteases, which are similar to MASPs, become
activated and can initiate complement activation. [Adapted from
A. L. DeFranco, R. M. Locksley, and M. Robertson, 2007, Immunity (Primers in
Biology), Sunderland, MA: Sinauer Associates.]
first complement component, C1q, and the structurally
related lectins MBL and ficolins, C-reactive protein, and
properdin), as well as by microbe-bound antibodies generated by the adaptive immune response. As we will see in
Chapter 6, phagocytosis is one of many important antimicrobial effects resulting from complement activation. The
importance of MBL’s roles as both an opsonizer and an
inducer of complement activation has been indicated by the
effects of MBL deficiencies, which affect about 25% of the
population. Individuals with MBL deficiencies are predisposed to severe respiratory tract infections, especially pneumococcal pneumonia. Interestingly, MBL deficiencies may
|
Innate Immunity
be protective against tuberculosis, probably reflecting MBL’s
opsonizing role in enhancing the phagocytosis of Mycobacterium tuberculosis, the route by which it infects macrophages, potentially leading to tuberculosis.
Phagocytosed Microbes are Killed
by Multiple Mechanisms
The binding of microbes—bacteria, fungi, protozoan parasites,
and viruses—to phagocytes via pattern recognition receptors
or opsonins and opsonin receptors (see Figure 5-6) activates
signaling pathways. These signaling pathways trigger actin
polymerization, resulting in membrane extensions around the
microbe particles and their internalization, forming phagosomes (see Figure 5-5). The phagosomes then fuse with lysosomes and, in neutrophils, with preformed primary and
secondary granules (see Figure 2-2a). The resulting phagolysosomes contain an arsenal of antimicrobial agents that then
kill and degrade the internalized microbes. These agents
include antimicrobial proteins and peptides (including defensins and cathelicidins), low pH, acid-activated hydrolytic
enzymes (including lysozyme and proteases), and specialized
molecules that mediate oxidative attack.
Oxidative attack on the phagocytosed microbes, which
occurs in neutrophils, macrophages, and dendritic cells,
employs highly toxic reactive oxygen species (ROS) and
reactive nitrogen species (RNS), which damage intracellular components (Figure 5-8). The reactive oxygen species are
generated by the phagocytes’ unique NADPH oxidase
enzyme complex (also called phagosome oxidase), which is
CHAPTER 5
151
activated when microbes bind to the phagocytic receptors.
The oxygen consumed by phagocytes to support ROS production by NADPH oxidase is provided by a metabolic
process known as the respiratory burst, during which oxygen
uptake by the cell increases severalfold. NADPH oxidase
converts oxygen to superoxide ion (•O2); other ROS generated by the action of additional enzymes are hydrogen
peroxide (H2O2), and hypochlorous acid (HClO), the active
component of household bleach.
The generation of RNS requires the transcriptional activation of the gene for the enzyme inducible nitric oxide
synthase (iNOS, or NOS2)—called that to distinguish it
from related NO synthases in other tissues. Expression of
iNOS is activated by microbial components binding to various PRRs. iNOS oxidizes L-arginine to yield L-citrulline and
nitric oxide (NO), a potent antimicrobial agent. In combination with superoxide ion (•O2) generated by NADPH oxidase, NO produces an additional reactive nitrogen species,
peroxynitrite (ONOO) and toxic S-nitrosothiols. Collectively the ROS and RNS are highly toxic to phagocytosed
microbes due to the alteration of microbial molecules
through oxidation, hydroxylation, chlorination, nitration,
and S-nitrosylation, along with formation of sulfonic acids
and destruction of iron-sulfur clusters in proteins. One
example of how these oxidative species may be toxic to
pathogens is the oxidation by ROS of cysteine sulfhydryls
that are present in the active sites of many enzymes, inactivating the enzymes. ROS and RNS also can be released from
activated neutrophils and macrophages and kill extracellular
pathogens.
Antimicrobial species generated from oxygen and nitrogen
Reactive oxygen species (ROS)
•O2 (superoxide anion)
OH• (hydroxyl radical)
H2O2 (hydrogen peroxide)
HClO (hypochlorous acid)
NADPH
phagosome
oxidase
O2
Oxygen
•O2
Superoxide
anion
Superoxide
dismutase
OH Hydroxyl
radicals
•OH
H2O2
Hydrogen
peroxide
Myeloperoxidase
HClO
Hypochlorous
acid
Cl
Chloride
ion
Reactive nitrogen species (RNS)
NO (nitric oxide)
NO2 (nitrogen dioxide)
ONOO (peroxynitrite)
Inducible nitric
oxide synthase
L-Arginine
L-Citrulline
NO
Nitric oxide
FIGURE 5-8 Generation of antimicrobial reactive oxygen and nitrogen species. In the cytoplasm of neutrophils,
macrophages, and dendritic cells, several enzymes, including
phagosome NADPH oxidase, transform molecular oxygen into
highly reactive oxygen species (ROS) that have antimicrobial
ONOO
Peroxynitrite
S-nitrosothiols
NO2
Nitrogen dioxide
activity. One of the products of this pathway, superoxide anion,
can interact with a reactive nitrogen species (RNS), generated by
inducible nitric oxide synthase (iNOS) to produce peroxynitrite,
another RNS. NO can also undergo oxidation to generate the RNS
nitrogen dioxide.
152
PA R T I I
|
Innate Immunity
Evidence from genetic defects in humans and mice
highlights the critical roles of these reactive chemical species in microbial elimination by phagocytic cells. The
importance to antimicrobial defense of phagosomal
NADPH oxidase and its products, ROS and RNS, is illustrated by chronic granulomatous disease (CGD). Patients
afflicted with this disease have dramatically increased susceptibility to some fungal and bacterial infections, caused
by defects in subunits of NADPH oxidase that destroy its
ability to generate oxidizing species. In addition, studies
with mice in which the genes encoding iNOS were knocked
out have shown that nitric oxide and substances derived
from it account for much of the antimicrobial activity of
macrophages against bacteria, fungi, and parasites. These
mice lost much of their usual ability to control infections
by such intracellular pathogens as M. tuberculosis and
Leishmania major, the intracellular protozoan parasite that
causes leishmaniasis.
Phagocytosis Contributes to Cell Turnover
and the Clearance of Dead Cells
Our discussion of phagocytosis thus far has focused on its
essential roles in killing pathogens. As the body’s main scavenger cells, macrophages also utilize their phagocytic receptors to take up and clear cellular debris, cells that have died
from damage or toxic stimuli (necrotic cell death) or from
apoptosis (programmed cell death), and aging red blood cells.
Considerable progress has been made in recent years in
understanding the specific markers and receptors that trigger macrophage phagocytosis of dead, dying, and aging cells.
Collectively the components of dead/dying cells and damaged tissues that are recognized by PRRs leading to their
clearance are sometimes referred to as damage-associated
molecular patterns (DAMPs). As the presence of these
components may also be an indicator of conditions harmful
to the body or may contribute to harmful consequences
(such as autoimmune diseases), the D in DAMP also can
refer to a “Danger” signal. Phagocytosis is the major mode of
clearance of cells that have undergone apoptosis as part of
developmental remodeling of tissues, normal cell turnover,
or killing of pathogen-infected or tumor cells by innate or
adaptive immune responses.
Apoptotic cells attract phagocytes by releasing the lipid
mediator lysophosphatidic acid, which functions as a chemoattractant. These dying cells facilitate their own phagocytosis by expressing on their surfaces an array of molecules
not expressed on healthy cells, including phospholipids
(such as phosphatidyl serine and lysophosphatidyl choline),
proteins (annexin I), and altered carbohydrates. These
DAMPs are recognized directly by phagocytic receptors such
as the phosphatidyl serine receptor and scavenger receptor
SR-A1. Other DAMPs are recognized by soluble pattern
recognition molecules that function as opsonins, including
the collectins MBL, SP-A, and SP-D mentioned earlier; various complement components; and the pentraxins C-reactive
protein and serum amyloid protein. These opsonins are then
recognized by opsonin receptors, activating phagocytosis
and degradation of the apoptotic cells.
An important additional activity of macrophages in the
spleen and those in the liver (known as Kupffer cells) is to
recognize, phagocytose, and degrade aging and damaged red
blood cells. As these cells age, novel molecules that are recognized by phagocytes accumulate in their plasma membrane. Phosphatidyl serine flips from the inner to the outer
leaflet of the lipid bilayer and is recognized by phosphatidyl
serine receptors on phagocytes. Modifications of erythrocyte
membrane proteins have also been detected that may promote phagocytosis.
Obviously it is important for normal cells not to be
phagocytosed, and accumulating evidence indicates that
whether or not a cell is phagocytosed is controlled by sets of
“eat me” signals—the altered membrane components
(DAMPs) described above—and “don’t eat me” signals
expressed by normal cells. Young, healthy erythrocytes avoid
being phagocytosed by not expressing “eat me” signals, such
as phosphatidyl serine, and also by expressing a “don’t eat
me” signal, the protein CD47. CD47, expressed on many cell
types throughout the body, is recognized by the SIRP-
receptor on macrophages, which transmits signals that
inhibit phagocytosis. Recent studies have shown that tumors
use elevated CD47 expression to evade tumor surveillance
and phagocytic elimination by the immune system. Increased
expression of CD47 on all or most human cancers is correlated with tumor progression, probably because the CD47
activates SIRP-1-mediated inhibition of the phagocytosis
of tumor cells by macrophages. This understanding of the
role of CD47 in preventing phagocytosis is being used to
develop novel therapies for certain cancers, such as using
antibodies to block CD47 on tumor cells, which should then
allow them to be phagocytosed and eliminated.
Induced Cellular Innate Responses
In addition to triggering their own uptake and killing by
phagocytic cells, microbes induce a broad spectrum of cellular innate immune responses by a wide variety of cell
types. Several families of pattern recognition receptors
(PRRs) other than those described earlier as mediating
phagocytosis (see Table 5-3) play major roles in innate
immunity. As we will see in this section, these PRRs bind
to PAMPs as well as to some endogenous (self) DAMPs
and trigger signal-transduction pathways that turn on
expression of genes with important functions in innate
immunity. Among the proteins encoded by these genes
are antimicrobial molecules such as antimicrobial peptides
and interferons, chemokines and cytokines that recruit and
activate other cells, enzymes such as iNOS (mentioned
earlier) that generate antimicrobial molecules, and proinflammatory mediators (i.e., components that promote
inflammation).
Innate Immunity
|
CHAPTER 5
153
Cellular Pattern Recognition Receptors Activate
Responses to Microbes and Cell Damage
Several families of cellular PRRs contribute to the activation
of innate immune responses that combat infections. Some of
these PRRs are expressed on the plasma membrane, while
others are actually found inside our cells, either in endosomes/lysosomes or in the cytosol. This array of PRRs
ensures that the cell can recognize PAMPs on both extracellular and intracellular pathogens. DAMPs released by cell
and tissue damage also can be recognized by both cell surface and intracellular PRRs. Many cell types in the body
express these PRRs, including all types of myeloid white
blood cells (monocytes, macrophages, neutrophils, eosinophils, mast cells, basophils, dendritic cells) and subsets of the
three types of lymphocytes (B cells, T cells, and NK cells).
PRRs are also expressed by some other cell types, especially
those commonly exposed to infectious agents; examples
include the skin, mucosal and glandular epithelial cells, vascular endothelial cells that line the blood vessels, and fibroblasts and stromal support cells in various tissues. While it is
unlikely that any single cell expresses all of these PRRs, subsets of the receptors are expressed by various cell subpopulations. The rest of this section describes the four main
families of mammalian PRRs and the signaling pathways
that they activate, leading to protective responses.
Toll-Like Receptors Recognize Many Types
of Pathogen Molecules
Toll-like receptors (TLRs) were the first family of PRRs to
be discovered and are still the best-characterized in terms of
their structure, how they bind PAMPs and activate cells, and
the extensive and varied set of innate immune responses that
they induce.
Discovery of Invertebrate Toll and Vertebrate
Toll-like Receptors
In the 1980s, researchers in Germany discovered that Drosophila
fruit fly embryos could not establish a proper dorsal-ventral
(back to front) axis if the gene encoding the Toll membrane
protein is mutated. (The name “Toll” comes from German
slang meaning “weird,” referring to the mutant flies’ bizarrely
scrambled anatomy.) For their subsequent characterization of
the toll and related homeobox genes and their role in the
regulation of embryonic development, Christiane NussleinVolhard, Eric Wieschaus, and Edward B. Lewis were awarded
the Nobel Prize in Physiology and Medicine in 1995. But what
does this have to do with immune system receptors? Many
mutations of the toll gene were generated, and in 1996 Jules
Hoffman and Bruno Lemaitre discovered that mutations in
toll made flies highly susceptible to lethal infection with
Aspergillus fumigatus, a fungus to which wild-type flies were
immune (Figure 5-9). This striking observation led to other
studies showing that Toll and related proteins are involved in
the activation of innate immune responses in invertebrates.
FIGURE 5-9 Impaired innate immunity in fruit flies with a
mutation in the Toll pathway. Severe infection with the fungus
Aspergillus fumigatus (yellow) results from a mutation in the signaling
pathway downstream of the Toll pathway in Drosophila that normally
activates production of the antimicrobial peptide drosomycin. [Electron micrograph adapted from B. Lemaitre et al., 1996, Cell 86:973; courtesy of
J. A. Hoffman, University of Strasbourg.]
For his pivotal contributions to the study of innate immunity
in Drosophila, Jules Hoffman was a co-winner of the 2011
Nobel Prize in Physiology and Medicine.
Characterization of the Toll protein surprisingly revealed
that its cytoplasmic signaling domain was homologous to
that of the vertebrate receptor for the cytokine IL-1 (IL-1R).
Through a search for human proteins with cytoplasmic
domains homologous to those of Toll and IL-1R, in 1997
Charles Janeway and Ruslan Medzhitov discovered a human
gene for a protein similar to Toll that activated the expression of innate immunity genes in human cells. Appropriately, this and other vertebrate Toll relatives discovered soon
thereafter were named Toll-like receptors (TLRs).
Through studies with mutant mice, in 1998 Bruce Beutler
obtained the important proof that TLRs contribute to normal immune functions in mammals. Mice homozygous for a
mutant form of a gene called lps were resistant to the harmful responses induced by lipolysaccharide (LPS; also known
as endotoxin), a major component of the cell walls of Gramnegative bacteria (Figure 5-10). In humans, a buildup of
endotoxin from severe bacterial infection can induce too
strong of an innate immune response, causing septic shock,
a life-threatening condition in which vital organs such as the
154
PA R T I I
(a) Gram negative bacteria
E. coli
|
Innate Immunity
Cell wall organization
Lipopolysaccharide
(endotoxin)
Outer
membrane
Peptidoglycan
Inner
membrane
(b) Gram positive bacteria
S. aureus
Cell wall organization
Glycolipids
Glycoproteins
Peptidoglycan
Membrane
FIGURE 5-10 Cell wall components of Gram-negative and
Gram-positive bacteria. Cell wall structures differ for Gramnegative (a) and Gram-positive (b) bacteria. Because of their thick
peptidoglycan layer, Gram-positive bacteria retain the precipitate
formed by the crystal violet and iodine reagents of the Gram stain,
whereas the stain is easily washed out of the less-dense cell walls
of Gram-negative bacteria. Thus the Gram stain identifies two
distinct sets of bacterial genera that also differ significantly in
other properties. [Photographs: E. coli photograph from Gary Gaugler/
brain, heart, kidney, and liver may fail. Each year, about
20,000 people die in the United States of septic shock caused
by Gram-negative bacterial infections, so it was striking that
some mutant strains of mice were resistant to fatal doses of
LPS. Beutler found that the defective mouse lps gene
encoded a mutant form of one TLR, TLR4, which differed
from the normal form by a single amino acid so that it no
longer was activated by LPS. This work provided an unequivocal demonstration that TLR4 is the cellular innate pattern
recognition receptor that recognizes LPS and earned Beutler
a share of the 2011 Nobel Prize.
Thus, this landmark series of experiments showed in
rapid succession that invertebrates respond to pathogens,
that they use receptors also found in vertebrates, and that
one of these receptors is responsible for LPS-induced innate
immune responses.
TLRs and Their Ligands
Visuals Unlimited. S. aureus photograph from Dr. Fred Hossler/Getty Images]
Intensive work over the last decade and a half has identified
multiple TLR family members in mice and humans—as of
2011, 13 TLRs that function as PRRs have been identified in
these species. TLRs 1-10 are conserved between mice and
humans, although TLR10 is not functional in mice, while
TLRs 11-13 are expressed in mice but not in humans. While
TLRs have not been shown to be involved in vertebrate development, unlike in fruit flies, the set of TLRs present in a
human or mouse can detect a wide variety of PAMPs from
bacteria, viruses, fungi, and parasites, as well as DAMPs from
damaged cells and tissues. Each TLR has a distinct repertoire
of specificities for conserved PAMPs; the TLRs and some of
their known PAMP ligands are listed in Table 5-4. Biochemical studies have revealed the structure of several TLRs and
Innate Immunity
TABLE 5-4
|
CHAPTER 5
155
TLRs and their microbial ligands
TLRs*
Ligands
Microbes
TLR1
Triacyl lipopeptides
Mycobacteria and Gram-negative bacteria
TLR2
Peptidoglycans
GPI-linked proteins
Lipoproteins
Zymosan
Phosphatidlyserine
Gram-positive bacteria
Trypanosomes
Mycobacteria and other bacteria
Yeasts and other fungi
Schistosomes
TLR3
Double-stranded RNA (dsRNA)
Viruses
TLR4
LPS
F-protein
Mannans
Gram-negative bacteria
Respiratory syncytial virus (RSV)
Fungi
TLR5
Flagellin
Bacteria
TLR6
Diacyl lipopolypeptides
Zymosan
Mycobacteria and Gram-positive bacteria
Yeasts and other fungi
TLR7
Single-stranded RNA (ssRNA)
Viruses
TLR8
Single-stranded RNA (ssRNA)
Viruses
TLR9
CpG unmethylated dinucleotides
Dinucleotides
Herpes virus components
Hemozoin
Bacterial DNA
Some herpesviruses
Malaria parasite heme byproduct
TLR10
Unknown
Unknown
TLR11
Unknown
Profilin
Uropathogenic bacteria
Toxoplasma
TLR12
Unknown
Unknown
TLR13
Unknown
Vesicular stomatitis virus
*All function as homodimers except TLR1, 2, and 6, which form TLR2/1 and TLR2/6 heterodimers. Ligands indicated for TLR2 bind to both; ligands indicated for TLR1
bind to TLR2/1 dimers, and ligands indicated for TLR6 bind to TLR2/6 dimers.
how they bind their specific ligands. TLRs are membranespanning proteins that share a common structural element
in their extracellular region called leucine-rich repeats
(LRRs); multiple LRRs make up the horseshoe-shaped
extracellular ligand-binding domain of the TLR polypeptide
chain (Figure 5-11a).
When TLRs bind their PAMP or DAMP ligands via their
extracellular LRR domains, they are induced to dimerize. In
most cases each TLR dimerizes with itself, forming a homodimer, but TLR2 forms heterodimers by pairing with either
TLR1 or TLR6. How TLRs bind their ligands was not known
until complexes of the extracellular LRR domain dimers with
bound ligands were analyzed by x-ray crystallography. Structures of TLR2/1 with a bound lipopeptide and TLR3 with
dsRNA are shown in Figure 5-11b; the characteristic
“m”-shaped conformation of TLR dimers is apparent.
As shown in Figure 5-12, TLRs exist both on the plasma
membrane and in the membranes of endosomes and lysosomes; their cellular location is tailored to enable them to
respond optimally to the particular microbial ligands they
recognize. TLRs that recognize PAMPs on the outer sur-
face of extracellular microbes (see Table 5-4) are found on
the plasma membrane, where they can bind these PAMPs
and induce responses. In contrast, TLRs that recognize
internal microbial components that have to be exposed by
the dismantling or degradation of endocytosed pathogens—
nucleic acids in particular—are found in endosomes and
lysosomes. Unique among the TLRs, TLR4 has been
shown to move from the plasma membrane to endosomes/
lysosomes after binding LPS or other PAMPs. As we will
see below, it activates different signaling pathways from
the two locations.
In addition to microbial ligands, TLRs also recognize
DAMPs, endogenous (self) components released by dead/
dying cells or damaged tissues. Among the DAMPS recognized by plasma membrane TLRs are heat shock and chromatin proteins, fragments of extracellular matrix components
(such as fibronectin and hyaluronin), and oxidized lowdensity lipoprotein and amyloid-. While self nucleic acids
usually do not activate the intracellular PRR, under certain
circumstances (such as when bound by anti-DNA or antichromatin antibodies in individuals with the autoimmune
PA R T I I
156
|
Innate Immunity
(a)
(b)
Lipopeptide
TLR2
TLR1
Leucine-rich repeats (LRRs)
Exterior
domain
N
N
C
Ribbon model of
exterior domain
C
TLR3
TLR3'
Cell membrane
TIR domain
N'
N
dsRNA
C
C'
FIGURE 5-11 Toll-like receptor (TLR) structure and binding of PAMP ligands. (a) Structure of a TLR polypeptide chain.
Each TLR polypeptide chain is made up of a ligand-binding exterior domain that contains many leucine-rich repeats (LRRs,
repeating segments of 24-29 amino acids containing the
sequence LxxLxLxx, where L is leucine and x is any amino acid), a
membrane-spanning domain (blue), and an interior Toll/IL-1R
(TIR) domain, which interacts with the TIR domains of other
members of the TLR signal-transduction pathway. Two such polypeptide chains pair to form Toll/IL-1R (TLR) dimers, the form that
binds ligands. (b) X-ray crystallographic structures of paired extracellular LRR domains of TLR dimers with bound ligands. Top:
TLR2/1 dimer with a bound lipopeptide molecule. Bottom: TLR3/3
dimer with bound double-stranded (ds) RNA molecule. [Part (b)
from Jin, M.S., and Lee, J-O. 2008. Immunity 29:182. PDB IDs 2Z7X
(top) and 3CIY (bottom).]
disease lupus erythematosus) they can be endocytosed and
activate endosomal/lysosomal TLRs, contributing to the
disease (see Chapter 16).
that induce the expression of subsets of proteins, some of
which are particularly effective in combating the type of
pathogen recognized by the particular TLR(s). An important
example is expression of the potent antiviral Type 1 interferons, IFNs- and -, induced by pathways downstream of the
TLRs that bind viral components. As described below, activation of the interferon regulatory factors (IRFs) is essential for inducing transcription of the genes encoding IFN-
and -. Combinations of transcription factors contribute to
inducing the expression of many of these genes; examples
include combinations of NF-B, IRFs, and/or transcription
factors downstream of MAP kinase (MAPK) pathways (see
Chapters 3 and 4), such as AP-1, that can be activated by
signaling intermediates downstream of certain TLRs.
The particular signal transduction pathway(s) activated by
a TLR dimer following PAMP binding are largely determined
by the TLR and by the initial protein adaptor (see Chapter 3)
that binds to the TLR’s cytoplasmic domain. As shown in Figure 5-11a, this region is called the TIR domain (from Toll/IL-1
receptor), referring to the similarity noted earlier between the
cytoplasmic domains of TLRs and IL-1 receptors (see Figure
4-5a). TIR domains of all TLR dimers serve as binding sites for
the TIR domains of adaptors that activate the downstream
Signaling Through TLRs: Overview
Given the wide variety of potential pathogens the innate
immune system needs to recognize and combat, how does
binding of a specific pathogen evoke an appropriate response
for that pathogen? Signaling through TLRs utilizes many of
the principles and some of the signaling molecules described
in Chapters 3 and 4, along with some unique to pathways
activated by TLRs (and by other PRRs, described below).
Detailed studies of the signaling pathways downstream of
all of the TLRs, summarized schematically in Figure 5-13,
have revealed that they include some shared components
and activate expression of many of the same genes. An
important example of a shared component is the transcription factor NF-B. NF-B is key for inducing many innate
and inflammatory genes, including those encoding defensins; enzymes such as iNOS; chemokines; and cytokines
such as the proinflammatory cytokines TNF-, IL-1, and
IL-6, produced by macrophages and dendritic cells. There
are also TLR-specific signaling pathways and components
Innate Immunity
Gram-negative
bacteria
Bacteria,
parasites
|
CHAPTER 5
157
Gram-positive Flagellated Uropathogenic
bacteria and
bacteria
bacteria
fungi
Plasma
membrane
TLR4/4
TLR2/1
TLR2/6
TLR5/5
TLR11/11
TLR4/4
TLR3/3
TLR7/7
TLR8/8
TLR9/9
Viral
Viral
proteins
dsRNA
Endosomes/lysosomes
Viral
ssRNA
Viral
ssRNA
Bacterial/viral
DNA
FIGURE 5-12 Cellular location of TLRs. TLRs that interact
with extracellular ligands reside in the plasma membrane; TLRs that
bind ligands generated from endocytosed microbes are localized to
endosomes and/or lysosomes. Upon ligand binding, the TLR4/4 dimer
moves from the plasma membrane to the endosomal/lysosomal
compartment, where it can activate different signaling components.
signaling pathways. The two key adaptors are MyD88 (Myeloid
differentiation factor 88) and TRIF (TIR-domain-containing
adaptor-inducing IFN- factor). As shown in Figure 5-13, most
TLRs, whether found on the cell surface or in endosomes/
lysosomes, bind the adaptor protein MyD88 (activating
MyD88-dependent signaling pathways). In contrast, TLR3
binds the alternative adaptor protein TRIF (activating TRIFdependent signaling pathways). TLR4 is unique in binding
both MyD88 (when it is in the plasma membrane, signaling its
endocytosis) and TRIF (when it is in endosomes, after internalization). Figure 5-13 also shows the presence of two additional adaptors associated with most TLRs: TIRAP (TIR-domain
containing adaptor protein) and TRAM (TRIF-related adaptor
molecule). They are TIR domain-containing adaptors that
serve as sorting receptors: TIRAP helps to recruit MyD88 to
TLRs 2/1, 2/6, and 4, and TRAM helps to recruit TRIF to both
TLR3 and endosomal/lysosomal TLR4.
vated IKK then phosphorylates the inhibitory IB subunit of
NF-B, releasing NF-B to enter the nucleus and activate
gene expression. TAK1 does double duty in this TLR signaling cascade. After separating from the IKK complex, it activates MAPK signaling pathways (see Chapter 3) that result
in the activation of transcription factors including Fos and
Jun, which make up the AP-1 dimer (see Figure 3-16).
In addition to activating NF-B and MAPK pathways via
the MyD88-dependent pathway, the endosomal TLRs 7, 8,
and 9, which bind microbial nucleic acids (especially viral
RNA and bacterial DNA), also trigger pathways that activate
IRFs. As shown in Figure 5-13, when triggered by these TLRs,
the MyD88/IRAK4/TRAF6 complex activates a complex containing TRAF3, IRAK1, and IKK, leading to the phosphorylation, dimerization, activation, and nuclear localization of
IRF7. IRF7 induces the transcription of genes for both Type 1
interferons, IFN- and -, which have potent antiviral
activity. Thus different TLRs may differentially activate distinct
transcription factors (NF-B, certain IRFs, and/or those activated by MAPK pathways), leading to variation in which genes
are turned on to help protect us against the invading pathogens.
MyD88-Dependent Signaling Pathways
After associating with a TLR dimer following ligand binding,
MyD88 initiates a signaling pathway that activates the
NF-B and MAPK pathways by essentially the same pathway as that activated by IL-1 (see Figure 4-6). As shown for
plasma membrane TLRs 2/1, 4, and 5 in Figure 5-13, MyD88
recruits and activates several IRAK protein kinases, which
then bind and activate TRAF6. TRAF6 ubiquitinates NEMO
and TAB proteins, leading to the activation of TAK1, which
then phosphorylates the IB kinase (IKK) complex. Acti-
TRIF-Dependent Signaling Pathways
For the two endosomal TLRs that recruit the TRIF adaptor
instead of MyD88—TLR3 and endosomal TLR4—the
downstream signaling pathways differ somewhat from those
activated by MyD88 (see Figure 5-13). TRIF recruits the RIP1
protein kinase that in turn recruits and activates TRAF6,
158
PA R T I I
|
Innate Immunity
LPS
TLR4/4
88
MyD
P
TIRA
dsRNA
TLR3/3
88
MyD
P
TIRA
Endosome
Lipopeptides
TLR2/1
Flagellin
TLR5/5
MyD88
Endosome
Viral proteins
TLR4/4
ssRNA
TLR7/7
CpG DNA
TLR9/9
IRAKs
TRAF6 Ub
TABs
TAK1
RIP1
8
D8
F
TRI
M
TRA
8
D8
Ub
My
My
F
TRI
M
TRA
PI3K
IRAK4
TRAF6 Ub
TRAF6 Ub
TRAF3
TBK-1 IKKε
NEMO Ub
P
IKKα IKKβ P
TRAF3
IRAK1 IKKα
P
P
IRF7 IRF3
MAP kinase
pathway
NF-κB
P
IκB
P
IRF7
Ub
Cytoplasm
Nucleus
P
P
IRF7
IRF3
NF-κB
AP-1
P
IRF7
IFN-β
IFN-α
Antimicrobials, cytokines,
and chemokines
FIGURE 5-13 TLR signaling pathways. Signaling pathways
downstream of TLRs are shown for three cell surface TLRs (TLR2/1,
TLR4, and TLR5; also activated by TLR2/6 and TLR 11 [not shown])
and for endosomal/lysosomal TLRs TLR7 and TLR 9 (in endosome
on right; also activated by TLR8 [not shown]) and TLR3 and TLR4
(in endosome on left). MyD88-dependent signaling pathways are
activated by the TIRAP and MyD88 receptor-binding adaptors.
TRIF-dependent signaling pathways are activated by the TRAM
and TRIF receptor-binding adaptors. Note: Not all signaling components are shown. See text for details. The MAPK pathway is
shown in Figure 3-16.
which then initiates the same steps as in the MyD88-dependent pathway. TLR3 also activates PI3K, which enhances
MAPK pathway activation. In addition, TRIF activates
TRAF3 and a complex of TBK-1 and IKK, which phosphorylates and activates IRF7 and a different IRF—IRF3.
IRF3 and IRF7 dimerize and move into the nucleus and
(together with NF-B and AP-1) induce the transcription of
the IFN- and - genes. Thus, all of the intracellular TLRs
that bind viral PAMPs in an infected cell induce the synthesis
and secretion of Type I interferons, which feed back to potently
inhibit the replication of the virus in that infected cell.
C-Type Lectin Receptors Bind Carbohydrates
on the Surfaces of Extracellular Pathogens
The second family of cell surface PRRs that activate innate and
inflammatory responses is the C-type lectin receptor (CLR)
family. CLRs are plasma membrane receptors expressed
Innate Immunity
-1
tin
CLR
Viral dsRNA
RLR
159
Fungal glucans
c
De
Plasma membrane
2
CHAPTER 5
ure 5-14). Dectin-1 contains a cytoplasmic domain with an
ITAM similar to those in the signaling chains of B-cell and
T-cell antigen receptors (see Chapter 3). After Dectin-1 binds
a ligand, its ITAM is phosphorylated and then recognized by
the tyrosine kinase Syk, also involved in the initial stages of
B-cell activation (see Figure 3-28). Syk triggers MAPK pathways, leading to the activation of the transcription factor
AP-1, and also activates CARD9, one of many signaling components with Caspase recruitment domains (CARD; see
Table 3-1 and additional examples below). CARD9 forms a
complex with additional signaling components, leading to
the activation of IKK and the nuclear translocation of NF-B
as described above for TLR. NF-B and AP-1 cooperate in
inducing expression of inflammatory cytokines and IFN-.
Signals through different PRRs can modulate each other’s
effects, enhancing or inhibiting expression of particular genes.
As the signaling pathways downstream of some CLRs and
TLRs are similar, signals coming into a cell from TLRs and
variably on monocytes, macrophages, dendritic cells, neutrophils, B cells, and T-cell subsets. CLRs generally recognize
carbohydrate components of fungi, mycobacteria, viruses,
parasites, and some allergens (peanut and dust mite proteins).
Humans have at least 15 CLRs that function as PRRs, most of
which recognize one or more specific sugar moieties such as
mannose (e.g., the mannose receptor and DC-SIGN), fucose
(e.g., Dectin-2 and DC-SIGN), and glucans (e.g., Dectin-1).
CLRs have a variety of functions. Unlike TLRs, which do
not promote phagocytosis, some CLRs function as phagocytic receptors (see Table 5-3), and all CLRs trigger signaling
pathways that activate transcription factors that induce effector gene expression. Some CLRs trigger signaling events that
initially differ from TLR signaling but generally lead to
downstream steps similar to the MyD88-dependent TLR
pathways that activate the transcription factors NF-B and
AP-1. An example of this is Dectin-1, which binds cell wall
-1,3 glucans on mycobacteria, yeast, and other fungi (Fig-
1
|
Syk P
Peptidoglyan
fragment
3
NLR
Mitochondrion
RIG-I
NOD1
CARD9
MAVS
RIP2
TRADD RIP1 Ub
TRAF3
TBK1 IKKε
P
P
IRF7
IRF3
NF-κB activation
pathway
MAP kinase
pathway
Cytoplasm
Nucleus
P
P
IRF7
IRF3
AP-1
NF-κB
IFN-β
IFN-α
Antimicrobials, cytokines,
and chemokines
FIGURE 5-14 CLR, RLR, and NLR signaling pathways.
Signaling pathways downstream of selected non-TLR PRRs are shown
for (1) the plasma membrane CLR Dectin-1 (2) the cytosolic RLR RIG-I
and (3) the cytosolic NLR NOD1. Note: Not all signaling components
are shown. CARD domains are shown in brown. See text for details.
The MAPK pathway is shown in Figure 3-16 and the NF-B pathway
is described in Figure 5-13.
160
PA R T I I
|
Innate Immunity
Dectin-1 (or related CLRs) that simultaneously recognize
PAMPs on the same or different microbes can combine to
enhance proinflammatory cytokine production. In addition,
other signaling pathways downstream of Dectin-1 can activate
and regulate different members of the NF-B family, which
control expression of genes for additional cytokines that regulate helper-T-cell differentiation, as we will see later. Other
CLRs have different cytoplasmic domains, downstream signaling pathways, and outcomes. For example, the binding of
PAMPs to one CLR, DC-SIGN, on dendritic cells curtails
inflammation by inducing expression of the anti-inflammatory
cytokine IL-10 or by reducing mRNA levels for proinflammatory cytokines. In some cases such modulatory effects are
beneficial, as responses can be switched from potentially
harmful inflammatory responses to the activation of protective T-cell populations. However, as is discussed later, pathogens can also take advantage of these control circuits to reduce
responses that would contribute to their elimination.
Retinoic Acid-Inducible Gene-I-Like Receptors
Bind Viral RNA in the Cytosol of Infected Cells
The Retinoic acid-inducible gene-I-like receptors (RLRs)
are soluble PRRs that reside in the cytosol of many cell types,
where they play critical roles as sensors of viral infection (see
Figure 5-14). The three known RLRs (RIG-I, MDA5, and
LGP2) are CARD-containing RNA helicases that recognize
viral RNAs, usually from RNA viruses such as influenza,
measles, and West Nile. These receptors appear to distinguish
viral from cellular RNA on the basis of particular structural
features not shared by normal cellular RNA, such as doublestranded regions of the RNA, virus-specific sequence motifs,
and, in the case of RIG-I, a 5 triphosphate modification that
arises during viral RNA synthesis and processing. Upon viral
RNA binding, RIG-I undergoes a conformational change that
leads to binding via CARD-CARD interactions to its downstream adaptor molecule located in mitochondrial membranes, the mitochondria-associated MAVS (mitochondrial
anti-viral signaling) protein. MAVS aggregates and recruits
additional proteins, including the adaptor TRADD, TRAF3,
and RIP1, which activate NEMO/IKKs that lead to the activation of IKK complexes, which activate IRFs 3 and 7 and the
NF-B pathway, leading to expression of IFNs and , other
antimicrobials, chemokines, and proinflammatory cytokines.
Nod-Like Receptors are Activated by a Variety of
PAMPs, DAMPs, and Other Harmful Substances
The final family of PRRs is the NLRs. (NLR is an acronym
that stands for both Nod-like receptors and nucleotide
oligomerization domain/leucine-rich repeat-containing
receptors). The NLRs are a large family of cytosolic proteins
activated by intracellular PAMPs and substances that alert
cells to damage or danger (DAMPs and other harmful substances). They play major roles in activating beneficial innate
immune and inflammatory responses, but, as we will see,
some NLRs also trigger inflammation that causes extensive
tissue damage and disease. The human genome contains
approximately 23 NLR genes, and the mouse genome up to
34. NLR proteins are divided into three major groups, based
largely on their domain structure, as shown in Figure 5-15:
NLRCs (which have caspase recruitment domains, CARD),
NLRBs (some of which have baculovirus inhibitory repeat,
BIR, domains), and NLRPs (which have pyrin domains, PYD).
The functions of many NLRs have not yet been well characterized; those of several NLRs are described below.
NOD1 and NOD2
The best-characterized members of the NLRC family are
NOD1 and NOD2. Important PAMPs recognized by NOD1
and NOD2 are breakdown products (such as muramyl dipeptides) produced during the synthesis or degradation of cell wall
peptidoglycans of either intracellular or extracellular bacteria—
peptides from the latter must enter the cell to activate NODs.
Studies in mice have shown that NOD1 also provides protection against the intracellular protozoan parasite Trypanosoma
cruzi, which causes Chagas disease in humans, and that NOD2
activates responses to some viruses, including influenza.
PAMP binding to the LRR regions of NODs initiates signaling
by activating NOD binding to the serine/threonine kinase
RIP2 through their CARDs (Figure 5-14). RIP2 then activates
IKK, leading to nuclear translocation of NF-B. RIP2 also
activates MAPK pathways, leading to AP-1 activation. The
activated NF-B and AP-1 initiate transcription of inflammatory cytokines—including TNF-, IL-1, and IL-6—and antimicrobial and other mediators. In addition, RIP2 activates the
TRAF3/TBK1/IKK complex, leading to phosphorylation of
IRFs 3 and 7 and to production of IFNs- and -.
NLR Inflammasomes
Whereas the NOD1 and NOD2 NLRCs activate the transcription of genes encoding inflammatory cytokines and
antimicrobial and other proteins, some other NLRs do not
trigger signaling pathways inducing expression of genes
involved in innate immune responses. Instead, these NLRs
assemble with other proteins into a complex that activates
proteases necessary for converting the inactive large precursor forms (procytokines) of IL-1 and IL-18 into the mature
forms that are secreted by activated cells. Because of the very
potent inflammatory effects of secreted IL-1 (and also to
some extent IL-18) (see Chapter 4), these complexes of
certain NLRs with other proteins, including key proteases,
are now referred to as inflammasomes. The discovery and
surprising properties of inflammasomes are described in
Advances Box 5-1.
Expression of Innate Immunity Proteins
is Induced by PRR Signaling
The PRR-activated signaling pathways described above
induce the transcription of genes encoding an arsenal of
Innate Immunity
|
CHAPTER 5
161
NLRP1
CARD
NOD1
NOD3
NOD4
NLRC4/
IPAF
NOD2
NLRX1/
NOD5
FIIND
NLRB
NLRP3
NLRP4-14
LRR
NAD
NBD
CARD
BIR
PYD
NLRCs
FIGURE 5-15 NLR families. NLRs are characterized by distinct
protein domains. All NLRs have a leucine-rich repeat (LRR) domain,
similar to those in TLRs, that functions in ligand binding in at least
some NLRs, and a nuclear-binding domain (NBD). The three main
classes are distinguished by their N-terminal domain (the lowest
domains in the figure): NLRC receptors have CARD domains, NLRP
proteins that help us to mount protective responses. Some of
the induced proteins are antimicrobial and directly combat
pathogens, while others serve key roles in activating and
enhancing innate and adaptive immune responses. There is
tremendous variation in what proteins are made in response
to pathogens, reflecting their PAMPs as well as the responding cell types and their arrays of PRRs. Some of the most
common proteins and peptides that are secreted by cells following PAMP activation of PRRs and that contribute to
innate and inflammatory responses are listed in Table 5-5.
Antimicrobial Peptides
Defensins and cathelicidins were mentioned earlier as being
important in barrier protection, such as on the skin and the
epithelial layers connected to the body’s openings (see
Table 5-2). Some cells and tissues constitutively (i.e., continually, without activation) express these peptides. For
example, human intestinal Paneth epithelial cells constitutively express -defensins and some -defensins, and some
defensins and the cathelicin LL-37 are constitutively synthesized and packaged in the granules of neutrophils, ready to
NRLPs
NLRB
receptors have pyrin domains (PYD), and NLRB receptors have BIR
(baculovirus inhibitor repeat) domains (although the related protein
IPAF does not). These domains function in protein-protein interactions, largely through homotypic domain interactions. [Adapted from
F. Martinon et al. 2007. Annual Review of Immunology 27:229; and E. Elinav
et al. 2011. Immunity 34:665.]
kill phagocytosed bacteria, fungi, viruses, and protozoan
parasites. However, in some other cell types, such as mucosal
and glandular epithelial cells, skin keratinocytes, and NK
cells, the expression of these antimicrobial peptides is
induced or enhanced by signaling through PRRs, in particular TLRs and the NOD NLRs. Macrophages do not produce
these antimicrobial peptides following PRR activation; interestingly, however, there is an indirect pathway by which
microbes induce cathelicidin in macrophages. Binding of
microbial ligands to macrophage TLRs induces increased
expression of receptors for vitamin D; binding of vitamin D
to these receptors activates the macrophages to produce
cathelicidin, which then can help the macrophages kill the
pathogens.
Type I Interferons
Another major class of antimicrobial proteins transcriptionally induced directly by PRRs is the Type I interferons
(IFN-,). Type I interferon production is generally activated by those cell surface TLRs and intracellular TLRs,
RLRs, and NLRs that recognize viral nucleic acids and other
162
PA R T I I
|
Innate Immunity
ADVANCES
Inflammasomes
The cytokine IL-1 has been recognized as
one of the potent inducers of inflammation.
While it was known that the IL-1 gene is transcriptionally activated following exposure to
pathogen, damage, or danger-associated
molecules, additional steps were needed to
generate the mature IL-1 protein from its
large intracellular pro-IL-1 precursor. An
enzyme, initially called IL-1 Converting
Enzyme (ICE), now known as caspase-1, was
shown to carry out the cleavage of pro-IL-1,
but the enzyme itself existed in most cells as
a large inactive precursor. The breakthrough
in understanding how mature IL-1 is produced came in 2002 when Jurg Tschopp and
others published biochemical studies show-
ing that activation of cells by LPS induced the
formation of a large multiprotein aggregate
containing an NLR and mature caspase-1
that cleaved pro-IL-1 into mature IL-1, allowing its release from cells. Because of the
importance of IL-1 in promoting inflammation, Tschopp and his colleagues coined the
term inflammasome for the large protein
complex that activates caspase-1 to generate
IL-1. Three NLRs (NLRP1, NLRP3, and NLRC4)
have been shown to form inflammasomes
that activate caspase-1 to cleave the large
precursors of both IL-1 and IL-18, generating
the mature proinflammatory cytokines.
The best-characterized inflammasome
is the NLRP3 inflammasome, which is
expressed by monocytes, macrophages,
neutrophils, dendritic cells, and some lymphocytes and epithelial cells. NLRP3 is a large
complex containing multiple copies each of
NLRP3, the adaptor protein ASC (which binds
to NLRP3 via homotypic PYD:PYD domain
interactions), and caspase-1 (Figure 1). NLRP3
can be activated in cells by a variety of components from bacteria, fungi, and some
viruses. In addition to microbial components, NLRP3 can also be activated by nonmicrobial (“sterile”) substances, including
several DAMPs released by damaged tissues
and cells, such as hyaluronan, -amyloid
(associated with Alzheimer’s plaques), and
extracellular ATP and glucose. Recent
NLRP3
NBD
PYD
NAD
LRR
PYD
CARD
Caspase-1
CARD
ASC
Pathogen activators
Sterile activators
Bacteria-derived
Pore-forming toxins
(S. aureus, Clostridia)
Flagellin
Peptidoglycan fragment
RNA
DNA
Self-derived
ATP
Cholesterol crystals
Urate crystals
Glucose
Amyloid β
Hyaluronan
Virus-derived
RNA
Influenza M2 protein
Environment-derived
Alum
Asbestos
Silica
Alloy particles
UV radiation
Skin irritants
Fungus-derived
β-glucans
Mannan
Zymosan
Hyphae
Protozoa-derived
Hemozoin
IL-1
IL-18
FIGURE 1
The NLRP3 inflammasome and its activators. Assembly of the
NLRs, ASC, and caspase-1 due to homotypic domain interactions leads to
the formation of a pentamer or heptamer structure: the inflammasome.
Activators of the inflammasome are divided into two categories: sterile
activators include non-microbial host- and environment-derived
molecules, and pathogen-associated activators include PAMPs derived
Cell death
from bacteria, virus, fungus, and protozoa. Activation of the inflammasome
leads to maturation and secretion of IL-1 and IL-18 and, in some instances, to
novel processes leading to inflammatory cell death. (Abbreviations: ASC,
apoptosis-associated speck-like protein containing a caspase recruitment
domain. For others, see legend to Figure 5-15 and text). [Adapted from B. K. Davis
et al. 2011. Annual Review of Immunology 29:707.]
Innate Immunity
|
CHAPTER 5
163
BOX 5-1
research has also implicated NLRP3 in mediating serious inflammatory conditions
caused by an unusual class of harmful substances: crystals. Crystals of monosodium
urate in individuals with hyperuricemia
cause gout, an inflammatory joint condition,
and inhalation of environmental silica or
asbestos crystals cause the serious, often
fatal, inflammatory lung conditions silicosis
and asbestosis. When these crystals are
phagocytosed, they damage lysosomal
membranes, releasing lysosomal components into the cytosol. Similar effects may be
responsible for the loosening (aseptic osteolysis) of artificial joints, caused by tiny metal
alloy particles from the prosthesis that also
activate NLRP3-mediated inflammation.
How these disparate PAMPs, DAMPs,
crystals, and metal particles all activate the
NLRP3 inflammasome is under intense
investigation. However, to date there is no
conclusive evidence for the binding of any
ligands directly to NLRP3; thus it may not act
as a PRR for PAMPs and DAMPs per se but
instead may sense changes in the intracellular milieu resulting from exposure to these
materials. Current models, summarized in
Figure 2, involve the activation by these varied stimuli of a common set of intracellular
signals, such as potassium ion efflux, reactive
oxygen species (ROS), and/or leakage of
lysosomal contents, all of which appear to
induce NLRP3 inflammasome assembly and
caspase-1 activation, leading to the processing and secretion of IL-1 and IL-18. It is clear,
however, that inflammasomes usually work
in tandem with other PRRs in a cell to generate these cytokines. Activation of many TLRs,
CLRs, RLRs, and NOD NLRs by PAMPs or
DAMPs primes the inflammatory response
by inducing transcription of the IL-1 and
IL-18 genes by the signaling pathways
described above. Once activated by as yet
unknown intracellular mediators, the NLRP3
inflammasome then processes the procytokine precursors to the mature, active IL-1 and
IL-18, which then are released by the cell by
an unknown atypical secretion process.
Somewhat less is known about the
other inflammasomes, although they are
also large multiprotein aggregates containing NLRs or related proteins and caspases
and generate mature IL-1 and IL-18. The
inflammasome discovered
by Tschopp contains the NLR
protein now called NLRP1
(see Figure 5-15), which is
expressed in a wide array of
hematopoietic and other
cell types. Its activation by
the lethal toxin of Bacillus
anthracis contributes to the
lethality of anthrax. The
NLRC4 (also called IPAF)
inflammasome, expressed
predominantly in hematopoietic cells, is activated by
certain Gram-negative bacteria and by the cytosolic
presence of bacterial flagellin. A fourth inflammasome
utilizes the AIM2 protein,
which does not contain an
NBD and hence is not an
NLR, but does contain a PYD
domain and hence can form
inflammasomes with ASC
and caspase-1. Interestingly,
AIM2, which is induced by
interferons, has been shown
recently to bind cytosolic
double-stranded DNA, such
as from intracellular viruses
and bacteria, and also to
generate mature IL-1 and
IL-18.
The recent discovery of
inflammasomes has provided the answer to one
long-standing question:
How is mature IL-1 generated in cells in response to
innate and inflammatory
stimuli? But other challenges remain, the most
important being to identify
the mechanism(s) by which
inflammasomes are activated, so that their key caspase-1 enzyme becomes
functional. In addition, as no
functions are known yet for
several other NLRs, more
inflammasomes may yet be
discovered.
DAMP/
PAMP
DAMP/PAMP
K+
ATP
P2X7
Pannexin-1
Crystalline/
particulate
substances
1
2
3
Reactive oxygen
species
Lysosomal
rupture
NLRP3 inflammasome
assembly
Activated
Caspase-1
Pro-IL-1
Pro-IL-18
IL-1
IL-18
“Priming” by
proinflammatory stimuli
e.g. TLR ligands, TNF, IL-1
IL-1
IL-18
FIGURE 2
Models for NLRP3 inflammasome activation. Three
major models for NLRP3 inflammasome activation have been
proposed, which may not be mutually exclusive. Model 1: The NLRP3
activator, ATP, triggers P2X7-dependent pore formation by the
pannexin-1 hemichannel, allowing potassium ion efflux and entry
into the cytosol of extracellular DAMPs and/or PAMPs which then
directly engage NLRP3. Model 2: Crystalline or particulate NLRP3
agonists are engulfed, and their physical characteristics lead to
lysosomal rupture. Released lysosomal contents, including enzymes,
then activate NLRP3 inflammasome in the cytoplasm by unknown
mechanisms; one possibility is cathepsin-B-dependent processing of
a direct NLRP3 ligand. Model 3: All DAMPs and PAMPs, including ATP
and particulate/crystalline activators, induce the generation of
reactive oxygen species (ROS). A ROS-dependent pathway triggers
NLRP3 inflammasome complex formation. Following these initial
activation steps, caspase-1 clustering in the inflammasome activates
caspase-1, which then cleaves pro-IL-1 and pro-IL-18 precursors
cytokines, generating IL-1 and IL-18, which are then secreted. In
some instances the death of the cell may also be induced. [Adapted
from K. Schroder and J. Tschopp. 2010. Cell 140:821.]
164
PA R T I I
TABLE 5-5
Antimicrobials
Cytokines
|
Innate Immunity
Secreted peptides and proteins induced by signaling through PRR
Peptides/proteins
Produced by
Act on
Immune/inflammatory effects
Defensins and
cathelicidin*
Epithelia (e.g.,
oro/nasal,
respiratory,
intestinal,
reproductive tracts;
skin keratinocytes,
kidney); NK cells
Pathogens
Inhibit, kill
Monocytes, immature
dendritic cells,
T cells
Mast cells
Chemoattractant; activate cytokine
production
Activate degranulation
Interferons -, -
Virus-infected cells,
macrophages,
dendritic cells, NK
cells
Virus-infected cells
NK cells
Macrophages, T cells
Inhibit virus replication
Activate
Regulate activity
IL-1
Monocytes,
macrophages,
dendritic cells,
keratinocytes,
epithelial cells,
vascular endothelial
cells
Lymphocytes
Bone marrow
Vascular endothelium
Enhances activity
Promotes neutrophil production
Activates; increases vascular
permeability
Induces acute-phase response
Fever
Monocytes,
macrophages,
dendritic cells, NK
cells, epithelial cells,
vascular endothelial
cells
Lymphocytes
Bone marrow
IL-6
TNF-
Monocytes,
macrophages,
dendritic cells, mast
cells, NK cells,
epithelial cells
Liver
Hypothalamus
Liver
Regulates activity
Promotes hematopoiesis
→ neutrophils
Activates; increases vascular
permeability
Induces acute-phase response
Hypothalamus
Fever
Macrophages
Vascular endothelium
Liver
Activates
Activates, increases vascular
permeability, fluid loss, local blood
clotting
Induces acute-phase response
Hypothalamus
Fever
Vascular endothelium
Tumors
Cytotoxic for many tumor cells
GM-CSF
Macrophages,
vascular endothelial
cells
Bone marrow
Stimulates hematopoiesis → myeloid
cells
IL-12, IL-18
Monocytes,
macrophages,
dendritic cells
Naïve CD4 T cells
Induce TH1 phenotype, IFN-
production
Activate
Naïve CD8 T cells, NK
cells
Chemokines
IL-10
Macrophages,
dendritic and mast
cells; NK, T, B cells
Macrophages, dendritic
cells
Antagonizes inflammatory response,
including production of IL-12 and
TH1 cells
Example: IL-8
(CXCL8)
chemokine**
Macrophages,
dendritic cells,
vascular endothelial
cells
Neutrophils, basophils,
immature dendritic cells,
T cells
Chemoattracts cells to infection site
* Defensins and cathelicidin LL-37 vary among tissues in expression and whether constitutive or inducible.
** Other chemokines that are induced by PRR activation of cells in certain tissues, including various epithelial layers, may also specifically recruit certain lymphoid and
myeloid cells to that site. See text, Chapter 14, and Appendix III.
Innate Immunity
|
CHAPTER 5
165
Type I interferons exert their antiviral and other effects by
binding to a specific receptor, frequently called IFNAR (IFNalpha receptor), expressed by most cell types (see Chapter 4).
Similar to signaling pathways downstream of many cytokine
receptors (see Table 4-3 and Figure 4-11), binding of IFN-
or - activates IFNAR to recruit and activate specific JAKs
(JAK1 and Tyk2), which then activate specific STATs (STATs 1
and 2; see Figure 5-16). STAT1/1 and STAT1/2 dimers then
induce expression of various genes, including those for the
three proteins that block viral replication: protein kinase R
(PKR), which inhibits viral (and cellular) protein synthesis
by inhibiting the elongation factor eIF2; 2,5-oligoadenlyate
A synthetase, which activates the ribonuclease RNase L that
degrades viral mRNA; and Mx proteins, which inhibit both
the transcription of viral genes into mRNAs and the assembly of virus particles. Reflecting the potent antiviral activities
of Type I interferons, they are used to treat some viral infections, such as hepatitis B and C. In addition to their key
components and activate the IRF transcription factors. As
summarized in Table 5-5, Type I interferons are produced in
two situations. When virus infected, many cell types are
induced to make IFN- and/or IFN- following binding of
cytosolic viral PAMPs, such as nucleic acids, to their RLRs or
NLRs (see Figure 5-14). In addition, many cells express cell
surface TLRs recognizing viral PAMPs and/or internalize
virus without necessarily being infected, allowing endosomal TLRs to recognize viral components. Signaling from
these TLRs activates the IRFs and IFN-, production. One
particular type of dendritic cell, called the plasmacytoid
dendritic cell (pDC) because of its shape, is a particularly
effective producer of Type I IFNs. The pDCs endocytose
virus that has bound to various cell surface proteins (including CLRs such as DC-SIGN, which, for example, binds
HIV). TLRs 7 and 9 in the endosomes then are activated by
viral PAMPs (ssRNA and CpG DNA, respectively), leading
to IRF activation.
INF-α/β
Cytosol
P
IFNAR1
IFNAR2
P
TYK2
P
JAK1
STAT2
P
STAT1
P
P
Nucleus
Dimerized STAT
Transcription
Protein
kinase R
2’, 5’-oligo A
synthetase
Mx proteins
GTP
GDP + P
dsRNA
dsRNA
Polymerized
Mx proteins
Oligo A
P
elF2α
Inhibition of translation
RNase L
mRNA degradation
FIGURE 5-16 Major antiviral activities induced by Type I
interferons. Interferons- and - bind to the IFNAR receptor, which
then recruits and activates the JAK1 and TYK2 protein kinases. They
bind and phosphorylate STATs 1 and 2, which dimerize, enter the
nucleus, and stimulate expression of three proteins that activate
antiviral effects. Protein kinase R (PKR) binds viral dsRNA and
Inhibition of virus
transcription and assembly
inhibits the activity of the eIF2 translation initiation factor. 2,5-oligoadenylate synthase synthesizes 2,5-oligoadenylate, which activates a ribonuclease, RNaseL, that degrades mRNAs. Mx proteins
self-assemble into ringlike structures that inhibit viral replication and
formation of new virus particles. [Adapted from Fig. 3-46, A. DeFranco et al.
2007. Immunity (Primers in Biology). Sunderland, MA: Sinauer Associates.]
166
PA R T I I
|
Innate Immunity
roles in controlling viral infections, Type I interferons have
other immune-related activities in that they activate NK cells
and regulate activities of macrophages and T cells. Treatment with IFN- has been shown to have beneficial effects
in some forms of multiple sclerosis, a T-cell-mediated autoimmune disease with inflammatory involvement, probably
by inhibiting production of certain proinflammatory T-cellderived cytokines (see Chapter 16).
Cytokines
Among the proteins transcriptionally induced by PRR activation are several key cytokines, which—while not directly
antimicrobial—activate and regulate a wide variety of cells
and tissues involved in innate, inflammatory, and adaptive
responses. As introduced in Chapter 4, cytokines function as
the protein hormones of the immune system, produced in
response to stimuli and acting on a variety of cellular targets.
Several key examples of cytokines induced by PRR activation during innate immune responses are listed in Table 5-5,
along with their effects on target cells and tissues.
Three of the most important cytokines are IL-1, IL-6, and
TNF-, the major proinflammatory cytokines that act locally
on blood vessels and other cells to increase vascular permeability and help recruit and activate cells at sites of infection;
they also have systemic effects (see below). IL-1, IL-6, and
GM-CSF also feed back on bone marrow hematopoiesis to
enhance production of neutrophils and other myeloid cells
that will contribute to pathogen clearance. Activation of
monocytes, macrophages, and dendritic cells through some
TLRs also induces production of IL-12 and IL-18, cytokines
that play key roles in inducing naïve helper T cells to become
TH1 cells, in particular by inducing production of IFN-. As
will be discussed in Chapter 11, the hallmark cytokine of
TH1 cells is IFN-, which stimulates cell-mediated immunity
and is an important macrophage-activating cytokine. Hence
IL-12 and IL-18 are also considered proinflammatory. IL-10
is another important cytokine specifically induced by some
TLRs in macrophages, dendritic cells, other myeloid cells,
and subsets of T, B, and NK cells. IL-10 is anti-inflammatory,
in that it inhibits macrophage activation and the production
of proinflammatory cytokines by other myeloid cells. IL-10
levels increase over time and contribute to controlling the
extent of inflammation-caused tissue damage.
Chemokines
These small protein chemoattractants (agents that induce
cells to move toward higher concentrations of the agent)
recruit cells into, within, and out of tissues (see Chapters 4
and 14 and Appendix III). Some chemokines are responsible
for constitutive (homeostatic) migration of white blood cells
throughout the body. Other chemokines, produced in
response to PRR activation, have key roles in the early stages
of immune and inflammatory responses in that they attract
cells that contribute both to clearing the infection or damage
and to amplifying the response. The first chemokine to be
cloned, IL-8 (also called CXCL8), is produced in response to
activation—by PAMPs, DAMPs, or some cytokines—of
a variety of cells at sites of infection or tissue damage, including macrophages, dendritic cells, epithelial cells, and vascular endothelial cells. One of IL-8’s key roles occurs in the
initial stages of infection or tissue damage; it serves as a
chemoattractant for neutrophils, recruiting them to sites of
infection. Other chemokines are specifically induced by PRR
activation of epithelial cells in certain mucosal tissues and
serve to recruit cells specifically to those sites, where they
generate immune responses appropriate for clearing the
invading pathogen. For example, B cells are recruited to the
lamina propria, the lymphocyte-rich tissue under the intestinal epithelium, by two chemokines, CCL28 and CCL20,
produced by intestinal epithelial cells activated by PAMP
binding to their TLRs. These activated epithelial cells (as well
as local dendritic cells activated by PAMPs) also produce
cytokines that stimulate the B cells to produce IgA, the class
of antibodies most effective in protecting against mucosal
infections (see Chapter 13).
Enzymes: iNOS and COX2
Among other genes activated in many cell types by PRRactivated signaling pathways are those for two enzymes that
contribute importantly to the generation of antimicrobial
and proinflammatory mediators: inducible nitric oxide synthase (iNOS) and cyclooxygenase 2 (COX2). As described
earlier, the iNOS enzyme catalyzes an important step in the
formation of nitric oxide, which kills phagocytosed microbes
(see Figure 5-8). COX2, whose synthesis is induced by PRR
activation in monocytes, macrophages, neutrophils, and
mast cells, is key to converting the lipid intermediate arachidonic acid to prostaglandins, potent proinflammatory
mediators. The next section provides an overview of the
main processes of inflammatory responses initiated by
innate immune responses. Inflammation will be covered in
more detail in Chapter 15.
Inflammatory Responses
When the outer barriers of innate immunity—skin and
other epithelial layers—are damaged, the resulting innate
responses to infection or tissue injury can induce a complex
cascade of events known as the inflammatory response.
Inflammation may be acute (short-term effects contributing
to combating infection, followed by healing)—for example,
in response to local tissue damage—or it may be chronic
(long term, not resolved), contributing to conditions such as
arthritis, inflammatory bowel disease, cardiovascular disease, and Type 2 diabetes.
The hallmarks of a localized inflammatory response were
first described by the Roman physician Celsus in the first
century AD as rubor et tumor cum calore et dolore (redness
and swelling with heat and pain). An additional mark of
inflammation added in the second century by the physician
Galen is loss of function ( functio laesa). Today we know that
|
Innate Immunity
these symptoms reflect an increase in vascular diameter
(vasodilation), resulting in a rise of blood volume in the area.
Higher blood volume heats the tissue and causes it to redden. Vascular permeability also increases, leading to leakage
of fluid from the blood vessels, resulting in an accumulation
of fluid (edema) that swells the tissue. Within a few hours,
leukocytes also enter the tissue from the local blood vessels.
These hallmark features of inflammatory responses reflect
the activation of resident tissue cells—macrophages, mast
cells, and dendritic cells—by PAMPs and DAMPs to release
chemokines, cytokines, and other soluble mediators into the
vicinity of the infection or wound. Recruited leukocytes are
activated to phagocytose bacteria and debris and to amplify
the response by producing additional mediators. Resolution
of this acute inflammatory response includes the clearance
of invading pathogens, dead cells, and damaged tissue; the
activation of the systemic acute phase response and additional physiological responses, including the initiation of
wound healing; and the induction of adaptive immune
responses. However, if the infection or tissue damage is not
resolved, it can lead to a chronic inflammatory state that can
cause more local tissue damage and potentially have systemic consequences for the affected individual.
CHAPTER 5
167
Inflammation Results from Innate Responses
Triggered by Infection, Tissue Damage,
or Harmful Substances
When there is local infection, tissue damage, or exposure to
some harmful substances (such as asbestos or silica crystals
in the lungs), sentinel cells residing in the epithelial layer—
macrophages, mast cells, and dendritic cells—are activated
by PAMPs, DAMPs, crystals, and so on to start phagocytosing the offending invaders (Figure 5-17). The cells are also
activated to release innate immunity mediators that trigger a
series of processes that collectively constitute the inflammatory response.
The recruitment of various leukocyte populations to the
site of infection or damage is a critical early component of
inflammatory responses. PRR signaling activates resident
macrophages, dendritic cells, and mast cells to release the
initial components of cellular innate immune responses,
including the proinflammatory cytokines TNF-, IL-1, and
IL-6; chemokines; prostaglandins (following the induced
expression of the COX2 enzyme); and histamine and other
mediators released by mast cells. These factors act on the
vascular endothelial cells of local blood vessels, increasing
1
Tissue damage and bacteria cause
resident sentinel cells to release
Tissue
chemoattractants and vasoactive
damage
factors that trigger a local increase in
blood flow and capillary permeability.
4
Phagocytes and antibacterial
substances destroy bacteria.
Bacteria
2
3
Permeable capillaries allow an
influx of fluid (exudate) and cells.
Exudate
(complement,
anti-microbial
proteins))
Neutrophils and other
phagocytes migrate to site of
inflammation (chemotaxis).
Neutrophils and
other phagocytes
Extravasation
Capillary
FIGURE 5-17 Initiation of a local inflammatory response.
Bacterial entry through wounds activates initial innate immune
mechanisms, including phagocytosis by and activation of resident
cells, such as macrophages and dendritic cells. Recognition of bacteria
by soluble and cellular pattern recognition molecules initiates an
inflammatory response that recruits antimicrobial substances
and phagocytes (first neutrophils and then monocytes) to the
site of infection.
168
PA R T I I
|
Innate Immunity
vascular permeability and the expression of cell adhesion
molecules (CAMs) and chemokines such as IL-8. The
affected epithelium is said to be inflamed or activated. Cells
flowing through the local capillaries are induced by chemoattractants and adhesion molecule interactions to adhere to
vascular endothelial cells in the inflamed region and pass
through the walls of capillaries and into the tissue spaces, a
process called extravasation that will be described in detail
in Chapter 14. Neutrophils are the first to be recruited to a
site of infection where they enhance local innate responses,
followed by monocytes that differentiate into macrophages
that participate in pathogen clearance and help initiate
wound healing.
In addition to these key events at the site of infection or
damage, the key cytokines made early in response to innate
and inflammatory stimuli—TNF-, IL-1, and IL-6—also
have systemic effects, which will be described in more detail
in Chapter 15. They induce fever (a protective response, as
elevated body temperature inhibits replication of some
pathogens) by inducing COX-2 expression, which activates
prostaglandin synthesis, as mentioned above. Prostaglandin
E2 (PGE2) acts on the hypothalamus (the brain center controlling body temperature), causing fever. These three proinflammatory cytokines also act on the liver, inducing the
acute phase response, which contributes to the resolution of
the inflammatory response.
Proteins of the Acute Phase Response Contribute
to Innate Immunity and Inflammation
During the 1920s and 1930s, before the introduction of antibiotics, much attention was given to controlling pneumococcal pneumonia. Researchers noted changes in the
concentration of several serum proteins during the acute
phase of the disease, the phase preceding recovery or death.
The serum changes were collectively called the acute phase
response (APR), and the proteins whose concentrations rise
during the acute phase are still called acute phase response
proteins (APR proteins). The physiological significance of
some APR proteins is still not understood, but we now know
that some contribute to the innate immune response to
infection.
The acute phase response (discussed more fully in
Chapter 15) is induced by signals that travel through the
blood from sites of injury or infection. The proinflammatory
cytokines TNF-, IL-1, and IL-6 are the major signals
responsible for induction of the acute phase response; they
act on hepatocytes in the liver, inducing them to secrete APR
proteins at higher levels. Among APR proteins are many
components of the complement system, which contribute to
both innate and adaptive immune responses, and other proteins that function as opsonins, enhancing phagocytosis (see
Table 5-3). As was mentioned briefly earlier and will be
described in more detail in Chapter 6, complement components and fragments contribute to pathogen inactivation and
clearance in a variety of ways, including serving as opsonins.
Several other proteins are produced at higher levels during the APR. Mannose-binding lectin (MBL), described
earlier, is a collectin that recognizes mannose-containing
molecular patterns on microbes and promotes phagocytosis
by blood monocytes. When MBL binds to the surface of
microorganisms it also initiates complement activation (see
Chapter 6). Another important APR made in the liver is
C-reactive protein (CRP), which belongs to a family of pentameric proteins called pentraxins that bind ligands in a
calcium-dependent reaction. Among the ligands recognized
by CRP are a polysaccharide found on the surface of pneumococcal bacteria and phosphorylcholine, which is present
on the surface of many microbes. CRP is an opsonin and
also activates a complement-mediated attack on the microbe.
Circulating CRP levels are considered an indicator of the
level of ongoing inflammation. Two other pentraxins, serum
amyloid protein and PTX, similarly function as both opsonins and activators of the complement pathway. Also released
into the blood at higher levels are the surfactant protein
opsonins SP-A and SP-D, mentioned earlier as providing
protection against lung infections, and a number of proteins
that participate in or regulate the coagulation (clotting)
pathway, such as fibrinogen. While most of these APR proteins are always present in the blood at low levels, their
increased concentrations during the acute phase response
provide enhanced protective functions during infections.
With the combined defenses mounted by innate and
inflammatory responses, together with those of the laterarising adaptive immune responses, most infections are
eliminated. Immune and inflammatory responses generally
are self-limiting, so once the pathogen and damaged tissue
are cleared, inflammation usually diminishes and the tissues
heal. However, persistence of pathogens (e.g., in tuberculosis) or other harmful substances (e.g., monosodium urate
crystals in gout) can cause chronic inflammation and continuing tissue damage and illness.
Natural Killer Cells
In addition to the mechanisms of innate immunity described
above, which largely are the responsibility of nonlymphoid
cell types, a population of lymphocytes also recognizes components associated with pathogens, damage, or stress and
generates rapid protective responses. Natural killer (NK)
cells constitute a third branch of lymphoid cells, along with
B and T lymphocytes of the adaptive immune system, all differentiating from the common lymphoid progenitor into
three separate lineages (see Chapter 2). Unlike B and T lymphocytes, whose receptors have tremendous diversity for
foreign antigens, NK cells express a limited set of invariant,
nonrearranging receptors that enable the cells to be activated
by indicators of infection, cancer, or damage that are
expressed by other cells. Again unlike B and T cells, which
require days of activation, proliferation, and differentiation
to generate their protective antibody and cell-mediated
Innate Immunity
immune responses, NK cells are preprogrammed to respond
immediately to appropriate stimuli, releasing from preformed secretory granules effector proteins that kill altered
cells by inducing apoptosis. This mechanism of cell-mediated
cytotoxicity is also carried out by cytotoxic T cells, which
appear days later (see Chapter 13). NK-cell-mediated cytotoxic activity is enhanced by IFN- produced early during
virus infections, an example of positive feedback regulation
in innate immunity. In addition to releasing cytotoxic mediators, activated NK cells also secrete cytokines—the proinflammatory cytokines IL-6 and TNF-, as well as Type I
IFNs and the Type II IFN, IFN-, a potent macrophage activator that also helps to activate and shape the adaptive
response. Thus, along with the Type I IFNs produced by virusinfected cells or induced during innate responses by viral
PAMPs, NK cells are an important part of the early innate
response to viral infections (as well as to malignancy and
other indicators of danger). These early innate responses control the infection for the days to week it takes for the adaptive
response (antibodies and cytotoxic T cells) to be generated.
How do NK cells sense that our cells have become
infected, malignant, or potentially harmful in other ways? As
will be described more fully in Chapter 13, NK cells express
a variety of novel receptors (collectively called NK receptors). Members of one group serve as activating receptors (of
which more than 20 have been described in humans and
mice) that have specificity for various cell surface ligands
that serve as indicators of infection, cancer, or stress. While
one of these activating ligands in mice has been shown to be
a protein component of a virus (mouse cytomegalovirus),
most NK activating receptors apparently have specificity not
for a pathogen-associated component but for proteins specifically up-regulated on infected, malignant, or stressed
cells, which serve as danger signals perceived by NK cells.
NK cells also express TLRs, and binding of PAMPs or
DAMPs can add to the activation signals.
As will be described later, to limit potential killing of normal cells in our body, NK cells also express inhibitory receptors that recognize membrane proteins (usually conventional
MHC proteins) on normal healthy cells and inhibit NK-cellmediated cytotoxic killing of those cells. Many virus-infected
or tumor cells lose expression of their MHC proteins and
thus do not send these inhibitory signals. Receiving an
excess of activating signals compared to inhibitory signals
tells an NK cell that a target cell is abnormal, and the NK cell
is activated to kill the target cell. Thus NK cells are part of
our innate sensing mechanisms that provide immediate protection, in this case recognizing and eliminating our own
body’s cells that have become harmful. Recent studies suggest that initial activation increases the population and/or
activity of responding NK cells, which therefore would generate a greater response when exposed later to the same
activating ligand(s). This resembles the immunological
memory in the adaptive immune system and represents the
best example of immunological memory in the vertebrate
innate immune system.
|
CHAPTER 5
169
Regulation and Evasion of Innate
and Inflammatory Responses
Innate immune responses and the inflammatory responses
they induce play key roles in clearing infections and healing
infected or damaged tissues. The importance of some of the
individual molecules involved in the generation of innate
and inflammatory responses is dramatically demonstrated
by the impact on human health of genetic defects and polymorphisms (genetic variants) that alter the expression or
function of these molecules. Some of these defects and their
clinical consequences are described in Clinical Focus box 5-2.
As illustrated by these conditions, and by the many known
roles (cited throughout this chapter) of innate and inflammatory mechanisms in protecting us against pathogens,
these responses are essential to keeping us healthy. However,
some other disorders show that innate and inflammatory
responses can also be harmful, in that overproduction of
various normally beneficial mediators and uncontrolled
local or systemic responses can cause illness and even death.
Therefore it is important that the occurrence and extent of
innate and inflammatory responses be carefully regulated to
optimize the beneficial responses and minimize the harmful
responses.
Innate and Inflammatory Responses
Can Be Harmful
To be optimally effective in keeping us healthy, innate and
inflammatory responses should use their destructive mechanisms to eliminate pathogens and other harmful substances
quickly and efficiently, without causing tissue damage or
inhibiting the normal functioning of the body’s systems. However, this does not always occur—a variety of conditions result
from excessive or chronic innate and inflammatory responses.
The most dangerous of these conditions is sepsis, a systemic response to infection that includes fever, elevated
heartbeat and breathing rate, low blood pressure, and compromised organ function due to circulatory defects. Several
hundred thousand cases of sepsis occur annually in the
United States, and mortality rates range from 20% to 50%,
but sepsis can lead to septic shock, circulatory and respiratory collapse that has a 90% mortality rate. Sepsis results
from septicemia, infections of the blood, in particular those
involving Gram-negative bacteria such as Salmonella,
although other pathogens can also cause sepsis.
The major cause of sepsis from Gram-negative bacteria is
the cell wall component LPS (also known as endotoxin),
which as we learned earlier is a ligand of TLR4. As we have
seen, LPS is a highly potent activator of innate immune
mediators, including the proinflammatory cytokines TNF-,
IL-1, and IL-6; chemokines; and antimicrobial components.
Systemic infections activate blood cells including monocytes
and neutrophils, vascular endothelial cells, and resident
macrophages and other cells in the spleen, liver, and other
CLINICAL FOCUS
Genetic Defects in Components of Innate and Inflammatory
Responses Associated with Disease
In combination with our rapidly
expanding understanding of the mechanisms by which innate and inflammatory
responses contribute to disease susceptibility and resistance, in recent years advances
in human genetics have helped to identify
a number of genetic defects that confer
greater susceptibility to infectious and
inflammatory diseases. The adverse effects
of mutations in genes encoding essential
components of innate and inflammatory
processes highlight the critical roles of
these proteins in keeping us healthy.
Since 2003, when the first mutations in
innate immune components that predispose individuals to recurrent bacterial
infections were discovered, a number of
mutations interfering with the generation
of protective innate immune responses
have been identified. Two examples were
mentioned earlier in this chapter: defects
in NADPH oxidase, which cause chronic
granulomatous disease, and MBL deficiencies, which predispose to respiratory
infections. Also leading to defects in
innate immunity are mutations in two
proteins—MyD88 and IRAK4—required
for the MyD88-dependent signaling pathway downstream of all TLRs except TLR3
(see Figure 5-13). Children with these
defects suffer from severe invasive S.
pneumococcus infections, some fatal, and
are also susceptible to Staphylococcus
aureus and Pseudomonas aeruginosa (Figure 1a). The MyD88 mutations completely
prevent cytokine and chemokine induction by ligands for TLRs 2/1, 2/6, 5, 7, and
8. Not surprisingly, the effects of these
mutations are less significant for TLR3
(which activates TRIF signaling pathways
instead of MyD88) and TLR4 (which activates both MyD88 and TRIF signaling
pathways). That the MyD88 mutations do
not leave these children more susceptible
to a wider variety of pathogens probably
reflects the induction of protective immunity by other PRRs as well as by the adaptive immune system. In fact, the children
become less susceptible to infections as
they get older (see Figure 1b), consistent
with the buildup of adaptive immunological memory to these pathogens.
Other genetic defects with clinical consequences have been identified in the
pathways by which Type 1 interferons (IFN,) are induced by viral nucleic acid
PAMPs and then block virus replication in
infected cells. As highlighted in Figure 2,
mutations that completely or partially
block these pathways (red symbols) have
been found in TLR3 and other components
of the pathway that induce IFN-,. Mutations have also been found in TYK2 and
STAT1, key components that activate the
(b)
100
100
Patients surviving (%)
Patients without infection (%)
(a)
antiviral effects of the interferons in
infected cells. Interestingly, these mutations were all discovered in children presenting with herpes simplex virus (HSV)
encephalopathy, a severe HSV infection of
the central nervous system. Cells in the
CNS express TLR3, and it may be that the
mutations in these pathways severely disable innate responses that are critical to
protection against CNS infection by this
virus. Children with TYK2 and STAT1 mutations are also very susceptible to other
infections, especially with mycobacteria,
probably because macrophages must be
activated by IFN- (which also utilizes TYK2
and STAT1 in its signaling pathway) to be
able to kill these intracellular bacteria.
The final set of genetic defects associated with disease states to be discussed
here involves the effects of genetic variants in NLRs (including inflammasomes) in
promoting inflammatory diseases.
Genome-wide genetic association studies
have indicated that a number of allelic
variants of TLRs and NLRs are associated
with inflammatory disorders. Several variants are associated with inflammatory
bowel disease, which includes ulcerative
colitis and Crohn’s disease. The most
impressive genetic association was of
Crohn’s disease with mutations in NOD2 in
or near its ligand-binding LRR region. As
75
50
25
0
0
10
20
30
Months
40
50
75
50
25
0
0
10
20
Years
30
40
FIGURE 1
Severe bacterial infection and mortality among 60 children with MyD88 or IRAK4 deficiencies. (a) Decline in the percentage of
children with these deficiencies who are asymptomatic reveals the incidence of the first severe bacterial infection during the first 50 months of life. (b) Survival
curve of children with deficiencies shows reduced mortality after 5 years of age. [Adapted from J-L. Casenova et al. 2011. Annual Review of Immunology 29:447; based on
data from C. Picard et al. 2010. Medicine 89:403]
170
BOX 5-2
Dendritic cell
Virus-infected cell
Virus
ER
IFNAR
Endosome
TRIF
TYK-2
TLR7,8,9
TLR3
Anti-viral
effects
MyD88
UNC-93B
IRAK4
JAK1
TRAF3
TRAF3
NEMO
TBK1 IKKε
IKK
STAT1
IRF3
IRF7
STAT2
NF-κB
IRF9
IFN-α, β
FIGURE 2
Genetic defects affecting the production and antiviral
effects of IFN-␣,. This schematic shows some protein components involved in Type I IFN production by dendritic cells and the Type
I IFN response in virus-infected cells. Proteins in which genetic mutations have been identified that result in defective functions and are
associated with greater susceptibility to viral diseases are shown in red.
Viruses are taken up by dendritic cells via specific receptors, and viral
nucleic acids are detected by the various TLRs expressed in endosomes. Transport to endosomes of TLRs 3, 7, 8, and 9 is dependent on
the ER protein UNC93B. Cytoplasmic signaling components activate
NOD2 is activated by bacterial cell wall
fragments, investigators have hypothesized that intestinal epithelial cells with the
mutant NOD2 PRRs are unable to activate
adequate protective responses to gut bacteria and/or to maintain appropriate balance between normal commensals and
pathogenic bacteria, and that these defective innate immune functions contribute
to Crohn’s disease pathogenesis. Consistent with this hypothesis, recent studies
have found that intestinal Paneth cells
from NOD2-defective individuals secrete
reduced amounts of -defensins which, as
mentioned earlier, are essential for maintaining normal commensal gut flora.
Genetic variants of the NLRP3 inflammasome have also been shown to be
transcription factors, including IRFs and NF-B, leading to synthesis and
secretion of IFN-,. TLR3, UNC93B, TRAF3, and NEMO deficiencies are
associated with impaired IFN production, particularly during herpesvirus
infection. The binding of IFN-, to their receptor induces the phosphorylation of JAK1 and TYK-2, activating the signal transduction proteins
STAT-1, STAT-2, and IRF9. This complex translocates as a heterotrimer to
the nucleus, where it acts as a transcriptional activator, binding to specific
DNA response elements in the promoter region of IFN-inducible genes.
TYK-2 and STAT-1 deficiencies are associated with impaired IFN responses.
[Adapted from J. Bustamante et al., 2008, Current Opinion in Immunology 20:39.]
associated with Crohn’s disease and other
inflammatory disorders. In fact, mutations
in NLRP3 (originally called cryopyrin) have
been show to be responsible for a set of
autoinflammatory diseases (i.e., noninfectious inflammatory diseases affecting the
body’s tissues) collectively known as CAPS
(for cryopyrin-associated periodic fever
syndromes); one example is NOMID (neonatal onset multisystem inflammatory disorder). These devastating syndromes
include many signs of systemic inflammation, including fever, rashes, arthritis, pain,
and inflammation affecting the nervous
system, with adverse effects on vision and
hearing.
More than 70 inherited and novel
mutations in NLRP3 associated with
CAPS have been identified. Most are in
the NRLP3 NBD domain, though some
are in the LRR domain. What many have
in common is their deregulating effect
on NLRP3 activation of caspase-1, which
may become constitutively active. Cells
from NOMID patients have recently
been shown to secrete higher levels of
IL-1 and IL-18, both spontaneously and
induced by PAMPs and DAMPs, promoting chronic inflammation. Another consequence of constitutive NLRP3
activation is the death of the activated
cells, releasing DAMPs that lead to more
inflammation. Fortunately, new therapeutic approaches that inhibit IL-1 activity seem to alleviate these symptoms in
some patients.
171
172
PA R T I I
|
Innate Immunity
tissues, to release these soluble mediators. They, in turn,
systemically activate vascular endothelial cells, inducing
them to produce cytokines, chemokines, adhesion molecules, and clotting factors that amplify the inflammatory
response. Enzymes and reactive oxidative species released by
activated neutrophils and other cells damage the vasculature;
this damage, together with TNF--induced vasodilation and
increased vascular permeability, results in fluid loss into the
tissues that lowers blood pressure. TNF also stimulates
release of clotting factors by vascular endothelial cells,
resulting in blood clotting in capillaries. These effects on the
blood vessels are particularly damaging to the kidneys and
lungs, which are highly vascularized. High circulating
TNF- and IL-1 levels also adversely affect the heart. Thus
the systemic inflammatory response triggered by septicemia
can lead to circulatory and respiratory failure, resulting in
shock and death. As high levels of circulating TNF-, IL-1,
and IL-6 are highly correlated with morbidity, considerable
effort is being invested in developing treatments that block
the adverse effects of these normally beneficial molecules.
While not as immediately dangerous as septic shock,
chronic inflammatory responses resulting from ongoing
activation of innate immune responses can have adverse
consequences for our health. For example, a toxin from Helicobacter pylori bacteria damages the stomach by disrupting
the junctions between gastric epithelial cells and also induces
chronic inflammation that has been implicated in peptic
ulcers and stomach cancer. Also, increasing evidence suggests that the noninfectious DAMPs cholesterol (as insoluble
aggregates or crystals) and amyloid contribute, respectively, to atherosclerosis (hardening of the arteries) and
Alzheimer’s disease. Other examples of harmful sterile
(noninfectious) inflammatory responses discussed earlier—
including gout, asbestosis, silicosis, and aseptic osteolysis—
are induced, respectively, by crystals of monosodium urate,
asbestos, and silica, and by metal alloy particles from artificial joint prostheses. These varied substances are all potent
inflammatory stimuli due to their shared ability to activate
the NLRP3 inflammasome, resulting in the release of the
proinflammatory cytokines IL-1 and IL-18.
Innate and Inflammatory Responses are
Regulated Both Positively and Negatively
As innate immune responses play essential roles in eliminating infections but also can be harmful when not adequately
controlled, it is not surprising that many regulatory processes have evolved that either enhance or inhibit innate and
inflammatory responses. These mechanisms control the
induction, type, and duration of these responses, in most
cases leading to the elimination of an infection without damaging tissues or causing illness.
Positive Feedback Mechanisms
Innate and inflammatory responses are increased by a variety of mechanisms to enhance their protective functions. To
amplify the early stages of protective innate responses, two
of the initial cytokines induced by PAMP or DAMP binding
to PRRs—TNF- and IL-1—activate pathways similar to
those downstream of TLRs (see Chapter 4) and hence induce
more of themselves, an example of positive feedback regulation. In other words, a cell’s response to TNF- or IL-1 signaling can include production of more TNF- and IL-1. In a
parallel fashion, in some cells Type I interferons can signal
cells through the IFNAR receptor to produce more IFN.
Other positive feedback mechanisms triggered by PRR activation include the activation of higher transcription rates of
the genes for some TLRs, for the subunits of NF-B itself,
and for NLRP3 and caspase-1. Collectively the increases in
these proteins amplify IL-1 secretion, as production of IL-1
requires not only the transcription of its gene but also the
processing of its large precursor protein by caspase-1, which
is activated by NLRP3.
Negative Feedback Mechanisms
On the other side of the equation, as uncontrolled innate and
inflammatory responses can have adverse consequences,
many negative feedback mechanisms are activated to limit
these responses. Several proteins whose expression or activity is increased following PRR signaling inhibit steps in the
signaling pathways downstream of PRRs. Examples include
production of a short form of the adaptor MyD88 that inhibits normal MyD88 function, the activation of protein phosphatases that remove key activating phosphate groups on
signaling intermediates, and the increased synthesis of IB,
the inhibitory subunit that keeps NF-B in the cytoplasm.
The activation of these and other intracellular negative feedback mechanisms can lead cells to become less responsive,
limiting the extent of the innate immune response. In a wellstudied example, when macrophages are exposed continually to the TLR4 ligand LPS, their initial production of
antimicrobial and proinflammatory mediators is followed by
the induction of inhibitors (including IB and the short
form of MyD88) that block the macrophages from continuing to respond to LPS. This state of unresponsiveness, called
LPS tolerance (or endotoxin tolerance), reduces the possibility that continued exposure to LPS from a bacterial infection will cause septic shock.
Other feedback pathways inhibit the inflammatory effects
of TNF- and IL-1. Each induces production of a soluble
version of its receptor or a receptor-like protein that binds
the circulating cytokine molecules (see Chapter 4), preventing them from acting on other cells. In addition, the antiinflammatory cytokine IL-10 is produced late in the
macrophage response to PAMPs; it inhibits the production
and effects of inflammatory cytokines and promotes wound
healing. Another anti-inflammatory cytokine, TGF-, is also
produced by macrophages, dendritic cells, and other cells,
especially after PRR activation by apoptotic cells. The inhibitory effects of TGF- reduce the likelihood that DAMPs
released by cells undergoing apoptosis will induce inflammatory responses. As a final example of a negative feedback
Innate Immunity
TABLE 5-6
|
CHAPTER 5
173
Pathogen evasion of innate and inflammatory responses
Type of evasion
Examples
Avoid detection by PRRs
Proteobacteria flagellin has a mutation that prevents it from being
recognized by TLR5.
Helicobacter, Coxiella, and Legionella bacteria have altered LPS that is
not recognized by TLR4.
HTLV-1 virus p30 protein inhibits transcription and expression of TLR4.
Several viruses (Ebola, influenza, Vaccinia) encode proteins that bind
cytosolic viral dsRNA and prevent it from binding and activating RLR.
Block PRR signaling pathways, preventing
activation of responses
Vaccinia virus protein A46R and several bacterial proteins have TIR
domains that block MyD88 and TRIF from binding to TLRs.
Several viruses block TBK1/IKK-activation of IRF3 and IRF7, required
for IFN production.
West Nile Virus NS1 protein inhibits NF-B and IRF transport into
the nucleus.
Yersinia bacteria produce Yop proteins that inhibit inflammasome
activity; the YopP protein inhibits transcription of the IL-1 gene.
Prevent killing or replication inhibition
Salmonella and Listeria bacteria rupture the phagosome membrane
and escape to the cytosol.
Mycobacteria tuberculosis blocks phagosome fusion with lysosomes.
Vaccinia virus encodes a protein that binds to Type I IFNs and
prevents them from binding to the IFN receptor.
Hepatitis C virus protein NS3-4A and Vaccinia virus protein E3L bind
protein kinase R and block IFN-mediated inhibition of protein synthesis.
loop, levels of circulating TNF- and IL-1 that could potentially induce harmful inflammatory responses act via the
hypothalamus to induce the adrenal medulla to secrete
glucocorticoid hormones, which have a variety of potent
anti-inflammatory effects, including inhibiting the production of TNF- and IL-1 themselves.
Pathogens Have Evolved Mechanisms to Evade
Innate and Inflammatory Responses
As it is advantageous to pathogens to evolve mechanisms
that allow them to evade elimination by the immune system,
many have acquired the ability to inhibit various innate and
inflammatory signaling pathways and effector mechanisms
that would clear them from the body. Most bacteria, viruses,
and fungi replicate at high rates and, through mutation, may
alter their components to avoid recognition or elimination
by innate immune effector mechanisms. Other pathogens
have evolved complex mechanisms that block normally effective
innate clearance mechanisms. A strategy employed especially
by viruses is to acquire genes from their hosts that have
evolved and function as inhibitors of innate and inflammatory
responses. Several strategies by which viruses evade immune
responses triggered by interferons and other cytokines are
listed in Table 4-5. Examples of mechanisms by which
pathogens more generally avoid detection by PRRs, activation of innate and inflammatory responses, or elimination
by those responses are described in Table 5-6.
Interactions Between the Innate
and Adaptive Immune Systems
The many layers of innate immunity are important to our
health, as illustrated by the illnesses seen in individuals lacking one or another component of the innate immune system
(see the Clinical Focus box and other examples mentioned
earlier in this chapter). However, innate immunity is not sufficient to protect us fully from infectious diseases, in part
because many pathogens have features that allow them to
evade innate immune responses, as discussed above. Hence
the antigen-specific responses generated by our powerful
adaptive immune system are usually needed to resolve infections successfully. While our B and T lymphocytes are the
key producers of adaptive response effector mechanisms—
antibodies and cell-mediated immunity—it is becoming
increasingly clear that our innate immune system plays
important roles in helping to initiate and regulate adaptive
174
PA R T I I
|
Innate Immunity
immune responses so that they will be optimally effective. In
addition, the adaptive immune system has co-opted several
mechanisms by which the innate immune system eliminates
pathogens, modifying them to enable antibodies to clear
pathogens.
The Innate Immune System Activates and
Regulates Adaptive Immune Responses
When pathogens invade our body, usually by penetrating
our epithelial barriers, the innate immune system not only
reacts quickly to begin to clear the invaders but also plays
key roles in activating adaptive immune responses. While
innate cells at the infection site (resident macrophages,
dendritic cells, mast cells, and newly recruited neutrophils)
are sensing the invading pathogens through their pattern
recognition receptors and generating antimicrobial and proinflammatory responses that will slow down the infection,
they are also initiating steps to bring the pathogens to the
attention of lymphocytes and help activate responses that,
days later, will generate the strong antigen-specific antibody
and cell-mediated responses that will resolve the infection.
The first step in the generation of adaptive immune
responses to pathogens is the delivery of the pathogen to
lymphoid tissues where T and B cells can recognize it and
respond. As is discussed in more detail in Chapter 14, dendritic cells (usually immature) that serve as sentinels in epithelial tissues bind microbes through various pattern-recognition
receptors. The dendritic cells carry the bound microbes—
either still attached to the cell surface or in phagosomes—via
the lymphatic vessels to nearby secondary lymphoid tissues,
such as the draining lymph nodes. There the dendritic cells
can transfer or present the microbes or microbial components
to other cells. In many cases the dendritic cell has internalized
and degraded the microbe, and microbe-derived peptides
come to the cell surface bound to its MHC class II proteins.
Meanwhile, the binding of microbial PAMPs to the dendritic
cell’s PRRs also has activated the dendritic cell to mature. It
now expresses higher levels of MHC class II proteins and has
turned on expression of costimulatory membrane proteins,
such as CD80 or CD86 (see Chapter 3), that are recognized by
receptors on TH cells. As a result of these processes of microbe
binding, processing, and maturation, mature dendritic cells
are the most effective antigen-presenting cells, particularly for
the activation of naïve (not previously activated) TH cells.
Pathogen-Specific Activation of TH Cell Subsets
Activation of dendritic cells by binding of various PAMPs to
PRRs has additional important consequences for adaptive
immune responses. Depending on the nature of the pathogen
and what PRRs and downstream signaling pathways are
activated, dendritic cells are induced to secrete specific cytokines that influence what cytokines a naïve TH (CD4 ) cell
will produce after activation, thus determining its functional
phenotype (Figure 5-18). These distinct phenotypes, or TH
cell subsets, will be discussed in detail in Chapter 11.
As one example, a Gram-negative bacteria may express
PAMPs that bind to TLR4, TLR5, and/or the endosomal/lysosomal TLRs that bind bacterial nucleic acids. This binding
stimulates dendritic cells to produce IL-12, which induces naïve
TH cells to become TH1 cells (see Chapter 4). The hallmark
cytokine of TH1 cells is IFN-, which activates macrophages to
kill intracellular bacteria, reinforces the TH1 phenotype, contributes to the activation of virus-specific cytotoxic T cells, and,
along with IL-2 (another TH1 cytokine), helps to activate B cells.
In contrast, binding of PAMPs from fungi, Gram-positive
bacteria, or helminths to TLR 2/1 or 2/6 activates production
of IL-10 and inhibits secretion of IL-12. This combination of
effects of stimulation through TLR2-containing receptors
supports naïve TH cell differentiation into TH2 cells, which
promote strong antibody responses that are more important
for providing protection against these pathogens than are
cell-mediated responses. Fungal PAMPs bind to the CLR
Dectin-1, which activates the dendritic cell to produce the
proinflammatory cytokines TNF-, IL-6, and IL-23; the latter two induce the TH to become TH17 cells, which are characterized by the secretion of proinflammatory cytokines
such as IL-17 and IL-6. These cytokines help recruit neutrophils to the site, which then phagocytose and kill the fungi.
It is important to recognize that the schematic shown in
Figure 5-18 is oversimplified, as there are many complexities to
the regulation of TH phenotype not shown. For example, IL-4,
a key cytokine for inducing formation of TH2 cells, may be
made by mast cells and/or basophils following PRR activation
by certain pathogens, including helminth worms. The resulting
TH2 cells induce production of IgG and IgE antibodies that are
particularly effective against some of these pathogens.
In summary, through their differential activation of specific
PRRs, various pathogens induce dendritic and other innate
cells to produce cytokines that activate TH cells to acquire distinct cytokine-producing phenotypes. These distinct TH subsets
generally elicit the types of adaptive immune responses that
will be most effective in eliminating the particular pathogen
invaders. The molecular mechanisms by which naïve CD4
TH cells are influenced to differentiate into various mature
TH subsets will be described in Chapter 11.
Pathogen-Specific Activation of TC Cells
Recent research has also revealed key roles of dendritic cells in
the activation of cytotoxic T-cell-mediated responses that are
needed to kill cells in our body that have become harmful
through infection with viruses or other intracellular pathogens
or through malignant transformation. As was the case with
naïve TH (CD4 ) cells, naïve cytotoxic TC (CD8 ) cells also are
most effectively activated by mature dendritic cells (see Chapter
13). Again, dendritic cell maturation (also referred to as licensing) is usually induced by the binding of microbial components
to a dendritic cell’s PRRs. These mature/licensed DCs also
are capable of novel pathways for processing internalized
|
Innate Immunity
CHAPTER 5
175
Bacteria
Naïve
TLR4 or
TLR5
Dectin-1
T H1
IFN-γ
IL-12
TLR3, 7, 9
Fungi
IL-6
IL-23
Virus
Naïve
TH17
IL-17
IL-10
TLR2/1
Naïve
Helminth
TH2
IL-4
IL-5
IL-13
TLR2/6
Fungi
IL-10
RA
TGF-β
CD28
CD80/86
TCR
MHC II with peptide
FIGURE 5-18 Differential signaling through dendritic cell
PRRs influences helper T cell functions. Microbial PAMPs bind
and activate distinct PRRs and signaling pathways that differentially
induce production of various cytokines and other mediators, such as
retinoic acid (RA). These cytokines interact with receptors on naïve
CD4 T cells that are in the process of being activated both by antigen-derived peptides bound to dendritic cell MHC class II proteins
pathogens so that peptides generated by endosomal/lysosomal
degradation are carried to the cell surface bound to MHC class
I proteins rather than to class II proteins (this process is called
cross-presentation and explained in Chapter 8). These peptide/
MHC class I complexes, together with costimulatory proteins
such as CD80 or CD86 on the surface of the mature/licensed
dendritic cell, activate the naïve TC to give rise to cytotoxic
T effector cells. As was the case with the activation of TH cells,
cytokines such as IL-12 and IFN- produced by innate cells
(dendritic cells, macrophages, NK cells, or other cells) contribute to the differentiation and activation of TC cells.
T-Independent Antigens
In addition to their indirect roles in promoting and regulating antibody and cell-mediated responses via their activation
of DC and other innate cells, described above, TLRs also can
be more directly involved in activating B and T cells. B cells
express TLRs, and the binding of PAMPs to these TLRs activates signaling pathways that can add to or substitute for the
signals from TH cells normally required for B-cell activation.
One example has been well studied in mouse B cells. In
combination with signals from the B cells’ BCR after antigen
binding, TLR4 binding of LPS (at low concentrations) can
activate sufficient signals to induce the B cell to proliferate
Naïve
Treg
IL-10
TGF-β
and by interactions with costimulatory molecules such as CD80 or
CD86. Note that activation of dendritic cells by PAMP binding to TLRs
stimulates increased expression of MHC class II and costimulatory
proteins. The particular cytokine(s) to which a naïve CD4 T cell is
exposed induces that cell turn on genes for certain cytokines, determining that cell’s functional phenotype (see text). [Adapted from B.
Pulendran et al. 2010. Nature Immunology 11:647.]
and differentiate into antibody-secreting plasma cells without
TH cells. At high concentrations of LPS, TLR4-activated signals are sufficient to activate all B cells (polyclonal activation),
regardless of their antigen-binding specificity; hence LPS has
for many years been called a T-independent antigen (see
Chapter 12). Human B cells do not express TLR4 and hence
do not respond to LPS; however, they do express TLR9 and
can be activated by microbial CpG DNA. The ability of TLR
signals to replace TH signals is often beneficial; co-binding of
bacteria to both BCR and TLR on a cell may activate that cell
more quickly than if it had to wait for signals from a TH cell.
Some T cells also express TLRs, which similarly function as
costimulatory receptors to enhance protective responses. For
example, mouse CD8 T cells express TLR2; costimulation of
CD8 T cells by TLR2 ligands along with TCR recognition of
peptide/MHC class I complexes reduces the T cells’ need for
costimulatory signals provided by dendritic cells and
enhances their proliferation, survival, and functions.
Adjuvants Activate Innate Immune Responses
to Increase the Effectiveness of Immunizations
Given these activating and potentiating effects of PRR ligands
on adaptive immune responses, can they be used to enhance
176
PA R T I I
|
Innate Immunity
the efficacy of vaccines in promoting protective immunity
against various pathogens? In fact, many of the materials—
known as adjuvants—that have been shown over the years by
trial and error to enhance immune responses in both laboratory animals and humans contain ligands for TLRs or other
PRRs (see Chapter 17). For example, Complete Freund’s Adjuvant, perhaps the most potent adjuvant for immunizations in
experimental animals, is a combination of mineral oil and
killed Mycobacteria. The mineral oil produces a slowly dispersing depot of antigen, a property of many effective adjuvants,
while fragments of the bacteria’s cell wall peptidoglycans serve
as activating PAMPs. Alum (a precipitate of aluminum hydroxide and aluminum phosphate) is used in some human vaccines;
it has recently been shown to activate the NRLP3 inflammasome, thereby enhancing IL-1 and IL-18 secretion and promoting inflammatory processes that enhance adaptive immune
responses. As the responding innate cells do not make IL-12,
immunizations with alum usually lead the activated T cells to
become TH2 cells, which promote strong antibody responses.
While many vaccines consist of killed or inactivated
viruses or bacteria and hence contain their own PAMPS,
which function as built-in adjuvants, some new vaccines
consist of protein antigens that themselves are not very
stimulatory to the immune system. Tumor antigens also tend
to induce weak responses. Hence, considerable effort is being
invested in developing new adjuvants based on current
knowledge about PRRs. LPS is a highly potent adjuvant but
generates too much inflammation to use; less harmful versions of LPS are being developed as potential adjuvants. The
ability of the TLR3 ligand poly I:C (synthetic double-stranded
RNA) to activate innate immunity and initiate inflammatory
responses has inspired clinical trials to test its ability to
enhance the effectiveness of weak vaccines, including antitumor vaccines. Vaccines using the TLR9 ligand CpG DNA
(which mimics bacterial DNA) as an adjuvant are in clinical
trials; there is great interest in CpG DNA as it preferentially
elicits a TH1 response, important for inducing cell-mediated
immunity (see Figure 5-18). Other approaches for generating
an effective vaccine for a pathogen protein include fusing the
protein to a TLR ligand using genetic engineering. For
example, fusions of pathogen proteins to the TLR5 ligand
flagellin are currently being tested. These are all examples of
how our expanding knowledge of the interactions between
the innate and adaptive immune systems may contribute to
the development of more effective vaccines.
Some Pathogen Clearance Mechanisms
are Common to Both Innate and Adaptive
Immune Responses
Adaptive immune responses—antibody responses in particular—have adopted and modified several effector functions
by which the innate immune system eliminates antigen, so
that they are also triggered by antibody binding to antigens.
While some will be discussed in more detail in Chapters 6
and 13, several examples are mentioned briefly here as illus-
trations of important interactions between innate and adaptive immune responses.
As discussed earlier in this chapter, several soluble proteins that recognize microbial surface components—including
SP-A, SP-D, and MBL—function as opsonins; when they are
bound to microbial surfaces they are recognized by receptors
on phagocytes, leading to enhanced phagocytosis. Antibodies
of the IgG and IgA classes also serve as opsonins; after binding to microbial surfaces these antibodies can be recognized
by Fc receptors that are expressed on macrophages and other
leukocytes, triggering phagocytosis (see Chapter 3). One
receptor for IgG Fc regions (FcR), called CD16, is also
expressed on NK cells, where it serves as an activating receptor. Thus, when IgG antibodies bind to foreign antigens on
the cell surfaces (such as viral or tumor antigens), the IgG
antibodies can be recognized by the FcR on the NK cell,
triggering NK-cell-mediated cytotoxic killing of the infected
or malignant cell (see Chapter 13).
Finally, as mentioned earlier and described fully in Chapter 6, the complement pathway can be activated by both
innate and antibody-mediated mechanisms. Components
on microbe surfaces can be directly recognized by soluble
pattern-recognition proteins, including MBL and C-reactive
protein, leading to activation of the complement cascade.
Similarly, when antibodies of the IgM class and certain IgG
subclasses bind microbe surfaces, their Fc regions are recognized by the C1 component of complement, also triggering
the complement cascade. As was shown in Figure 5-7, MBL
and C1 are related structurally. Once the complement pathway is activated by any of these microbe-binding proteins, it
generates a common set of protective activities; various
complement components and fragments promote opsonization, lysis of membrane-bounded microbes, and generation
of fragments that have proinflammatory and chemoattractant activities. Thus, the adaptive immune system makes
good use of mechanisms that initially evolved to contribute
to innate immunity, co-opting them for the elimination of
pathogens.
Ubiquity of Innate Immunity
Determined searches among plant and invertebrate animal
phyla for the signature proteins of the highly efficient adaptive immune system—antibodies, T cell receptors, and MHC
proteins—have failed to find any homologs of these important vertebrate proteins. Yet without them multicellular
organisms have managed to survive for hundreds of millions
of years. The interior spaces of organisms as diverse as the
tomato, fruit fly, and sea squirt (an early chordate, without a
backbone) do not contain unchecked microbial populations.
Careful studies of these and many other representatives of
nonvertebrate phyla have found arrays of well-developed
processes that carry out innate immune responses. The accumulating evidence leads to the conclusion that multiple
immune mechanisms protect all multicellular organisms
Innate Immunity
TABLE 5-7
|
CHAPTER 5
177
Immunity in multicellular organisms
Taxonomic
group
Higher plants
Innate
immunity
(nonspecific)
Adaptive
immunity
(specific)
Invasioninduced
protective
enzymes
and enzyme
cascades
Phagocytosis
Anti
microbial
peptides
Pattern
recognition
receptors
Lymphocytes
Variable
lymphocyte
receptors
Antibodies
Invertebrate animals
Porifera (sponges)
?
Annelids (earthworms)
?
Arthropods (insects,
crustaceans)
Vertebrate animals
Jawless fish (hagfish,
lamprey)
Elasmobranchs
(cartilaginous
fish; e.g.,
sharks, rays)
Bony fish (e.g.,
salmon, tuna)
Amphibians
Reptiles
Birds
Mammals
Sources: M. J. Flajnik and L. Du Pasquier, 2008, “Evolution of the Immune System,” in Fundamental Immunology (6th ed.), W. E. Paul, ed., Philadelphia: Lippincott; J. H.
Wong et al., 2007, “A review of defensins of diverse origins,” Current Protein and Peptide Science 8:446.
from microbial infection and exploitation (Table 5-7). Some
of the innate immune system components occur across the
plant and animal kingdoms. For example, as mentioned
early in this chapter, virtually all plant and animal species,
and even some fungi, have antimicrobial peptides similar to
defensins. Most multicellular organisms have patternrecognition receptors containing leucine-rich repeats
(LRRs), although many organisms also have other families of
PRRs. While the innate immune responses activated by
these receptors in plants and invertebrates show both similarities and differences compared to those of vertebrates,
innate immune response mechanisms are essential for the
health and survival of these varied organisms.
Plants Rely on Innate Immune Responses
to Combat Infections
Despite plants’ tough outer protective barrier layers, such as
bark and cuticle, and the cell walls surrounding each cell,
plants can be infected by a wide variety of bacteria, fungi,
and viruses, all of which must be combated by the plant
innate immune system. Plants do not have phagocytes or
other circulating cells that can be recruited to sites of infection to mount protective responses. Instead, they rely on
local innate immune responses for protection against
infection. As described in Box 5-3, some resemble innate
responses of animals, while others are quite distinct.
Invertebrate and Vertebrate Innate Immune
Responses Show Both Similarities
and Differences
Additional innate immune mechanisms evolved in animals,
and we vertebrates share a number of innate immunity
features with invertebrates. PRRs (including relatives of
Drosophila Toll and vertebrate TLRs) that have specificities
for microbial carbohydrate and peptidoglycan PAMPs are
found in organisms as primitive as sponges. Together with
soluble opsonin proteins (including some related to complement components), some of these early invertebrate PRRs
function in promoting phagocytosis, an early mechanism for
clearing pathogens. Innate signaling has been well studied in
Drosophila, and signaling proteins have been identified in
the flies that are homologous to several of those downstream
of vertebrate TLRs (including MyD88 and IRAK homologs).
A difference is that fly Toll does not bind to PAMPs directly
but is indirectly activated by the soluble protein product of
an enzyme cascade triggered by pathogen binding to soluble
PA R T I I
178
|
Innate Immunity
EVOLUTION
Plant Innate Immune Responses
In the plasma membrane
under the cell wall, plant cells express
pattern recognition receptors with LRR
domains reminiscent of animal TLRs.
These PRRs recognize what plant biologists refer to as microbe-associated
molecular patterns (MAMPs), including
bacterial flagellin, a highly conserved
bacterial translation elongation factor,
and various bacterial and fungal cell wall
components (Figure 1). As is true for animal TLRs, some plant PRRs respond to
danger-associated molecular patterns
(DAMPs), which usually are created by
pathogen enzymes that attack and fragment cell wall components. Some bacterial and fungal pathogens directly inject
into plant cells toxin effector proteins
that inhibit signaling through the plasma
membrane PRRs. These toxins are recognized by a distinct class of LRR receptors
in the cytoplasm called R proteins, which,
like animal NLR proteins, have both LRR
and
nucleotide-binding
domain
domains. After ligand binding, plant PRRs
activate signaling pathways and transcription factors distinct from those of
Oomycetes
Fungi
MAMPs
DAMPs
Bacteria
PRRs
Plant
cell
wall
PRRs
Cell wall
degrading
enzymes
Extracellular
effectors
PRRs
Intracellular
effectors
Cytoplasm
Innate immune
responses
R-protein
PRRs
FIGURE 1
Activation of plant innate immune responses. Microbe-associated molecular patterns (MAMPs), danger-associated molecular patterns
(DAMPs) generated by microbial enzymes (such as by degrading the cell wall), and microbial effectors (such as toxins) are recognized by plasma
membrane LRR-containing pattern recognition receptors (PRRs). Microbial effectors that enter the cytoplasm are recognized by a class of PRRs called
resistance (R) proteins. Recognition of MAMPs, DAMPs, and effectors by PRRs induces innate immune responses. [Adapted from T. Boller and G. Felix, 2009, Annual
Review of Plant Biology 60:379.]
|
Innate Immunity
CHAPTER 5
179
BOX 5-3
vertebrate cells (plants do not have
NF-B or IRF homologs), triggering
innate responses.
The primary protective innate immune
response mechanisms of plants to infection are the generation of reactive oxygen and nitrogen species, elevation of
internal pH, and induction of a variety of
antimicrobial peptides (including defensins) and antimicrobial enzymes that can
digest the walls of invading fungi (chitin-
ases) or bacteria (-1,3 glucanase). Plants
may also be activated to produce organic
molecules, such as phytoalexins, that
have antibiotic activity. In some cases, the
responses of plants to pathogens even
goes beyond these protective substances
to include structural responses. For example, to limit infection of leaves, PAMP
binding to PRRs induces the epidermal
guard cells that form the openings (stomata) involved in leaf gas exchange to
Tomato
close, preventing further invasion (Figure
2). Other protective mechanisms include
the isolation of cells in the infected area
by strengthening the walls of surrounding noninfected cells and the induced
death (necrosis) of cells in the vicinity of
the infection to prevent the infection
from spreading to the rest of the plant.
Mutations that disrupt any of these processes usually result in loss of the plant’s
resistance to a variety of pathogens.
Tobacco
Open stomata
PAMPs
or bacteria
PAMPs
or bacteria
Closed stomata
FIGURE 2
Induced closure of leaf stomata following exposure to bacterial PAMPs. In normal light conditions, the openings (stomata) formed
by pairs of guard cells in the leaf epidermis are open (visible as the football-shaped openings in the upper pair of leaf photographs). However, as shown
in the lower panels, exposure of the tobacco leaf to the bacterial plant pathogen Pseudomonas syringae and exposure of the tomato leaf to bacterial LPS
induce the closure of the stomata (no openings visible). [From M. Melotto et al. 2008. Annual Review of Phytopathology 46:101–122.]
180
PA R T I I
|
Innate Immunity
pattern-recognition proteins. Through this pathway bacterial
and fungal infections lead to the degradation of an IB homolog and activation of NF-B family members, Dif and Dorsal, which induce the production of drosomycin, an insect
defensin, and other antimicrobial peptides. In addition to
these and other pathways activated by PRRs, Drosophila and
other arthropods employ other innate immune strategies,
including the activation of phenoloxidase cascades that result
in melanization—the deposition of a melanin clot around
invading organisms that prevents their spread. Thus invertebrates and vertebrates have common as well as distinct
innate immune response mechanisms.
S U M M A R Y
■
■
■
■
■
■
The receptors of the innate immune system recognize
conserved pathogen-associated molecular patterns (PAMPs),
which are molecular motifs found in microbes, and damage-associated molecular patterns (DAMPs) from aging,
damaged, or dead cells. Therefore these receptors are
called pattern-recognition receptors (PRRs).
The epithelial layers that insulate the interior of the body
from outside pathogens—the skin and epithelial layers of
the mucosal tracts and secretory glands—constitute an
anatomical barrier that is highly effective in protecting
against infection. These epithelial layers produce a variety
of protective substances, including acidic pH, enzymes,
binding proteins, and antimicrobial proteins and peptides.
Phagocytosis—engulfment and internalization of particulate materials such as microbes—is mediated by receptors
on phagocytes (monocytes, macrophages, neutrophils,
and dendritic cells) that either directly recognize PAMPs
on the surface of microbes or recognize soluble proteins
(opsonins) that bind to the microbes. PAMP binding triggers microbe uptake into phagosomes, which fuse with
lysosomes or prepackaged granules, leading to their killing through enzymatic degradation, antimicrobial proteins
and peptides, and toxic effects of reactive oxygen species
(ROS) and reactive nitrogen species (RNS).
Families of PRRs (TLRs, CLRs, RLRs, and NLRs) recognize a wide variety of PAMP (and DAMP) ligands and
trigger signaling pathways that activate genes encoding
proteins that contribute to innate and inflammatory
responses.
Vertebrate Toll-like receptors (TLRs, homologous to the
fruit fly Toll receptor) are dimers of chains with extracellular leucine-rich (LRR) domains that bind PAMPs and
DAMPs. Of the 13 TLRs found in mice and humans, some
are on the cell surface, while others are in endosomes and
lysosomes, where each binds PAMPs revealed by the disassembly or degradation of pathogens, such as microbial
nucleic acids.
TLR binding of PAMPs activates signaling pathways; the
particular pathway varies depending on the TLR and the
adaptor protein that binds to its cytoplasmic TIR domain
(either MyD88 or TRIF). The signaling pathways activate
transcription factors NF-B, various interferon regulating
factors (IRFs), and transcription factors (such as AP-1)
downstream of MAP kinase (MAPK) pathways.
■
■
■
■
■
■
C-type lectin receptors (CLRs) comprise a heterogeneous
family of cell surface PRRs that recognize cell wall components, largely sugars and polysaccharides, of bacteria and
fungi. They trigger a variety of distinct signaling pathways
that activate transcription factors, of which some are
similar to those activated by TLRs.
RIG-I-like receptors (RLRs) are RNA helicases that function as cytosolic PRRs. Most PAMPs that they recognize
are viral double-stranded RNAs; after PAMP binding,
RLRs trigger signaling pathways that activate IRFs and
NF-B.
NOD-like receptors (NLRs) are a large family of cytosolic
PRRs activated by intracellular PAMPs, DAMPs, and
other harmful substances. The NOD NLRs bind intracellular microbial components such as cell wall fragments.
Other NLRs, such as NLRP3, function as inflammasomes;
they are activated by a wide variety of PAMPs, DAMPs,
and crystals by sensing changes in the intracellular milieu.
Inflammasomes then activate the caspase-1 protease,
which cleaves the inactive large precursors of the proinflammatory cytokines IL-1 and IL-18, so that active cytokines
can be released from cells.
The signaling pathways downstream of PRRs activate
expression of a variety of genes, including those for antimicrobial peptides, Type 1 interferons (potent antiviral
agents), cytokines (including proinflammatory IL-1,
TNF-, and IL-6), chemokines, and enzymes that help to
generate antimicrobial and inflammatory responses.
Inflammatory responses are activated by the innate
immune response to local infection or tissue damage—in
particular, by the proinflammatory cytokines and certain chemokines that are produced. Key early components of inflammatory responses are increased vascular
permeability, allowing soluble innate mediators to reach
the infected or damaged site, and the recruitment of
neutrophils and other leukocytes from the blood into
the site.
A later component of inflammatory responses is the acute
phase response (APR), induced by certain proinflammatory cytokines (IL-1, TNF-, and IL-6). The APR involves
increased synthesis and secretion by the liver of several
antimicrobial proteins, including MBL, CRP, and complement components, which activate a variety of processes
that contribute to eliminating pathogens.
Innate Immunity
■
■
■
■
■
Natural killer (NK) cells are lymphocytes with innate
immune functions. They express a set of activating receptors that recognize surface components of the body’s cells
induced by infection, malignant transformation, or other
stresses. Activated NK cells can kill the altered self cell
and/or produce cytokines that help to induce adaptive
immune responses to the altered cell.
The importance of innate and inflammatory responses is
demonstrated by the impact of a variety of genetic defects
in humans. Defects in PRRs and signaling pathways activating innate responses lead to increased susceptibility to
certain infections, while other defects that constitutively
activate inflammasomes contribute to a variety of inflammatory disorders.
As innate and inflammatory responses can be harmful as
well as helpful, they are highly regulated by positive and
negative feedback pathways.
To escape elimination by innate immune responses,
pathogens have evolved a wide array of strategies to block
antimicrobial responses.
There are many interactions between the innate and adaptive immune systems. The adaptive immune system has
|
CHAPTER 5
181
co-opted several pathogen-clearance mechanisms, such as
opsonization and complement activation, so that they
contribute to antibody-mediated pathogen elimination.
■
■
■
B cells and some T cells express TLRs, which can serve
as costimulatory receptors. PAMP binding helps activate
these cells to generate adaptive immune responses.
TLRs expressed by NK cells can serve as activating
receptors.
Dendritic cells are a key cellular bridge between innate
and adaptive immunity. Microbial components acquired
by dendritic cells via PRR binding during the innate
response are brought from the site of infection to lymph
nodes. Activation of a dendritic cell by PAMPs stimulates
the cell to mature, so that it acquires the abilities to activate naïve T cells to become mature cytotoxic and helper
T cells and to influence the sets of cytokines that the
mature TH will produce.
Innate immunity appeared early during the evolution of
multicellular organisms. While TLRs are unique to animals,
PRRs with leucine-rich repeat (LRR) ligand-binding domains
that induce innate immune responses are found in virtually all plants and animals.
R E F E R E N C E S
Areschoug, T., and S. Gordon. 2009. Scavenger receptors: role
in innate immunity and microbial pathogenesis. Cellular
Microbiology 11:1160.
Fukata, M., et al. 2009. Toll-like receptors (TLRs) and Nod-like
receptors (NLRs) in inflammatory disorders. Seminars in
Immunology 21:242.
Beutler, B., and E. T. Rietschel. 2003. Innate immune sensing
and its roots: The story of endotoxin. Nature Reviews
Immunology 3:169.
Geijtenbeek, T. B. H., and S. I. Gringhuis. 2009. Signaling
through C-type lectin receptors: Shaping immune repertoires.
Nature Reviews Immunology 9:465.
Bowie, A. G., and L. Unterholzner. 2008. Viral evasion and subversion of pattern-recognition receptor signalling. Nature
Reviews Immunology 8:911.
Grimsley, C., and K. S. Ravichandran. 2003. Cues for apoptotic
cell engulfment: Eat-me, don’t-eat-me, and come-get-me signals. Trends in Cellular Biology 13:648.
Bulet, P., et al. 2004. Anti-microbial peptides: from invertebrates to vertebrates. Immunology Review 198:169–184.
Herrin, B. R., and M. D. Cooper. 2010. Alternative adaptive
immunity in jawless vertebrates. Journal of Immunology
185:1367.
Chroneos, Z. C., et al. 2010. Pulmonary surfactant: An immunological perspective. Cellular Physiology and Biochemistry
25:13.
Coffman, R., et al. 2010. Vaccine adjuvants: Putting innate
immunity to work. Immunity 33:492–503.
Davis, B. K., et al. 2011. The inflammasome NLRs in immunity,
inflammation, and associated diseases. Annual Review of
Immunology 11:707.
deFranco, A., et al. 2007. Immunity (Primers in Biology).
Sunderland, MA: Sinauer Associates.
Elinav, E., et al. 2011. Regulation of the antimicrobial response
by NLR proteins. Immunity 34:665.
Froy, O. 2005. Regulation of mammalian defensin expression by
Toll-like receptor-dependent and independent signalling
pathways. Cellular Microbiology 7:1387.
Kawai, T., and S. Akira. 2010. The role of pattern recognition
receptors in innate immunity: Update on Toll-like receptors.
Nature Immunology 11:373.
Kerrigan, A. M., and G. D. Brown. 2009. C-type lectins and
phagocytosis. Immunobiology 214:562.
Lemaitre, B. 2004. The road to Toll. Nature Reviews Immunology
4:521.
Litvack, M. L., and N. Palaniyar. 2010. Soluble innate immune
pattern-recognition proteins for clearing dying cells and cellular components: Implications on exacerbating or resolving
inflammation. Innate Immunity 16:191.
Loo, Y-M., and M. Gale. 2011. Immune signaling by RIG-I-like
receptors. Immunity 34:680.
Medzhitov, R., et al. 1997. A human homologue of the Toll protein
signals activation of adaptive immunity. Nature 388:394.
182
PA R T I I
|
Innate Immunity
Nathan, C., and A. Ding. 2010. SnapShot: Reactive oxygen
intermediates (ROI). Cell 140:952.
Palm, N. W., and R. Medzhitov. 2009. Pattern recognition receptors and control of adaptive immunity. Immunology Review
227:221.
Poltorak, A., et al. 1998. Defective LPS signaling in C3Hej and
C57BL/10ScCr mice: Mutations in TLR4 gene. Science
282:2085.
Salzman, N. H., et al. 2010. Enteric defensins are essential regulators of intestinal microbial ecology. Nature Immunology 11:76.
Schroder, K., and J. Tschopp. 2010. The inflammasome. Cell
140:821.
Sun, J. C., et al. 2011. NK cells and immune “memory.” Journal
of Immunology 186:1891.
Takeuchi, O., and S. Akira. 2010. Pattern recognition receptors
and inflammation. Cell 140:805.
Tecle, T., et al. 2010. Defensins and cathelicidins in lung immunity. Innate Immunity 16:151.
Tieu, D. T., et al. 2009. Alterations in epithelial barrier function
and host defense responses in chronic rhinosinusitis. Journal
of Clinical Immunology 124:37.
van de Vosse, E., et al. 2009. Genetic deficiencies of innate
immune signaling in human infectious disease. The Lancet
Infectious Diseases 9:688.
Willingham, S. B., et al. 2012. The CD47-signal regulatory
protein alpha (SIRP1a) interaction is a therapeutic target for
human solid tumors. Proceedings of the National Academy of
Sciences USA 109:6662.
Yu, M., and S. J. Levine. 2011. Toll-like receptor 3, RIG-I-like
receptors and the NLRP3 inflammasome: Key modulators
of innate immune responses to double-stranded viruses.
Cytokine & Growth Factor Reviews 22:63.
Useful Web Sites
cpmcnet.columbia.edu/dept/curric-pathology/
pathology/pathology/pathoatlas/GP_I_menu.
html Images are shown of the major inflammatory cells
involved in acute and chronic inflammation, as well as
examples of specific inflammatory diseases.
portal.systemsimmunology.org/portal/web/guest/
homepage Systems Approach to Immunology (systemsimmunology.org) is a large collaborative research program
formed to study the mechanisms by which the immune system responds to infectious disease by inciting innate inflammatory reactions and instructing adaptive immune responses.
www.biomedcentral.com/1471-2172/9/7 Web site
of The Innate Immunity Database, an NIH-funded multiinstitutional project that assembled microarray data on
expression levels of more than 200 genes in macrophages
stimulated with a panel of TLR ligands. The database is
intended to support systems biology studies of innate
responses to pathogens.
www.immgen.org/index_content.html The Immunological Genome Project is a new cooperative effort for
deep transcriptional profiling of all immune cell types.
www.ncbi.nlm.nih.gov/PubMed PubMed, the National
Library of Medicine database of more than 15 million publications, is the world’s most comprehensive bibliographic
database for biological and biomedical literature. It is also
highly user friendly, searchable by general or specific topics,
authors, reviews, and so on. It is the best resource to use for
finding the latest research articles on innate immunity or
other topics in the biomedical sciences.
animaldiversity.ummz.umich.edu/site/index.
html The Animal Diversity Web (ADW), based at the University of Michigan, is an excellent and comprehensive
database of animal classification as well as a source of information on animal natural history and distribution. Here is
a place to find information about animals that are not
humans or mice.
www.biolegend.com/media_assets/.../InnateImmunityResourceGuide_3.pdf Presents a concise sum-
en.wikipedia.org/wiki/Innate_immune_system The
mary of innate immunity, including a detailed table on the
cells of the innate immune system, including their locations
in the body, functions, products, receptors, and modes of
activation.
Wikipedia Web site presents a detailed summary of the
innate immune system in animals and plants; it contains
numerous figures and photographs illustrating various
aspects of innate immunity, plus links to many references.
S T U D Y
Q U E S T I O N S
CLINICAL FOCUS QUESTION What infections are unusually
prevalent in individuals with genetic defects in TLRs or the
MyD88-dependent TLR signaling pathway? in individuals with
defects in pathways activating the production or antiviral activities of IFN-,? Why is it thought that these individuals aren’t
susceptible to a wider range of diseases, and what evidence supports this hypothesis?
1. Use the following list to complete the statements that fol-
low. Some terms may be used more than once or not at all.
Innate Immunity
Acute phase response
NK cells
Antibodies
NLRs
Arginine
NO
Caspase-1
O2
C-reactive protein (CRP)
PAMPs
Complement
Phagocytosis
Costimulatory molecules
Proinflammatory cytokines
Defensins
PRRs
Dendritic cells
Psoriasin
Ficolins
RLRs
IL-1
ROS
Inflammasomes
RNS
iNOS
Surfactant proteins (SP) A, D
Interferons-,
TRIF
IRFs
T-cell receptors
Lysozyme
TLR2
Mannose-binding lectin (MBL)
TLR3
MyD88
TLR4
NADPH
TLR7
NADPH phagosome oxidase
TLR9
NF-B
TNF-
a. Examples of proteins and peptides with direct antimi-
b.
c.
d.
e.
f.
g.
crobial activity that are present on epithelial surfaces are
_____, _____, and _____.
Soluble pattern-recognition proteins that function as
opsonins, which enhance ____, include _____, _____,
_____, and _____; of these, the ones that also activate
complement are _____ and _____.
The enzyme _____ uses _____ to generate microbekilling _____; one of these, plus the antimicrobial gas
____ generated by the _____ enzyme from the amino
acid _____, are used to generate ____, which are also
antimicrobial.
As components of innate immune responses, both
_____ (secreted proteins) and _____ (a type of lymphocyte) defend against viral infection.
The _____ occurs when _____, such as _____ and
_____, are generated by innate immune responses and
act on the liver.
_____, receptors of innate immunity that detect
_____, are encoded by germline genes, whereas the
signature receptors of adaptive immunity, _____ and
_____, are encoded by genes that require gene rearrangements during lymphocyte development to be
expressed.
Among cell surface TLRs, _____ detects Gram-positive
bacterial infections while _____ detects Gram-negative
infections.
|
CHAPTER 5
183
h. Some cells use the intracellular TLRs _____ and _____
to detect RNA virus infections and _____ to detect
infections by bacteria and some DNA viruses.
i. _____ is unique among PRRs in that it functions both
on the plasma membrane and in endosomes and binds
both the _____ and _____ adaptor proteins.
j. _____ are receptors that detect cytosolic viral nucleic
acids, while _____ include cytosolic receptors that
detect intracellular bacterial cell wall components.
k. Key transcription factors for inducing expression of
proteins involved in innate immune responses are
_____ and _____.
l. The production of the key proinflammatory cytokine
_____ is complex, as it requires transcriptional activation by signaling pathways downstream of _____ followed by cleavage of its large precursor protein by
_____, which is activated by _____, members of the
_____ family of innate receptors.
m. After maturation induced by binding of _____ to their
_____, cells known as _____ become efficient activators
of naïve helper and cytotoxic T cells.
n. _____ and _____ are innate immune system components common to both plants and animals.
2. What were the two experimental observations that first
linked TLRs to innate immunity in vertebrates?
3. What are the hallmark characteristics of a localized inflam-
matory response? How are they induced by the early innate
immune response at the site of infection, and how do these
characteristics contribute to an effective innate immune
response?
4. In vertebrates, innate immunity collaborates with adaptive
immunity to protect the host. Discuss this collaboration,
naming key points of interaction between the two systems.
Include at least one example in which the adaptive immune
response contributes to enhanced innate immunity.
5. As adaptive immunity evolved in vertebrates, the more
ancient system of innate immunity was retained. Can
you think of any disadvantages to having a dual system
of immunity? Would you argue that either system is more
essential?
ANALYZE THE DATA A variety of studies have shown that hel-
minth worm parasites have developed numerous mechanisms
for inhibiting innate and inflammatory responses that otherwise would contribute to their elimination. One particular
parasite, the filarial nematode Acanthocheilonema viteae, has
been shown to secrete a lipoglycoprotein, ES-62, that inhibits
proinflammatory responses. Based on these initial observations, a group of investigators has been studying how ES-62
achieves this anti-inflammatory effect and whether ES-62
might inhibit the development of septic shock in individuals
with sepsis. Interestingly, they found that ES-62 binds to TLR4
(but not to any other TLR). Based on what you have learned
in this chapter and the information and data provided below
from these investigators’ recent paper (P. Puneet et al., 2011,
Nature Immunology 12:344), answer the following questions.
(Note: The various methods used to obtain the data below are
described in Chapter 20.)
PA R T I I
184
|
Innate Immunity
Basal
LPS
LPS + ES-62
BLP
BLP + ES-62
Mediator release (ng/ml)
2.8
2.4
2.0
1.6
1.2
0.8
0.4
0
*
*
*
IL-1β
IL-6
*
*
*
*
*
IL-8
*
TNF
HMGB1
(a) Effects of ES-62 on proinflammatory mediator production
Macrophages
100
100
Events
Events
150
50
101
102
103
50
0 0
10
104
Macrophages
RANTES
101
102
103
104
Neutrophils
100
Events
Events
*
TLR4
100
50
0 0
10
MCP-1
*
Isotype
Normal
Sepsis
Sepsis
+ ES-62
TLR4
(b)
*
Neutrophils
150
0 0
10
*
flammatory cytokines IL-1, IL-6, and TNF-; the
chemokines IL-8 (CXCL8), MCP-1 (CCL2), and
RANTES (CCL5); and the nuclear protein HMGB1
(which is released by activated or damaged cells and
has proinflammatory activities).
Macrophages from patients with sepsis were incubated
with nothing (“Basal”), or with the TLR4 ligand LPS or
the TLR2 lipoprotein ligand BLP with or without
ES-62. After 24 hours, tissue culture fluids were harvested and assayed by ELISA for levels of the proin(a)
**
101
102
103
104
TLR2
(b) Effects of ES-62 on cell surface expression of TLR4 and
TLR2
Levels of TLR4 and TLR2 were measured on macrophages and neutrophils from normal people and from
patients with sepsis incubated with or without ES-62 for
3 hours. Cells were stained with fluorescent antibodies to
TLR4 or TLR2, or with control (nonreactive) antibodies
Isotype
Normal
Sepsis
Sepsis
+ ES-62
50
0 0
10
101
102
103
104
TLR2
with the same immunoglobulin heavy-chain class
(“Isotype”). The stained cells were analyzed using a
flow cytometer for the levels of their fluorescence,
indicated on the X-axis. The fluorescence obtained
with the Isotype control antibody represents no specific staining. The number of cells (“events”) with different levels of fluorescence is indicated on the Y-axis.
(c) Effect of ES-62 on MyD88 levels
Cell extracts were prepared from macrophages and neutrophils from patients with sepsis that were incubated
without (0) or for 3 or 6 hours with ES-62. The extract
proteins were separated by size by SDS-PAGE and subjected to Western blots to assess levels of MyD88 and, as a
control, the ubiquitous cellular protein -tubulin. The gels
were blotted onto filter paper and the filters stained with
antibodies to MyD88 and to -tubulin. Bound antibodies
were visualized to reveal the levels of MyD88 and tubulin.
Innate Immunity
a. What is septic shock, and how is it induced by bacterial
infections?
b. The investigators tested whether ES-62 inhibits the production of proinflammatory mediators by macrophages
from normal people and from patients hospitalized with
sepsis. Shown in panel (a) above are the results with macrophages from patients with sepsis; cells from normal
individuals gave similar results. What effect does ES-62
have on the responses induced by the two PAMPs? Given
that ES-62 binds to TLR4 but not TLR2, what do these
results say about the mechanism of ES-62 inhibition?
The investigators compared the levels of TLR4 and
TLR2 expressed by macrophages and neutrophils from
normal people with those from patients with sepsis. As had
been seen by others, panel (b) shows that cells from patients
with sepsis express higher levels of these TLRs than do cells
from normal individuals. Of what kind of feedback regulation is this an example? Why is this feedback response
normally beneficial in individuals with limited infections
but disadvantageous in individuals with sepsis?
|
CHAPTER 5
185
In the same experiment shown in panel (b), some
cells from patients with sepsis were incubated with
ES-62. What effect does ES-62 have on TLR4 expression? on TLR2 expression?
The investigators also looked at the levels of MyD88
in macrophages and neutrophils from sepsis patients
after no exposure or 3 or 6 hours of exposure to ES-62
(panel [c]). What does this Western blot show?
Returning to the initial finding that ES-62 inhibits
the induction of proinflammatory mediators by PAMPs
that activate either TLR4 or TLR2 [panel (a)], how
might that finding be explained by the data shown in
panels (b) and (c) (i.e., the effects of ES-62 on TLR and
MyD88 expression)?
Given these results, do you think injection of ES-62
might be used to cure septic shock once it has occurred?
Do you think ES-62 might prevent septic shock from
occurring in patients with sepsis? Explain your answers.
This page lelt intentionally blank.
6
The Complement System
T
he term complement (spelled with an e) refers to
a set of serum proteins that cooperates with both
the innate and the adaptive immune systems to
eliminate blood and tissue pathogens. Like the
components of the blood clotting system, complement
proteins interact with one another in catalytic cascades.
As mentioned in Chapters 3 and 5, various complement
components bind and opsonize bacteria, rendering them
susceptible to receptor-mediated phagocytosis by
macrophages, which express membrane receptors for
complement proteins. Other complement proteins elicit
inflammatory responses, interface with components of the
adaptive immune system, clear immune complexes from
the serum, and/or eliminate apoptotic cells. Finally, a
Membrane Attack Complex (MAC) assembled from
complement proteins directly kills some pathogens by
creating pores in microbial membranes. The biological
importance of complement is emphasized both by the
pathological consequences of mutations in the genes
encoding complement proteins as well as by the broad
range of strategies that have evolved in microorganisms to
evade it.
Research on complement began in the 1890s when
Jules Bordet at the Institut Pasteur in Paris showed that
sheep antiserum to the bacterium Vibrio cholerae caused
lysis (membrane destruction) of the bacteria, and that
heating the antiserum destroyed its bacteriolytic activity.
Surprisingly, the ability to lyse the bacteria was restored to
the heated serum by adding fresh serum that contained
no antibacterial antibodies. This finding led Bordet to
reason that bacteriolysis required two different
substances: the heat-stable specific antibodies that bound
to the bacterial surface, and a second, heat-labile
(sensitive) component responsible for the lytic activity.
In an effort to purify this second, nonspecific
component, Bordet developed antibodies specific for red
blood cells and used these, along with purified serum
fractions, to identify those fractions that cooperated with
the antibodies to induce lysis of the red blood cells
(hemolysis). The famous immunologist Paul Ehrlich,
working independently in Berlin, carried out similar
Sheep red blood cells (red) are phagocytosed
by macrophages (green) after opsonization
by complement.
■
The Major Pathways of Complement Activation
■
The Diverse Functions of Complement
■
The Regulation of Complement Activity
■
Complement Deficiencies
■
Microbial Complement Evasion Strategies
■
The Evolutionary Origins of the
Complement System
experiments and coined the term complement, defining it
as “the activity of blood serum that completes the action
of antibody.”
In the ensuing years, researchers have discovered that
the action of complement is the result of interactions
among a complex group of more than 30 glycoproteins.
Most complement components are synthesized in the liver
by hepatocytes, although some are also produced by other
cell types, including blood monocytes, tissue macrophages,
fibroblasts, and epithelial cells of the gastrointestinal and
genitourinary tracts. Complement components constitute
approximately 15% of the globulin protein fraction in
plasma, and their combined concentration can be as high
as 3 mg/ml. In addition, several of the regulatory
components of the system exist on cell membranes, so the
term complement therefore now embraces glycoproteins
distributed among the blood plasma and cell membranes.
Complement components can be classified into seven
functional categories (Overview Figure 6-1):
187
188
PA R T I I
|
Innate Immunity
6-1
OVERVIEW FIGURE
Proteins Involved in the Complement System
1
2
Convertase activators
(C1r, C1s, C4b, C2a)
and enzymatic mediators
(C3 convertase, C5 convertase)
Inititators
(C1q, MBL, ficolins)
+
C5b
C6
C7
Phagocyte
3
Opsonins
4
5
Anaphylatoxins
C8C9
C9
9
C9
C9 C9
Membrane
attack complex
6
Complement receptors
Inflammation
Degrade complement
components
7
(1) The complement pathways are initiated by proteins that bind
to pathogens, either directly or via an antibody or other pathogenspecific protein. After a conformational change, (2) enzymatic
mediators activate other enzymes that generate the central proteins of the complement cascade, the C3 and C5 convertases,
which cleave C3 and C5, releasing active components that
mediate all functions of complement, including (3) opsonization,
(4) inflammation, and (5) the generation of the membrane attack
Prevent
deposition of
components
Regulatory proteins
complex (MAC). Effector complement proteins can label an
antibody-antigen complex for phagocytosis (opsonins), initiate
inflammation (anaphylatoxins), or bind to a pathogen and nucleate the formation of the MAC. Often, these effectors act through
(6) complement receptors on phagocytic cells, granulocytes, or
erythrocytes. (7) Regulatory proteins limit the effects of complement by promoting their degradation or preventing their binding to host cells.
The Complement System
1.
2.
3.
Initiator complement components. These proteins initiate
their respective complement cascades by binding to particular soluble or membrane-bound molecules. Once bound
to their activating ligand, they undergo conformational
alterations resulting in changes in their biological activity.
The C1q complex, Mannose Binding Lectin (MBL), and the
ficolins are examples of initiator complement components.
Enzymatic mediators. Several complement components,
(e.g., C1r, C1s, MASP2, and factor B) are proteolytic
enzymes that cleave and activate other members of the
complement cascade. Some of these proteases are activated by binding to other macromolecules and undergoing a conformational change. Others are inactive until
cleaved by another protease enzyme and are thus termed
zymogens: proteins that are activated by proteolytic cleavage. The two enzyme complexes that cleave complement
components C3 and C5, respectively, are called the C3
and C5 convertases and occupy places of central importance in the complement cascades.
Membrane-binding components or opsonins. Upon activation of the complement cascade, several proteins are
cleaved into two fragments, each of which then takes on a
particular role. For C3 and C4, the larger fragments, C3b
and C4b, serve as opsonins, enhancing phagocytosis by
binding to microbial cells and serving as binding tags for
phagocytic cells bearing receptors for C3b or C4b. As a
general rule, the larger fragment of a cleaved complement
component is designated with the suffix “b,” and the
smaller with the suffix “a.” However, there is one exception to this rule: the larger, enzymatically active form of
the C2 component is named C2a.
4.
Inflammatory mediators. Some small complement fragments act as inflammatory mediators. These fragments
enhance the blood supply to the area in which they are
released, by binding to receptors on endothelial cells lining the small blood vessels and inducing an increase in
capillary diameter. They also attract other cells to the site
of tissue damage. Because such effects can be harmful in
excess, these fragments are called anaphylatoxins, meaning substances that cause anaphylaxis (“against protection”). Examples include C3a, C5a, and C4a.
5.
Membrane attack proteins. The proteins of the membrane
attack complex (MAC) insert into the cell membranes of
invading microorganisms and punch holes that result in
lysis of the pathogen. The complement components of the
MAC are C5b, C6, C7, C8, and multiple copies of C9.
6.
Complement receptor proteins. Receptor molecules on cell
surfaces bind complement proteins and signal specific cell
functions. For example, some complement receptors such as
CR1 bind to complement components such as C3b on the
surface of pathogens, triggering phagocytosis of the
C3-bound pathogen. Binding of the complement component C5a to C5aR receptors on neutrophils stimulates
|
CHAPTER 6
189
neutrophil degranulation and inflammation. Complement
receptors are named with “R,” such as CR1, CR2, and C5aR.
7.
Regulatory complement components. Host cells are protected from unintended complement-mediated lysis by
the presence of membrane-bound as well as soluble regulatory proteins. These regulatory proteins include factor I,
which degrades C3b, and Protectin, which inhibits the
formation of the MAC on host cells.
This chapter describes the components of the
complement system, their activation via three major
pathways, the effector functions of the molecules of the
complement cascade, as well as their interactions with other
cellular and molecular components of innate and adaptive
immunity. In addition, it addresses the mechanisms that
regulate the activity of these complement components, the
evasive strategies evolved by pathogens to avoid destruction
by complement, and the evolution of the various
complement proteins. This chapter’s Classic Experiment
tells the story of the scientist who discovered the alternative
pathway of complement. In the Clinical Focus segment, we
address various therapies that target elements of the
complement cascades. Finally, an Advances Box describes
some of the many tactics used by Staphylococcus aureus
bacteria to escape complement-mediated destruction.
The Major Pathways of
Complement Activation
Complement components represent some of the most evolutionarily ancient participants in the immune response. As
viruses, parasites, and bacteria have attacked vertebrate
hosts and learned to evade one aspect of the complement
system, alternative pathways have evolved in an endless
dance of microbial attack and host response. In the complex
pathways that follow, we can gain a glimpse of the current
state of what is actually an ongoing evolutionary struggle on
the part of host organisms to combat microbial infection,
while minimizing damage to their own cells.
The major initiation pathways of the complement cascade
are shown in Figure 6-2. Although the initiating event of each of
the three pathways of complement activation is different, they
all converge in the generation of an enzyme complex capable of
cleaving the C3 molecule into two fragments, C3a and C3b. The
enzymes that accomplish this biochemical transformation are
referred to as C3 convertases. As illustrated in Figure 6-2, the
classical and lectin pathways use the dimer C4b2a for their C3
convertase activity, while the alternative pathway uses C3bBb to
achieve the same end; however, the final result is the same: a
dramatic increase in the concentration of C3b, a centrally
located and multifunctional complement protein.
The second set of convertase enzymes of the cascade, C5
convertases, are formed by the addition of a C3b component
190
PA R T I I
|
Innate Immunity
Classical pathway
Lectin pathway
Alternative pathway
Antigen-antibody
immune complexes
PAMP recognition
by lectins
Spontaneous hydrolysis
or pathogenic surfaces
Initiation
MBL
C1q
C3(H2O)Bb
C3
C3
Antibody
Carbohydrates
C3a
MASPs
C1r2s2
C3b
C3bBb
Factors B, D,
Properdin
C2
C4
Amplification
C4a
C3 convertases
C5 convertases
C3bBb
C3bBbC3b
C4b2a
C4b2a3b
C3a
C3b
C5
C3
Termination
Inflammation
FIGURE 6-2 The generation of C3 and C5 convertases by
the three major pathways of complement activation. The
classical pathway is initiated when C1q binds to antigen-antibody
complexes. The antigen is shown here in dark red and the initiating
antibody in green. The C1r enzymatic component of C1 (shown in
blue) is then activated and cleaves C1s, which in turn cleaves C4 to
C4a (an anaphylatoxin, bright red) and C4b. C4b attaches to the
membrane, and binds C2, which is then cleaved by C1s to form C2a
and C2b. (C2b is then acted upon further to become an inflammatory mediator.) C2a remains attached to C4b, forming the classical
pathway C3 convertase (C4b2a). In the lectin pathway, mannosebinding lectin (MBL, green) binds specifically to conserved carbohy-
to the C3 convertases. C5 convertases cleave C5 into the
inflammatory mediator, C5a, and C5b, which is the initiating
factor of the MAC.
We will now describe each of these pathways in more
detail. The proteins involved in each of these pathways are
listed in Table 6-1.
The Classical Pathway Is Initiated
by Antibody Binding
The classical pathway of complement activation is considered part of the adaptive immune response since it begins with
the formation of antigen-antibody complexes. These complexes may be soluble, or they may be formed when an antibody binds to antigenic determinants, or epitopes, situated on
viral, fungal, parasitic, or bacterial cell membranes. Soluble
Opsonization
C5b C5a
Lysis
Inflammation
drate arrays on pathogens, activating the MBL-associated serine
proteases (MASPs, blue). The MASPs cleave C2 and C4 generating the
C3 convertase as in the classical pathway. In the alternative pathway,
C3 undergoes spontaneous hydrolysis to C3(H2O), which binds
serum factor B. On binding to C3(H2O), B is cleaved by serum factor
D, and the resultant C3(H2O)Bb complex forms a fluid phase C3 convertase. Some C3b, released after C3 cleavage by this complex, binds
to microbial surfaces. There, it binds factor B, which is cleaved by
factor D, forming the cell-bound alternative pathway C3 convertase,
C3bBb. This complex is stabilized by properdin. The C5 convertases
are formed by the addition of a C3b fragment to each of the C3
convertases.
antibody-antigen complexes are often referred to as immune
complexes, and only complexes formed by IgM or certain subclasses of IgG antibodies are capable of activating the classical
complement pathway (see Chapter 13). The initial stage of
activation involves the complement components C1, C2, C3,
and C4, which are present in plasma as zymogens.
The formation of an antigen-antibody complex induces
conformational changes in the nonantigen-binding (Fc)
portion of the antibody molecule. This conformational
change exposes a binding site for the C1 component of
complement. In serum, C1 exists as a macromolecular complex consisting of one molecule of C1q and two molecules
each of the serine proteases, C1r and C1s, held together in a
Ca2⫹-stabilized complex (C1qr2s2) (Figure 6-3a). The C1q
molecule itself is composed of 18 polypeptide chains that
associate to form six collagen-like triple helical arms, the tips
The Complement System
TABLE 6-1
|
CHAPTER 6
191
Proteins of the three major pathways of complement activation
Molecule
Biologically
active
fragments
Biological function
Active in which
pathway
IgM, IgG
Binding to pathogen surface and initiating
complement cascade
Classical pathway
Mannose-binding
lectin, or ficolins
Binding to carbohydrates on microbial surface
and initiating complement cascade
Lectin pathway
Initiation of the classical pathway by binding Ig
Binding to apoptotic blebs and initiating
phagocytosis of apoptotic cells
Serine protease, cleaving C1r and C1s
Serine protease, cleaving C4 and C2
Classical pathway
MASP-1
MBL-Associated Serine Protease 1. MASP-2
appears to be functionally the more relevant MASP-protease.
Lectin pathway
MASP-2
Serine protease. In complex with MBL/ficolin
and MASP-2, cleaves C4 and C2.
Lectin pathway
C2a*
Serine protease. With C1 and C4b, is a C3 convertase.
Classical and lectin
pathways
C2b*
Inactive in complement pathway. Cleavage of C2b by
plasmin releases C2 kinin, a peptide that stimulates vasodilation.
C4b
Binds microbial cell membrane via thioester bond.
With C1 and C2a, is a C3 convertase.
Has weak anaphylatoxin (inflammatory) activity
Proteolytic cleavage products generated by factor I
C1
C1q
(C1r)2
(C1s)2
C2
C4
C4a
C4c, C4d
C3
C3a
C3b
C3(H2O)
iC3b and C3f
C3c and C3d
or C3dg
Factor B
Ba
Bb
Factor D
Anaphylatoxin. Mediates inflammatory signals via C3aR.
Potent in opsonization, tagging immune complexes,
pathogens, and apoptotic cells for phagocytosis
With C4b and C2a, forms the C5 convertase
With Bb, forms the C3 convertase
With Bb and one more molecule of C3b
(C3bBb3b) acts as a C5 convertase
C3 molecule in which the internal thioester bond
has undergone hydrolysis
With Bb, acts as a fluid phase C3 convertase.
Proteolytic fragments of C3b, generated by factor I.
iC3b binds receptors CR3, CR4, and CRIg; CR2 binds weakly.
Proteolytic fragments of iC3b generated by factor I. C3d/dg
bind antigen and to CR2, facilitating antigen-binding to B cells.
C3c binds CRIg on fixed tissue macrophages
Binds C3(H2O) and is then cleaved by factor D into
two fragments: Ba and Bb
Smaller fragment of factor D-mediated cleavage of factor B.
May inhibit proliferation of activated B cells.
Larger fragment of factor D-mediated cleavage of factor B
With C3(H2O), acts as fluid phase C3 convertase
With C3b, acts as cell-bound C3 convertase
With two molecules of C3b, acts as C5 convertase
Proteolytic enzyme that cleaves factor B into Ba and Bb
only when it is bound to either C3(H2O) or to C3b
Classical and lectin
pathways
Classical pathway
Tickover and properdin
alternative pathways
Tickover and properdin
alternative pathways
Alternative pathway
Alternative pathway
Alternative pathway
Alternative pathway
Alternative pathway
Alternative pathway
(continued)
192
TABLE 6-1
PA R T I I
|
Innate Immunity
(continued)
Factor P (properdin)
C5
Stabilizes the C3bBb complex on microbial cell surface
C5a
C5b
Alternative pathway
Anaphylatoxin binding to C5aR induces inflammation
Component of Membrane Attack Complex (MAC).
Binds cell membrane and facilitates binding of other
components of the MAC.
All
C6
Component of MAC. Stabilizes C5b. In the absence of C6,
C5b is rapidly degraded.
All
C7
Component of MAC. Binds C5bC6 and induces
conformational change allowing C7 to insert into interior
of membrane.
All
C8
Component of MAC. Binds C5bC6C7 and creates a small
pore in membrane.
All
C9
Component of MAC. 10–19 molecules of C9 bind
C5bC6C7C8 and create large pore in membrane.
All
* Because C2a is the larger, active fragment of C2, some writers have tried to alter the nomenclature in order to make C2 conform to the convention that the “b” fragment is the larger, active fragment and the “a” fragment is smaller and may be an anaphylatoxin. However, this effort does not appear to be making headway. Note
that the smaller fragment of C2, which we name C2b, is inactive in the complement pathway.
(a)
(b)
C1q
2
C1
complex
3
C1r2s2
s
r
IgG
antibody
(c)
Collagen-like
triple helix
Binds CH2 domain
of antibody
FIGURE 6-3 Structure of the C1 macromolecular complex. (a) C1q interacts with two
molecules each of C1r and C1s to create the C1 complex. (b) The C1q molecule consists of 18
polypeptide chains in six collagen-like triple helices, each of which contains one A, B, and C chain,
shown in three different shades of green. The head group of each triplet contains elements of all
three chains. (c) Electron micrograph of C1q molecule showing stalk and six globular heads. The
dark circle is a gold-labeled molecule of fibromodulin, a component of cartilage that binds the
head groups of C1q. [Sjoberg, A., et al. September 16, 2005 The J. Biological Chem, Vol. 280, Issue 37, 3230132308. ?2005 by the Merican Society for Biochemistry and Molecular Biology.]
The Complement System
|
CHAPTER 6
193
(a)
(b)
FIGURE 6-4 Models of pentameric IgM in planar form (a) and “staple” form (b). Several C1q-binding sites in the Fc region are
accessible in the staple form, whereas none are exposed in the planar form. [From A. Feinstein et al., 1981, Monographs in Allergy 17:28, and 1981, Annals
of the New York Academy of Sciences 190:1104.]
of which bind the CH2 domain of the antigen-bound antibody molecule (Figure 6-3b, c).
Each C1 macromolecular complex must bind by its C1q
globular heads to at least two Fc sites for a stable C1-antibody
interaction to occur. Recall from Chapter 3 that serum IgM
exists as a pentamer of the basic four-chain immunoglobulin structure. Circulating, nonantigen-bound IgM adopts
a planar configuration (Figure 6-4a), in which the C1qbinding sites are not exposed. However, when pentameric
IgM is bound to a multivalent antigen, it undergoes a conformational change, assuming the so-called “staple” configuration (Figure 6-4b), in which at least three binding sites for
C1q are exposed. Thus, an IgM molecule engaged in an
antibody-antigen complex can bind C1q, whereas circulating, nonantigen-bound IgM cannot.
In contrast to pentameric IgM, monomeric IgG contains
only one C1q binding site per molecule, and the conformational changes IgG undergoes on antigen binding are much
more subtle than those experienced by IgM. There is therefore
a striking difference in the efficiency with which IgM and IgG
are able to activate complement. Less than 10 molecules of
IgM bound to a red blood cell can activate the classical complement pathway and induce lysis, whereas some 1000 molecules of cell-bound IgG are required to ensure the same result.
The intermediates in the classical activation pathway are
depicted schematically in Figure 6-5. Proteins of the classical
pathway are numbered in the order in which the proteins
were discovered, which does not quite correspond with the
order in which the proteins act in the pathway (a disconnect
that has troubled generations of immunology students).
Note that binding of one component to the next always
induces either a conformational change or an enzymatic
cleavage, which allows for the next reaction in the sequence
to occur.
Binding of C1q to the CH2 domains of the Fc regions of
the antigen-complexed antibody molecule induces a conformational change in one of the C1r molecules that converts it
to an active serine protease enzyme. This C1r molecule then
cleaves and activates its partner C1r molecule. The two C1r
proteases then cleave and activate the two C1s molecules
(see Figure 6-5, part 1).
C1s has two substrates, C4 and C2. C4 is activated when
C1s hydrolyzes a small fragment (C4a) from the amino terminus of one of its chains (see Figure 6-5, part 2). The C4b
fragment attaches covalently to the target membrane surface
in the vicinity of C1, and then binds C2. C4b binding to the
membrane occurs when an unstable, internal thioester on
C4b, exposed upon C4 cleavage, reacts with hydroxyl or
amino groups of proteins or carbohydrates on the cell membrane. This reaction must occur quickly, otherwise the
thioester C4b is further hydrolyzed and can no longer make a
covalent bond with the cell surface (Figure 6-6). Approximately
90% of C4b is hydrolyzed before it can bind the cell surface.
On binding C4b, C2 becomes susceptible to cleavage by
the neighboring C1s enzyme, and the smaller C2b fragment
diffuses away, leaving behind an enzymatically active C4b2a
complex. In this complex, C2a is the enzymatically active
fragment, but it is only active when bound by C4b. This
C4b2a complex is called C3 convertase, referring to its role
in converting C3 into an active form. The smaller fragment
generated by C4 cleavage, C4a, is an anaphylatoxin, and its
function is described below.
The membrane-bound C3 convertase enzyme, C4b2a,
now hydrolyzes C3, generating two unequal fragments; the
small anaphylatoxin C3a and the pivotal fragment C3b. A
single C3 convertase molecule can generate over 200 molecules of C3b, resulting in tremendous amplification at this
step of the classical pathway.
PA R T I I
194
|
Innate Immunity
6-5
OVERVIEW FIGURE
Intermediates in the Classical Pathway of Complement Activation
up to the Formation of the C5 Convertase
1
C1q binds antigen-bound antibody, and induces a conformational
change in one C1r molecule, activating it. This C1r then activates
the second C1r and the two C1s molecules.
3
C3 convertase hydrolyzes many C3 molecules. Some combine
with C3 convertase to form C5 convertase.
C1qr2s2
C1q
Antibody
C1r2s2
Antibody binding sites
on antigen (epitopes)
FC
+
C3b
C3
C4b2a3b
C5 convertase
C4b2a
2
C1s cleaves C4 and C2. C4 is cleaved first and C4b binds to the
membrane close to C1. C4b binds C2 and exposes it to the action
of C1s. C1s cleaves C2, creating the C3 convertase, C4b2a.
4
C3a
The C3b component of C5 convertase binds C5, permitting
C4b2a to cleave C5.
C4
C4a
C5
C5 convertase
C5a
C5b
C4b
C4bC2
C4b2a
C3 convertase
C2b
Antigenic determinants are shown in dark red, initiating components (antibodies and C1q) are shown in green, active enzymes are shown
in blue, and anaphylatoxins in bright red.
The generation of C3b is centrally important to many of
the actions of complement. Deficiencies of complement
components that act prior to C3 cleavage leave the host
extremely vulnerable to both infectious and autoimmune
diseases, whereas deficiencies of components later in the
pathway are generally of lesser consequence. This is because
C3b acts in three important and different ways to protect the
host.
First, in a manner very similar to that of C4b, C3b binds
covalently to microbial surfaces, providing a molecular “tag”
and thereby allowing phagocytic cells with C3b receptors to
engulf the tagged microbes. This process is called opsonization. Second, C3b molecules can attach to the Fc portions of
antibodies participating in soluble antigen-antibody complexes. These C3b-tagged immune complexes are bound by
C3b receptors on phagocytes or red blood cells, and are
|
The Complement System
S
S
S
S
S
S
SH
S
C
OH
NH2
NH
Cell membrane
O
S
S
S
S
SH
S
C
S
S
S
NH2
O
195
(c) Hydrolyzed C4b
S
S
S
S
S
S
C
S
S
O
Thioester bond
S
(b) Bound C4b
S
(a) Unbound C4b
CHAPTER 6
Cell membrane
Cell membrane
FIGURE 6-6 Binding of C4b to the microbial membrane
surface occurs through a thioester bond via an exposed
amino or hydroxyl group. (a) Both C3b and C4b contain highly
reactive thioesters, which are subject to nucleophilic attack by
hydroxyl or amino groups on cell membrane proteins and carbohy-
drates. (b) Breakage of the thioester leads to the formation of covalent bonds between the membrane macromolecules and the
complement components. (c) If this covalent bond formation does
not occur quickly after generation of the C3b and C4b fragments, the
thioester will be hydrolyzed to a nonreactive form.
either phagocytosed, or conveyed to the liver where they are
destroyed.
Finally, some molecules of C3b bind the membranelocalized C4b2a enzyme to form the trimolecular, membranebound, C5 convertase complex C4b2a3b. The C3b component
of this complex binds C5, and the complex then cleaves C5
into the two fragments: C5b and C5a (see Figure 6-5 part 4).
C4b2a3b is therefore the C5 convertase of the classical pathway. This trio of tasks accomplished by the C3b molecule
places it right at the center of complement attack pathways.
As we will see, C5b goes on to form the MAC with C6, C7,
C8, and C9.
dimensional array. Thus, one can think of MBL as a classic
pattern recognition receptor (see Chapter 5). Consistent
with MBL’s place at the beginning of an important immune
cascade, many individuals with low levels of MBL suffer
from repeated, serious bacterial infections.
MBL is constitutively expressed by the liver and, like C1q,
which it structurally resembles, MBL belongs to the subclass
of lectins known as collectins (see Chapter 5). More recently,
The Lectin Pathway Is Initiated When Soluble
Proteins Recognize Microbial Antigens
The lectin pathway, like the classical pathway, proceeds
through the activation of a C3 convertase composed of C4b
and C2a. However, instead of relying on antibodies to recognize the microbial threat and to initiate the complement
activation process, this pathway uses lectins—proteins that
recognize specific carbohydrate components primarily
found on microbial surfaces—as its specific receptor molecules (Figure 6-7; see also Figure 5-7a). Mannose-binding
lectin (MBL), the first lectin demonstrated to be capable of
initiating complement activation, binds close-knit arrays of
mannose residues that are found on microbial surfaces such
as those of Salmonella, Listeria, and Neisseria strains of bacteria; Cryptococcus neoformans and Candida albicans strains
of fungi; and even the membranes of some viruses such as
HIV-1 and respiratory syncytial virus. The complement
pathway that it initiates is referred to as the lectin pathway
of complement activation.
Further characterization of the sugars recognized by MBL
demonstrated that MBL also recognizes structures in addition to mannose, including N-acetyl glucosamine, D-glucose,
and L-fucose. All those sugars, including mannose, present
their associated hydroxyl groups in a particular three-
MBL
MASP-1
or MASP-2
Polysaccharide
antigen
FIGURE 6-7 Initiation of the lectin pathway relies on lectin receptor recognition of microbial cell surface carbohydrates. Lectin receptors, such as MBL, bind microbial cell surface
carbohydrates. Once attached to the carbohydrates, they bind the
MASP family serine proteases, which cleave C2 and C4 with the formation of a lectin-pathway C3 convertase.
196
PA R T I I
|
Innate Immunity
the ficolins (see Chapter 5), another family of proteins structurally related to the collectins, have been recognized as
additional initiators of the lectin pathway of complement
activation. L-ficolin, H-ficolin, and M-ficolin each bind specific types of carbohydrates on microbial surfaces. We will
use MBL as our example in subsequent paragraphs.
MBL is associated in the serum with MBL-Associated Serine Proteases, or MASP proteins. Three MASP proteins—
MASP1, MASP2, and MASP3—have been identified, but most
studies of MASP function point to the MASP2 protein as being
the most important actor in the next step of the MBL pathway.
MASP-2 is structurally related to the serine protease C1s,
and can cleave both C2 and C4 (see Figure 6-2), giving rise to
the C3 convertase, C4b2a, that we first encountered in our
discussion of the classical pathway. Thus, the lectin pathway
utilizes all the same components as the classical pathway with
the single exception of the C1 complex. The soluble lectin
receptor replaces the antibody as the antigen-recognizing component, and MASP proteins take the place of C1r and C1s in
cleaving and activating the C3 convertase. Once the C3 convertase is formed, the reactions of the lectin pathway are the same
as for the classical pathway; the C5 convertase of the lectin
pathway, like that of the classical pathway, is also C4b2a3b.
binds another serum protein, factor B (Figure 6-8a). When
bound to C3(H2O), factor B becomes susceptible to cleavage
by a serum protease, factor D. Factor D cleaves B, releasing a
smaller Ba subunit, which diffuses away and leaves a catalytically active Bb subunit that remains bound to C3(H2O).
This C3(H2O)Bb complex, which at this point is still in
the fluid phase (i.e., in the plasma, not bound to any cells),
has C3 convertase activity. It rapidly cleaves many molecules
of C3 into C3a and C3b (Figure 6-8b). This initiating C3
convertase is constantly being formed in plasma, and breaking down a few C3 molecules, but it is then just as rapidly
degraded. However, if there is an infection present, some of
(b)
(a)
C3
C3
Initiation of the alternative pathway of complement activation is independent of antibody-antigen interactions, and so
this pathway, like the lectin pathway, is also considered to be
part of the innate immune system. However, unlike the lectin
pathway, the alternative pathway uses its own set of C3 and C5
convertases (see Figure 6-2). As we will see, the alternative pathway C3 convertase is made up of one molecule of C3b and one
molecule unique to the alternative pathway, Bb. A second C3b
is then added to make the alternative pathway C5 convertase.
Recent investigations have revealed that the alternative pathway can itself be initiated in three distinct ways. The first mode
of initiation to be discovered, the “tickover” pathway, utilizes the
four serum components C3, factor B, factor D, and properdin
(see Table 6-1). The term tickover refers to the fact that C3 is
constantly being made and spontaneously inactivated and is
thus undergoing “tickover.” Two additional modes of activation
for the alternative pathway have also been identified: one is
initiated by the protein properdin, and the other by proteases
such as thrombin and kallikrein. The story of the discovery of
properdin is addressed in Classic Experiment Box 6-1.
The Alternative Tickover Pathway
The alternative tickover pathway is initiated when C3, which
is at high concentrations in serum, undergoes spontaneous
hydrolysis at its internal thioester bond, yielding the molecule
C3(H2O). The conformation of C3(H2O) has been demonstrated by spectrophotometric means to be different from that
of the parent protein, C3. C3(H2O) accounts for approximately 0.5% of plasma C3 and, in the presence of serum Mg2⫹
C3a
C3b
C3(H2O)
Factor B
Factor B
C3bB
C3(H2O)B
Factor D
Factor D
Ba
The Alternative Pathway Is Initiated
in Three Distinct Ways
Fluid phase
C3 convertase
Ba
C3bBb
Membrane-bound
C3 convertase
C3(H2O)Bb
Fluid phase
C3 convertase
(c)
C3 convertase
Properdin
(factor P)
+
C3
C3b
C3a
C3bBbC3b
C5 convertase
C5
C5b C5a
FIGURE 6-8 The initiation of the alternative tickover
pathway of complement. (a) Spontaneous hydrolysis of soluble
C3 to C3(H2O) allows the altered conformation of C3(H2O) to bind
factor B, rendering it susceptible to cleavage by factor D. The resulting complex C3(H2O)Bb forms a fluid phase convertase capable of
cleaving C3 to C3a and C3b. (b) Some of the C3b molecules formed
by the fluid phase convertase bind to cell membranes. C3b, like
C3(H2O) binds factor B in such a way as to make B susceptible to factor D-mediated cleavage. (c) However, membrane-bound C3bBb is
unstable in the absence of properdin (factor P), which binds the
C3bBb complex on the membrane and stabilizes it. Addition of a
second C3b molecule to the C3bBb complex forms the C5 convertase, which is also stabilized by factor P.
The Complement System
the newly formed C3b molecules bind nearby microbial
surfaces via their thioester linkages.
Since factor B is capable of binding to C3b as well as to
C3(H2O), factor B now binds the newly attached C3b molecules on the microbial cell surface (see Figure 6-8b), and
again becomes susceptible to cleavage by factor D, with the
generation of C3bBb complexes. These C3bBb complexes are
now located on the microbial membrane surface. Like the
C4b2a complexes of the classical pathway, the cell-bound
C3bBb complexes have C3 convertase activity, and this complex now takes over from the fluid phase C3(H2O)Bb as the
predominant C3 convertase.
To be clear, there are two C3 convertases in the alternative
tickover pathway: a fluid phase C3(H2O)Bb, which kicks off
the pathway, and a membrane-bound C3bBb C3 convertase.
The cell-bound C3bBb C3 convertase is unstable until it is
bound by properdin, a protein from the serum (Figure 6-8c).
Once stabilized by properdin, these cell-associated, C3bBb C3
convertase complexes rapidly generate large amounts of C3b
at the microbial surface; these, in turn, bind more factor B,
facilitating its cleavage and activation and resulting in a dramatic amplification of the rate of C3b generation. This amplification pathway is rapid and, once the alternative pathway has
been initiated, more than 2 106 molecules of C3b can be
deposited on a microbial surface in less than 5 minutes.
Just as the C5 convertase of the classical and lectin
pathways was formed by the addition of C3b to the C4b2a
C3 convertase complex, so the C5 convertase of the alternative pathway is formed by the addition of C3b to the
alternative pathway C3 convertase complex. The C5 convertase complex therefore has the composition C3bBbC3b,
and, like the C3 convertase, is also stabilized by binding to
factor P. Like the classical and lectin pathway C5 convertase, C3bBbC3b cleaves C5, which goes on to form the
MAC (see Figure 6-8c).
The Alternative Properdin-Activated Pathway
In the previous section, we introduced properdin as a regulatory factor that stabilizes the C3bBb, membrane-bound C3
convertase. However, recent data suggest that, in addition to
stabilizing the ongoing activity of the alternative pathway,
properdin may also serve to initiate it.
In vitro experiments demonstrated that if properdin molecules were attached to an artificial surface and allowed to
interact with purified complement components in the presence of Mg2, the immobilized properdin bound C3b and
factor B (Figure 6-9). This bound factor B proved to be susceptible to cleavage by factor D, and the resultant C3bPBb
complex acted as an effective C3 convertase, leading to the
amplification process discussed above. Thus, it seemed that
properdin could initiate activation of the alternative pathway
on an artificial substrate.
However, proving that a set of reactions can occur in vitro
does not necessarily mean that it actually does occur in vivo.
Investigators next proved properdin’s ability to bind specifically to certain microbes, including Neisseria gonorrhoeae, as
Properdin
C3b
|
CHAPTER 6
Factor B
197
Factor D
C3bPB
C3bPBb
Ba
FIGURE 6-9 Initiation of the alternative pathway by specific, noncovalent binding of properdin to the target membrane. Properdin binds to components of microbial membranes,
and stabilizes the binding of C3bBb complexes of the alternative
complement pathway. The difference between this and the tickover
pathways is that properdin binds first and initiates complement
deposition on the membrane.
well as to apoptotic and necrotic cell surfaces. Once bound,
properdin was indeed able to initiate the alternative pathway,
as indicated above.
Support for the physiological relevance of this properdininitiated pathway is provided by the observation that patients
with a deficiency in properdin production are uniquely susceptible to meningococcal disease induced by the Neisseria
gonorrhoeae bacterium. These findings suggest that properdin
has the capacity to act as a pattern recognition receptor (PRR),
specifically directing the activation of the alternative pathway
onto the surface of Neisseria and other microbial cells.
Note that this pathway relies on the preexistence of low
levels of C3b, which must be generated by mechanisms such
as the tickover pathway. However, the specific binding of properdin to Neisseria membranes demonstrates how the properdin pathway can provide greater selectivity than that available
from the nonspecific C3b binding of the tickover pathway.
The Alternative Protease-Activated Pathway
The complement and blood-coagulation pathways both use
protease cleavage and conformational alterations to modify
enzyme activities, as well as amplification of various steps of the
pathways by feed-forward loops. Recently, some elegant work
has revealed the existence of functional interactions as well as
theoretical parallels between these two proteolytic cascades.
Several decades ago, it was shown that protein factors
involved in blood clotting, such as thrombin, could cleave the
complement components C3 and C5, in vitro, with the
release of the active anaphylatoxins C3a and C5a. Since these
cleavage reactions required relatively high thrombin concentrations, they were at first thought not to be physiologically
meaningful. More recently, however, it has been demonstrated in a mouse disease model that initiation of the coagulation cascade may result in the cleavage of physiologically
relevant amounts of C3 and C5 to produce C3a and C5a.
Specifically, in an immune-complex model of acute lung
inflammation, thrombin was shown to be capable of cleaving
C5, with the release of active C5a. This thrombin-mediated
C5 cleavage was also demonstrated in C3 knockout mice, in
which the canonical C5 convertases could not possibly have
been generated. Thus, strong inflammatory reactions can
result in the activation of at least a part of the alternative
198
PA R T I I
|
Innate Immunity
CLASSIC EXPERIMENT
The Discovery of Properdin
The study of the history of science
shows us that scientists, like any other
professionals, are often tempted to think
about problems only in ways that are
already well trodden and familiar. Science,
like art and fashion, has its fads and its “in
crowds.” Indeed, sometimes those whose
work moves in directions too different
from that of the mainstream in their field
have difficulty gaining credibility until the
thinking of the rest of their colleagues
catches up with their own. Such was the
case for Louis Pillemer, the discoverer of
properdin; by the time others in the burgeoning field of immunochemistry appreciated the power of his discovery, it was
quite simply too late.
Louis Pillemer (Figure 1) was born in
1908 in Johannesburg, South Africa. In
1909 the family emigrated to the United
States and settled in Kentucky. Pillemer
completed his bachelor’s degree at Duke
University and started medical school at
the same institution. However, in the middle of his third year, the emotional problems that would plague him for the rest of
his life surfaced for the first time, and he
left. At that time, deep in the Depression,
individuals in Kentucky who could pass an
exam in the rudiments of medical care
were encouraged to care for patients not
otherwise served by a physician. Pillemer
dutifully passed this examination and
began to travel through Kentucky on
horseback, visiting the sick and tendering
whatever treatments were then available.
In 1935, he quit this wandering life and
entered graduate school at what was
then Western Reserve University (now
Case Western Reserve University).
There, Pillemer earned his Ph.D. and,
except for some time at Harvard and a
tour of duty at the Army Medical School in
Washington D.C., he remained at Case
Western Reserve for the rest of his life,
developing a reputation as an excellent
biochemist. Among his more noteworthy
accomplishments were the first purifications of both tetanus and diphtheria toxins, which were used, along with killed
pertussis organisms, in the development
of the standard DPT vaccine.
FIGURE 1
Louis Pillemer, the discoverer of properdin. [Presidential Address to the AAI, 1980, by Lepow; reprinted
in J. Immunol. 125, 471, 1980. Fig.1. 1980. The American Association of Immunologists, Inc.]
After these successes, Pillemer turned
his attention to the biochemistry of the
complement system, which he had initially encountered during his graduate
work. The antibody-mediated classical
pathway had already been established.
However, Pillemer was intrigued by some
more recent experiments that had shown
that mixing human serum with zymosan,
an insoluble carbohydrate extract from
yeast cell walls, at 37C resulted in the
selective loss of the vital third component
of complement, C3. He was curious about
the mechanism of this loss of C3, and initially interpreted the result to suggest that
the C3 was being selectively adsorbed
onto the zymosan surface. He reasoned
that, if this were true, adsorption to zymosan might be used as a method to purify
C3 from plasma. However, that first idea
was proven incorrect, and instead he
began to investigate whether the loss of
C3 resulted from C3 cleavage that was
occurring at the zymosan surface.
Pillemer succeeded in demonstrating
that such was indeed the case and went
on to show that C3 cleavage only happened when his experiments were run at
pH 7.0 and at 37C. This suggested to him
that perhaps an enzyme in the serum
was binding to the zymosan and causing
the inactivation of the C3 component.
Consistent with this hypothesis, when he
mixed the serum and the zymosan at
17C, no cleavage occurred. However, if
he allowed the serum and zymosan to
mix at 17C and then warmed up the
mixture to 37C, the C3 was cleaved as
effectively as if it had been incubated at
37C all along.
Next, he incubated the serum and
zymosan together at 17C, spun down
and removed the zymosan from the mixture, and then added fresh zymosan to
the remaining serum containing the C3.
He then raised the temperature to 37C.
Nothing happened. The C3 was untouched.
Whatever enzymic activity in the serum
was responsible for the breakdown of C3
The Complement System
|
CHAPTER 6
199
BOX 6-1
had been adsorbed by the zymosan and
removed from the serum.
Pillemer concluded that a factor present in serum and adsorbed onto the
zymosan was necessary for the cleavage
of C3. With his students and collaborators,
he purified this component and named it
properdin, from the Latin perdere meaning “to destroy.” His flow sheet for these
initial experiments, and published as part
of his landmark Science paper of 1954, is
shown in Figure 2. He also identified a
heat-labile factor in the serum that was
required for C3 cleavage to occur.
Pillemer and colleagues went on to
characterize properdin as a protein that
represented less than 0.03% of serum
proteins and whose activity was absolutely dependent on the presence of
magnesium ions. In a brilliant series of
experiments, Pillemer demonstrated the
importance of his newly discovered factor
in complement-related antibacterial and
antiviral reactions, as well as its role in the
disease known as paroxysmal nocturnal
hemoglobinemia.
In light of today’s knowledge, we can
interpret exactly what was happening in
his experiments. Properdin bound to the
zymosan and stabilized the C3 convertase
of the alternative pathway, resulting in C3
cleavage. Indeed, experiments performed
as recently as 2007 show that properdin
binds to zymosan in a manner similar to
its binding to Neisseria membranes.
Pillemer’s discovery of properdin coupled with his purification of the tetanus
and diphtheria toxins should have sealed
his reputation as a world-class biochemist.
Indeed, his findings were deemed of sufficient general interest at the time that
they were publicized in the lay press as
well as in the scientific media; articles and
editorials about them appeared in the
New York Times, Time magazine, and Collier’s.
Nor did Pillemer oversell his case. Scientists writing about this period are at pains
to note that Pillemer did not make or
broadcast any claims for his molecule
beyond what appeared in the reviewed
scientific literature.
However, in 1957 and 1958, scientist
Robert Nelson offered an alternative
explanation for Pillemer’s findings. He
pointed out that what Pillemer had
described as a new protein could simply
be a mixture of natural antibodies specific for zymosan. If that were the case,
then all Pillemer had succeeded in doing
was describing the classical pathway a
FIGURE 2
The flow sheet of Pillemer’s experiments, which showed that a substance in
serum that is adsorbed onto zymosan, the yeast cell wall extract, is capable of
catalyzing the cleavage of C3. [Pillemer, L., et al. The properdin system and immunity. I. Demonstration and
isolation of a new serum protein, properdin, and its role in immune phenomena. Science. 1954 Aug 20;120(3112):279-85.
Reprinted with permission from AAAS.]
second time. Sensitive biochemical
experiments indeed demonstrated the
presence of low levels of antizymosan
antibodies in properdin preparations, and
the immunological community began to
doubt the relevance of properdin to the
complement activation that Pillemer
described.
Pillemer was devastated. Never completely stable emotionally, Pillemer’s
“behavior became erratic, he occasionally
abused alcohol and he appeared to be
experimenting with drugs,” according to
his former graduate student, Irwin H.
Lepow, who went on to become a distinguished immunologist in his own right.
On August 31, 1957, right in the midst of
the controversy over properdin, Pillemer
died of acute barbiturate intoxication. His
death was ruled a suicide, although
nobody can know whether he was merely
seeking short-term relief from stress and
his death was therefore accidental, or
whether he truly meant to bring an end to
his life.
Subsequent experiments demonstrated that antibodies to zymosan could
be removed from partially purified properdin without loss of the ability of the
preparation to catalyze the cleavage of
C3, thus confirming Pillemer’s finding.
Furthermore, the heat-labile factor identified by Pillemer’s experiments was identified as a previously unknown molecule,
which was subsequently named factor B.
By the late 1960s other laboratories
entered the arena, and slowly Pillemer’s
discovery was confirmed and extended
into what we now know as the alternative
pathway of complement activation. It is
one of the great tragedies of immunology
that Pillemer did not live to enjoy the validation of his elegant work.
Lepow, I. H. 1980. Louis Pillemer, properdin and
scientific controversy. Presidential address
to American Association of Immunologists
in Anaheim, CA, April 16, 1980. Journal of
Immunology 125:471–478.
Pillemer, L., et al. 1954. The properdin system
and immunity. I. Demonstration and isolation of a new serum protein, properdin, and
its role in immune phenomena. Science
120:279–285.
PA R T I I
200
|
Innate Immunity
complement pathway, via the enzymes of the coagulation
cascade. Given the proinflammatory roles of the anaphylatoxins, this would result in further amplification of the
inflammatory state. Additional experiments have since
revealed that other coagulation pathway enzymes, such as
plasmin, are capable of generating both C3a and C5a.
Furthermore, when blood platelets are activated during a
clotting reaction, they release high concentrations of ATP
and Ca2⫹ along with serine/threonine kinases. These
enzymes act to phosphorylate extracellular proteins, including C3b. Phosphorylated C3b is less susceptible to proteolytic degradation than its unphosphorylated form, and thus,
by this route, activation of the clotting cascade enhances all
of the complement pathways.
The Three Complement Pathways Converge at
the Formation of the C5 Convertase
The end result of the three initiation pathways is the formation
of a C5 convertase. For the classical and lectin pathways, the C5
convertase has the composition C4b2a3b; for the alternative
pathway, the C5 convertase has the formulation C3bBbC3b. (In
the thrombin-initiated pathway, the anaphylatoxin C5a is
formed by cleavage of C5 by the blood clotting enzymes, but
functionally meaningful C5b concentrations are not generated
by this route.) However, the end result of all types of C5 convertase activity is the same: the cleavage of the C5 molecule into two
fragments, C5a and C5b. The large C5b fragment is generated
on the surface of the target cell or immune complex and provides a binding site for the subsequent components of the
MAC. However, the C5b component is extremely labile, is not
covalently bound to the membrane like C3b and C4b, and is
rapidly inactivated unless it is stabilized by the binding of C6.
C5 Initiates the Generation of the MAC
Up to this point in the complement cascades, all of the complement reactions take place on the hydrophilic surfaces of
microbes or on immune complexes in the fluid phase of blood,
lymph, or tissues. In contrast, when C5b binds C6 and C7, the
resulting complex undergoes a conformational change that
exposes hydrophobic regions on the surface of the C7 component capable of inserting into the interior of the microbial
membrane (Figure 6-10a). If, however, the reaction occurs on
(a)
C7
C9
C9
C8
C9
C5b
C6
C5b
C5b
C6
C7
C5b
C6
C7 C8
C5b
C6
C7 C8
C9
C6
C7
C8
C9
C9 C9
C9 C9
9
Cell membrane
(b)
FIGURE 6-10 Formation of the membrane attack
complex (MAC). (a) The formation of the MAC showing the
addition of C6, C7, C8, and C9 components to the C5b component. (b) Photomicrograph of poly-C9 complex formed by in
vitro polymerization of C9 (inset) and complement-induced
lesions on the membrane of a red blood cell. These lesions result
from formation of membrane-attack complexes. [Part b from E. R.
Podack, 1986, Immunobiology of the Complement System, G. Ross, ed.,
Academic Press, Orlando, FL; part b inset from J. Humphrey and R. Dourmashkin, 1969, The lesions in cell-membranes caused by complement,
Advances in Immunology, 11:75.]
The Complement System
an immune complex or other noncellular activating surface,
then the hydrophobic binding sites cannot anchor the complex
and it is released. Released C5b67 complexes can insert into the
membrane of nearby cells and mediate “innocent bystander”
lysis. Under physiologic conditions, such lysis is usually minimized by regulatory proteins (see below).
C8 is made up of two peptide chains: C8 and C8. Binding of C8 to the C5b67 complex induces a conformational
change in the C8 dimer such that the hydrophobic domain of
C8 can insert into the interior of the phospholipid membrane. The C5b678 complex can create a small pore, 10 Å in
diameter, and formation of this pore can lead to lysis of red
blood cells, but not of nucleated cells. The final step in the
formation of the MAC is the binding and polymerization of
C9 to the C5b678 complex. As many as 10 to 19 molecules of
C9 can be bound and polymerized by a single C5b678 complex. During polymerization, the C9 molecules undergo a
transition, so that they, too, can insert into the membrane. The
completed MAC, which has a tubular form and functional
pore diameter of 70 Å to 100 Å, consists of a C5b678 complex
surrounded by a poly-C9 complex (Figure 6-10b). Loss of
plasma membrane integrity leads irreversibly to cell death.
The Diverse Functions
of Complement
Table 6-2 lists the main categories of complement function,
each of which is discussed below. In addition to its longunderstood role in antibody-induced lysis of microbes,
TABLE 6-2
|
CHAPTER 6
201
complement also has important functions in innate immunity, many of which are mediated by the soluble innate
immune receptors such as MBL and the ficolins. The pivotal importance of C3b-mediated responses such as opsonization has been clearly demonstrated in C3/ knockout
animals, which display increased susceptibility to both
viral and bacterial infections. In addition, recent experiments have explored the roles of various complement components at the interface of innate and adaptive immunity,
and identified multiple mechanisms by which the release of
active complement fragments acts to regulate the adaptive
immune system. Complement also plays an important role
in the contraction phase of the adaptive immune response,
and recent work has even suggested that it is important in
the elimination of excess synapses during the development
of the nervous system. These various functions are
described below.
Complement Receptors Connect ComplementTagged Pathogens to Effector Cells
Many of the biological activities of the complement system
depend on the binding of complement fragments to cell
surface complement receptors. In addition, some complement receptors play an important role in regulating complement activity by mediating proteolysis of biologically active
complement components. The levels of a number of the
complement receptors are subject to regulation by aspects of
the innate and adaptive immune systems. For example, activation of phagocytic cells by various agents, including the
The three main classes of complement activity in the service of host defense
Activity
Responsible complement component
Innate defense against infection
Lysis of bacterial and cell membranes
Opsonization
Induction of inflammation and chemotaxis by anaphylatoxins
Membrane attack complex (C5b–C9)
Covalently bound C3b, C4b
C3a, C4a, C5a (anaphylatoxins) and their receptors on leukocytes
Interface between innate and adaptive immunity
Augmentation of antibody responses
Enhancement of immunologic memory
Enhancement of antigen presentation
Potential effects on T cells
C3b and C4b and their proteolyzed fragments bound to immune
complexes and antigen; C3 receptors on immune cells
C3b and C4b and their fragments bound to antigen and immune
complexes; receptors for complement components on follicular
dendritic cells
MBL, C1q, C3b, C4b, and C5a
C3, C3a, C3b, C5a
Complement in the contraction phase of the immune response
Clearance of immune complexes from tissues
Clearance of apoptotic cells
Induction of regulatory T cells
C1, C2, C4; covalently bound fragments of C3 and C4
C1q; covalently bound fragments of C3 and C4. Loss of CD46
triggers immune clearance.
CD46
This table has been considerably modified from the superb formulation of Mark Walport, Complement, New England Journal of Medicine 344:1058–1066, Table 1.
202
TABLE 6-3
PA R T I I
|
Innate Immunity
Receptors that bind complement components and their breakdown products
Receptor
Alternative
name(s)
Ligand
Cell surface binding or
expression
Function
CR1
CD35
C3b, C4b,
C1q,
iC3b
Erythrocytes, neutrophils, monocytes,
macrophages, eosinophils, FDCs,
B cells, and some T cells
Clearance of immune complexes,
enhancement of phagocytosis, regulation
of C3 breakdown
CR2
CD21,
Epstein-Barr
virus receptor
C3d, C3dg
(human), C3d
(mouse) iC3b
B cells and FDCs
Enhancement of B-cell activation, B-cell
coreceptor, and retention of C3d-tagged
immune complexes
CR3
CD11b/CD18,
Mac-1
iC3b and
factor H
Monocytes, macrophages, neutrophils,
NK cells, eosinophils, FDCs, T cells
Binding to adhesion molecules on
leukocytes, facilitates extravasation; iC3b
binding enhances opsonization of
immune complexes
CR4
CD11c/CD18
iC3b
Monocytes, macrophages, neutrophils,
dendritic cells, NK cells, T cells
iC3b-mediated phagocytosis
CRIg
VSIG4
C3b, iC3b,
and C3c
Fixed-tissue macrophages
iC3b-mediated phagocytosis and inhibition
of alternative pathway
C1qRp
CD93
C1q, MBL
Monocytes, neutrophils, endothelial
cell, platelets, T cells
Induces T-cell activation; enhances
phagocytosis
SIGN-R1
CD209
C1q
Marginal zone and lymph node
macrophages
Enhances opsonization of bacteria by MZ
macrophages
C3aR
None
C3a
Mast cells, basophils, granulocytes
Induces degranulation
C5aR
CD88
C5a
Mast cells, basophils, granulocytes,
monocytes, macrophages, platelets,
endothelial cells, T cells
Induces degranulation; chemoattraction;
acts with IL-1 and/or TNF- to induce acute
phase response; induces respiratory burst
in neutrophils
C5L2
None
C5a
Mast cells, basophils, immature
dendritic cells
Uncertain, but most probably downregulates proinflammatory effects of C5a
After Zipfel, P. F., and C. Skerka, 2009, Complement regulators and inhibitory proteins. Nature Reviews Immunology 10:729–740; Kemper, C., and J. P. Atkinson, 2007,
T-cell regulation: With complements from innate immunity, Nature Reviews Immunology 7:9–18; Eggleton, P., A. J. Tenner, and K. B. M. Reid, 2000, C1q receptors, Clinical
and Experimental Immunology 120:406–412; Ohno, M., et al., 2000, A putative chemoattractant receptor, C5L2, is expressed in granulocyte and immature dendritic
cells, but not in mature dendritic cells, Molecular Immunology 37(8):407–412; and Kindt, T. J., B. A. Osborne, and R. A. Goldsby, 2006, Kuby Immunology, 6th ed., New
York: W. H. Freeman, Table 7-4.
anaphylatoxins of the complement system, has been shown
to increase the number of complement receptors by as much
as tenfold.
Therefore, before we venture into a discussion of the biological functions of complement, we should first become
familiar with the receptors for complement components and
their activities. The complement receptors and their primary
ligands are listed in Table 6-3. Where receptors have more
than one name, we have offered both on first introduction,
and subsequently refer to that receptor by the more common
name.
CR1 (CD35) is expressed on both leukocytes and erythrocytes and binds with high affinity to C4b, C3b, and smaller
C3b breakdown products. CR1 receptors on erythrocytes
bind immune complexes and take them to the liver where
they are picked up by phagocytes and cleared. Binding of
complement-opsonized microbial cells via CR1 on phagocytes results in receptor-mediated phagocytosis and the
secretion of proinflammatory molecules such as IL-1 and
prostaglandins. CR1 on B cells mediates antigen uptake of
C3b-bound antigen, leading to its degradation in the B-cell
lysosomal system and subsequent presentation to T cells.
This process makes CR1 an actor in the adaptive, as well as
in the innate immune response. As we will see later, CR1 also
mediates the protection of host cells against the ravages of
complement attack, by serving as a cofactor for the destructive cleavage of C3b and C4b on host cell membranes by
factor I, as well as acting as an accelerator of the decay of the
C3 and C5 convertases.
C3b, either in solution or bound to the surface of cells, is
subject to breakdown by endogenous proteases. CD21
(CR2) is expressed on B cells and binds specifically to the
The Complement System
|
CHAPTER 6
203
C3d
Antigen
CD21
(CR2)
mIgM
CD19
CD81
(TAPA-1)
FIGURE 6-11 Coligation of antigen to B cells via IgM and
CD21. The B-cell coreceptor is a complex of the three cell membrane
molecules, CD21 (CR2), CD81 (TAPA-1), and CD19. The CD19 component is important in BCR antigen receptor signaling. Antigen that has
been covalently bound to fragments of the C3 complement compo-
breakdown products of C3b: iC3b, C3d, and C3dg. Since
C3b can form covalent bonds with antigens, the presence of
CD21 on B cells enables the B cell to bind antigen via both
the B-cell receptor and CD21 (Figure 6-11). This ability to
simultaneously coengage antigen through two receptors has
the effect of reducing the antigen concentration necessary
for B-cell activation by up to a hundredfold (see Chapter 12).
Deficiencies in CD21 have been identified in patients suffering from autoimmune diseases such as systemic lupus erythematosus (see below).
CR3 (a complex of CD11b and CD18) and CR4 (a complex of CD11c and CD18) are important in the phagocytosis
of complement-coated antigens. CR3 and CR4 bind C3b
and several of its breakdown products, including iC3b, C3c,
and C3dg.
CRIg also binds C3b. It is expressed on macrophages
resident in the fixed tissues, including on the Kupffer cells of
the liver. Its importance in clearing C3b-opsonized antigens
by facilitating their removal from circulation in the liver is
emphasized by the finding that CRIg-deficient mice are
unable to efficiently clear C3-opsonized particles. Animals
with this deficiency are therefore subject to higher mortalities during infections.
C3aR, C5aR, and C5L2 are all members of the G protein
coupled receptor (GPCR) family first encountered in Chap-
nent is bound by both the immunoglobulin BCR and by the CD21
complement receptor, thus significantly increasing the avidity of the
cell receptors for the antigen and allowing lower concentrations of
antigen to trigger B-cell activation.
ter 4. C3aR and C5aR mediate inflammatory functions after
binding the small anaphylatoxins, C3a and C5a, respectively
(Figure 6-12). The C5L2 receptor also binds C5a, is structurally similar to C5aR, and is expressed on some of the
same cells. However, C5L2 is not functionally coupled to the
G protein signaling pathway used by C5aR; instead, signaling through C5L2 appears to modulate C5a signaling
through C5aR and C5L2 knockout animals express
enhanced inflammatory responses upon C5a binding to its
receptor.
More recently, a transmembrane lectin, SIGN-R1, able to
bind C1q has been shown to be expressed on the surface of
macrophages located in the marginal zone of the spleen.
SIGN-R1 exists on the macrophage cell surface in aggregated
form and is also able to bind carbohydrates present on the
coat of the Gram-positive bacterium Staphylococcus pneumoniae. When SIGN-R1 binds to bacterial polysaccharides,
the C1q-binding capacity is activated in the same, or in
nearby SIGN-R1 molecules, eventually resulting in the opsonization of the bacterium with C3b. The opsonized bacteria
are then released from these macrophages and bound by
nearby phagocytes, B cells, or dendritic cells. This unusual
mechanism explains a problem long noted with patients who
have undergone splenectomy: an increased susceptibility to
infection by S. pneumoniae.
PA R T I I
204
|
Innate Immunity
Anaphylatoxins and inflammatory response
C3a
Chemotaxis
Cytokine production
Phagocytosis
C5a
Macrophage
NH2
NH2
or
C3aR
C5aR
Neutrophil
Chemotaxis
Oxidative burst
Phagocytosis
Degranulation
Degranulation
Basophil
COOH
COOH
Degranulation
Chemotaxis
α
β
α
γ
GDP
GTP
GDP
Eosinophil
GTP
β
Degranulation
Chemotaxis
γ
Mast cell
FIGURE 6-12 Binding of the anaphylatoxins C3a and C5a to the G-protein-coupled receptors C3aR and C5aR. The C3aR and
C5aR receptors are members of the G-protein-coupled receptor family described in Chapter 4. Binding of the anaphylatoxins to these receptors
stimulates the release of proinflammatory mediators from macrophages, neutrophils, basophils, eosinophils, and mast cells, as indicated.
[Adapted from J. R. Dunkelberger and W.-C. Song, 2010, Complement and its role in innate and adaptive immune responses, Cell Research 20:34–50, Figure 3B.]
Complement Enhances Host Defense
Against Infection
Complement proteins actively engage in host defense against
infection by forming the MAC, by opsonizing potentially
pathogenic microbes, and by inducing an inflammatory
response that helps to guide leukocytes to the site of infection.
MAC-Induced Cell Death
The first function of complement to be described was its role
in inducing cell death following insertion of the MAC into
target cell membranes. Early experiments on MAC formation used erythrocytes as the target membranes, and in this
cellular system large pores involving 17 to 19 molecules of
C9 were reported. Formation of these holes in the cell membrane (see Figure 6-10b) facilitated the free movement of
small molecules and ions. The penetrated red blood cell
membranes were unable to maintain osmotic integrity, and
the cells lysed after massive influx of water from the extracellular fluid.
However, subsequent studies using nucleated eukaryotic
cells indicated that smaller pores can be generated using
only a few molecules of C9, and that death under these circumstances occurs via a type of apoptosis (programmed cell
death), following calcium influx into the cytoplasm. Nuclear
fragmentation, a hallmark of apoptotic death, was observed
during MAC-induced lysis of nucleated cells, which further
supported the notion that at least some MAC-targeted cells
succumb to apoptosis. More careful observations indicated
that the apoptosis induced by the MAC does not share all the
biochemical characteristics normally associated with pro-
grammed cell death, and so this MAC-induced apoptosis
has been termed apoptotic necrosis.
Killing eukaryotic cells with complement is actually quite
difficult, as the membranes of such cells have a number of
factors that act together to inactivate the complement proteins, and thus protect the host cells from collateral damage
during a complement-mediated attack on infectious microorganisms (see below). However, when high concentrations
of complement components are present, MACs can overwhelm the host defenses against MAC attack, and the resulting cell fragments, if present in sufficiently high concentrations,
can induce autoimmunity. Indeed, complement-mediated
damage is a problem in several autoimmune diseases, and
the complement system is considered a target for therapeutic
intervention in autoimmune syndromes.
Can a eukaryotic cell recover from a MAC attack? Welldocumented studies have demonstrated that MACs can be
removed from the cell surface, either by shedding MACcontaining membrane vesicles into the extracellular fluid, or
by internalizing and degrading the MAC-containing vesicles
in intracellular lysosomes. If the MAC is shed or internalized
soon enough after its initial expression on the membrane, the
cell can repair any membrane damage and restore its osmotic
stability. An unfortunate corollary of this capacity to recover
from MAC attack is that complement-mediated lysis directed
by tumor-specific antibodies may be rendered ineffective by
endocytosis or shedding of the MAC (see Chapter 19).
Different types of microorganisms are susceptible to
complement-induced lysis to varying degrees. Antibody and
complement play an important role in host defense against
viruses and can be crucial both in containing viral spread
The Complement System
during acute infection and in protecting against reinfection.
Most enveloped viruses, including herpesviruses, orthomyxoviruses such as those causing measles and mumps,
paramyxoviruses such as influenza, and retroviruses are susceptible to complement-mediated lysis.
However, some bacteria are not susceptible to MAC
attack. Gram-positive bacteria efficiently repel complement
assault, as the complement proteins cannot penetrate the
bacterial cell wall to gain access to the membrane beyond.
In contrast, when attacking Gram-negative bacteria, complement first permeabilizes the outer layer and, following
destruction of the thin cell wall, lyses the inner bacterial
membrane. In some cases, the MAC has been found to
localize at regions of apposition of the inner and outer cell
walls, and breaches them both simultaneously. Neisseria
(a)
Opsonization and phagocytosis
|
CHAPTER 6
205
meningitidis is a Gram-negative bacterium that is susceptible to MAC-induced lysis, and patients who are deficient in
any of the complement components of the MAC are particularly vulnerable to potentially fatal meningitis caused by
this bacterial species.
Promotion of Opsonization
As described in Chapter 5, the term opsonization refers to
the capacity of antibodies and complement components (as
well as other proteins) to coat dangerous antigens that can
then be recognized by Fc receptors (for antibodies) or complement receptors (for complement components) on phagocytic cells. Binding of complement-coated antigen by
phagocytic cells results in complement receptor-mediated
phagocytosis and antigen destruction (Figure 6-13). In
(b)
Coated
virus
particle
C1q
MBLmannose
Antibody: C3b/
antigen
C4b
C3b cleavage
products
(iC3b, C3c, C3dg)
FIGURE 6-13 Opsonization of microbial cells by comple-
CR1
(CD35)
CR3
(CD11b/
CD18)
CR4
(CD11c/
CD18)
CRlg
Fc
receptors
SIGNR1
ment components and antibodies. (a) Phagocytosis is mediated
by many different complement receptors on the surface of macrophages and neutrophils, including CR1, CR3, CR4, and CrIg. Phagocytes, using their Fc receptors, also bind to antigens opsonized by
antibody binding. (b) Electron micrograph of Epstein-Barr virus
coated with antibody and C3b and bound to the Fc and C3b receptor
(CR1) on a B lymphocyte. [Part b from N. R. Cooper and G. R. Nemerow, 1986,
Immunobiology of the Complement System, G. Ross, ed., Orlando, FL: Academic Press.]
206
PA R T I I
|
Innate Immunity
addition, complement receptors on erythrocytes also serve
to bind immune complexes, which are then transported to
the liver for phagocytosis by macrophages (Figure 6-14).
Although less visually compelling than MAC formation,
opsonization may be the most physiologically important
function assumed by complement components.
Opsonization with antibody and complement provides
critical protection against viral infection. Antibody and
complement can create a thick protein coat around a virus
that neutralizes viral infectivity by preventing the virus from
binding to receptors on the host cell. It also promotes phagocytosis via the complement receptors, followed by intracellular destruction of the digested particle.
Promotion of Inflammation
So far, we have focused on the roles of the larger products of
complement factor fragmentation: C3b and C4b in opsonization and C5b in the formation of the MAC. However, the
smaller fragments of C3, C4, and C5 cleavage—C3a, C4a,
and C5a—are no less potent and mediate critically important events in immune responses, acting as anaphylatoxins
or inducers of inflammation. We will focus here specifically
on the activities of C3a and C5a.
C3a and C5a are structurally similar, small proteins
(about 9 kDa, or 74–77 amino acid residues in size) that
promote inflammation and also serve as chemoattractants
for certain classes of leukocytes. C3a and C5a bind G-proteincoupled, activating receptors (C3aR for C3a, C5aR for C5a)
on granulocytes, monocytes, macrophages, mast cells, endothelial cells, and some dendritic cells (see Figure 6-12). Binding of these anaphylatoxins to their receptors on some cells
triggers a signaling cascade that leads to the secretion of
soluble mediators such as IL-6 and TNF-. These cytokines
in turn induce localized increases in vascular permeability
that enable leukocyte migration into the site of infection,
and a concomitant increase in smooth muscle motility that
helps to propel the released fluid to the site of damage. In
addition, these proinflammatory mediators promote phagocytosis of offending pathogens and localized degranulation
of granulocytes (neutrophils, basophils, and eosinophils)
with the resultant release of a second round of inflammatory
mediators, including histamines and prostaglandins. Inflammatory mediators expedite the movement of lymphocytes
into neighboring lymph nodes where they are activated by
the pathogen. This localized inflammatory response is
further supported by systemic effects, such as fever, that
decrease microbial viability.
The central role of the anaphylatoxins in the promotion of
physiologically important inflammatory responses has
brought them to the attention of clinicians seeking ways to
tamp down the pathological levels of inflammation experienced by patients suffering from diseases such as rheumatoid arthritis and systemic lupus erythematosus (SLE).
Clinical Focus Box 6-2 describes some approaches that
investigators are taking to explore the complement system as
a potential therapeutic target.
BLOOD
Ig
Ag
Soluble immune complex
Complement activation
C3b
CR1
Erythrocyte
LIVER
AND
SPLEEN
Phagocyte
FIGURE 6-14 Clearance of circulating immune complexes
by reaction with receptors for complement products on
erythrocytes and removal of these complexes by receptors
on macrophages in the liver and spleen. Because erythrocytes
have fewer receptors than macrophages, the latter can strip the complexes from the erythrocytes as they pass through the liver or spleen.
Deficiency in this process can lead to renal damage due to accumulation of immune complexes.
The Complement System
Complement Mediates the Interface Between
Innate and Adaptive Immunities
Experiments in the past two or three decades have revealed
multiple mechanisms by which components of the complement system modulate adaptive immunity. Many of these
findings are extremely recent, and the study of how the binding of complement components and regulatory proteins may
affect antigen-presenting cells, T cells, and B cells is still in its
early phases.
Complement and Antigen-Presenting Cells
Dendritic cells (DCs), follicular dendritic cells (FDCs), and
macrophages express all of the known complement receptors. When bound to antigens, MBL, C1q, C3b, and C4b are
each capable of engaging their respective receptors on antigenpresenting cells during the process of antigen recognition,
and signaling through their respective receptors acts to
enhance antigen uptake.
In addition, signaling of antigen-presenting cells through
the C5aR anaphylatoxin receptor has been shown to modulate their migration and affect interleukin production, particularly that of the cytokine IL-12. Production of IL-12 by an
antigen-presenting cell normally skews the T-cell response
toward the TH1 phenotype (see Chapter 11). Since both
induction and suppression of IL-12 production have been
detected after activation of the anaphylatoxin receptors,
depending upon the route of antigen delivery, the nature of
the antigen, and the maturation status of the antigen-presenting
cell, we must infer that many signaling pathways are being
integrated to arrive at the eventual biological response to
antigen and complement components.
Complement and B-Cell-Mediated
Humoral Immunity
In 1972, Pepys showed that depleting mice of C3 impaired
their T-dependent antigen-specific B-cell responses, thus
implying that complement may participate in the initiation
of the B-cell response. It now appears that this seminal
observation was the first description of the complement
receptor CD21 acting as a coreceptor in antigen recognition.
Recall that the C3b molecule is capable of covalently “tagging” antigen by interaction between its reactive thioester
group and hydroxyl or amino groups on the surface of the
antigen. As described in Chapters 3 and 12, the immunoglobulin receptor is expressed on the B-lymphocyte membrane along with the complement receptor CD21, the
signaling protein CD19, and the transmembrane CD81 (tetraspanin, or TAPA-1) molecule (see Figure 6-11). CD21 is a
receptor with specificity for both C3b and the products of
C3b proteolytic breakdown, C3d (human) or C3dg (mouse).
If C3b was originally bound to a microbial surface, the C3d
or C3dg fragments remain bound to that surface after C3b
proteolysis. A B cell bearing an immunoglobulin receptor
specific for that microbe will therefore be able to bind the
microbe with both its immunoglobulin receptor and with
|
CHAPTER 6
207
the associated CD21 complement receptor, thus enhancing
the interaction between the B cell and antigen and reducing
the amount of antigen that is required to signal a B-cell
response by a factor of tenfold to a thousandfold.
Complement and T-Cell-Mediated Immunity
The mechanisms by which complement affects T-cell responses
are not as well characterized as those for B cells and antigenpresenting cells. However, some recent, interesting results have
been generated by the study of mice either genetically deficient
in one or more complement components or following treatment with complement inhibitors. For example, mice lacking
the C3 gene have reduced CD4⫹ and CD8⫹ T cell responses,
implying that signaling through C3 enhances T-cell function.
Furthermore, mice treated with inhibitors of C5aR signaling
produced fewer antigen-specific CD8⫹ T cells, following influenza infection, than wild-type mice, indicating that C5a may
act as a costimulator during CD8⫹ T-cell activation, possibly
by increasing IL-12 production by antigen-presenting cells, as
described above. Both these experiments demonstrate that
signaling through complement components may have positive
effects on T-cell-mediated adaptive immune responses.
The study of T-cell regulation by complement components is in its infancy, but these provocative findings suggest
that there is much to learn and that complement may have
more powerful effects on adaptive immunity than previously
thought. As these effects are further explored, the question
of whether complement is acting directly on the T cells, or
indirectly, via effects on antigen-presenting cells, is an issue
that will need to be clearly defined.
Complement Aids in the Contraction Phase
of the Immune Response
At the close of an adaptive immune response, most of the
lymphocytes that were generated in the initial proliferative
phase undergo apoptosis (programmed cell death), leaving
only a few antigen-specific cells behind to provide immunological memory (see Chapters 11 and 12). In addition, soluble antigen-antibody complexes may still be present in the
bloodstream and immune organs. If autoimmune disease is
to be avoided, these excess lymphocytes and immune complexes must be disposed of safely, without the induction of
further inflammation, and complement components play a
major role in these processes.
Disposal of Apoptotic Cells and Bodies
Apoptotic cells express the phospholipid phosphatidyl serine
on the exterior surface of their plasma membranes. In healthy
cells, this phospholipid is normally restricted to the cytoplasmic side of the membrane, and this change in its location
serves to signal the immune system that the cell is undergoing
programmed cell death. Surprisingly, experiments have also
demonstrated the presence of nucleic acids on the surface
membrane of apoptotic cells. Nuclear fragmentation and
208
PA R T I I
|
Innate Immunity
CLINICAL FOCUS
The Complement System as a Therapeutic Target
The involvement of complementderived anaphylatoxins in inflammation
makes complement an interesting therapeutic target in the treatment of inflammatory diseases, such as arthritis. In addition,
autoimmune diseases that potentially result
in complement-mediated damage to host
cells, such as multiple sclerosis and agerelated macular degeneration, are also
potential candidates for complementfocused therapeutic interventions.
The last three decades have seen the
crystallization and molecular characterization of several complement components,
a necessary precursor to the development
of designer drugs targeted to specific proteins. At the time of writing, however, just
two therapies directed at interfering with
complement components are approved
for clinical use, although at least one more
is currently in clinical trials.
Purified C1 esterase inhibitor is currently in use in the clinic as a therapeutic
option for the treatment of hereditary
angioedema. However, its beneficial
effects are most likely the result of the
inhibition of excess esterase activation
external to the complement system per se.
The other successful complementrelated treatment is more specifically
directed toward a dysregulation of the
complement cascade. Paroxysmal noctur-
nal hemoglobinuria (PNH) manifests as
increased fragility of erythrocytes, leading
to chronic hemolytic anemia, pancytopenia (loss of blood cells of all types), and
venous thrombosis (formation of blood
clots). The name PNH derives from the
presence of hemoglobin in the urine, most
commonly observed in the first urine
passed after a night’s sleep. The cause of
PNH is a general defect in the synthesis of
cell-surface proteins, which affects the
expression of two regulators of complement: DAF (CD55) and Protectin (CD59).
DAF and Protectin are cell-surface proteins that function as inhibitors of complement-mediated cell lysis, acting at
different stages of the process. DAF
induces dissociation and inactivation of
the C3 convertases of the classical, lectin,
and alternative pathways (see Figure
6-16). Protectin acts later in the pathway
by binding to the C5b678 complex and
inhibiting C9 binding, thereby preventing
formation of the membrane pores that
destroy the cell under attack. Deficiency
in these proteins leads to increased sensitivity of host cells to complement-mediated lysis and is associated with a high risk
of thrombosis.
Protectin and DAF are attached to the
cell membrane via glycosylphosphatidyl
inositol (GPI) anchors, rather than by
DNA cleavage follow quickly after outer membrane phosphatidyl serine expression. Once apoptosis begins, the dying cell
is broken down into membrane-bound vesicles termed apoptotic bodies, which express phosphatidyl serine and/or surfacebound DNA on their exterior surfaces. Recent work has
demonstrated that C1q binds specifically to DNA.
The complement component C1q specifically binds apoptotic bodies and assists in their clearance. When light-sensitive
keratinocytes (skin cells) are treated with UVB to initiate apoptosis and then stained with anti-C1q antibody, C1q staining is
restricted to the apoptotic blebs, where DNA, exposed on
apoptotic membranes, is specifically recognized by C1q (Figure 6-15). With C1q deposition, the classical pathway is initiated and the apoptotic cells are then opsonized by C3b. This, in
turn, leads to clearance of the apoptotic cells by phagocytes.
stretches of hydrophobic amino acids, as is
the case for many membrane proteins.
Patients with PNH lack an enzyme, Phosphatidyl Inositol Glycan Class A (PIGA), that
attaches the GPI anchors to the appropriate proteins. Patients thus lack the expression of DAF and Protectin needed to
protect erythrocytes against lysis. The term
paroxysmal refers to the fact that episodes
of erythrocyte lysis are often triggered by
stress or infection, which result in increased
deposition of C3b on host cells. The gene
encoding the PIGA enzyme is X-linked.
The defect identified in PNH occurs
early in the enzymatic pathway leading to
formation of a GPI anchor and resides in
the pig-a gene. Transfection of cells from
PNH patients with an intact pig-a gene
restores the resistance of the cells to host
complement lysis. Further genetic analysis revealed that the defect is not encoded
in the germline genome (and therefore is
not transmissible to offspring), but rather
is the result of mutations that occurred
within the hematopoietic stem cells
themselves, such that any one individual
may have both normal and affected cells.
Those patients who are most adversely
affected display a preferential expansion
of the affected cell populations.
PNH is a chronic disease with a mean
survival time between 10 and 15 years.
In the absence of C1q, the apoptotic membrane blebs
are released from the dying cells as apoptotic bodies,
which can then act as antigens and initiate autoimmune
responses. As a consequence, mice deficient in C1q show
increased mortality and higher titers of auto-antibodies
than control mice and also display an increased frequency
of glomerulonephritis, an autoimmune kidney disease.
Analysis of the kidneys of C1q knockout mice reveal
deposition of immune complexes as well as significant
numbers of apoptotic bodies.
Disposal of Immune Complexes
As mentioned earlier, the coating of soluble immune complexes with C3b facilitates their binding by CR1 on erythrocytes. Although red blood cells express lower levels of CR1
The Complement System
|
CHAPTER 6
209
BOX 6-2
The most common causes of mortality in
PNH are venous thromboses affecting
hepatic veins and progressive bone marrow failure.
A breakthrough in treatment of PNH
was reported in 2004 using a humanized
monoclonal antibody that targets complement component C5 and thus inhibits
the terminal steps of the complement
cascade and formation of the membraneattack complex. This antibody—eculizumab (Soliris)—was infused into patients,
who were then monitored for the loss of
red blood cells. A dramatic improvement
was seen in patients during a 12-week
period of treatment with eculizumab (Figure 1). Treatment of PNH patients with
eculizumab relieves hemoglobinuria,
reverses the kidney damage resulting
from high levels of protein in the urine,
and significantly reduces the frequency of
thromboses. In 2007, eculizumab was
approved for the treatment of PNH in the
general population.
Since the control of infections with
Meningococcal bacteria (Neisseria meningitidis) relies upon an intact and functioning membrane attack complex (MAC),
patients being treated with eculizumab
are routinely vaccinated against meningococcus.
The pathology of PNH underscores the
potential danger to the host that is inherent in the activation of the complement
Days with paroxysms
per patient per month
3.0
2.5
2.0
1.5
1.0
0.5
0
Before
administration
of eculizumab
During
administration
of eculizumab
FIGURE 1
Treatment of PNH patients with eculizumab relieves hemoglobinuria. The number
of days with paroxysm (onset of attack) per patient per month in the month prior to treatment (left bar)
and for a 12-week period of treatment with eculizumab (right bar) is shown. [From P. Hillmen, et al., 2004.
Eculizumab in patients with paroxysmal nocturnal hemoglobinuria. New England Journal of Medicine 350:6, 552–559.]
(100–1000 molecules per cell, depending on the age of the
cell and the genetic constitution of the donor) than do
granulocytes (5 104 per cell), there are about 1000 erythrocytes for every white blood cell, and therefore erythrocytes account for about 90% of the CR1 in blood.
Erythrocytes also therefore play an important role in clearing C3b-coated immune complexes by conveying them to
the liver and spleen where the immune complexes are
stripped from the red blood cells and phagocytosed (see
Figure 6-14).
The importance of the complement system in immune
complex clearance is emphasized by the finding that patients
with SLE, the autoimmune disease, have high concentrations
of immune complexes in their serum that are deposited in
system. Complex systems of regulation
are necessary to protect host cells from
the activated complement complexes
generated to lyse intruders, and alterations in the expression or effectiveness of
these regulators has the potential to result
in a pathological outcome.
The centrality of C3 within the complement cascades may suggest that it
would be an excellent target for therapeutic intervention in complementmediated disease. However, its very position
at the crossroads of the three pathways
means that interference with C3 activity
places patients at increased risk for infectious diseases normally curtailed by the
activities of the innate as well as the adaptive immune systems, and so systemic
drugs specifically targeting C3 may prove
to have too high a risk-benefit ratio.
However, in one success story, targeted to a particular organ system, the
drug Compstatin was developed as a
result of phage display experiments
designed to discover compounds that
bound to C3. Chemical refinements based
on structural and computational studies
followed, and a Compstatin derivative,
POT-4, is now in clinical trials for the treatment of age-related macular degeneration, a progressive and debilitating eye
disease that leads inexorably and quickly
to total blindness. Because POT-4 can be
delivered directly into the vitreous humor
of the eye, systemic side effects on the
patient’s immune system are minimized.
the tissues. Complement is activated by these tissue-deposited
immune complexes, and pathological inflammation is induced
in the affected tissues (see Chapter 15).
Since complement activation is implicated in the pathogenesis of SLE, it may therefore seem paradoxical that the
incidence of SLE is highly correlated to C4 deficiency.
Indeed, 90% of individuals who completely lack C4 develop
SLE. The resolution to this paradox lies in the fact that deficiencies in the early components of complement lead to a
reduction in the levels of C3b that are deposited on the
immune complexes. This reduction, in turn, inhibits their
clearance by C3b-mediated opsonization and allows for the
activation of the later inflammatory and cytolytic phases of
complement activation.
210
PA R T I I
|
Innate Immunity
levels of complement components in the adult retina during
the early phases of the disease, suggesting that inappropriate,
complement-mediated synaptic pruning may be a causal
factor in that disease. C1q up-regulation has also been
observed in animal models of both ALS and Alzheimer’s
disease, potentially implicating inappropriate complement
activation in the etiology of some clinically important neurodegenerative illnesses.
The Regulation of
Complement Activity
FIGURE 6-15 Binding of C1q to apoptotic keratinocytes
(skin cells) is limited to surface apoptotic blebs. Apoptosis was
induced in human skin cells by exposure to ultraviolet light. The cells
were then incubated in human serum and stained with an antibody
to C1q. This is an image derived from confocal microscopy that
shows the C1q staining in green, clearly localizing to the apoptotic
blebs. The cell itself is visualized by phase contrast microscopy, and
the image was generated by merging the phase contrast and immunofluorescent frames. [Korb and Ahearn. 1997. J. Immunol. 158: 4525–4528.]
Complement Mediates CNS Synapse
Elimination
Complement has also been shown to play an important role
outside the boundaries of the immune system. In the developing nervous system, growing neurons first form a relatively high number of connections (synapses) with one
another; the number of those synapses is then pruned as the
nervous system matures. Scientists studying the development of the mouse eye have demonstrated that animals
deficient in C1q or C3 fail to eliminate these early, excess
synapses, and display anatomical abnormalities in the visual
nervous system, indicating that complement may play an
important role in the process of synaptic remodeling.
In healthy animals, the expression of complement components in mouse neurons is up-regulated early in development, in response to signals provided by immature astrocytes
(glial cells that assist in the function and maintenance of the
nervous system). Complement component expression is
then down-regulated as the animals mature and astrocyte
function is dampened. However, patients suffering from
glaucoma have been shown to display erroneously high
All biological systems with the potential to damage the host,
be they metabolic pathways, cytotoxic cells, or enzyme cascades, are subject to rigorous regulatory mechanisms, and
the complement system is no exception to this general rule.
Especially in light of the potent positive feedback mechanisms and the absence of antigen specificity of the alternative pathway, the host must ensure that the destructive
potential of complement proteins is confined to the appropriate pathogen surfaces and that collateral damage to
healthy host tissue is minimized.
Below, we discuss the different mechanisms by which the
host protects itself against the potential ravages of inadvertent complement activation.
Complement Activity Is Passively Regulated by
Protein Stability and Cell-Surface Composition
Protection of vertebrate host cells against complementmediated damage is achieved by both general, passive and
specific, active regulatory mechanisms. The relative instability of many complement components is the first means by
which the host protects itself against inadvertent complement activation. For example, the C3 convertase of the
alternative pathway, C3bBbC3b, has a half-life of only 5 minutes before it breaks down, unless it is stabilized by reaction
with properdin. A second passive regulatory mechanism
depends upon the difference in the cell surface carbohydrate
composition of host versus microbial cells. For example, fluid
phase proteases that destroy C3b bind much more effectively to host cells, which bear high levels of sialic acid,
than to microbes that have significantly lower levels of this
sugar. Hence, any C3b that happens to alight on a host cell
is likely to be destroyed before it can effect significant
damage.
In addition to these more passive environmental brakes
on inappropriate complement activation, a series of active
regulatory proteins act to inhibit, destroy, or tune down the
activity of complement proteins and their active fragments.
The stages at which complement activity is subject to regulation are illustrated in Figure 6-16, and the regulatory proteins are listed in Table 6-4.
The Complement System
|
CHAPTER 6
211
(a) Dissociation of C1 components
C1INH
C1r2s2
C1r2s2
C1q
(b) Decay-accelerating activity for C3 convertases
C4b2a
DAF (CD55)
CR1 (CD35)
C4BP
Classical
pathway
C3bBb
DAF (CD55)
CR1 (CD35)
Factor H
Alternative
pathway
C4b
C2a
C3b
Bb
(c) Factor I cofactor activity
Factor I
C3b
MCP (CD46)
CR1 (CD35)
Factor H
Factor I
MCP (CD46)
CR1 (CD35)
C4BP
C4b
C3c
iC3b
C4c
C4d
(d) Inhibition of lysis
C5bC6
C7 C8
C3dg
C9
Protectin (CD59)
Vitronectin/S protein
MAC inhibition
prevents C9 binding
and polymerization
(e) Cleavage of the anaphylatoxins
C3a des Arg
C3a
C5a
Carboxypeptidase N
Carboxypeptidase B
Carboxypeptidase R
C5a des Arg
FIGURE 6-16 The regulation of complement activity. The various stages at which complement activity is subject to regulation (see
text for details). [Adapted from J. R. Dunkelberger and W.-C. Song, 2010. Cell Research 20:34–50.]
The C1 Inhibitor, C1INH, Promotes Dissociation
of C1 Components
to control the time period during which they can remain
active.
C1INH, the C1 inhibitor, is a plasma protein that binds in
the active site of serine proteases, effectively poisoning
them. C1INH belongs to the class of proteins called serine
protease inhibitors (serpins), and it acts by forming a complex with the C1 proteases, C1r2s2, causing them to dissociate from C1q and preventing further activation of C4 or
C2 (see Figure 6-16a). C1INH inhibits both C3b and the
serine protease MASP2. It is the only plasma protease
capable of inhibiting the initiation of both the classical and
lectin complement pathways. Its presence in plasma serves
Decay Accelerating Factors Promote Decay
of C3 Convertases
Since the reaction catalyzed by the C3 convertase enzymes
is the major amplification step in complement activation,
the generation and lifetimes of the C3 convertases C4b2a
and C3bBb are subject to rigorous control. Several membrane-bound factors accelerate the decay of the C4b2a on
the surface of host cells, including decay accelerating factor, or DAF (CD55), CR1, and C4BP (C4 binding protein)
212
TABLE 6-4
PA R T I I
|
Innate Immunity
Proteins involved in the regulation of complement activity
Protein
Fluid phase or
membrane
Pathway affected
Function
C1 inhibitor (C1INH)
Fluid phase
Classical and lectin
Induces dissociation and inhibition of
C1r2s2 from C1q; serine protease inhibitor
Decay Accelerating
Factor (DAF) CD55
Membrane bound
Classical, alternative,
and lectin
Accelerates dissociation of C4b2a and
C3bBb C3 convertases
CR1 (CD35)
Membrane bound
Classical, alternative,
and lectin
Blocks formation of, or accelerates dissociation
of, the C3 convertases C4b2a and C3bBb by
binding C4b or C3b
Cofactor for factor I in C3b and C4b
degradation on host cell surface
C4BP
Soluble
Classical and lectin
Blocks formation of, or accelerates dissociation
of, C4b2a C3 convertase
Cofactor for factor I in C4b degradation
Factor H
Soluble
Alternative
Blocks formation of, or accelerates dissociation
of, C3bBb C3 convertase
Cofactor for factor I in C3b degradation
Factor I
Soluble
Classical, alternative,
and lectin
Serine protease: cleaves C4b and C3b using
cofactors shown in Figure 6-16
Membrane cofactor
of proteolysis, MCP
(CD46)
Membrane bound
Classical, alternative,
and lectin
Cofactor for factor I in degradation of C3b
and C4b
S protein or
Vitronectin
Soluble
All pathways
Binds soluble C5b67 and prevents insertion
into host cell membrane
Protectin (CD59)
Membrane bound
All pathways
Binds C5b678 on host cells, blocking binding
of C9 and the formation of the MAC complex
Carboxypeptidases
N, B, and R
Soluble
Anaphylatoxins produced
by all pathways
Cleave and inactivate the anaphylatoxins C3a
and C5a
(see Figure 6-16b). These decay accelerating factors cooperate to accelerate the breakdown of the C4b2a complex
into its separate components. The enzymatically active
C2a diffuses away, and the residual membrane-bound C4b
is degraded by another regulatory protein, factor I (see
Figure 6-16c).
In the alternative pathway, DAF and CR1 function in a
similar fashion and are joined by factor H in separating the
C3b component of the alternative pathway C3 convertase
from its partner, Bb. Again, Bb diffuses away, and residual
C3b is degraded (see Figure 6-16b).
Whereas DAF and CR1 are membrane-bound components and their expression is therefore restricted to host
cells, factor H and C4BP are soluble regulatory complement
components. Host-specific function of factor H is ensured
by its binding to polyanions such as sialic acid and heparin,
essential components of eukaryotic, but not prokaryotic, cell
surfaces. Similarly, C4BP is preferentially bound by host cell
membrane proteoglycans such as heparan sulfate. In this
way, host cells are protected from deposition of complement
components, whereas it is permitted on the membranes of
microbial invaders that lack DAF and CR1 expression and
fail to bind factor H or C4BP.
Factor I Degrades C3b and C4b
Factor H, C4BP, and CR1 also figure in a second and potentially more critical pathway of complement regulation: that
catalyzed by factor I. Factor I is a soluble, constitutively
(always) active serine protease that can cleave membraneassociated C3b and C4b into inactive fragments (Figure 6-16c).
However, if factor I is indeed soluble, constitutively
active, and designed to destroy C3b and C4b, one might
wonder how the complement cascades described above can
ever succeed in destroying invading microbes? The answer is
that factor I requires the presence of cofactors in order to
function and that two of these cofactors, Membrane Cofactor
of Proteolysis (MCP, or CD46) and CR1, are found on the
The Complement System
surface of host cells, but not on the surfaces of the microbial
cells (see Figure 6-16c). The other two cofactors, factor H
and C4BP, described above, bind host cell-surface peptidoglycans. Hence, cleavage of membrane-bound C3b on host
cells is conducted by factor I in collaboration with the
membrane-bound host cell proteins, CD46 and CR1, and
the soluble cofactor H. Similarly, cleavage of membranebound C4b is effected again by factor I, this time in collaboration with membrane-bound CD46 and CR1 and soluble
cofactor C4BP. Since these membrane-bound, or membranebinding cofactors are not found on microbial cells, C3b and
C4b are thus allowed to remain on microbial cells and effect
their specific functions.
CD46 has recently been implicated as a factor in the control of apoptosis of dying T cells. When a T cell commits to
apoptosis, it expresses DNA on its cell membrane that binds
circulating C1q, as described above. It then begins to shed
CD46 from the cell surface. Only after CD46 is lost can progression of the classical pathway occur, with increased C3b
binding and phagocytosis of the apoptotic T cells. As the cell
continues to deteriorate, it may also begin to release nucleotides and other molecules into the tissue fluids that serve as
chemotactic signals for phagocytes.
Protectin Inhibits the MAC Attack
In the case of a particularly robust antibody response, or of an
inflammatory response accompanied by extensive complement activation, inappropriate assembly of MAC complexes
on healthy host cells can potentially occur, and mechanisms
have evolved to prevent the resulting inadvertent host cell
destruction. A host cell surface protein, Protectin (CD59),
binds any C5b678 complexes that may be deposited on host
cells and prevents their insertion into the host cell membrane
(see Figure 6-16d). Protectin also blocks further C9 addition
to developing MACs. In addition, the soluble complement S
protein, otherwise known as Vitronectin, binds any fluid
phase C5b67 complexes released from microbial cells, preventing their insertion into host cell membranes.
Carboxypeptidases Can Inactivate the
Anaphylatoxins C3a and C5a
Anaphylatoxin activity is regulated by cleavage of the
C-terminal arginine residues from both C3a and C5a by
serum carboxypeptidases, resulting in rapid inactivation of
the anaphylatoxin activity (Figure 6-16e). Carboxypeptidases are a general class of enzymes that remove amino acids
from the carboxyl termini of proteins; the specific enzymes
that mediate the control of anaphylatoxin concentrations are
carboxypeptidases N, B, and R. These enzymes remove arginine residues from the carboxyl termini of C3a and C5a to
form the so-called des-Arg (“without Arginine”), inactive
forms of the molecules. In addition, as mentioned above,
binding of C5a by C5L2 also serves to modulate the inflammatory activity of C5a.
|
CHAPTER 6
213
Complement Deficiencies
Genetic deficiencies have been described for each of the
complement components. Homozygous deficiencies in any
of the early components of the classical pathway (C1q, C1r,
C1s, C4, and C2) result in similar symptoms, notably a
marked increase in immune-complex diseases such as SLE,
glomerulonephritis, and vasculitis. The effects of these deficiencies highlight the importance of the early complement
reactions in generating C3b and C3b’s role in the solubilization and clearance of immune complexes. In addition, as
described above, C1q has been shown to bind apoptotic
blebs. In the absence of C1q binding, cells bearing these
apoptotic blebs, or the blebs themselves, can act as autoantigens and lead to the development of autoimmune diseases such as SLE. Individuals with deficiencies in the early
complement components may also suffer from recurrent
infections by pyogenic (pus-forming) bacteria such as streptococci and staphylococci. These organisms are Gram-positive
and therefore are normally resistant to the lytic effects of the
MAC. Nevertheless, the early complement components can
help to mitigate such infections by mediating a localized
inflammatory response and opsonizing the bacteria.
A deficiency in MBL, the first component of the lectin
pathway, has been shown to be relatively common, and
results in serious pyrogenic (fever-inducing) infections in
babies and children. Children with MBL deficiency suffer
from respiratory tract infections. MBL deficiency is also
found with a frequency two to three times higher in SLE
patients than in normal subjects, and certain mutant forms of
MBP are found to be prevalent in chronic carriers of hepatitis
B. Deficiencies in factor D and properdin—early components
of the alternative pathway—appear to be associated with
Neisseria infections but not with immune-complex disease.
People with C3 deficiencies display the most severe clinical manifestations of any of the complement deficiency
patients, reflecting the central role of C3 in opsonization and
in the formation of the MAC. The first person identified with
a C3 deficiency was a child suffering from frequent, severe
bacterial infections leading to meningitis, bronchitis, and
pneumonia, who was initially diagnosed with agammaglobulinemia. After tests revealed normal immunoglobulin levels, a deficiency in C3 was discovered. This case highlighted
the critical function of the complement system in converting
a humoral antibody response into an effective defense mechanism. The majority of people with C3 deficiency have
recurrent bacterial infections and may also present with
immune-complex diseases.
Levels of C4 (one of the proteins operative very early in
the classical pathway) vary considerably in the population.
Specifically, the genes encoding C4 are located in the major
histocompatibility locus (see Chapter 8), and the number of
C4 genes may vary among individuals from two to six. Low
gene copy numbers are associated with lower levels of C4 in
plasma and with a correspondingly higher incidence of SLE,
for the reasons described above. Patients with complete
214
PA R T I I
|
Innate Immunity
deficiencies of the components of the classical pathway such
as C4 suffer more frequent infections with bacteria such as
S. pneumoniae, Haemophilus influenza, and N. meningitidis.
However, even patients with low copy numbers of C4 appear
to be relatively well protected against such infections. Interestingly, C4 exists in two isoforms: C4A and C4B. C4B is
more effective in binding to the surfaces of the three bacterial
species mentioned above, and homozygous C4B deficiency
has been shown to result in a slightly higher incidence of
infection than is experienced by individuals with C4A deficiency, or those who are heterozygous for the two isoforms.
Individuals with deficiencies in members of the terminal
complement cascade are more likely than members of the
general population to suffer from meningococcal infections,
indicating that cytolysis by complement components C5-C9
is of particular relevance to the control of N. meningitidis.
This has resulted in the release of public health guidelines
that highlight the need for vaccinations against N. meningitidis for individuals deficient in the terminal complement
components.
Deficiencies of complement regulatory proteins have also
been reported. As described above, C1INH, the C1 inhibitor,
regulates activation of the classical pathway by preventing
excessive C4 and C2 activation by C1. However, as a serine
protease inhibitor, it also controls two serine proteases in the
blood clotting system. Therefore, patients with C1INH deficiency suffer from a complex disorder that includes excessive
production of vasoactive mediators (molecules that control
blood vessel diameter and integrity), which in turn leads to
tissue swelling and extracellular fluid accumulation. The
resultant clinical condition is referred to as hereditary angioedema. It presents clinically as localized tissue edema that
often follows trauma, but sometimes occurs with no known
TABLE 6-5
cause. The edema can be in subcutaneous tissues; within the
bowel, where it causes abdominal pain; or in the upper respiratory tract, where it can result in fatal obstruction of the
airway. C1INH deficiency is an autosomal dominant condition with a frequency of 1 in 1000 in the human population.
Studies in humans and experimental animals with homozygous deficiencies in complement components have provided important information regarding the roles of individual
complement components in immunity. These initial observations have been significantly enhanced by studies using
knockout mice, genetically engineered to lack expression of
specific complement components. Investigations of in vivo
complement activity in these animals has allowed dissection
of the complex system of complement proteins and the
assignment of precise biologic roles to each.
Microbial Complement
Evasion Strategies
The importance of complement in host defense is clearly
illustrated by the number and variety of strategies that have
evolved in microbes, enabling them to evade complement
attack (see Table 6-5). Gram-positive bacteria have developed thick cell walls and capsules that enable them to shrug
off the insertion of MAC complexes, while others escape into
intracellular vacuoles to escape immune detection. However,
these two general strategies are energy intensive for the
microbe, and many microbes have adopted other tactics to
escape complement-mediated destruction.
In Advances Box 6-3, we describe the multiple ways in
which Staphylococcus aureus, a major human pathogen, protects itself against complement-mediated destruction. In this
Some microbial complement evasion strategies
Complement evasion strategy
Example
Interference with antibody-complement interaction
Antibody depletion by Staphylococcal protein A
Removal of IgG by Staphylokinase
Binding and inactivation of complement proteins
S. aureus protein SCIN binds and inactivates the C3bBb C3
convertase
Parasite protein C2 receptor trispanning protein disrupts the
binding between C2 and C4
Protease-mediated destruction of complement component
Elastase and alkaline phosphatase from Pseudomonas degrade
C1q and C3/C3b
ScpA and ScpB from Streptococcus degrade C5a
Microbial mimicry of complement regulatory components
Streptococcus pyogenes M proteins bind C4BP and factor H to the
cell surface, accelerating the decay of C3 convertases bound to
the bacterial surface
Variola and Vaccinia viruses express proteins that act as cofactors
for factor I in degrading C3b and C4b
The Complement System
section, we address complement microbial evasion at a more
conceptual level, categorizing the approaches that microbes
have evolved to elude this effector arm of the immune
response.
Some Pathogens Interfere with the First Step
of Immunoglobulin-Mediated Complement
Activation
Many classes of viruses act at the beginning of the classical
complement pathway by synthesizing proteins and glycoproteins that specifically bind the Fc regions of antibodies, thus
preventing complement binding and the generation of the
classical pathway reactions. Some viruses also effect enhanced
clearance of antibody-antigen complexes from the surfaces of
virus-infected cells and/or manufacture proteins that induce
rapid internalization of viral protein-antibody complexes.
Microbial Proteins Bind and Inactivate
Complement Proteins
Since highly specific protein-protein interactions between
complement components are central to the functioning of
the cascade, it is logical that microbes have evolved mechanisms that interfere with some of these binding reactions.
Advances Box 6-3 describes those used by S. aureus, but the
strategy of inhibiting the interactions between complement
proteins is not restricted to bacteria. The production of molecules that inhibit the interactions between complement
components has also been detected in certain human parasites. Specifically, a protein generated by some species of
both Schistosoma and Trypanosoma, the complement C2
receptor trispanning protein, disrupts the interaction
between C2a and C4b and thus prevents the generation of
the classical pathway C3 convertase.
Microbial Proteases Destroy
Complement Proteins
Some microbes produce proteases that destroy complement
components. This strategy is utilized primarily by bacteria,
and numerous bacterial proteases exist that are capable of
digesting a variety of complement components. For example, elastase and alkaline protease proteins from Pseudomonas aeruginosa target the degradation of C1q and C3/C3b,
and two proteases derived from Streptococcal bacteria, scpA
and scpB, specifically attack the anaphylatoxin C5a. Streptococcal pyrogenic exotoxin B was found to degrade the complement regulator properdin, with resultant destabilization
of the alternative pathway C3 convertase on the bacterial
surface.
Fungi can also inactivate complement proteins. The
opportunistic human pathogen Aspergillus fumigatus
secretes an alkaline protease, Alp1, that is capable of cleaving
C3, C4, C5, and C1q, as well as IgG. Since this pathogen
|
CHAPTER 6
215
tends to attack patients who are already immunocompromised, its capacity to further damage the immune system is
particularly troubling to the clinician.
Some Microbes Mimic or Bind Complement
Regulatory Proteins
Several microbial species have exploited the presence of cellsurface or soluble regulators of complement activity for their
own purposes, either by mimicking the effects of these regulators, or by acquiring them directly. Indeed, recent work
suggests that sequestration of host-derived regulators of
complement activity may be one of the most widely utilized
microbial mechanisms for complement evasion.
Many microbes have developed the ability to bind one or
another of the fluid phase inhibitors of complement, C4BP
or factor H. For example, Streptococcus pyogenes, an important human pathogen, expresses a family of proteins called
“M proteins” capable of binding to C4BP and factor H.
Expression of these regulatory proteins on the bacterial surface results in the inhibition of the later steps of complement
fixation.
More surprisingly, some microbes appear to acquire
membrane-bound regulators from host cells. Helicobacter
pylori, obtained from patients suffering from gastric ulcers,
was found to stain positive for Protectin, a potent host inhibitor of MAC complex formation. Protectin is normally
attached to the host membrane by a glycosylphosphatidyl
inositol anchor, and this anchor must be transferable in
some way from the host to the bacterial cell membrane.
Many viruses have evolved the capacity to produce proteins that closely mimic the structure and function of complement regulatory proteins. For example, the Variola
(smallpox) and Vaccinia (cowpox) viruses express complement inhibitory proteins that bind C3b and C4b and serve as
cofactors for factor I, thus preventing complement activation
at the viral membrane. In addition to generating complement regulatory proteins, it should be noted that some
viruses actively induce regulatory complement components
within the cells that harbor them. Other viruses camouflage
themselves during budding from the host cell by hiding
within regions of the host membrane on which regulatory
complement components are expressed. And finally, some
viruses mimic eukaryotic membranes by incorporating high
levels of sialic acid into the viral membrane, thus facilitating
the binding of the cofactors for factor I that normally bind
only to host cell membranes. This therefore results in the
inhibition of complement activation on the viral surface.
The Evolutionary Origins of the
Complement System
Although the complement system was initially characterized
by its ability to convert antibody binding into pathogen
lysis, studies of complement evolution have identified the
PA R T I I
216
|
Innate Immunity
ADVANCES
Staphylococcus aureus Employs Diverse Methods to Evade
Destruction by the Complement System
Staphylococcus aureus has
polysaccharide itself provides the bacterium with some mechanical inhibition of
opsonization (1 in Figure 1). Although
complement factors can assemble on the
cell-wall surface underneath the capsule,
most are then inaccessible to the complement receptors on phagocyte sur-
developed an impressive variety of mechanisms that inhibit both the classical and
the alternative pathways of complement
activation.
S. aureus is an encapsulated, Grampositive bacterial strain. The capsular
faces. However, if the C3 convertases
C4b2a and C3bBb do succeed in assembling at the bacterial surface, they are
then bound by a small, 9.8 kDa protein
called Staphylococcus Complement
INhibitor (SCIN). Blockage of further convertase activity prevents amplification of
1
4
Capsule
Cell wall
Sak
Plasminogen
5 ClfA binding
fibrinogen
Plasmin
IgG
C3b
2
3 SpA binding IgG
Efb
C3
C3b C3a
components of the lytic MAC as relatively late additions to
the animal genome. Long before the emergence of the adaptive immune system, complement components provided
invertebrate organisms with an important advance on the
peptide-based humoral immunity previously available to
them (see Chapter 5). Activation of complement cascades in
invertebrates and early vertebrates most likely culminated in
opsonization and phagocytosis by primitive hemocytes. It
appears probable that, among the non-MAC complement
components, the proteins of the alternative pathway were the
first to appear, followed by those of the lectin recognition
systems. A fully fledged MAC emerged only around the
same time as the appearance of the adaptive immune system
(see Figure 6-17 and Table 6-6).
FIGURE 1
Mechanisms by which S. aureus avoids
opsonophagocytosis. (1) The capsular polysaccharides denies access of neutrophils to opsonized
bacteria. (2) The extracellular fibrinogen binding protein
(Efb) binds C3, preventing it from reaching the cell
surface and inhibiting further activation of the
complement cascade. (3) Protein A (SpA) binds IgG in a
conformation that does not permit Fc receptor binding.
(4) Staphylokinase (Sak), secreted by the bacterium,
activates plasminogen, a protease capable of cleaving
and inactivating IgG and C3b. (5) Clumping factor A
binds factor I and localizes it to the microbial surface,
where it cleaves and inactivates any C3b that binds
there. [Adapted from Foster, T. J., 2005. Immune evasion by
Staphylococci. Nature Reviews Microbiology 3: 948–958, Figure 3.]
Genomic analysis has classified complement components
into five gene families, each of which possesses unique
domain structures that have enabled investigators to trace
their phylogenetic origins. The first of these gene families
encodes C1q, mannose-binding lectin (MBL), and ficolins.
Genes for prototypical C1q molecules have been identified
in species as primitive as lampreys (the jawless fishlike vertebrate; Phylum Chordata, Subphylum Vertebrata), sea
urchins (Phylum Echinodermata, Subphylum Echinozoa),
and ascidians (Phylum Chordata, Subphylum Urochordata).
Analysis of C1q gene clusters in different species has demonstrated that C1q genes appeared prior to the generation of
immunoglobulin genes, and that at least some of these C1q
proteins bind to specific carbohydrates, implying that they
The Complement System
|
CHAPTER 6
217
BOX 6-3
the complement cascade. A second protein secreted by S. aureus, the Extracellular Fibrinogen-Binding protein (Efb),
binds C3, thus preventing its deposition
on the bacterial cell surface (2). But if the
immune system succeeds in depositing
C3, in spite of all these evasive actions by
the bacterium, S. aureus has evolved the
capacity to manufacture other proteins
to destroy it. Recently, a metalloproteinase enzyme, aureolysin, has been shown
to be secreted by S. aureus. Aureolysin
cleaves C3 in such a manner that it is then
further degraded and inactivated by host
regulatory proteins.
Other mechanisms interfere with the
ability of antibodies to initiate the complement cascade. Staphylococcal protein
A (SpA) is anchored into the carbohydrate cell wall, and possesses four or five
extracellular domains, each of which is
capable of binding to the Fc portion of a
molecule of IgG, thus effectively blocking effector functions of the antibody
molecules, including complement activation (3). In addition, S. aureus secretes
an enzyme called Staphylokinase (SAK)
that attaches to the bacterial surface and
binds to plasminogen, the host cell
plasma zymogen, activating it upon
binding (4). The activated plasminogen
has a serine protease activity that cleaves
and releases any surface-bound IgG and
C3b.
The S. aureus protein Sbi, which is
found both in association with the bacterial cell and in solution, contains two
immunoglobulin-binding domains and
two other domains that bind to the C3
complement component. The cellassociated Sbi has been demonstrated
to be capable of preventing Fc-mediated
protective functions such as opsonization and complement activation, by
blocking complement, or Fc receptor
binding, to the IgG constant regions. In
addition, the portion of the Sbi protein
that is released into the extracellular
fluid has also been shown to be protective for the bacterium. Although the
precise nature of the protective mechanism is still being investigated, it is
thought to result from the depletion of
complement components in the vicinity of the bacterium.
S. aureus has also evolved mechanisms that take advantage of host complement control strategies. A bacterial
cell-surface protein, Clumping factor A
(ClfA) is able to bind factor I, thus localizing it to the bacterial surface, where it
cleaves and inactivates any C3b that has
bound there (5). Another complement
inhibitory protein secreted by S. aureus
acts in a slightly different way. The Chemotaxis Inhibitory Protein of S. aureus
(CHIPs) binds two chemotactic receptors
on neutrophils, the C5a receptor, and
are potentially able to discriminate between host and pathogen. Thus, C1q may have expressed the capacity to recognize
foreign molecules at a very early point in time, independent
of antibody binding and in a manner similar to that of MBL.
Mannose-binding lectins have been identified in lampreys,
thus placing the origin of the MBLs at least as far back as the
early vertebrates. Indeed, functional lectin pathways have
been characterized in ascidians. However, although genes
encoding MBL-like proteins have been characterized in
ascidian genomes, the nature of their relationship to vertebrate MBL proteins is still unclear.
The next three gene families to be considered all encode
proteins that are implicated in the cleavage of the C3, C4,
and C5 components of complement. The first of these fami-
the formylated peptide receptor, thus
effectively eliminating the capacity of
neutrophils to respond to C5a-mediated
chemotactic signals. Other Staphylococcal proteins are active in the inhibition of
neutrophil binding to the membrane
surfaces of the endothelial cells lining
blood capillaries, and thus retard the
extravasation of neutrophils in the vicinity of a local infection.
The variety and effectiveness of these
inhibitory mechanisms that have evolved
in just one strain of bacteria emphasize for
us the importance of the complement
system in the normal control of bacterial
infections, and the complexity that
researchers and clinicians face when
attempting to develop the next generation of antibacterial therapies.
Foster, T .J. 2005. Immune evasion by staphylococci. Nature Reviews Microbiology 3:948–
958.
Laarman, A., F. Milder, J. van Strijp, and S. Rooijakkers. 2010. Complement inhibition by
gram-positive pathogens: Molecular mechanisms and therapeutic implications. Journal of Molecular Medicine (Berlin) 88:115–120.
Serruto, D., R. Rappuoli, M. Scarselli, P. Gros, and
J. A. van Strijp. 2010. Molecular mechanisms
of complement evasion: Learning from
staphylococci and meningococci. Nature
Reviews Microbiology 8:393–399.
lies is composed of factor B (Bf) and the serine protease C2;
the second family comprises the serine proteases MASP-1,
MASP-2, MASP-3, C1r, and C1s; and the third family contains the C3 family members C3, C4, and C5, which are not
themselves proteases, but which share a common internal
thioester and are subject to protease cleavage and activation.
Studies of a number of invertebrate species have indicated
that the genome of each species contains only a single copy
representative of each of these three gene families. In contrast,
one or more gene duplication events have occurred within each
gene family in most vertebrates, suggesting that the gene duplications found in vertebrates probably occurred in the early
stages of jawed vertebrate evolution. Genes for each of the
Bf, C3, and MASP families have been identified from all
218
PA R T I I
|
Innate Immunity
Alternative pathway
Lectin pathway
Classical pathway
C3/C4/C5
Bf/C2
MASP/C1r,s
TCC (MAC)
Mammalia
Reptilia/Aves
Amphibia
Teleostei
Chondrichthyes Deuterostomes
Agnatha
Urochordata
Cephalochordata
Hemichordata
Echinodermata
Protstomia
Cnidaria
1000
500
Million years ago
0
FIGURE 6-17 Evolution of complement components. The appearance of each of the three major complement pathways is illustrated
by gray arrows, and the timing of the gene duplications that gave rise to the classical complement pathway is indicated by double-headed
arrows. [Adapted from Nonaka, M., and A. Kimura, 2006, Genomic view of the evolution of the complement system, Immunogenetics 58:701.]
TABLE 6-6
Complement system pathways in the major groups of deuterostome animals
Animal
group
Alternate
pathways
Classical
pathway
Lectin
pathway
TCC complex
Antibodies
present?
Mammals
⫹
⫹
⫹
⫹
⫹
Birds
⫹
⫹
⫹
⫹
⫹
Reptiles
⫹
⫹
⫹
⫹
⫹
Amphibians
⫹
⫹
⫹
⫹
⫹
Teleost fish
⫹
⫹
⫹
⫹
⫹
Cartilaginous fish
⫹
⫹
⫹
⫹
⫹
Agnathan fish
⫹
⫹
?
Tunicates
⫹
⫹
?
Echinoderms
⫹
?
?
After J. O. Sunyer et al., 2003. Evolution of complement as an effector system in innate and adaptive immunity. Immunologic Research 27:549–564, Figure 2.
The Complement System
invertebrate deuterostomes that have so far been analyzed,
with the single exception of the MASP family gene, which is
missing from the sea urchin (Phylum Echinodermata) genome.
The situation in protostomes is somewhat more complicated. Whereas members of the C3 and Bf families have been
identified in some very early protostomes—for example, the
horseshoe crab and the carpet-shell clam (Phyla: Arthropoda and Mollusca respectively)—analysis of whole genome
sequences of insects such as Drosophila melanogaster
(Arthropoda) or the nematode Caenorhabditis elegans
(Nematoda) have failed to locate any proto-complement
protein sequences. This contrasts starkly with the presence
of all three family genes in sea anemones and corals (Phylum, Cnidaria), and suggests that the ability to encode proteins of the complement system has been lost from the
genome of the common, model-system protostomes D.
melanogaster and C. elegans. The absence of genes encoding
any of the later, cytolytic components of the complement
cascade in protostomes and cnidarians supports the hypothesis that the proto-complement systems of cnidarians and
protostomes function by opsonization.
The proteins encoded by the fifth gene family, C6, C7,
C8, C8, and C9, together make up the terminal complement complex (TCC). They share a unique domain structure
that enables molecular phylogenetic analysis. In particular,
the mammalian TCCs, C6, C7, C8, C8, and C9 all share
the MAC/perforin (MACPF) domain, in addition to other
domains common among two or more members of this protein group. Mammals, birds, and amphibians share all TCC
genes (with the exception of the bird C9 gene, which is miss-
|
CHAPTER 6
219
ing from the chicken genome sequence). Although the full
complement of MAC proteins has not yet been documented
in cartilaginous fishes such as the sharks (the earliest known
organisms to possess an adaptive immune system), a gene
encoding a C8 subunit orthologous to mammalian C8
has been cloned and characterized. Furthermore, the serum
of sharks has been known for decades to express hemolytic
activity, and microscopic analysis of the pores formed on
target cell surfaces reveals a transmembrane pore structure
indistinguishable from that induced by a MAC attack. It thus
seems likely that the functionality of the MAC emerged not
long after the adaptive immune system.
Genomic analysis has traced the origin of the TCC genes
back to before the divergence of the urochordates, cephalochordates, and vertebrates (see Figure 6-17) as ascidians
(urochordates) and amphioxus (cephalochordates) possess
primitive copies of TCC genes. However, the ascidian and
amphioxus TCC proteins could not have functioned in a
manner similar to that of later vertebrate species because
they lack the domain responsible for interacting with the C5
protein. Intriguingly, toxins of the venomous sea anemone,
which express a very high hemolytic potential, are closest in
structure to the TCCs of the complement pathway.
In summary, complement evolution is based on the diversification and successive duplications of members of five
gene families that evolved first in response to the need for
microbial recognition and opsonization and then responded
to the appearance of the adaptive immune system by additional gene duplication and functional adaptation to form
the complement system as we know it today.
S U M M A R Y
■
■
■
■
■
■
The complement system comprises a group of serum proteins, many of which circulate in inactive forms that must
first be cleaved or undergo conformational changes prior
to activation.
Complement proteins include initiator molecules, enzymatic mediators, membrane-binding components or opsonins, inflammatory mediators, membrane attack proteins,
complement receptor proteins, and regulatory components.
Complement activation occurs by three pathways—classical,
lectin, and alternative—which converge in a common
sequence of events leading to membrane lysis.
The classical pathway is initiated by antibodies of the IgM
or IgG classes binding to a multivalent antigen. This allows
binding of the first component of complement, C1q, and
begins the process of complement deposition.
The lectin pathway is initiated by binding of lectins such as
mannose-binding lectin or members of the ficolin family
to microbial surface carbohydrates.
The alternative pathway is initiated when the third component of complement, C3, undergoes nonspecific break-
down. C3b is formed and adheres to a cell surface.
Inadvertent initiation is prevented by the presence of control proteins on host cell membranes.
■
■
■
■
■
The end result of the initiating sequence of all three pathways is the generation of enzymes that cleave C3 into C3a
and C3b, and C5 into C5a and C5b.
C3b opsonizes microbial cells and immune complexes,
rendering them suitable for phagocytosis.
Activation of the terminal components of the complement
cascade C5b, C6, C7, C8 and C9 results in the deposition
of a membrane attack complex (MAC) onto the microbial
cell membrane. This complex introduces large pores in the
membrane, preventing it from maintaining osmotic integrity and resulting in the death of the cell.
Binding of the complement component C1q to apoptotic
cells, apoptotic bodies, and immune complexes results in
their opsonization with C3b and subsequent phagocytosis.
Patients deficient in the early complement components do
not clear apoptotic cells efficiently and suffer disproportionally from autoimmune disease such as SLE.
220
■
■
■
PA R T I I
|
Innate Immunity
In addition to acting during host defense against infection,
complement proteins also bind to receptors on the surfaces
of antigen-presenting cells, T cells, and B cells, inducing
interleukin production and aiding in their activation.
A system of regulatory complement proteins ensures that
inadvertent complement activation on the surface of host
cells is prevented and controlled, by deactivating complement components on the surface of host cells, and ensuring that regulatory proteins are bound specifically to host,
but not to microbial cell membranes.
Patients suffering from complement deficiencies often
present with immune complex disorders and suffer disproportionally from infections by encapsulated bacteria
such as Neisseria meningitidis. Animal models exist for
most complement deficiencies.
■
■
Underscoring the importance of complement in the
immune system, a broad variety of complement evasion
strategies has evolved in viruses, bacteria, fungi, and
parasites, including mimicking regulatory proteins, interfering with the interactions between antibodies and
complement components, or between proteins of the
complement pathways, or by destruction of the complement components.
The genes encoding complement components belong to
five families. Genes of the alternative pathway components appear first in evolution, and those encoding the
terminal complement components appear last. Thus,
prior to the emergence of adaptive immunity, complement served its protective functions by mediating
phagocytosis.
R E F E R E N C E S
Alexander, J. J., A. J. Anderson, S. R. Barnum, B. Stevens, B., and
A. J. Tenner. 2008. The complement cascade: Yin-Yang in
neuroinflammation—neuro-protection and -degeneration.
Journal of Neurochemistry 107:1169–1187.
Al-Sharif, W. Z., J. O. Sunyer, J. D. Lambris, and L. C. Smith.
1998. Sea urchin coelomocytes specifically express a homologue of the complement component C3. Journal of Immunology 160:2983–2997.
Azumi, K., et al. 2003. Genomic analysis of immunity in a Urochordate and the emergence of the vertebrate immune system:
“Waiting for Godot.” Immunogenetics 55:570–581.
Behnsen, J., et al. 2010. Secreted Aspergillus fumigatus protease
Alp1 degrades human complement proteins C3, C4, and C5.
Infection and Immunity 78:3585–3594.
Blom, A. M., T. Hallstrom, and K. Riesbeck, K. 2009. Complement evasion strategies of pathogens—acquisition of inhibitors and beyond. Molecular Immunology 46:2808–2817.
Botto, M., et al. 1998. Homozygous C1q deficiency causes glomerulonephritis associated with multiple apoptotic bodies.
Nature Genetics 19:56–59.
Kang, Y. S., et al. 2006. A dominant complement fixation pathway for pneumococcal polysaccharides initiated by SIGN-R1
interacting with C1q. Cell 125:47–58.
Kemper, C., and J. P. Atkinson. 2007. T-cell regulation: With
complements from innate immunity. Nature Reviews Immunology 7:9–18.
Kemper, C., and D. E. Hourcade. 2008. Properdin: New roles in
pattern recognition and target clearance. Molecular Immunology 45:4048–4056.
Lambris, J. D., D. Ricklin, and B. V. Geisbrecht. 2008. Complement evasion by human pathogens. Nature Reviews Microbiology 6:132–142.
Litvack, M. L. and N. Palaniyar. 2010. Soluble innate immune
pattern-recognition proteins for clearing dying cells and cellular components: Implications on exacerbating or resolving
inflammation. Innate Immunity 16:191–200.
Longhi, M. P., C. L. Harris, B. P. Morgan, and A. Gallimore, A.
2006. Holding T cells in check: A new role for complement
regulators? Trends in Immunology 27:102–108.
Dunkelberger, J. R., and W. C. Song. 2010. Complement and its
role in innate and adaptive immune responses. Cell Research
20:34–50.
Markiewski, M. M., B. Nilsson, K. N. Ekdahl, T. E. Mollnes,
and J. D. Lambris. 2007. Complement and coagulation:
Strangers or partners in crime? Trends in Immunology
28:184–192.
Elward, K., et al. 2005. CD46 plays a key role in tailoring innate
immune recognition of apoptotic and necrotic cells. Journal of
Biological Chemistry 280:36342–36354.
Miller, D. J., et al. 2007. The innate immune repertoire in Cnidaria: Ancestral complexity and stochastic gene loss. Genome
Biology 8:R59.
Fearon, D. T., and M. C. Carroll. 2000. Regulation of B lymphocyte responses to foreign and self-antigens by the CD19/CD21
complex. Annual Review of Immunology 18:393–422.
Nonaka, M., and A. Kimura. 2006. Genomic view of the evolution of the complement system. Immunogenetics 58:701–713.
Gros, P., F. J. Milder, and B. J. Janssen. 2008. Complement driven by
conformational changes. Nature Reviews Immunology 8:48–58.
Price, J. D., et al. 2005. Induction of a regulatory phenotype in
human CD4⫹ T cells by streptococcal M protein. Journal of
Immunology 175:677–684.
Jensen, J. A., E. Festa, D. S. Smith, and M. Cayer. 1981. The
complement system of the nurse shark: hemolytic and comparative characteristics. Science 214:566–569.
Ricklin, D., and J. D. Lambris. 2008. Compstatin: A complement
inhibitor on its way to clinical application. Advances in Experimental and Medical Biology 632:273–292.
The Complement System
|
CHAPTER 6
221
Roozendaal, R., and M. C. Carroll. 2007. Complement receptors
CD21 and CD35 in humoral immunity. Immunological
Reviews 219:157–166.
Zipfel, P. F., and C. Skerka. 2009. Complement regulators
and inhibitory proteins. Nature Reviews Immunology
9:729–740.
Rosen, A. M., and B. Stevens. 2010. The role of the classical complement cascade in synapse loss during development and glaucoma. Advances in Experimental and Medical Biology 703:75–93.
Useful Web Sites
Stevens, B., et al. 2007. The classical complement cascade mediates CNS synapse elimination. Cell 131:1164–1178.
Sunyer, J. O., et al. 2003. Evolution of complement as an effector
system in innate and adaptive immunity. Immunologic
Research 27:549–564.
Suzuki, M. M., N. Satoh, and M. Nonaka. 2002. C6-like and
C3-like molecules from the cephalochordate, amphioxus, suggest a cytolytic complement system in invertebrates. Journal of
Molecular Evolution 54:671–679.
Ward, P. A. 2009. Functions of C5a receptors. Journal of Molecular Medicine 87:375–378.
Zhu, Y., S. Thangamani, B. Ho, and J. L. Ding. 2005. The ancient
origin of the complement system. European Molecular Biology
Organization Journal 24:382–394.
S T U D Y
www.complement-genetics.uni-mainz.de The
Complement Genetics Homepage from the University of
Mainz gives chromosomal locations and information on
genetic deficiencies of complement proteins.
www.cehs.siu.edu/fix/medmicro/cfix.htm A clever
graphic representation of the basic assay for complement
activity using red blood cell lysis, from D. Fix at the University of Southern Illinois, Carbondale.
www.youtube.com/watch?vⴝy2ep6j5kHUc
www.youtube.com/watch?vⴝAIjaiJV4m2g There
are a number of animations of the complement cascade
available on the Internet. These links are to clear animations.
Q U E S T I O N S
1. Indicate whether each of the following statements is true or
5. Complement activation can occur via the classical, alterna-
false. If you think a statement is false, explain why.
tive, or lectin pathway.
a. A single molecule of bound IgM can activate the C1q
a. How do the three pathways differ in the substances that
b.
c.
d.
e.
f.
component of the classical complement pathway.
The enzymes that cleave C3 and C4 are referred to as
convertases.
C3a and C3b are fragments of C3 that are generated by
proteolytic cleavage mediated by two different enzyme
complexes.
Nucleated cells tend to be more resistant to complement-mediated lysis than red blood cells.
Enveloped viruses cannot be lysed by complement
because their outer envelopes are resistant to pore formation by the membrane attack complex (MAC).
MBL has a function in the lectin pathway analogous
to that of IgM in the classical pathway, and MASP-1
and MASP-2 take on functions analogous to C1
components.
2. Explain why serum IgM cannot activate complement prior
to antigen binding.
3. Genetic deficiencies have been described in patients for all
of the complement components except factor B. Particularly severe consequences result from a deficiency in C3.
Describe the consequences of an absence of C3 for each of
the following:
a. Initiation of the classical and alternate pathways
b. Clearance of immune complexes
c. Phagocytosis of infectious bacteria
4. Describe three ways in which complement acts to protect
the host during an infection.
can initiate activation?
b. Which parts of the overall activation sequence differ
among the three pathways, and which parts are similar?
c. How does the host ensure that inadvertent activation of
the alternative pathway on its own healthy cells does not
lead to autoimmune destruction?
6. Briefly explain the mechanism of action of the following
complement regulatory proteins. Indicate which pathway(s)
each protein regulates.
a.
b.
c.
d.
e.
f.
C1 inhibitor (C1INH)
C4b-binding protein (C4bBP)
Decay-accelerating factor (DAF)
Membrane cofactor of proteolysis protein, MCP (CD46)
Protectin (CD59)
Carboxypeptidase N
7. Explain why complement deficiencies in the early compo-
nents of complement give rise to immune-complexmediated disorders such as systemic lupus erythematosus.
8. You have prepared knockout mice with mutations in the
genes that encode various complement components. Each
knockout strain cannot express one of the complement
components listed across the top of the table below. Predict
the effect of each mutation on the steps in complement
activation and on the complement effector functions indicated in the table using the following symbols: NE
no
effect; D process/function decreased but not abolished;
A process/function abolished.
PA R T I I
222
|
Innate Immunity
Complement component knocked out
C1q
C4
C3
C5
C9
Factor B
MASP-2
Formation of classical pathway C3 convertase
Formation of alternative pathway C3 convertase
Formation of classical pathway C5 convertase
Formation of lectin pathway C3 convertase
C3b-mediated opsonization
Neutrophil chemotaxis and inflammation
Cell lysis
As shown in the figure below, two flow cytometric histograms
of red blood cells were obtained from a patient, stained with
antibodies toward CD59 (Protectin) (part A) and CD55 (Decay
Accelerating Factor or DAF) (part B). On the right of each histogram is a large population of cells expressing relatively high
levels of each antigen (population I). Populations II express
midrange levels of the two antigens, and populations III express
levels of antigen below statistical detectability. (Detecting laser
voltages on flow cytometers are normally set such that negative
staining populations show levels of fluorescence below 101 on
the lower logarithmic scale. Note also the creep of cells up the
100 axis on each plot.)
a. Can you offer an explanation as to why this patient may
b. Why do you think a single patient can generate red blood
have red blood cells expressing low levels of these two
particular antigens? From what disease do you think
this patient is suffering?
c. Now look at parts C and D of the same figure, which
CLINICAL FOCUS QUESTION
cells expressing three different levels of CD59 and CD55?
display the expression of two other markers on red
The Complement System
blood cells from the same patient. In these flow cytometric profiles, the investigators have shown you quadrant markers to indicate whether the cells they are
investigating are considered to be positive or negative
for the proteins labeled on the axes. For example, the
cell population high and right in part C is considered to
express both CD66abce and CD16, whereas the cell
population low and left is negative for both of these
markers.
Using your answer for part (a) as a starting point,
what can you deduce about the biochemistry of the
membrane proteins CD16, CD66abce, CD46, and CD14?
CHAPTER 6
a. In the figure below, a flow cytometric histogram
describes the numbers of human Jurkat T cells expressing low and high levels of CD46 after treatment with an
apoptosis-inducing reagent. On the figure, label the cell
population that you think is undergoing apoptosis and
explain your reasoning.
200
CD 46
Counts
160
120
80
40
0
100
101
102
FL2-H
103
104
223
On the flow cytometric dot plot shown below, indicate the following and explain your labeling:
a. Where you would expect to find a healthy T-cell
population?
b. Where you would expect to find a T-cell population
undergoing apoptosis?
C1q
ANALYZE THE DATA
|
CD46
This page lelt intentionally blank.
Site of DNA
cleavage
The Organization and
Expression of Lymphocyte
Receptor Genes
3'
T
o protect its host, the immune system must
recognize a vast array of rapidly evolving
microorganisms. To accomplish this, it must
generate a diverse and flexible repertoire of
receptor molecules, while minimizing the expression of
receptors that recognize self antigens. As described in
Chapter 3, each B or T lymphocyte expresses a unique
antigen-specific receptor. When these receptors bind to
their corresponding antigens under the appropriate
conditions (described in Chapters 11 and 12), T and B
lymphocytes proliferate and differentiate into effector
cells that eliminate the microbial threat (Chapter 13).
In Chapter 3, we described the biochemistry of the T
and B lymphocyte receptors and the secreted antibodies
formed by B lymphocytes following antigen stimulation.
We also outlined the experiments which demonstrated
that secreted antibodies are identical in antigen-binding
specificity to the B-cell receptors of the secreting cell. In
this chapter, we address the question of how an organism
can encode and express receptors capable of recognizing a
constantly evolving universe of microbial threats using a
finite amount of genetic information.
The production of specific lymphocyte receptors
employs a number of genetic mechanisms that are unique
to the immune system. In 1987, Susumu Tonegawa won
the Nobel Prize for Physiology or Medicine “for his
discovery of the genetic principle for generation of
antibody diversity,” a discovery that challenged the
fundamental concept that one gene encoded one
polypeptide chain. Tonegawa and his colleagues showed
that the antibody light chain was encoded in the germ line
by not one but three families of gene segments separated by
kilobases of DNA (Figure 7-1). (The germ-line DNA is the
genetic information encoded in the sperm and egg, which
can be passed on to future generations.) Their work
demonstrated that two DNA segments, one from each
family, are conjoined, only in B lymphocytes, to create the
mature form of the light-chain variable region of the
immunoglobulin (Ig) gene. A third segment encodes the
constant region of the gene. Different combinations of
RSS
Antigen receptor
gene segment
5'
7
12 bp
ACAAAAACC–3'
5 '– C A C A G T G
23 bp
Heptamer
Spacer
Nonamer
The Recombination Signal Sequence (RSS)
serves as the site of DNA cleavage. The RSS is
composed of conserved heptamer and
nonamer sequences, separated by a spacer
region of 12 or 23 bp, which is conserved in
length, but not in sequence. Cleavage occurs
at the junction of the heptamer and the
variable region coding segment. [Adapted from
D. G. Schatz and Y. Ji, 2011, Recombination centres
and the orchestration of V(D)J recombination,
Nature Reviews. Immunology 11:251–263.]
■
The Puzzle of Immunoglobulin Gene Structure
■
Multigene Organization of Ig Genes
■
The Mechanism of V(D)J Recombination
■
B-Cell Receptor Expression
■
T-Cell Receptor Genes and Expression
segments are used in each B cell, to create the diverse
repertoire of light-chain receptor genes. Subsequent
experiments have shown that all of the B- and T-cell
receptor genes are assembled from multiple gene
segments by similar rearrangements.
We describe below the unique genetic arrangements of
T- and B-cell receptor gene segments, and the
mechanisms by which they are rearranged and expressed.
We will address here only those mechanisms that shape
the receptor repertoires of mouse and human naïve B and
T cells, which have not yet been exposed to antigen.
225
PA R T I I I
226
5'
V
V
|
Adaptive Immunity: Antigen Receptors and MHC
V
3'
J
23
kb
5'
3'
V J
C
2.5
kb
C
(a) There are many variable regions,
but just a few constant regions.
V region
(thousands,
heavy and
light chain)
C region
(four, light
chain; or eight,
heavy chain)
(b) The same V region can be found
connected to different C regions.
V region
C region
FIGURE 7-1 The antibody light-chain gene encodes three
families of DNA segments. During B-cell development, one V segment and one J segment (which encode contiguous parts of the
light-chain variable region) join together with the C (constant) region
to form the gene for the antibody light chain. This gene rearrangement occurs in the DNA, prior to gene transcription into mRNA.
Additional layers of diversity are generated in B cells
following antigen exposure and those will be addressed in
Chapter 12. These powerful mechanisms include the
generation of antibodies of different heavy-chain classes,
each capable of mediating a discrete set of effector
functions, as well as the creation of modified antigenbinding regions by somatic hypermutation, followed by
antigen-driven selection. Both of these processes are
triggered in human and mouse B cells only after antigen
contact.
The Puzzle of Immunoglobulin
Gene Structure
The immune system relies on a vast array of B-cell receptors
(BCRs) that possess the ability to bind specifically to a correspondingly large number of potential pathogens. The first
indication of the immense size of the antibody repertoire was
provided by immunologists using synthetic molecules to
stimulate antibody production. They discovered that antibodies can discriminate between small synthetic molecules
differing in as little as the position of an amino or hydroxyl
group on a phenyl ring. If the immune system can discriminate between small molecules that it had presumably never
encountered during evolutionary selection and that differ in
such subtle ways, then, it was reasoned, the number of potential antibodies must be very large indeed. A series of experiments conducted in the late 1970s and early 1980s estimated
the number of different BCRs generated in a normal mouse
immune system to be at least 107, but we now know that estimate was many orders of magnitude too small.
Investigators trying to make sense of Ig genetics were also
faced with an additional puzzle: protein sequencing of mouse
and human antibody heavy and light chains revealed that the
first (amino terminal) 110 amino acids of antibody heavy and
light chains are extremely variable among different antibody
molecules. This region was therefore defined as the variable
(V) region of the antibody molecule. In contrast, the remainder of the light and heavy chains could be classified into one
FIGURE 7-2 Early sequencing studies indicated that the
light chain may be encoded in more than one segment.
(a) Many Ig variable regions in both heavy and light chains could be
found in association with few constant regions. (b) The same variable
region can be found in contiguity with several different heavy-chain
constant regions.
of only four sequences (light chain) or eight sequences (heavy
chain) (Figure 7-2a) and was, therefore, named the constant
(C) region. This raised an intriguing question: If each of 105
to 107 antibodies is encoded by a separate gene, how could
the constant region part of the gene remain constant in
sequence in the face of evolutionary drift?
Furthermore, antibodies could be found in which the
same antibody variable (V) region was associated with more
than one heavy-chain constant (C) region (Figure 7-2b),
lending further support to the possibility that the expression
of the variable and constant regions of each antibody chain
were independently controlled. (These additional constant
regions are generated by the antigen-induced process of class
switch recombination, discussed in Chapter 12. In the current context, the only point of relevance is the independence
with which the variable and constant regions appear to be
expressed.) It was rapidly becoming clear that the solution to
the antibody gene puzzle was more complex than had previously been imagined.
Investigators Proposed Two Early Theoretical
Models of Antibody Genetics
Classical germ-line theories of antibody diversity suggested
that the genetic information for each antibody is encoded, in
its entirety, within the germ-line genome. This means that
the genes encoding the entire sequence of every antibody
heavy or light chain that the animal could ever make must be
present in every cell. However, a quick calculation is sufficient to demonstrate that if there are 107 or more antibodies,
each of which requires approximately 2,000 nucleotides, this
would require massive expenditure of genetic information—
indeed, more DNA would be required to encode the receptors of the immune system than is available to the organism.
Although arguments were initially made that the dedication
of a considerable fraction of the genome to the immune system may represent a reasonable evolutionary strategy, it
The Organization and Expression of Lymphocyte Receptor Genes
became clear, as estimates of the size of the antibody repertoire were revised upward, that there simply was not enough
DNA to go around and new ideas must be found.
In 1965, William Dreyer and J. Claude Bennett proposed
that antibody heavy and light chains are each encoded in two
separate segments in the germ-line genome and that one of
each of the V region- and C region-encoding segments are
brought together in B-cell DNA to form complete antibody
heavy and light-chain genes. The idea that DNA in somatic
cells might engage in recombinatorial activity was revolutionary. However, germ-line theorists began to modify their
ideas to embrace this possibility.
Others suggested the equally innovative idea that the
number of variable region genes in the genome might be
extremely limited, and proposed the somatic hypermutation theory. According to this hypothesis, a limited number
of antibody genes is acted upon by unknown mutational
mechanisms in somatic cells to generate a diverse receptor
repertoire in mature B lymphocytes. This latter idea had the
advantage of explaining how a large repertoire of antibodies
could be generated from a relatively small number of genes,
but the disadvantage that such a process, like the somatic cell
gene recombination suggested by others, had never been
observed. There was additional argument over whether
mutation would occur prior to, or only after, antigen contact
and, therefore, whether mutation was responsible for generation of the so-called “primary repertoire” that exists prior
to antigen binding by the B cell.
|
CHAPTER 7
227
Heated debate continued between the proponents of
modified germ-line versus somatic mutation theories
throughout the early 1970s, until a seminal set of experiments revealed that both sides were correct. We now know
that each antibody molecule is encoded by multiple, germline, variable-region gene segments, which are rearranged
differently in each naïve immune cell to produce a diverse
primary receptor repertoire. These rearranged genes are
then further acted upon after antigen encounter by somatic
hypermutation and antigenic selection, resulting in an
expanded and exquisitely honed repertoire of antigenspecific B cells (see Chapter 12).
Breakthrough Experiments Revealed
That Multiple Gene Segments Encode
the Light Chain
From the mid 1970s until the mid 1980s, a small group of
brilliant immunologists completed a series of experiments
that fundamentally altered the way in which scientists think
about the genetics of immune receptor molecules. The first
breakthrough occurred when Susumu Tonegawa showed, as
Dreyer and Bennett had predicted, that multiple gene segments encode the antibody light chain. His achievement is
all the more impressive because the modern tools of molecular biology were not yet available. The experiment he performed with Nobumichi Hozumi is described in Classic
Experiment Box 7-1.
BOX 7-1
CLASSIC EXPERIMENT
Hozumi and Tonegawa’s Experiment: DNA Recombination Occurs
in immunoglobulin Genes in Somatic Cells
The
paradigm-shifting
experiment of Hozumi and Tonegawa was
designed to determine whether the DNA
encoding Ig light-chain constant and variable regions existed in separate segments
in non antibody-producing cells, such that
a single constant region gene could associate with different variable region genes in
different B cells. They hypothesized that
this might be the case, because of Ig amino
acid sequencing data showing that the
sequences of the constant regions of many
Ig light chains were identical. They reasoned that if multiple copies of a constant
region gene existed, then each copy would
be expected to accumulate silent or neutral mutations over time. The most likely
explanation for the finding of a single con-
stant region amino acid sequence arranged
in tandem with many different variable
regions was, therefore, that a single constant region gene segment was cooperating with multiple variable region gene
segments in different cells or situations to
generate the light-chain gene. However,
this notion was heretical to a generation of
scientists brought up on the concept that
one gene encodes one polypeptide chain.
Scientists are accustomed to thinking
about the concept of alternative RNA
splicing, wherein different proteins may
be encoded by the same piece of DNA by
differentially using particular RNA segments, cut and spliced following transcription of a long precursor RNA
transcript. However, this experiment
asked an entirely new question: Can a
piece of DNA change its place on a chromosome in a somatic cell?
To serve as their source of germ-line
DNA, Hozumi and Tonegawa used DNA
from an organ in which Ig genes would not
be expressed, embryonic liver in this case.
(Sperm and egg DNA would have been
much more difficult to obtain.) For B-cell
DNA, they used DNA from an antibodyproducing plasma cell tumor line, MOPC
321, which secretes fully functional light
chains. They separately cut both sets of DNA
with the same restriction enzyme and used
radioactive probes to determine the sizes of
the DNA fragments on which the variable
and constant regions of the light-chain
gene were found in the two sets of cells.
(continued)
228
PA R T I I I
CLASSIC EXPERIMENT
|
Adaptive Immunity: Antigen Receptors and MHC
(continued)
Below, we describe how Hozumi and
Tonegawa conducted their experiment,
and indicate how the experimental protocol would be modified using the reagents
and techniques available to a modern
molecular biologist. We then display their
data as it would appear both in the modern gel format as well as in its original form.
Hozumi and Tonegawa:
a. Purified genomic DNA from embryonic liver cells and from the MOPC
321 tumor cell line.
b. Cut the two sets of genomic DNA
with a restriction endonuclease
(BamHI) and separated the DNA
fragments using electrophoresis.
(Today we would use a polyacrylamide gel and electrophorese submicrogram quantities of samples
over the course of a few hours;
Hozumi and Tonegawa used a footlong agarose gel that needed 2
liters of agarose. They loaded 5 mg
of DNA and ran the gel for 3 days.)
c. Made 125I-labeled nucleic acid
probes specific for the two regions
of mouse light chain. One of
Tonegawa’s probes was a full-length
radiolabeled piece of mRNA encoding the entire -chain sequence. The
other probe was the 3 half of the
sequence, which would hybridize to
the constant, but not to the variable
region of the -chain gene. Today,
making enzymatically labeled, stable
DNA probes for any particular
sequence is a safe and relatively
straightforward task. In Hozumi and
Tonegawa’s day, this task was significantly more challenging.
d. Probed the nuclease-generated
DNA fragments to determine the
size of the fragments carrying the
variable and constant region
sequences. We would normally use
a Southern blot procedure, electrophoresing the fragments, blotting
the gel with nitrocellulose paper,
and then probing the paper with
enzyme-labeled fragments complementary to the sequences of
interest, developing the blots with
Antibody-producing
tumor cells from same
strain of mouse
Embryonic liver cells
(Germline DNA)
Extract DNA from cells
Cut DNA with Restriction Endonuclease
Gel electrophoresis
1
2
3
4
Blot onto nitrocellulose paper and
probe with radioactive sequences
complementary to antibody
variable and constant regions
Germline DNA
C-region
probe
1
DNA from antibody producing cells
Probe complementary
to whole light chain
2
C-region
probe
3
Probe complementary
to whole light chain
4
FIGURE 1
The light-chain gene is formed by DNA recombination between variable and constant
region gene segments. DNA from embryonic liver cells was used as a source of germ-line DNA, and
DNA from a B-cell tumor cell line was used as an example of DNA from antibody-producing cells. In
embryonic liver, the DNA sequences encoding the variable and constant regions, respectively, were
located on different restriction endonuclease fragments. However, these two sequences were colocated on a single restriction fragment in the myeloma DNA. (See text for details of the experiment.)
luminescent or fluorescent substrates. Hozumi and Tonegawa’s
approach was much more timeconsuming. They cut the gel into
about 30 slices, melted the agarose,
and separately eluted the DNA
from each slice. To each DNA sample, they added radiolabeled RNA,
allowed it to anneal, and removed
the un-annealed RNA using RNase.
The radioactivity remaining in each
fraction was then plotted against
the size of the DNA in the slice.
Figure 1 shows the results that would
be obtained from a modern-day Southern
The Organization and Expression of Lymphocyte Receptor Genes
|
CHAPTER 7
229
BOX 7-1
blot of Tonegawa’s fragments. The germline DNA blot probed with the constant
region mRNA sequence probe (lane 1)
shows only a single band. This indicates
that the restriction endonuclease is not
cutting in the middle of the light-chain
constant region, but rather that the whole
constant region sequence is encoded
within a single fragment.
In contrast, probing the blot with the
whole light-chain sequence (V and C,
lane 2) yields two bands. Since the difference between lanes 1 and 2 is the presence of the variable region sequence in
the probe used for lane 2, the presence of
two bands in lane 2 indicates that the
restriction endonuclease used to generate
the DNA fragments is cutting somewhere
in between the variable and constant
regions in the germ-line DNA, such that
the variable region lies on a DNA fragment distinct from that bearing the constant region.
Analyzing the blots from the plasma
cell tumor, we first note that the large
fragment containing the constant region
DNA and the midsized fragment bearing
the variable region DNA have disappeared. Both the variable and constant
region gene segments are now located
on smaller fragments. This implies that a
new pair of restriction endonuclease sites
now forms the boundaries of each of the
constant and variable region fragments.
Right away, we can tell that the DNA environment around the light-chain genes
changes as the B cell differentiates.
Next, we note that the sizes of the
DNA fragments on which the constant
and variable region gene segments are
located are apparently the same. This
implies, although it does not yet prove,
that the movement of the constant and/
or variable region gene segments has
brought them into close proximity with
one another, such that the constant and
variable region gene segments co-locate
on the same fragment. An alternative
explanation is that they have both altered
their locations and the similarity in the
size of the fragments is coincidental.
DNA sequencing supported the former
FIGURE 2
The original data from Hozumi and Tonegawa’s classic experiment
proving immunoglobulin gene recombination occurs in B cells.
DNA from the sources shown was run out on an agarose gel for 3 days at
4C. The gel was sliced, the DNA eluted, and then each sample was hybridized with radiolabeled probes for either the constant region of the light
chain, or the whole chain. The plot shows the amount of radioactivity in
each fraction as a function of the migration distance, which reflects the
molecular weight of the DNA fragment. (See text for details.) [ From N.
Hozumi and S. Tonegawa, 1976, Evidence for somatic rearrangement of immunoglobulin genes coding for variable and constant regions, Proceedings of the National
Academy of Sciences USA 73:3628–3632.]
interpretation: as the B cell differentiates,
the variable and constant region gene segments are moved from distant regions of the
chromosome into close apposition with one
another.
It is all too easy, with modern molecular biology technologies, to forget what a
tour de force this experiment actually
represented. Few of the reagents or pieces
of apparatus we commonly encounter in
the molecular biology laboratory today
were available to Hozumi and Tonegawa—
they had to make their own.
The original paper speaks of purifying
their own restriction endonuclease
(BamHI) from a bacterial sample obtained
from a colleague, and of performing the
electrophoresis at 4oC for 3 days. Furthermore, technical difficulties precluded their
being able to make a good 5 (V-region)
probe. They therefore probed each sample
with a 3 (C-region) probe, as well as one
that bound to the whole light chain, and
then inferred 5 (V-region) binding by
subtraction. It was an extraordinary piece
of work. Their original data are shown in
Figure 2.
What happened to the DNA on the
alternative allele that did not encode the
tumor cell secreted light chain? Subsequent analysis of the DNA from this tumor
showed that the DNA from both chromosomes had undergone rearrangement.
Hozumi and Tonegawa were fortunate
that the variable regions used by both
rearrangements were close to one another
and so the fragment patterns overlapped.
Hozumi, N., and S. Tonegawa. 1976. Evidence
for somatic rearrangement of Ig genes
coding for variable and constant regions.
Proceedings of the National Academy of
Sciences USA 73:3628–3632.
230
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
Tonegawa’s experiment showed that, in the DNA from
non-antibody-producing, embryonic liver cells, there is a
BamHI endonuclease site between the variable and constant
regions. We know this because probes for the variable and
constant regions each recognized a different DNA fragment
in BamHI-digested germ-line DNA. However, in the antibodyproducing tumor cell, the DNA encoding the variable and
constant regions appeared to be combined in just one fragment, thus demonstrating that DNA rearrangement must
have occurred during the formation of an antibody lightchain gene (see Classic Experiment Box 7-1 for further
details).
Although this experiment demonstrated that the V and C
regions of antibody genes were located in different contexts
in the DNA of non-antibody-producing cells, it did not
speak to their relative locations; indeed, the initial experiment did not rule out the possibility that the V and C fragments could be encoded on different chromosomes in the
embryonic cells. However, sequencing experiments subsequently showed that the segments encoding the V and C
genes of the light chains are on the same chromosome and
that, in non B cells, the two segments are separated by a long
non-coding DNA sequence.
The impact of this result on the biological community
was profound. For the first time, DNA was shown to be cut
and rejoined during the process of cell differentiation. This
experiment not only provided the experimental proof of
Dreyer and Bennett’s prediction; it also paved the way for the
next surprising finding.
Tonegawa’s experiment had identified the V and C segments encoding the kappa light chain. However, when scientists in his group sequenced the antibody light-chain DNA,
they encountered another unexpected result. As expected,
the embryonic (unrearranged) V region segment had, at its
5 terminus, a short hydrophobic leader sequence, a common feature of membrane proteins necessary to guide the
nascent protein chain into the membrane (Figure 7-3a). A 93
bp sequence of non-coding DNA separated the leader
sequence from a long stretch of DNA that encoded the first
97 amino acids of the V region. But the light-chain V region
domain is approximately 110 amino acids long. Where was
the coding information for the remaining 13 amino acids?
Sequencing of the light-chain constant region fragment
from embryonic DNA provided the answer. Upstream from
the constant region coding sequence, and separated from it
by a non-coding DNA segment of 1250 bp, were the 39 bp
encoding the remaining 13 amino acids of the V region. This
additional light-chain coding segment was named the joining (J) gene segment (shown in red in Figure 7-3a). Further
sequencing of mouse and human light-chain variable and
constant region genes confirmed Tonegawa’s second, astonishing finding. Not only are the variable and constant
regions encoded in two separate segments, but the lightchain variable regions themselves are encoded in two separate gene segments, the V and J segments, that are made
contiguous only in antibody-producing cells.
The heavy-chain variable region gene was then shown to
require yet a third genetic segment. Adopting a similar
strategy of cloning and sequencing Ig heavy-chain genes,
Lee Hood’s group identified a germ-line Heavy-chain Variable Region (VH) gene fragment that encoded amino acid
residues 1–101 of the antibody heavy chain and a second
fragment that included a Heavy-chain Joining gene segment (JH) that determined the sequence of amino acid residues 107–123. Neither of these segments contained the
DNA sequence necessary to encode residues 102–106 of the
heavy chain. Significantly, these missing residues were
included in the third complementarity-determining region
of the antibody heavy chain, CDR3, which provides contact
residues in the binding of most antigens (see Chapter 3).
Gene segments encoding this part of the antibody heavy
chain were eventually located 5 of the J region in mouse
embryonic DNA by Hood and colleagues (Figure 7-3b). The
importance of the contribution made by this gene segment
to the diversity of antibody specificities is denoted by its
name; the Diversity (D) region. (Because there is no D
region in the light chain, immunologists usually drop the
subscript denoting the heavy chain.)
Thus, the variable region of the heavy chain of the antibody molecule is encoded by three discrete gene segments,
(a) Light chain V region gene segments in embryo (germline DNA)
5'
Leader
V
J
Downstream to
C region gene
segment
3'
V segment
(amino acids 1-97)
J segment
(amino acids 98-110)
(b) Heavy chain V region gene segments in embryo (germline DNA)
5'
Leader
V
D
V segment
(amino acids 1-101)
D segment
(amino acids 102-106
approx.)
J
3'
J segment
(amino acids 107-123
approx.)
Downstream to
C region gene
segments
FIGURE 7-3 Antibody light chains are encoded in two segments—V and J, whereas antibody heavy chains are encoded in
three segments—V, D, and J.
|
The Organization and Expression of Lymphocyte Receptor Genes
TABLE 7-1
231
Kappa Light-Chain Genes Include
V, J, and C Segments
Chromosomal locations of
immunoglobulin genes in humans
and mice
Human
chromosome
Mouse chromosome
light chain
22
16
light chain
2
6
Heavy chain
14
12
Gene
CHAPTER 7
The mouse Ig locus spans 3.5 Mb and includes 91 potentially
functional V genes, which have been grouped into 18 V gene
families based on sequence homology (Figure 7-4a). (If two
sequences are greater than 80% identical, they are classified as
belonging to the same V gene family.) These V gene families
can be further grouped into V gene clans, based again on
sequence homology. Each V segment includes the leader
exon encoding the signal peptide. Individual V segments are
separated by non-coding gaps of 5 to 100 kb. The transcriptional orientation of particular V segments may be in the
same or in the opposite direction as the constant region segment. The relative orientations of the variable and constant
region segments do not affect the use of the segments, but do
alter some details of the recombinational mechanism that creates the complete light-chain gene, as will be discussed later.
Downstream of the V region cluster are four functional
J segments and one pseudogene, or other nonfunctional
open reading frame (Table 7-2). A similar arrangement is
found in the human V locus, although the numbers of V
and J gene segments vary. A single C segment is found
downstream of the J region, and all kappa light-chain constant regions are encoded by this segment.
and the variable region of the light chain by two segments, in
the germ-line genome. These segments are brought together
by a process of DNA recombination that occurs only in the
B lymphocyte lineage to create the complete variable region
gene. Furthermore, the DNA at the junction between the
V and J segments for light chains, and at the VD and DJ
junctions in heavy chains, accounts for the extraordinary
diversity that was first observed by Kabat and Wu in the
CDR3 regions of both heavy and light chains (see Chapter 3).
Below, we will describe how further genetic diversity is generated at these junctions by additional processes unique to
immune system genetics.
Multigene Organization of Ig Genes
Lambda Light-Chain Genes Pair Each
J Segment with a Particular C Segment
Recall that Ig proteins consist of two identical heavy chains
and two identical light chains (see Chapter 3). The light chains
can be either kappa () light chains or lambda () light
chains. The heavy-chain, kappa, and lambda gene families are
each encoded on separate chromosomes (Table 7-1).
The mouse Ig locus spans a region of approximately 190 kb.
Lambda light chains are only found in 5% of mouse antibodies
because of a deletional event in the mouse genome that has
eliminated most of the lambda light-chain variable region
segments. It was therefore not surprising to discover that
(a) κ-chain DNA
5'
3'
L Vκ1
L Vκ(n)
L Vκ2
Cκ
Jκ
ψ
23
kb
2.5
kb
(b) λ-chain DNA
5'
3'
L Vλ2
L Vλ3
Jλ2
Cλ2
Jλ4
Jλ3
L Vλ1
Cλ4
Cλ3
Jλ1
Cλ1
ψ
19
kb
55
kb
1.2
kb
2.0 1.3
kb kb
19
kb
1.4
kb
1.7 1.3
kb kb
Cγ3
Cγ 1
(c) Heavy-chain DNA
5'
3'
L VH1
L VH(n)
DH1 DH14 JH1
JH4
6.5
kb
Cμ
Cδ
4.5
kb
55
kb
34
kb
Cγ2b
21
kb
Cγ2a
15
kb
Cε
14
kb
Cα
12
kb
FIGURE 7-4 Organization of immunoglobulin germ-line gene segments in the mouse. The (a) light chain and (b) light chains
are encoded by V, J, and C gene segments. The (c) heavy chain is encoded by V, D, J, and C gene segments. The distances in kilobases (kb)
separating the various gene segments in mouse germ-line DNA are shown below each diagram.
232
TABLE 7-2
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
Immunoglobulin variable region gene numbers in humans and mice
Human
Mouse
VK
34–48 functional; 8 ORFs; 30 pseudogenes.
91 functional; 9 ORFs; 60 pseudogenes
JK
5 functional; multiple alleles
4 functional; 1 ORF
V
33 functional; 6 ORFs; 36 pseudogenes
3–8 functional
J
5 functional; 2 ORFs
4 functional; 1 pseudogene
VH
38–44 functional; 4 ORFs; 79 pseudogenes
101 functional; 69 pseudogenes*
D
23 functional; 4 ORFs
19 functional; 7 ORFs; 6 pseudogenes
JH
6 functional; 3 pseudogenes; several different alleles found
4 functional; again, multiple haplotypes
* Gene numbers of the mouse VH locus refer to the sequenced chromosome 14 of the C57 BL/6 mouse.
A germ-line gene is considered to be functional if the coding region has an open reading frame (ORF) without a stop codon, and if there is no described defect in the
splicing sites, RSS, and/or regulatory elements.
A germ-line entity is considered to be an ORF if the coding region has an open reading frame but alterations have been described in the splicing sites, RSSs, and/
or regulatory elements, and/or changes of conserved amino acids have been suggested by the investigators to lead to incorrect folding and/or the entity is an orphon,
a nonfunctional gene located outside the main chromosomal locus.
A germ-line entity is considered to be a pseudogene if the coding region has a stop codon(s) and/or a frameshift mutation. In the case of V gene segments, these
mutations may be either in the V gene coding sequence, or in the leader sequence.
SOURCE: Gene numbers and definitions summarized from the International Immunogenetics Information System Web site: http://imgt.org. Accessed November 16, 2011.
there are only three fully functional V gene segments,
although this number varies somewhat by strain (see Table
7-2). Interestingly, there are also three fully functional
chain constant regions, each one associated with its own J
region segment (Figure 7-4b). The J-C4 segments are not
expressed because of a splice site defect.
Since recombination of Ig gene segments always occurs in
the downstream direction (V to J), the location of the V1
variable region sequence upstream of the JC3 and JC1 segments but downstream from JC2 means that V1 is always
expressed with either JC3 or JC1 but never with JC2. For
the same reason, V2 is usually associated with the JC2
pairing, although occasional pairings of V2 with the 190 kb
distant JC1 segments have been observed.
In humans, 40% of light chains are of the type and
about 30 V-chain gene segments are used in mature antibody light chains (see Table 7-2). Downstream from the
human V locus is a series of seven JC pairs, of which four
pairs are fully functional.
mouse strains). Just 0.7 kb downstream of the most 3 D segment is the JH region cluster, which contains four functional
JH regions. A further gap separates the last JH segment from
the first constant region exon, C1.
The human VH locus has a similar arrangement with
approximately 30 functional D segments, and 6 functional JH
segments. Human D regions can be read in all three reading
frames, whereas mouse D regions are mainly read in reading
frame 1, because of the presence of stop codons in reading
frames 2 and 3 in most mouse D regions.
The eight constant regions of antibody heavy chains are
encoded in a span of 200 kb of DNA downstream from the
JH locus, as illustrated in Figure 7-4c. Recall that the constant
regions of antibodies determine their heavy-chain class and,
ultimately, their effector functions (see Chapter 13).
Heavy-Chain Gene Organization Includes VH,
D, JH, and CH Segments
In V(D)J recombination, the DNA encoding a complete
antibody V-region is assembled from V, D, and J (heavy
chain) or V and J (light chain) segments that are initially
separated by many kilobases of DNA. Each developing B cell
generates a novel pair of variable region genes by recombination at the level of genomic DNA. Recombination is catalyzed by a set of enzymes, many of which are also involved
in DNA repair functions (Table 7-3), and is directed to the
appropriate sites on the Ig gene by recognition of specific
DNA sequence motifs called Recombination Signal
Sequences (RSSs). These sequences ensure that one of each
Multiple VH gene segments lie across a region of approximately 3.0 Mb in both mice and humans (Figure 7-4c).
These segments can be classified, like V segments, into
families of homologous sequences. Humans express at least
38 functional VH segments and mice approximately 101 (see
Table 7-2). Downstream from the cluster of mouse VH
region segments is an 80 kb region containing approximately
14 D regions (the actual number varies among different
The Mechanism of V(D)J
Recombination
The Organization and Expression of Lymphocyte Receptor Genes
CHAPTER 7
233
type of segment (V and J for the light chain, or V, D, and J
for the heavy chain) is included in the recombined variable
region gene. During cleavage and ligation of the segments,
the DNA is edited in various ways, adding further variability
to the recombined gene.
Proteins involved in V(D)J
recombination
TABLE 7-3
|
Protein
Function
RAG-1/2
Lymphoid-specific complex of two proteins that catalyze DNA strand breakage
and rejoin to form signal and coding joints
TdT
Lymphoid-specific protein that adds N
region nucleotides to the joints between
gene segments in the Ig heavy chain and
at all joints between TCR gene segments
HMG1/2 proteins
Stabilize binding of RAG1/2 to Recombination Signal Sequences (RSSs), particularly
to the 23-bp RSS; stabilize bend introduced into the 23-bp spacer DNA by the
RAG1/2 proteins
Ku70 and Ku80
heterodimers
Binds DNA coding and signal ends and
holds them in protein-DNA complex
DNA PKcs
In complex with Ku proteins, recruits and
phosphorylates Artemis
Artemis
Opens the coding end hairpins
XRCC4
Stabilizes and activates DNA ligase lV
DNA ligase lV
In complex with XRCC4, and Cernunnos
ligates DNA ends
Cernunnos
With XRCC4, activates DNA ligase lV
Recombination Is Directed by Signal Sequences
If recombination is to occur in the DNA of every lymphocyte, then a mechanism must exist to ensure that it only
occurs in antibody and T-cell receptor genes, and that the
moving DNA segments end up exactly where they should be
in the genome. Otherwise, dire consequences, including
malignancy, can ensue.
In the late 1970s, investigators working with light-chain
genes, described two blocks of conserved sequences—a
nonamer (a set of nine base pairs) and a heptamer (a set of
seven base pairs)—that are highly conserved and occur in
the noncoding regions upstream of each J segment. The heptamer appeared to end exactly at the J region coding
sequence. Further sequencing showed that the same motif
was repeated in an inverted manner on the downstream side
of the V region coding sequences, again with the heptamer
sequence ending flush with the V-region gene segment
(Figure 7-5a).
A further noteworthy feature of these sequences was the
presence of a spacer sequence of either 12 or 23 bp between
the heptamer and the nonamer. The significance of the
spacer lengths was clear; they represented one, or two, turns
(a) Nucleotide sequence of RSSs
CACAGTG
23 bp
ACAAAAACC
GGTTTTTGT
12 bp
CACTGTG
GTGTCAC
Heptamer
23 bp
TGTTTTTGG
Nonamer
CCAAAAACA
Nonamer
12 bp
GTGACAC
Heptamer
23 base pair spacer
12 base pair spacer
FIGURE 7-5 Two conserved sequences
(b) Location of RSSs in germ-line immunoglobulin DNA
κ-chain DNA
λ-chain DNA
Heavy-chain DNA
5'
L Vκ
3'
5'
12
5'
3'
23
L VH
Cκ
Jλ
Cλ
23
L Vλ
3'
Jκ
12
DH
23
12 12
JH
23
CH
in light-chain and heavy-chain DNA
function as recombination signal
sequences (RSSs). (a) Both signal sequences
consist of a conserved heptamer and conserved AT-rich nonamer; these are separated
by nonconserved spacers of 12 or 23 bp. (b)
The two types of RSS have characteristic
locations within -chain, -chain, and heavychain germ-line DNA. During DNA rearrangement, gene segments adjacent to the 12-bp
RSS can join only with segments adjacent to
the 23-bp RSS.
234
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
(a)
V cluster
Germline DNA
5'
1
3'
3
1
Rearrangement between V3 and J3
5'
3'
Recombined VJ
gene in B cell
2
1
Constant
region gene
J cluster
2
3 3
2
3
4
Constant
region gene
4
(b)
D cluster
(12-14 segments)
V cluster
Germline DNA
J cluster
(4 functional segments)
Constant
region gene
5'
3'
D-J rearrangement
5'
3'
Recombined VDJ
gene in B cell
5'
V-D rearrangement
3'
FIGURE 7-6 Recombination between gene segments is
required to generate a complete light chain gene. (a) Recombination between a V region (in this case, V3) and a J region (in this
case J3) generates a single VC light-chain gene in each B cell. The
recombinase enzymes recognize the RSS downstream of the V
region (orange triangle) and upstream of the J region (brown triangle). In every case, an RSS with a 12-bp (one-turn) spacer is paired
with an RSS with a 23-bp (two-turn) spacer. This ensures that there is
no inadvertent V-V or J-J joining. (b) Recombination of V (blue), D
(purple), and J (red) segments creates a complete heavy-chain variable region gene. Again, the recombinase enzyme recognizes the
RSS sequences downstream of the V region, up- and downstream of
the D region, and upstream of the J region, pairing 23-bp spacers
with 12-bp spacers.
of the double helix. Thus, the spacer sequence ensures that
the ends of the nonamer and heptamer closest to the spacers
would be on the same side of the double helix and, thus,
accessible to binding by the same enzyme. The investigators
correctly concluded that they had discovered the signal that
directs recombination between the V and J gene segments
and termed this heptamer 12/23 nonamer motif, the recombination signal sequence (RSS).
To summarize, the RSS consists of three elements:
sequence with a 12-bp spacer with one sequence with a
23-bp spacer, something we now know to be the case.
Figures 7-6a and 7-6b illustrate the generation of complete light-chain and heavy-chain genes from individual V
and J and V, D, and J segments, respectively.
•
Two proteins, encoded by closely linked genes, RAG1
(Recombination Activating Gene 1) and RAG2 (Recombination Activating Gene 2), were shown to be necessary
for recombining antibody genes. The RAG1 and RAG2
genes are encoded just 8 kilobases apart and are transcribed together. The expression of RAG1 and RAG2 is
developmentally regulated in both T and B cells (see Chapters 9 and 10) and, although RAG1 is expressed at all
phases of the cell cycle, RAG2 is stable only in G0- or
G1-phase cells. RAG1 is the predominant recombinase; it
forms a complex with the RSS that is stabilized by the binding of RAG2. RAG2 by itself does not exhibit detectable
RSS binding activity.
•
•
An absolutely conserved, 7-bp (heptamer) consensus
sequence 5-CACAGTG –3
A less conserved spacer of either 12 or 23 bp
A second conserved, 9-bp (nonamer) consensus
sequence 5-ACAAAAACC-3
In the heavy-chain gene segments, a similar pattern was
noted. The spacer regions separating the heptamer and
nonamer pairs were 23 bp in length following the V and
preceding the J segments and 12 bp in length before and
after the D segments. The relative locations of the 12 and 23
base pair spacers (Figure 7-5b) suggested that the VDJ
recombinase enzyme is designed to bring together one
Gene Segments Are Joined by the
RAG1/2 Recombinase
The Organization and Expression of Lymphocyte Receptor Genes
V
Binding of RAG1/2,
HMG proteins
Synapsis
|
CHAPTER 7
235
J
5'
3'
5'
3'
5'
3'
RAG1/2
HMG proteins
Cleavage and processing
of signal and coding
joints
5'
3'
5'
3'
Artemis
TdT
DNA ligase IV
NHEJ proteins
Generation of functional
Ig variable region gene
5'
3'
Coding joint
Signal joint
FIGURE 7-7 Overview of recombination of immunoglobulin variable region genes. The RAG1/2 complex (represented together
by the green oval) binds the RSSs and catalyzes recombination. Other enzymes fill in or cleave nucleotides at the coding end, and ligase completes the process. See text for details. [Adapted from Krangel, M. Nature Immunology 4, p. 625. 2003]
Only three of the proteins implicated in V(D)J recombination are unique to lymphocytes: RAG1, RAG2, and Terminal
deoxynucleotidyl Transferase (TdT), which is responsible
for the generation of additional diversity in the CDR3 region
of the antibody heavy chain (as we will see below). TdT is
expressed only in developing lymphocytes and adds untemplated “N” nucleotides to the free 3 termini of coding ends
following their cleavage by RAG1/2 recombinases.
Other enzymes participating in the recombination process are not lymphoid specific. Whereas binding of the RSS
by RAG1/2 can occur in the absence of any other proteins,
other cellular factors, most of which are part of the NonHomologous End Joining (NHEJ) pathway of DNA repair
are necessary for completion of V(D)J recombination. These
other non-lymphocyte-specific proteins known to participate in V(D)J joining are described in Table 7-3.
V(D)J Recombination Results in a Functional
Ig Variable Region Gene
The process of V(D)J recombination occurs in several
phases (Figure 7-7). The end product of each successful
rearrangement is an intact Ig gene, in which V and J (light
chains) or V, D, and J (heavy chains) segments are brought
together to create a complete heavy or light chain gene. The
new joints in the antibody V region gene, created by this
recombination process are referred to as coding joints. The
joints between the heptamers from the RSSs are referred to
as signal joints.
The first phase of this process, DNA recognition and
cleavage, is catalyzed by the RAG1/2 proteins. The second
phase, end processing and joining, requires a more complex
set of enzymatic activities in addition to RAG1/2, including
Artemis, TdT, DNA ligase lV, and other NHEJ proteins. The
individual steps involved in the process of recombination
between V and J segments are shown sequentially in
Figure 7-8.
Step 1 Recognition of the heptamer-nonamer Recombination Signal Sequence (RSS) by the RAG1/RAG2 enzyme
complex. The RAG1/2 recombinase forms a complex with
the heptamer-nonamer RSSs contiguous with the two gene
segments to be joined. Complex formation is initiated by
recognition of the nonamer RSS sequences by RAG1 and the
12-23 rule is followed during this binding.
Step 2 One-strand cleavage at the junction of the coding and
signal sequences. The RAG1/2 proteins then perform one of
their unique functions: the creation of a single-strand nick, 5 of
the heptameric signal sequence on the coding strand of each V
segment and a similar nick on the non-coding strand exactly at
the heptamer-J region junction. (Figure 7-8 shows this process
for the V segment only.)
Step 3 Formation of V and J region hairpins and blunt
signal ends. The free 3 hydroxyl group at the end of the coding strand of the VK segment now attacks the phosphate
group on the opposite, non-coding VK strand, forming a
new covalent bond across the double helix and yielding a
DNA hairpin structure on the V-segment side of the break
(coding end). Simultaneously, a blunt DNA end is formed at
Step 1. RAG1/2 and HMG
proteins bind to the RSS
and catalyze synapse
formation between a
V and a J gene segment.
V
J
5'
3'
1
Step 5. Opening of the
hairpin can result in a
5' overhang, a 3' overhang, or a blunt end.
5'
3'
V
2
3
Artemis
5'
V
5'
3'
3'
5'
3'
Step 2. RAG1/2 performs a single
stranded nick at the exact 5' border
of the heptameric RSSs bordering
both the V and the J segments.
5'
Opening at 1
yields 5' overhang
5'
3'
5'
3'
Coding end
5'
5'
3'
Step 6. Cleavage of
the hairpin generates
sites for P nucleotide
addition.
V
3'
5'
V
5'
3'
V
HO:
P
P
AG
AT
J
3'
5'
J
TCGA
3'
5'
ATAT
P
:OH
V
5'
3'
Signal end
OH
J
23 bp G T G A C A C
23 bp C A C T G T G
P
Nonamer
P
OH
P
OH
Sequence at the signal junction
results from the joining of the
two heptameric regions
Nonamer12 bp C A C T G T G C A C A G T G 23 bp Nonamer
Nonamer12 bp G T G A C A C G T G T C A C 23 bp Nonamer
TCGA
TATA
AGCT
ATAT
J
3'
5'
Ligation of completed
segments by DNA
Ligase IV and NHEJ
proteins
3'
Nonamer
3'
TA
Filling in of complementary
strands by DNA repair enzymes
Step 4. Ligation of the signal ends
P
G T G T C A C 12 bp
C A C A G T G 12 bp
OH
3'
5'
TC
3'
5'
V
5'
3'
Hairpin cleavage
by Artemis
5'
Step 7. Ligation of
light chain V and J
regions
5'
Opening at 3
yields 3' overhang
J
3'
Step 3. The hydroxyl group attacks
the phosphate group on the noncoding strand of the V segment to
yield a covalently-sealed hairpin
coding end and a blunt signal end.
3'
Opening at 2
yields blunt end
V
TCGATATA
AGCTATAT
J
3'
5'
FIGURE 7-8 The mechanism of V(D)J recombination. 1. The
RAG1/2 complex (pale green oval) binds to the RSSs and catalyzes
synapse formation. The coding (5 → 3) strand of DNA is drawn in a
thick line, the non-coding (3 → 5) strand in a thin line. For steps 1
to 5 we show only the events associated with the V region gene
segment, although the single-strand cleavage, hairpin formation, and
templated nucleotide addition simultaneously occur at the borders
of the V and the J segments. In this example, the V and J regions are
encoded in the same direction on the chromosome, and so the DNA
encoding the RSSs and the intervening DNA is released into the
nucleus as a circular episome and will be lost on cell division. The
DNA that was on the coding strand of the V region prior to rearrangement is emboldened. The actual signal joint is between the
residues that were in contiguity with the V and J regions, respectively.
Only the heptamer sequence is written out to preserve clarity.
Nucleotides encoded in the germ-line genome are shown in black; P
nucleotides are in blue, and non-templated nucleotides added by
TdT at heavy chain VD and DJ joints are shown in red. Steps 8, 9, and
10, shown on the facing page, only occur in heavy chain loci. (See
text for details.)
The Organization and Expression of Lymphocyte Receptor Genes
5'
D
J
TCGA
3'
3'
5'
ATAT
Step 8. In heavy chain VD and DJ joints
only: Exonuclease cleavage results in
loss of coding nucleotides at joint - can
occur on either or both sides of joint
5'
D
J
T
3'
3'
5'
TAT
Step 9. Non-templated nucleotides (in
red) are added to the coding joint by TDT
Step 10. Ligation of
heavy chain by DNA
ligase IV and NHEJ
proteins
5'
3'
D
TCGTCTATA
AGCAGATAT
J
3'
5'
the edge of the heptameric signal sequence. The same process occurs simultaneously on the J side of the incipient
joint. At this stage, the RAG1/2 proteins and HMG proteins
are still associated with the coding and signal ends of both
the V and J segments in a postcleavage complex.
Step 4 Ligation of the signal ends. DNA ligase IV then
ligates the free blunt ends to form the signal joint. The involvement of particular enzymes in this process was deduced from
observations of V(D)J recombination in natural and artificially
generated systems lacking one or more enzymes (see Table 7-3).
Step 5 Hairpin cleavage. The hairpins at the ends of the
V and J regions are now opened in one of three ways. The
identical bond that was formed by the reaction described in
step 3 above, may be reopened to create a blunt end at the
coding joint. Alternatively, the hairpin may be opened asymmetrically on the “top” or on the “bottom” strand, to yield a
5 or a 3 overhang, respectively. A 3 overhang is more common in in vivo experiments. Hairpin opening is catalyzed by
Artemis, a member of the NHEJ pathway.
Step 6 Overhang extension, leading to palindromic
nucleotides. In Ig light-chain rearrangements, the resulting
overhangs can act as substrates for extension DNA repair
enzymes, leading to double stranded palindromic (P)
nucleotides at the coding joint. For example, the top row of
bases in the V region in the 5 to 3 direction reads TCGA.
Reading backward on the bottom strand from the point of
ligation also yields TCGA. The palindromic nature of the
bases at this joint is a direct function of an asymmetric hairpin opening reaction. P nucleotide addition can also occur
at both the VD and DJ joints of the heavy-chain gene segments, but, as described below, other processes can intervene to add further diversity at the VH-D and D-JH junctions.
Step 7 Ligation of light-chain V and J Segments. Members
of the NHEJ pathway repair both the signal and the coding
joints, but the precise roles of each, and potentially other
enzymes in this process, have yet to be fully characterized.
|
CHAPTER 7
237
During B-cell development, Ig heavy-chain genes are
rearranged first, followed by the light-chain genes. This temporal dissociation of the two processes enables two additional diversifying mechanisms to act on heavy-chain V
region segments. The enzymes responsible for these mechanisms are usually turned off before light-chain rearrangements begin.
Comparative sequence analysis of germ-line and mature
B-cell Ig genes demonstrated that both loss of templated
nucleotides (nucleotides found in the germ-line DNA) and
addition of untemplated nucleotides (nucleotides not found
in germ-line DNA) could be identified in heavy-chain
sequences. Two distinct enzyme-catalyzed activities are
responsible for these findings.
Step 8 Exonuclease trimming. An exonuclease activity,
which has yet to be identified, trims back the edges of the
V region DNA joints. Since the RAG proteins themselves
can trim DNA near a 3 flap, it is possible that the RAG
proteins may cut off some of the lost nucleotides. Alternatively, Artemis has also been shown to have exonuclease, as
well as endonuclease activity, and could be the enzyme
responsible for the V(D)J-associated exonuclease function.
Exonuclease trimming does not necessarily occur in sets of
three nucleotides, and so can lead to out-of-phase joining.
V segment sequences in which trimming has caused the
loss of the correct reading frame for the chain cannot
encode antibody molecules, and such rearrangements are
said to be unproductive. As mentioned above, such exonuclease trimming is more common at the two heavychain V gene joints (V-D and D-J) than at the light-chain
V-J joint. In cases where trimming is extensive, it can lead
to the loss of the entire D region as well as the elimination
of any P nucleotides formed as a result of asymmetric hairpin cleavage.
Step 9 N nucleotide addition. Non-templated (N)
nucleotides are added by TdT to the coding joints of heavychain genes after hairpin cleavage. This enzyme can add up
to 20 nucleotides to each side of the joint. The two ends are
held together throughout this process by the enzyme complex, and again, loss of the correct phase may occur if
nucleotides are not added in the correct multiples of three
required to preserve the reading frame.
Step 10 Ligation and repair of the heavy-chain gene. This
occurs as for the light-chain genes.
In describing V(D)J recombination, investigators must
explain not only the mechanism of RSS recognition, cleavage, and ligation but also address the question of how two
RSSs, located many kilobases distant from one another in
the linear DNA sequence, are brought into close apposition.
Furthermore, once successful recombination has occurred
on one heavy-chain and one light-chain allele, this information must be communicated to the homologous chromosome, so that the other alleles can be silenced. How does
this occur?
Recent research indicates that both the structure and the
location of recombinationally active, Ig V region DNA
PA R T I I I
238
|
Adaptive Immunity: Antigen Receptors and MHC
change significantly as B-cell development proceeds, and
that some of these changes are signaled by epigenetic alterations in the chromatin structure, mediated by specific methylation reactions on associated histone residues. DNA close
to the nuclear membrane (pericentric DNA) does not
include recombinationally active V regions; rather, DNA
undergoing recombination moves away from the nuclear
membrane, toward the center of the nucleus. The structure
of the chromatin undergoing recombination also alters,
such that long stretches of DNA are condensed into loops
that allow the recombining sequences to come into closer
contact with one another. Once successful recombination
has occurred, the inactive allele has been shown to migrate
to the pericentric regions. Thus, the recombination process
is occurring within an actively modulating nucleoplasmic
context, and is controlled by enzymes that alter chromatin
structure, in addition to those that cleave and recombine
the DNA.
(a)
V1
V2
V3
J1
J2
J3
J2
J3
J2
J3
J2
J3
V2 J2
J3
5'
3'
Nicking and hairpin formation
V1
V2
V3
J1
5'
3'
Signal and coding joints form. Formation
of signal joint results in circularizing of
intervening DNA into an episome, that
is lost upon subsequent cell divisions.
V1 J2
J3
5'
J1
V2
3'
V3
(b)
V1
V2
V3
J1
5'
3'
Nicking and hairpin formation
V1
V2
V3
J1
5'
3'
Invert and ligate central fragment
V1
5'
Signal
joint
J1
V3
3'
Note that the signal joint remains
in the DNA upstream of the
recombined antibody gene.
FIGURE 7-9 Recombination can occur between DNA segments aligned in the same, or opposite, transcriptional
direction on the chromosome. (a) Recombination is shown
occurring between V1 and J2, which are both encoded in the same
transcriptional orientation, from left to right. The intervening DNA is
The direction of transcription is
now consistent between the
recombined V and J segments.
excised as a circular episome. (b) Recombination is shown occurring
between V2 and J2, which are encoded in opposite transcriptional
orientations. In this case, the DNA containing the signal joint remains
inverted in the DNA upstream of the recombined pair.
The Organization and Expression of Lymphocyte Receptor Genes
TABLE 7-4
|
CHAPTER 7
239
Combinatorial antibody diversity in humans
Nature of
segment
Number of heavy-chain
segments (estimated)
Number of -chain
segments (estimated)
Number of -chain
segments (estimated)
V
41
41
33
D
23
J
6
5
5
41 23 6 5658
41 5 205
30 5 165
Possible number of
combinations
Possible number of heavy-light chain combinations in the human 5658 (205
V(D)J Recombination Can Occur between
Segments Transcribed in Either the Same or
Opposite Directions
Sequencing of chromosomal fragments containing multiple
antibody V region genes showed that transcription of some V
region genes occurs in a direction opposite from that of
downstream D and J segments. However, work with artificial
recombinase substrates has demonstrated that the same general mechanism of recombination is used, regardless of the
transcriptional direction of the gene segments involved.
When recombination occurs between two gene segments that
are transcribed in the same direction, the intervening DNA is
deleted and lost (Figure 7-9a). When they are transcribed in
opposite directions, the DNA that was located between the V
and J segments is retained, in an inverted orientation, in the
DNA upstream of the rearranged VJ region (Figure 7-9b).
Five Mechanisms Generate Antibody Diversity
in Naïve B Cells
This description of the complex and sophisticated apparatus
by which Ig genes are created allows us to understand how
such an immensely diversified antibody repertoire can be
generated from a finite amount of genetic material. To summarize, the diversity of the naïve BCR repertoire is shaped
by the following mechanisms:
1.
2.
3.
4.
Multiple gene segments exist at heavy (V, D, and J) and
light-chain (V and J) loci. These can be combined with
one another to provide extensive combinatorial diversity
(Table 7-4).
P nucleotide addition results when the DNA hairpin at the
coding joint is cleaved asymmetrically. Filling in the singlestranded DNA piece resulting from this asymmetric
cleavage generates a short palindromic sequence.
Exonuclease trimming sometimes occurs at the VDJ and
VJ junctions, causing loss of nucleotides.
Non-templated N nucleotide addition in heavy chains
results from TdT activity. Mechanisms 2, 3, and 4 give rise
to extra diversity at the junctions between gene segments,
which contribute to CDR3.
5.
165) 2.09 106
In addition to these four mechanisms for generating antibody diversity that operate on individual
heavy- or light-chain variable segments, the combination of different heavy and light-chain pairs to form a
complete antibody molecule provides further opportunities for increasing the number of available antibody
combining sites (Table 7-4).
Combinatorial diversity: The same heavy chain can combine with different light chains, and vice versa.
These five mechanisms are responsible for the creation of
the diverse repertoire of BCRs and antibodies that is available to the organism before any contact with pathogens or
antigen has occurred. Following antigenic stimulation, B
cells are able to use yet another mechanism, unique to the
immune system, to further diversify and refine the antigenspecific receptors and antibodies: somatic hypermutation.
As described in Chapter 12, a specialized enzyme complex
targets the genes encoding the variable regions of Ig genes
only in those B cells that have undergone antigen-specific
activation in the presence of T-cell help. B cells are exposed
to successive cycles of mutation at the BCR loci, followed by
antigen-mediated selection in specialized regions of the
lymph node and spleen. The end result of this process is that
the average affinity of antigen-specific BCRs and antibodies
formed at the end of an immune response is considerably
higher than that at its instigation. This process is referred to
as affinity maturation.
This section has described the process of the generation
of the primary Ig variable region repertoire as it occurs in
humans and rodents. Although the same principles apply
to most vertebrate species, different species have evolved
their own variations. For example, the process of gene
conversion is used in chickens, and some species, such as
sheep and cows, use somatic hypermutation in the generation of the primary as well as the antigen-experienced
repertoire.
In Evolution Box 7-2, we describe the evolution of this
system of recombined lymphocyte receptors, addressing the
current hypothesis that the key event was the introduction of
the RAG1/2 gene segment into the early vertebrate genome
as a transposon.
240
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
EVOLUTION
Evolution of Recombined Lymphocyte Receptors
Scientists are slowly beginning to
•
•
•
The DNA-binding region of RAG1 is
strikingly homologous to that of
known transposable elements and is
evolutionarily related to members of
the Transib transposase family, currently expressed in species such as
fruit flies, mosquitoes, silkworm, and
the red flour beetle.
The inverted repeats in the RSSs of the
Ig and TCR gene segments are structurally similar to those found in other
transposons.
The mechanism of action of the RAG
proteins involves formation of a DNA
hairpin intermediate reminiscent of the
action of certain transposases. For
example, another phylogenetically con-
•
Current thinking supports the hypothesis that the appearance of the transposon in a primordial antigen receptor gene
might have separated the receptor gene
into two or more pieces. This hypothesis is
supported by the presence, in lower chordates, of genes that are related to BCR and
TCR V regions, and that therefore could
RAG1 and RAG2
VLR
TCR, BCR, and MHC
Adaptive immunity based on
rearranging antigen receptors
TLR, NLR, and SR
Million years ago
Mammals (humans)
Birds (chickens)
Reptiles (snakes)
Amphibians (frogs)
Bony fish (sturgeons)
Bony fish (zebrafish, medaka fish)
Cartilaginous fish (sharks)
Cyclostomes (lampreys)
Cyclostomes (hagfish)
Urochordates (sea squirts)
Cephalochordates (amphioxi)
Hemichordates (acorn worms)
Innate immunity based on pattern-recognition receptors
Echinoderms (sea urchins)
understand how the process of V(D)J
recombination may have evolved. With
that understanding has come an appreciation for the fact that the BCRs and TCRs of
the vertebrate adaptive immune system
may not be the only immune receptor
molecules generated by recombinatorial
genetics.
Ig-like receptors have been identified
in species as ancient as the earliest jawed
vertebrates. However, extensive analyses
of the only surviving jawless vertebrates
(Agnathans), the hagfish and lamprey,
show no evidence of TCR or BCR V(D)J
segments, RAG1/2 genes, or genes encoding elements of a primitive Major HistoCompatibility Complex (MHC) system.
This suggests that the RSS-based, recombinatorially generated set of adaptive
immune receptors first emerged in a
common ancestor of jawed vertebrates,
most probably more than 500 million
years ago (Figure 1).
The driving force in the development
of the adaptive immune system was most
probably the incorporation of an ancestor
of the RAG1 genes into the ancestral
genome in the form of a transposable element. This hypothesis is supported by
several observations and experiments:
served transposon in flies, the HERMES
transposon, has been shown to induce
a double-strand break via the hairpinformation mechanism described above
for V(D)J recombination.
RAG proteins have the ability to transpose an RSS-containing segment of
DNA to an unrelated target DNA in vitro.
222
326
370
476
525
652
794
891
896
FIGURE 1
The emergence of the RSS-dependent recombination-based adaptive immune
system coincides with the first appearance of the jawed vertebrates. SR scavenger
receptor. [Adapted from Figure 1, M. F. Flajnik and M. Kasahara, 2010, Origin and evolution of the
adaptive immune system: Genetic events and selective pressures, Nature Reviews Genetics 11:47–59.]
|
The Organization and Expression of Lymphocyte Receptor Genes
CHAPTER 7
241
BOX 7-2
have provided the substrate for the transposon invasion. When the transposon
subsequently jumped out of the gene,
the RSSs would have been left behind.
Multiple such rounds of transposition followed by gene duplication would have
given rise to the TCR and BCR loci we
know today. Interestingly, our knowledge
of phylogeny suggests that two waves of
gene duplications occurred at the time of
vertebrate origin, right around the time of
the proposed transposon entry.
Much still remains unknown about the
evolution of the adaptive immune receptor molecules. What was the nature of the
primordial antigen receptor that was the
target for the first transposition? At what
stage did the RAG2 gene become associated with the RAG1 component? And
how do we account for the evolution of
the Ig receptors of species such as the
shark, in which fully rearranged V(D)J
genes exist in the germ-line genome?
Recent evidence has suggested that
the recombinatorial strategy for generating antigen receptors is not limited to the
RSS-based Ig and TCR systems of vertebrates. Jawless fish, such as lampreys and
hagfish, possess lymphocyte-like cells
that can be stimulated to divide and differentiate. Moreover, these cells release
specific agglutinins after immunization
with antigens, and higher serum levels of
these agglutinins are secreted after a secondary, than after a primary, immunization, suggesting that immunological
memory exists in these fish. These lymphocyte-like cells therefore appear to
express all the hallmarks of adaptive
immunity, and yet the agglutinins they
secrete do not appear to have an Ig-like
structure.
Analysis of the agglutinins has revealed
that activated lamprey “lymphocytes”
express abundant quantities of leucinerich-repeat (LRR)-containing proteins.
LRRs are protein motifs that are frequently
associated with protein-protein recognition, and such motifs have already been
encountered in the context of the Toll-like
receptors described in Chapter 5. The
lamprey agglutinin receptors are gener-
(a)
Amino acids/cassette
SP LRRNT LRR1
27-34
24
LRRV
(0-8)
GPI
Stalk
LRRVe CP LRRCT
24
24
HP
16 48-63
(b)
3
2
4
1
5' LRR cassettes
5
3' LRR cassettes
Functional VLR
LRRNT
LRR1
LRRV1
LRRV2
LRRV3
LRRVe
LRRCP
LRRCT
ated by recombination of gene segments
during lymphocyte development, and the
LRR receptor repertoires of these fish are
remarkably diverse. These receptors have
therefore been named variable lymphocyte receptor (VLR) molecules.
Figure 2a shows an example of the
arrangement of the protein modules in
one of these VLR molecules. At the amino
terminal of the protein is an invariant signal peptide, followed by a 27–34 residue
N-Terminal Leucine-Rich Repeat (LRRNT).
This is followed by the first of several
FIGURE 2
Agnathans such as lampreys and
hagfish have evolved a recombination-based system of recognition
molecules that does not depend on
RSSs. Agnathans, such as lamprey and
hagfish, have lymphocyte-like cells which carry
antigen receptors that are remarkably diverse
and are created by a system of recombination
of small gene cassettes. (a) The modules that
make up a complete VLR receptor. (See text for
details.) (b) The assembly of a complete VLR
gene from the basic skeleton germ-line and
flanking cassettes. (c) The three-dimensional
structure of a VLR, showing the repeating
arrangement of the randomly recombined LRR
protein modules. [Adapted from Figures 1 and 2,
Herrin, B. R., and M. D. Cooper. 2010. Journal of
Immunology 185:1367.]
24-residue Leucine-Rich Repeat (LRR)
modules, LRR1, connected to a series of up
to eight 24-residue LRR modules with variable sequences (LRRVs). At the C-terminal
of the molecule, a 24-residue end LRRV
segment (LRRVe) is attached to a short
16-residue connecting peptide (CP),
which culminates in a 48–63 residue
C-terminal LRR (LRRCT). The VLR molecules are attached to the lymphocyte
membrane by an invariant stalk, rich in
threonine and proline residues, connecting to a glycosyl-phosphatidyl-inositol
(continued)
242
PA R T I I I
EVOLUTION
|
Adaptive Immunity: Antigen Receptors and MHC
BOX 7-2
(continued)
(GPI) anchor and a hydrophobic peptide.
Lymphocyte activation leads to phospholipase cleavage at the GPI anchor, enabling
soluble forms of the VLRs to be released
from the lymphocyte following antigen
stimulation.
The assembly process that generates
the completed protein occurs only in
lymphocytes. In the germ-line DNA, LRR
cassettes flank the skeleton LRR genes,
which consist initially only of sequences
coding for parts of the LRRNT and LRRCT,
separated by non-coding DNA (Figure 2b).
During lymphocyte development, the
non-coding sequence is replaced with
variable LRRs. Gene segments are copied
from one part of the genome into another
in a one-way process, similar to gene
conversion. Each lamprey lymphocyte
expresses a unique VLR gene from a single allele, and the diversity of the repertoire is limited only by the number of
lymphocytes.
Figure 2c shows a cartoon of the structure of one of these receptors, generated
from an x-ray crystallographic model.
Comparison of the sequences of several
of these primarily -stranded structures
has shown that their sequence diversity is
concentrated on the concave (left) side of
the molecule and can be attributed to the
inherent diversity of the LRR cassettes.
Did the two LRR- and V(D)J-based,
recombinatorially derived sets of immune
receptors exist side by side in an ancestral
species? This we do not know. Agnathans
other than lampreys and hagfish were
extinguished 400 million years ago, and
B-Cell Receptor Expression
The expression of a receptor on the surface of a B cell is the
end result of a complex and tightly regulated series of events.
First, the cell must ensure that the various gene recombination
events culminate in productive rearrangements of both the
heavy- and light-chain loci. Second, only one heavy-chain
and one light-chain allele must be expressed in each B cell.
Finally, the receptor must be tested to ensure it does not bind
self antigens, in order to protect the host against the generation of an autoimmune response.
Pre-BCR
J
reB
Vp
λ5
The random nature of Ig heavy- and light-chain gene rearrangement means that more than one heavy-light-chain
pair could potentially be expressed on the surface of individual B cells. Furthermore, since each heavy chain could
potentially combine with both light chains and vice versa,
this could result in the creation of B cells bearing a variety
of different antigen-binding sites. Whereas this opportunity to increase the number of available receptors may initially sound advantageous to the organism, in practice the
presence of more than one receptor per B cell creates prohibitive difficulties for those mechanisms that protect
against the generation of autoimmunity. The mechanism by
which B cells ensure that only one heavy- and one lightchain allele are transcribed and translated is referred to as
allelic exclusion.
The rearrangement of Ig genes occurs in an ordered way,
and begins with recombination at one of the two homologous chromosomes carrying the heavy-chain loci. The production of a complete heavy chain and its expression on the
B-cell surface in concert with a surrogate light chain, made
up of two proteins, VpreB and 5 (Figure 7-10), signals the
end of heavy-chain gene rearrangement and, thus, only one
antibody heavy chain is allowed to complete the rearrangement process. The mechanism by which this allelic exclusion
occurs is still under investigation. However, we do know that
following successful arrangement at one allele, the VH gene
locus on the other chromosomes is methylated and recruited
to heterochromatin.
VD
Allelic Exclusion Ensures That Each B Cell
Synthesizes Only One Heavy Chain and One
Light Chain
we have access only to the fossil remains
of ostracoderms (agnathans with dermal
skeletons), which are thought to be the
ancestors of the gnathostomes. One theory would hold that the VLR recombinatorial system evolved in an ancestor
common only to hagfish and lampreys,
and could have been causal in their ability to survive when other agnathan species succumbed to environmental insult.
Alternatively, the VLR and Ig-based systems may have co-existed for a time, with
lymphocytes bearing both types of
receptors, until one of the systems was
lost.
Combinatorial genetics in the generation of the vertebrate antigen receptors
may not be as unique a mechanism as
immunologists had previously thought.
FIGURE 7-10 Ig heavy-chain expression on the cell surface
in concert with VpreB and 5 signals, via Ig␣/Ig, the end of
heavy-chain rearrangement.
|
The Organization and Expression of Lymphocyte Receptor Genes
Pre-BCR, made up of μ heavy chain plus the
surrogate light chain, inhibits rearrangement
of μ allele #2 and induces κ rearrangement
CHAPTER 7
243
μ + κ chains inhibit
rearrangement of κ allele #2
and λ rearrangement
IgM
μ + κ chains inhibit
λ rearrangement
Pre-BCR
μ+κ
μ
VH
DH JH
μ + λ chains inhibit
rearrangement of λ
allele #2
JH
Productive
allele #1
Vκ Jκ
Productive
allele #1
μ+κ
Vκ Jκ
μ
μ
Productive
allele #2
Nonproductive
allele #1
VH
DH
Progenitor
B cell
DH
Productive
allele #2
μ+λ
JH
JH
DH
VH
VH
JH
DH
Nonproductive
allele #1
μ
Vλ Jλ
Nonproductive
allele #2
Productive
allele #1
μ
Vλ Jλ
Nonproductive
allele #1
μ+λ
Productive
allele #2
μ
Nonproductive
allele #2
Nonproductive
allele #2
Cell death
Cell death
FIGURE 7-11 The generation of a functional immunoglobulin receptor requires productive rearrangement at
heavy- and light-chain alleles. Heavy-chain genes rearrange first,
and once a productive heavy-chain gene rearrangement occurs, the
pre-BCR containing the protein product prevents rearrangement of
the other heavy-chain allele and initiates light-chain rearrangement.
In the mouse, rearrangement of light-chain genes precedes rear-
If the first attempt at a heavy-chain rearrangement is
unproductive (i.e., results in the formation of out-of-frame
joints caused by P and N nucleotide addition or by exonuclease trimming at the joint), rearrangement will initiate at the
second heavy-chain allele. If this is also unsuccessful, the B
cell will die. Once a complete heavy chain is expressed, lightchain rearrangement begins. In mice, light-chain rearrangement always begins on one of the alleles, and continues until
a productive light-chain rearrangement is completed. In
humans, it may begin at either a or a light-chain locus.
Because successive arrangements (using upstream V regions
and downstream J regions) can occur on the same light-chain
chromosome (see below) and there are four alleles (two and
two ) to choose from, once a B cell has made a complete
heavy chain, it will normally progress to maturity. Successful
completion of light-chain rearrangement results in the expression of the BCR on the cell surface, and this signals the end of
further Ig receptor gene rearrangement events (Figure 7-11).
rangement of the genes, as shown here. In humans, either or
rearrangement can proceed once a productive heavy-chain rearrangement has occurred. Formation of IgM inhibits further light-chain
gene rearrangement. If a nonproductive rearrangement occurs for
one allele, then the cell attempts rearrangement of the other allele.
[Adapted from G. D. Yancopoulos and F. W. Alt, 1986, Regulation of the assembly
and expression of variable-region genes, Annual Review of Immunology 4:339.]
The process of generating a fully functional set of B cells
is energetically expensive for the organism because of the
amount of waste involved. On average, two out of three
attempts at the first heavy-chain chromosomal locus will
result in an unproductively rearranged heavy chain, and the
same is true for the rearrangement process at the second
heavy-chain locus. Therefore the probability of successfully
generating a functional heavy chain is only 1/3 (at the first
allele)
{2/3 (probability the first rearrangement was not
successful) 1/3 (probability that the second rearrangement is successful)} 0.55 or 55%.
Receptor Editing of Potentially Autoreactive
Receptors Occurs in Light Chains
The immune system can still make some changes even after
the completed receptor is on the cell surface. If the receptor
is found to be auto-reactive, the cell can swap it out for
244
PA R T I I I
Germline DNA
|
5'
3'
Adaptive Immunity: Antigen Receptors and MHC
V cluster
1
2
Constant
region
J cluster
1
3
2
3
4
First rearrangement between V3 and J3 is productive, but combination of the
resultant light chain with Ab heavy chain results in an autoimmune antibody
Recombined VJ
gene in B cell
5'
1
3'
3 3
2
4
Second rearrangement between V2 and J4 is productive, and combination
with the Ab heavy chain results in an non autoimmune antibody
5'
3'
1
2 4
FIGURE 7-12 light-chain receptor editing. In the event of a productive light-chain rearrangement that leads to the formation of an
auto-antibody, additional rounds of rearrangement may occur between upstream V and downstream J gene segments. Here, the primary
rearrangement between V3 and J3 is edited out with a secondary rearrangement between V2 and J4.
a different one, in a process termed receptor editing. Since
the receptor’s binding site contains elements of both the
heavy and the light chain, changing just one of these is often
sufficient to alter the specificity, and light-chain receptor
editing has been shown to occur quite frequently.
In this process, the DNA rearrangement machinery is
switched back on after the first completed receptor in the
immature B cell has been found to have autoimmune specificity. For example, as shown in Figure 7-12, the initial rearrangement of V3 to J3 may result in an Ig receptor molecule
that bound to a self antigen. The B cell can then rearrange a
light-chain gene at the second -chain allele, it can rearrange
a -chain allele, or it can edit the original -chain rearrangement. All of these choices have been demonstrated in different B cells. However, the term receptor editing refers to the
process in which a B cell uses the same allele more than
once, and engages in a second rearrangement process in
which a variable region gene segment upstream of the original segment (in this case V2) is recombined with a J segment
downstream of the initially utilized J region (in this case, J4).
Switching out the rearranged VH segments is less common. Since rearrangement of the heavy-chain variable
region deletes the RSSs on either side of the D region
sequences, it was at first thought to be impossible. However,
recent advances have demonstrated that cryptic RSSs exist
that can be used in heavy-chain editing, and so some measure of heavy-chain receptor editing does occur.
Ig Gene Transcription Is Tightly Regulated
Immunoglobulin genes are expressed only in cells of the B
lymphocyte lineage, and within this lineage, expression of Ig
genes is developmentally regulated (see Chapter 10). The
control of Ig gene transcription is therefore necessarily complex, although it involves the familiar paradigm of transcription factors that bind to sequence elements in the DNA
(Table 7-5 and Figure 7-13).
Expression of the Ig genes requires the coordination of
two types of cis regulatory elements: promoters and
enhancers. Like other promoters, the Ig heavy and lightchain gene promoters contain a highly conserved, AT-rich
sequence called the TATA box, which serves as the binding site for RNA polymerase. RNA polymerase II starts
transcribing the DNA from the initiation site, about 25
base pairs (bp) downstream of the TATA box. Each VH and
VL gene segment has a promoter located just upstream
from its leader sequence. The promoters upstream of Ig VH
and VL gene segments bind to RNA polymerase II quite
weakly, and the rate of transcription of VH and VL coding
regions is very low in unrearranged germ-line DNA. Rearrangement of Ig genes brings the promoters close enough
to enhancers located in the intron sequences and in
regions downstream of the C region genes for them to
influence the V region promoters. Consequently, the rate
of transcription is increased by a factor of 104 upon V gene
rearrangement.
However, although the transcription of unrearranged
genes is low, it is not absent in developing B cells; evidence
now suggests that, as the Ig genes move out of the pericentric regions of the nucleus and assume more euchromatic
configurations prior to recombination, a low level of transcription occurs which indicates to the recombination
machinery that they are in a state of readiness for rearrangement to occur.
|
The Organization and Expression of Lymphocyte Receptor Genes
TABLE 7-5
CHAPTER 7
245
Some of the cis regulatory elements that control immunoglobulin gene transcription
Elements
Function
Location(s) in Ig genes
Promoters
Promote initiation of transcription of
neighboring gene in a specific direction. Bind RNA polymerase and direct
the formation of the pre-initiation
complex.
Upstream of VH, D, V, V1, and V2.
Stimulate transcription of associated
genes/gene segments when bound by
transcription factors (TFs). Can function
at variable distance from promoter,
and in either orientation.
Intronic enhancers lie between JH and
C and between J and C.
Enhancers
Symbol in Figure 7-13
Additional promoters are found
upstream of all constant regions.
Arrow denotes direction of
transcription.
3 enhancers lie to the 3 end of the
C gene and may be part of a locus
control region.
Light-chain enhancers lie 3 of C and
C genes.
Locus control
regions
Collection of smaller elements that
may each have enhancer function.
Complex regulatory region, RR, 3 of
murine C gene. Contains four enhancers in two separate units.
The J cluster and each of the D genes of the heavy-chain
locus are also preceded by promoters. These D and J promoters allow for transient transcription of DJ and J segments, which appears to be important for efficient VDJ
rearrangement. So far, there is no evidence of conventional
promoter sequences upstream of any of the J loci, although
some germ-line transcription of J sequences has been
detected.
The control of Ig gene transcription depends on the interplay between a large number of transcription factors that
bind to the promoter and enhancer regions of the Ig genes,
and binding of different combinations of these proteins to
VH promoters
IgH
5'
3'
Constant region promoters
μ
Jκ regions
κ intron enhancer
γ1
3' enhancer/LCR/RR
α
3' κ enhancer
5'
3'
Vλ2 promoter
Igλ
μ intron enhancer
D region promoters
Vκ promoters
Igκ
promoters and enhancers can lead to very different outcomes. In Chapter 10, we discuss how transcription of Ig
genes is controlled by varying combinations of transcription
factors during B cell development.
Occasionally, the powerful B-cell Ig enhancers can engage
in activities that are less than benign. Translocation of the
c-myc oncogene close to the Ig enhancer regions results in
constitutive expression of the c-Myc protein and the generation of an aggressive, highly proliferative B-cell lymphoma.
Similarly, the translocation of the bcl-2 gene close to the Ig
enhancer leads to suspension of programmed cell death in B
cells, resulting in follicular B-cell lymphoma.
5'
3'
Vλ2
JCλ2
λ2-4
JCλ4
enhancer
Vλ1 promoter
Vλ1
JCλ3
λ3-1
enhancer
JCλ1
FIGURE 7-13 Locations of the cis regulatory control elements on immunoglobulin genes. The expression of Ig genes is controlled by transcription factor binding to promoter and enhancer regions. Promoters are designated by rectangular arrows and enhancer
regions by green ovals. Accessibility to recombination is often preceded by low levels of transcription from downstream D and J segments.
246
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
Mature B Cells Express Both IgM and IgD
Antibodies by a Process That Involves
mRNA Splicing
As described above, the same Ig heavy-chain variable region
can be expressed in association with more than one heavychain constant region. Furthermore, each of the heavy-chain
constant regions can be synthesized as both membranebound and secreted forms.
Immature B cells in the bone marrow express only membrane IgM receptors. However, as the B cells mature, they
also place IgD receptors on their cell surface. These IgD
receptors bear the identical antigen-binding V region, but
carry a , rather than a constant region. In this section, we
describe how cells create the membrane-bound and secreted
forms of IgM and IgD by RNA splicing. The generation of
antibodies belonging to the heavy-chain classes other than
IgM and IgD occurs by an additional DNA recombination
process, referred to as class switch recombination (CSR),
and results in the loss of the DNA encoding the and gene
segments. CSR is described in Chapter 12, because it is a
process that occurs only after antigen-mediated activation of
B cells.
The primary heavy-chain transcript in a B cell that has
not yet been exposed to antigen encodes both the C and
the C constant regions, and is approximately 15 kb in
length. The determination of whether a cell will synthesize
the membrane-bound or secreted forms of and/or heavy
chains is made by nuclear proteins that select particular
choices of polyadenylation and RNA splicing sites. During
mRNA processing, polyadenylation sites signal for the cleavage and addition of a poly-A sequence at a particular site on
the primary transcript, that then becomes the 3 terminus of
the mature, processed mRNA. (Poly-A tails are characteristically found at the 3 termini of many eukaryotic mRNA
species.) A nuclear complex called the spliceosome determines the sites at which the primary transcript will be
cleaved and spliced to form the mature mRNA.
Figure 7-14 illustrates the splicing and polyadenylation
choices made when a cell generates membrane bound IgM
(the only Ig species expressed by immature B cells) versus
membrane-bound IgD, which is expressed upon B-cell
maturation. Both of these species are present on the cell
membrane prior to antigen encounter.
Analysis of the primary Ig transcript revealed that the
C4 exon (the 3-most exon of the heavy chain) ends in a
nucleotide sequence called S (secreted), which encodes the
hydrophilic portion of the CH4 domain of secreted IgM. The
S portion of the mRNA sequence includes a polyadenylation
site, polyadenylation site 1. Two additional exons, M1 and
M2 (M for membrane), are located 1.8 kb downstream of the
3 end of the C4 exon. M1 encodes the membranespanning region of the heavy chain. M2 encodes the three
cytoplasmic amino acids at the C-terminus of IgM, followed
by polyadenylation site 2 (see Figure 7-14a). Production of
mRNA encoding the membrane-bound form of the chain
occurs when cleavage of the primary transcript and addition
of the poly-A tail occurs at polyadenylation site 2 (see Figure 7-14b). RNA splicing then removes the S sequence at the
3 end of the C4 exon, and joins the remainder of the C4
exon to the M1 and M2 exons. The RNA sequence between
the M1 and M2 exons is also spliced out, resulting in the
production of mRNA encoding the membrane-bound form
of the IgM heavy chain. This is the only splicing pattern that
occurs in immature B cells.
As the B cell matures, it begins to express membranebound IgD in addition to membrane IgM (see Figure 7-14c).
The exons encoding the membrane-bound and secreted
forms of IgD are arranged similarly to those of IgM, with
polyadenylation sites 3 and 4 at the 3 termini of the
sequences encoding the secreted and membrane-bound
forms of IgD respectively. If polyadenylation occurs at site 4,
and the C RNA is spliced out, the cell will make the
membrane-bound chain, and express IgD as well as IgM
on the cell surface. Since mature B cells bear both IgM and
IgD on their surfaces, it is clear that both splicing patterns
can occur simultaneously.
Following antigen stimulation, the B cell begins to generate the secreted form of IgM (Figure 7-15a). Polyadenylation
shifts to site 1, and the mRNA sequences downstream of site
1 are degraded. (Similarly, the use of splice site 3 generates
secreted IgD, although this is not shown for purposes of clarity.) The secreted form of IgD is used less frequently than the
secreted form of the other antibody heavy-chain classes, and
so splice site 3 is used relatively rarely.
At the protein level, the carboxyl-terminus of the membrane-bound form of the Ig chain consists of a sequence of
about 40 amino acids, which is composed of a hydrophilic
segment that extends outside the cell, a hydrophobic transmembrane segment, and a very short hydrophilic segment
at the carboxyl end that extends into the cytoplasm (Figure 7-15b). In the secreted form, these 40 amino acids are
replaced by a hydrophilic sequence of about 20 amino acids
in the carboxyl terminal domain.
What mechanism determines which splice sites are
used? This question is not yet fully resolved. However,
work with artificial constructs containing different combinations of the various splice sites suggests that there is an
intrinsic “strength” to each of the splice sites that determines how frequently it is used, and that this strength
reflects the binding affinity between the DNA sequence at
the splice site and the spliceosome. Analysis of the
sequences of the splice sites in Ig mRNA suggests that
splice site 2 is considerably “stronger” than splice site 1,
which explains why immature B cells make a preponderance of the membrane form of the chain. B-cell stimulation appears to cause an increase in the concentration of
the splice site recognition proteins, as well as the early termination of transcription prior to the M1 and M2
sequences, thereby facilitating the use of splice site 1 and
|
The Organization and Expression of Lymphocyte Receptor Genes
CHAPTER 7
247
(a) H-chain primary transcript
Cδ
Cµ
L
VDJ
µ1
J
µ2
µ3
µ4 S
δ1
M1 M2
δ2
δ3 S
M1M2
5'
3'
- 6.5
kb
Poly-A
site 1
Poly-A
site 3
Poly-A
site 2
Poly-A
site 4
(b) Polyadenylation of primary transcript at site 2 → µm
Cµ
L
VDJ
J
S
M1 M2
µm transcript 5'
(A)n
Splicing
L
µm mRNA
VDJ
µ1 µ2 µ3 µ4 M1 M2
(A)n
5'
(c) Polyadenylation of primary transcript at site 4 → δm
Cµ
VDJ
L
J
µ1 µ2 µ3 µ4 S
δm transcript 5'
Cδ
M1 M2
δ1
δ2
δ3 S
M1M2
(A)n
Splicing
L
δm mRNA
5'
VDJ
δ1 δ2 δ3 M1 M2
(A)n
FIGURE 7-14 Differential expression of membrane forms of and ␦ heavy chains by alternative RNA processing. (a) Structure
of rearranged heavy-chain gene showing C and C exons and poly-adenylation sites. (b) Structure of m primary transcript and m mRNA
resulting from polyadenylation at site 2, followed by RNA splicing. (c) Structure of m primary transcript and m mRNA resulting from polyadenylation at site 4 and RNA splicing. Both processing pathways can proceed in any given B cell.
generation of the secreted form of the protein. A similar
mechanism may control the differential splicing of the
membrane-bound and secreted forms of the chain.
T-Cell Receptor Genes and Expression
The initial characterization of the BCR molecule was facilitated by the fact that the secreted Ig product of the B cell
shares the antigen-binding region with the membrane receptor. The secreted Ig proteins could be used as antigens to
generate antibodies in other species that recognized the
B-cell surface receptor and these anti-receptor antibodies
could then be employed as reagents to isolate and characterize the BCR. In addition, a set of monoclonal B-cell tumors
secreting high concentrations of soluble antibodies was
available.
Investigators attempting to purify and analyze the T-cell
receptor (TCR) enjoyed no such advantages, and therefore it
should not be surprising to learn that several decades
elapsed between the characterization of the BCR and that of
its T-cell counterpart. Indeed, characterization of the TCR
was so difficult and vexing that one of the distinguished writers for the scientific journal Nature, Miranda Robertson,
referred to it as “The Hunting of the Snark,” in reference to
the Lewis Carroll poem of the same name, which described
the search for a mythical creature.
In Chapter 3, we described the TCR protein. Below, we
describe the parallel story of the discovery of the TCR genes.
Understanding the Protein Structure
of the TCR Was Critical to the Process
of Discovering the Genes
In Chapter 3 we described how investigators were able to use
a monoclonal antibody (mAb) specific for the TCR to purify
and characterize it as an
heterodimeric protein. Each
chain was shown to consist of one variable and one constant
region domain, and both chains contained regions arranged
248
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
(a)
Cµ
L
Primary
H-chain 5'
transcript
VDJ
J
µ1
µ2
µ3
Cδ
µ4 S
M1
M2
3'
Poly-A
site 1
Poly-A
site 2
Poly-A
site 3
Poly-A
site 4
Polyadenylation
Site 1
RNA transcript for secreted µ
L V D J J µ1 µ2 µ3
Site 2
RNA transcript for membrane µ
µ1 µ2 µ3 µ4 S M1 M2
L V DJ J
µ4 S
(A)n
5'
(A)n
5'
RNA splicing
L V D J µ1 µ2 µ3 µ4 M1 M2
L V D J µ1 µ2 µ3 µ4 S
5'
5'
(A)n
mRNA encoding secreted µ chain
(A)n
mRNA encoding membrane µ chain
Cμ4
(b)
CHO
SS bridge
T
G
K
P
T
L
Y
N
V
S
L
I
M
S
D
T
G
G
T
C
Y
556
Key:
Cμ4
Cμ 4
Hydrophilic
+
Hydrophobic
556
556
563
Encoded
by S exon
of Cμ
576
–
COOH
Encoded by
M1 and M2
exons of Cμ
575
576
COOH
Outside
568
576
Membrane
594
597
Cytoplasm
COOH
T 556
– E
G
– E
V
–
A N
E
– E
– E
G
F
568
– E
N
L
F I W
V
T
T
S A T L
F
S T T
L
Y
LV
F L S
T
L
F 594
+ K
V
+ K 597
COOH
Secreted μ
FIGURE 7-15 Differential expression of the secreted and
membrane-bound forms of immunoglobulin chains is
regulated by alternative RNA processing. (a) Structure of the
primary transcript of a rearranged heavy chain, showing the C
exons and poly-A sites. Polyadenylation of the primary transcript
at either site 1 or site 2, and subsequent splicing (indicated by
V-shaped lines) generates mRNAs encoding either secreted or
Membrane μ
membrane chains, as shown. (b) Amino acid sequences of the
carboxyl terminal ends of secreted and membrane heavy
chains. Residues are described by the single-letter amino acid
code. Hydrophilic (pink) and hydrophobic (yellow) residues and
regions are shown, charged amino acids are indicated with a
or , and the N-linked glycosylation site on the secreted form is
labeled “CHO.”
The Organization and Expression of Lymphocyte Receptor Genes
•
•
Vα
Cα
Vβ
Cβ
•
•
[PDB ID 1TCR.]
in the characteristic Ig fold (Figure 7-16). The presence of
these variable and constant regions provided clues as to the
potential arrangement of the genes that encode the TCR
proteins, and enabled researchers to design the experiments
that ultimately led to the characterization of the TCR genes.
The -Chain Gene Was Discovered
Simultaneously in Two Different Laboratories
Two different research groups published papers in the same
issue of Nature in 1984 that described the discovery of the
TCR genes in mice and in humans respectively. Mark Davis,
Stephen Hedrick, and their collaborators utilized a uniquely
creative approach to isolate the receptor genes from several
mouse T-cell hybridoma cell lines, which we describe below.
The other group, headed by Tak Mak, used more classical
methods of genetic analysis to isolate the TCR genes from
human cell lines.
The Davis-Hedrick group strategy was based on four
assumptions about TCR genes:
•
•
TCR genes will be expressed in T cells but not B cells.
Since the genes encode a membrane-bound receptor, the
transcribed mRNA will be found associated with
membrane-bound polyribosomes.
CHAPTER 7
249
Like Ig genes, TCR genes will code for a variable and a
constant region.
Like Ig genes, the genes that encode the TCR will
undergo rearrangement in T cells.
Prior experiments had revealed that only about 2% of
the genes expressed by lymphocytes differed between T
and B cells. Furthermore, only a small proportion (about
3%) of T-cell mRNA had been shown to be associated with
membrane-bound polyribosomes. Davis and Hedrick reasoned that if they could generate 32P-labeled DNA copies
(cDNA) of the membrane-bound polyribosomal fraction
of antigen-specific T-cell mRNA, and remove from that
population those sequences that were also expressed in B
cells, they would be left with labeled, T-cell-derived cDNA
that would be greatly enriched in sequences encoding the
TCR (Figure 7-17).
They therefore performed the following steps:
•
FIGURE 7-16 The heterodimeric ␣ T-cell receptor. The
alpha chain is shown here in green and the beta chain in blue.
|
•
•
Extract membrane-bound polysomal mRNA (RNA
sequences that encode membrane-bound proteins) from
several T hybridoma lines.
Reverse transcribe the polysomal mRNA in the presence
of 32P-labeled nucleotides, to generate radioactive cDNA
copies of each of the mRNA species.
Mix the 32P-labeled T-cell cDNA with B-cell mRNA and
remove all the cDNA that hybridized with B-cell mRNA.
(This leaves behind only the T-cell mRNA that encodes
proteins not expressed in B cells.)
Recover the “T-cell-specific cDNAs” and use them to
identify hybridizing DNA clones from a T-cell specific
library.
Radioactively label those clones that hybridized to the
“T-cell-specific” cDNAs, and use cloned probes
individually to probe Southern blots of DNA from liver
cells and B-cell lymphomas (which would not be
expected to rearrange TCR genes) and DNA from
different T-cell hybridomas (which would be expected to
show different patterns of rearrangements in different
T-cell clones).
Probing with the TM90 cDNA probe (see bottom of Figure 7-17) resulted in a pattern of bands that did not vary
depending on the origin of the cellular DNA. This indicated
that TM90 recognizes DNA found in T cells but not B cells,
that does not rearrange in different ways in different cells,
and therefore is probably not complementary to a TCR gene.
In contrast, the TM86 cDNA probe hybridized to bands in
the T-cell DNA of different sizes than in either the liver or
the B-cell DNA. Furthermore, TM86 showed different patterns of hybridization in different T-cell hybridomas, indicating that the sequence recognized by the TM86 probe is
found in varying contexts within the DNA of different
hybridomas. Note that this probe also shows the same pattern of bands in the liver cells and B cells. This is consistent
with the notion that the genes encoding the TCR recombine in
T cells, but not in any other cell type. These results indicated
PA R T I I I
250
|
TH-cell clone
FIGURE 7-17 Production and identification of a cDNA
B cell
clone encoding the T-cell receptor gene. The flowchart outlines the procedure used by S. M. Hedrick, M. M. Davis, and colleagues
to obtain 32P-labeled cDNA clones corresponding to T-cell-specific
mRNAs. DNA subtractive hybridization was used to isolate T-cellspecific cDNA fragments that were cloned and labeled with 32P. The
labeled cDNA clones were used to probe Southern blots of DNA
isolated from embryonic liver (TCR genes should be in the germ-line
configuration), from a B-cell lymphoma (TCR genes should also be in
the germ-line configuration), and from a panel of T-cell clones. (TCR
genes should be differently arranged in each clone.) Their clone,
TM86, encodes a gene that rearranged differently in each T-cell clone
analyzed. Comparison of its sequence to that of the protein sequence
of the
TCR isolated by Kappler and Marrack revealed TM86 to
encode the chain of the TCR. TM90 cDNA identified the gene for
another T cell membrane molecule that does not undergo rearrangement. [Based on S. M. Hedrick et al., 1984, Isolation of cDNA clones encoding
mRNA
mRNA
97% in free
cytoplasmic
polyribosomes
Adaptive Immunity: Antigen Receptors and MHC
3% in
membrane-bound
polyribosomes
32P
Reverse
transcriptase
[32P] cDNA
T cell-specific membrane-associated proteins, Nature 308:149.]
Hybridize
A Search for the -Chain Gene Led to the
-Chain Gene Instead
Hybrids with cDNAs
common to T cells
and B cells
Discard
cDNAs
specific
to T cells
Liver cells
B cell lymphoma
Separate on
hydroxyapatite
column
T-cell clones
a b c d e f
10 different
cDNA clones
Use as probes in
Southern blots of
genomic DNA
Probed with cDNA TM86
Probed with cDNA TM90
that the TM86 cDNA probe was binding to DNA that is rearranged specifically only in T cells and thus most probably
encodes a gene for a TCR protein. In this way the first mouse
TCR gene was isolated.
Comparison of the DNA sequence of Mak’s human clone
with the amino acid sequence from a human hybridoma
confirmed that Davis and Hedrick et al., and Mak et al., had
isolated the gene for the mouse and human forms of the
TCR gene, respectively.
Focus now switched to the search for the TCR chain. But
here, immunology took one of its strange and wonderful
twists.
Tonegawa’s lab, this time using the subtractive hybridization approach pioneered by Hedrick and Davis, succeeded in
cloning a gene that appeared at first to have all the hallmarks
of the TCR chain gene. It was expressed in T cells, but not
in B cells; it was rearranged in T cell clones; and on sequencing, it revealed regions corresponding to a signal peptide,
two Ig family domains, a transmembrane region, and a short
cytoplasmic peptide. Furthermore, the predicted molecular
weight of the encoded chain appeared to be very close to that
of the -chain protein. Tonegawa’s lab initially suggested
that “while direct evidence is yet to be produced, it is very
likely the pHDS4/203 codes for the subunit of the T cell
receptor.”
However, biochemical analysis of the TCR and chain
proteins by another laboratory had clearly demonstrated
that both the and the chains of the TCR heterodimer
were glycosylated. A search for potential sequences corresponding to sites of carbohydrate attachment on Tonegawa’s
putative -chain sequence came up short. It therefore
seemed that the gene Tonegawa’s lab had isolated was not the
chain, but something very similar. Further work demonstrated that they had discovered a gene encoding an hitherto
unknown receptor that contained both variable and constant
regions and recombined only in T cells. They named this
gene (and the receptor chain it encoded) .
Davis and colleagues next isolated a genomic sequence
expressed in T and not B cells that encoded a protein with
four potential N-linked glycosylation sites, and a molecular
weight consistent with that of the chain of the TCR. With
|
The Organization and Expression of Lymphocyte Receptor Genes
CHAPTER 7
251
Mouse TCR α-chain and δ-chain DNA
5'
L Vα1
L Vα2
L Vαn
L Vδ 1
L Vδ n
Dδ 1 Dδ 2 Jδ 1 Jδ 2
Cδ
L Vδ 5 Jα1 Jα2 Jα3
Jαn
Cα
3'
Mouse TCR β-chain DNA
5'
L Vβ 1
L Vβ 2
L Vβn
Dβ 1
Jβ 1.1-Jβ 1.7
Cβ 1
Jβ 2.1-Jβ 2.7
Dβ 2
ψ
Cβ 2
L Vβ 14
ψ
3'
Mouse TCR γ-chain DNA
5'
L V γ5
L V γ2
L V γ4
L Vγ 3
J γ1
Cγ1
L
3'
Vγ1.3
J γ3
Cγ3
ψ
ψ
Cγ2
J γ2
L
L
Jγ4
Cγ4
Vγ1.2 Vγ1.1
= Enhancer
ψ = pseudogene
FIGURE 7-18 Germ-line organization of the mouse TCR ␣-,
-, ␥-, and ␦- chain gene segments. Each C gene segment is
composed of a series of exons and introns, which are not shown. The
organization of TCR gene segments in humans is similar. Approximate numbers of the gene segments are shown in Table 7-7. [Adapted
the genes for the and chains of the TCR fully characterized, those attempting to understand TCR genetics were left
with a series of intriguing questions. Was the chain
expressed by the same T cells that expressed the
receptor? Did the chain that Tonegawa had discovered also exist
as a heterodimer? If so, what was its partner, and who would
find it first?
After 3 years of searching, Chien and colleagues, working
in the Davis lab, described a fourth TCR gene that they
named . Intriguingly, the V chain gene was found to lie
downstream of the previously described V genes and just 5
to the J C coding regions. Rearrangement of this locus was
found to occur very early in thymic differentiation, and
expression of -chain RNA paralleled that of the expression
of the chain in thymic subpopulations, occurring in the
same T cells. Most striking about the location of the genes
was the observation that, if a T cell rearranged the chain,
the intervening genes would be lost (Figure 7-18). Study of
the T cells expressing these , receptors demonstrated that
, -bearing T cells represented an entirely new T-cell subset
TABLE 7-6
Gene
Chromosomal locations of TCR
genes in humans and mice
Human
Mouse
chain
14
14
chain
7
6
chain
7
13
chain
14
14
from D. Raulet, 1989, The structure, function, and molecular genetics of
the gamma/delta T cell receptor, Annual Review of Immunology 7:175; and
M. M. Davis, 1990, T cell receptor gene diversity and selection, Annual Review
of Biochemistry 59:475.]
with an antigen repertoire distinct from that of , T cells
and a different anatomical distribution within the lymphoid
system in the host.
TCR Genes Undergo a Process of Rearrangement
Very Similar to That of Ig Genes
Table 7-6 shows the chromosomal locations of the TCR
receptor genes in mice and humans and Figure 7-18 shows
the arrangement of the TCR genes in the mouse genome.
The four TCR loci are organized in the germ line in a manner remarkably similar to the multigene organization of the
Ig loci, shown in Figure 7-4, and as in the case of Ig genes,
functional TCR genes are produced by rearrangements of V
and J segments in the - and -chain families and between
V, D, and J segments in the - and -chain genes.
The approximate numbers of different V, D, and J segments in each of the TCR gene families of mice and humans
are shown in Table 7-7. Interestingly, some of the V and V
gene segments have been shown to be used in both and
TCR chains, which further adds to the diversity of the TCR
repertoire.
Rearrangement of TCR variable region genes follows the
same general outline as that of Ig genes. One point of contrast however is that, whereas Ig light chains incorporate few
N nucleotides because of the developmental down-regulation
of expression of the TdT enzyme in B cells by the time of
light-chain rearrangement, N nucleotides are seen at similar
frequencies in all of the TCR chains.
The - and -chain variable genes, like the Ig light-chain
gene, are generated from one V and one J segment. The TCR
and chain variable region genes are assembled from V,
252
TABLE 7-7
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
TCR gene segments in humans and mice
Human
Mouse
V
40–42 functional; 6 ORFs; 12–13 pseudogenes
21 functional; 1 ORF; 12 pseudogenes
D
2 functional
2 functional
J
6 J 1; 7 J 2
5 J 1 functional; 2 J 1 ORFs
6 J 2 functional; 1 J 2 pseudogene
V
43; some used in conjunction with
or constant regions
J
50 functional; 8 ORFs; 3 pseudogenes
71 functional; 2 ORFs; 14 pseudogenes
39 functional; 12 ORFs; 10 pseudogenes
V
3*
15 functional; 1 pseudogene
D
3
2
J
4
2
V
4–6 functional; 3 ORFs; 5 pseudogenes
7 functional
J
5
4
* Additional gene segments (5 for human and 9 for mouse) have been found able to associate with downstream or gene segments.
A germ-line gene is considered to be functional if the coding region has an ORF without a stop codon and, if there is no described defect in the splicing sites, RSS and/
or regulatory elements.
A germ-line entity is considered to be an ORF if the coding region has an open reading frame but alterations have been described in the splicing sites, RSSs, and/or
regulatory elements, and/or changes of conserved amino acids have been suggested by the authors to lead to incorrect folding and/or the entity is an orphan.
A germ-line entity is considered to be a pseudogene if the coding region has a stop codon(s) and/or a frameshift mutation. In the case of V gene segments, these
mutations may be either in the V gene coding sequence, or in the leader sequence.
SOURCE: Gene numbers and definitions summarized from the International Immunogenetics Information System Web site: http://imgt.org. Accessed November 16, 2011.
D, and J segments, like the Ig heavy chains. In contrast to
the Ig genes, however, the TCR families do not have a functionally diverse set of C regions. There is one constant region
gene for each of the and gene families, and although
there is more than one gene encoding each of the C and C
regions, they are highly homologous and do not appear to
differ in function.
Animals deficient in RAG1/2 or the other enzymes affecting Ig gene recombination, such as Artemis or TdT, are
similarly deficient in the generation of both their BCRs and
their TCRs, and so it is thought that the essential features of
the rearrangement processes are similar in B and T cells
(Clinical Focus Box 7-3). The heptamer-nonamer recombination signal sequences are found at the 3 termini of the V
sequences, the 5 and 3 termini of the D sequences, and the
5 termini of the J sequences, just as for the Ig variable region
gene segments.
Recall that the 12-23 rule specifies that recombination
can only occur between gene segments contiguous in one
case with an RSS containing a 12 (one-turn) bp spacer and
in the other with an RSS containing 23 (two-turn) bp spacer
(see Figure 7-5). By this rule, recombination between VH
and JH immunoglobulin gene segments is disallowed, and
hence all properly rearranged heavy-chain V regions must
have all three segments represented.
In contrast, the D regions of the TCR and chain
genes are bordered on the one side by a 12-bp spacer and on
the other by a 23-bp spacer (Figure 7-19). By the 12-23 rule,
V and J segments from the and gene families can theoretically recombine directly, without an intervening D segment. Furthermore, the rule would also allow for D-D
segment recombination. Do these unusual recombinations
occur in vivo?
In the case of the -chain gene, the answer appears to be
that they do not. Evidence obtained from sequencing cDNA
fragments containing precisely known V, D, and J sequences
has demonstrated that functioning TCR -chain genes
always contain one each of the V, D, and J segments. However, the situation is not quite as simple for the TCR -chain
genes.
Scientists noted that many TCR V genes were considerably longer than TCR V genes; furthermore, the increased
sequence length seemed primarily to occur in those parts of
the gene encoding the CDR3 residues that make antigen
contact. Subsequent analysis showed that many TCR V
genes have incorporated not one but two D-region segments, and this additional segment is primarily responsible
for the observed increase in gene length. In some V genes,
further length was added to the TCR -chain gene by N
nucleotide addition at the D-D joint.
In contrast to the dramatic levels of diversity facilitated by
the incorporation of N nucleotides in all chains of the TCR,
TCRs do not appear to undergo somatic hypermutation following antigen contact at any appreciable rate.
Table 7-8 compares the mechanisms of the generation
and expression of diversity in BCR versus TCR molecules.
The Organization and Expression of Lymphocyte Receptor Genes
Gene family
α
β
γ
δ
V region segments
D region segments
|
CHAPTER 7
253
J region segments
5'
5'
5'
3'
3'
3'
5'
5'
5'
3'
3'
3'
5'
5'
5'
3'
3'
3'
5'
5'
5'
3'
3'
3'
FIGURE 7-19 The locations of the 12-bp and 23-bp RSS spacers in TCR genes. Note that the spacers on either side of the D and
D gene segments are different, allowing for the occurrence of recombined genes bearing two copies of the D . Duplication of D segments
has not been observed in vivo.
TCR Expression Is Controlled by Allelic Exclusion
Allelic exclusion occurs among TCR genes almost as efficiently
as among BCR genes, with one exception: some fraction of
T cells bear more than one TCR chain. However, it would be
unlikely, although not impossible, that a T cell bearing two
chains would be able to recognize more than one antigen, given
the complexity of T-cell antigen recognition. In Chapter 8,
we explain how T cells recognize an antigen made up of a
foreign peptide presented to the immune system in a specialized groove in a Major Histocompatibility Complex (MHC)
antigen. As we will learn in Chapter 9, TCRs are tested twice
within the thymus to determine their functionality before the
T cells are released into the periphery. First, T cells are tested to
ensure that the and chains can pair into a functioning TCR
that is capable of recognizing self MHC antigens bearing self
peptides with low to moderate affinity (positive selection). This
is necessary because the T cell must have some minimal capacity to recognize self MHC if it is to be able to bind to a peptide
presented on the surface of the MHC antigen. Then, T cells
bearing receptors that recognize self peptides in association
with MHC antigens at high affinity are eliminated in order to
eliminate the generation of autoreactive receptors (negative
selection). The likelihood that a single TCR chain could pair
with more than one TCR chain and meet all these stringent
criteria is extremely small. The presence of even one
TCR
capable of binding self antigens at high affinity would result in
the elimination of the cell, irrespective of the specificity of the
receptor formed using an alternative chain.
Although B and T cells use very similar mechanisms for
variable region gene rearrangements, complete rearrangement of Ig genes does not occur in T cells and complete rearrangement of TCR genes does not occur in B cells. Not only
is the recombinase enzyme system differently regulated in
each lymphocyte lineage, but chromatin is uniquely reconfigured in B cells and T cells to allow the recombinase access
only to the appropriate specific antigen receptor genes.
TCR Gene Expression Is Tightly Regulated
As at Ig loci, expression of TCR genes requires the coordinated use of transcriptional promoters and enhancers that
serve at specific stages in T-cell development to make the
receptor chromatin open and accessible for recombination
(see Chapter 9). T cells begin to rearrange the TCR gene
at the pro-T-cell stage. The mouse TCR locus (see Figure
7-18) consists of approximately 40 V genes located upstream
of 2 distinct D J clusters. Immediately upstream of each of
the two D genes is a promoter region (PD 1 and PD 2).
These promoters require the activity of a single enhancer, E ,
located at the 3end of the locus. Activation of the enhancer
serves to facilitate the opening of the chromatin throughout
the D J locus, and activation of each of the two PD promoters localizes the rearrangement to their own set of DJ
regions.
Several TCR enhancer-binding proteins have been identified and, as for B-cell Ig genes, some of these are shared
among cells of many different types. Among the T-cellspecific proteins is GATA-3, which binds in a sequencespecific manner to the enhancer elements of all four TCR
genes. Tissue- and stage-specific expression of enhancerbinding proteins at the TCR loci helps to facilitate locus
accessibility that in turn allows the complex processes of
recombination and transcription of the TCR genes and,
eventually, cell-surface expression of the TCR proteins.
254
PA R T I I I
TABLE 7-8
|
Adaptive Immunity: Antigen Receptors and MHC
Comparison of the mechanisms of the generation and expression of diversity
among B-cell and T-cell receptor molecules
Mechanism
Used in B cells
Used in T cells
Comments
Multiple germ-line V(D)J
genes
Yes
Yes
The mouse V locus has
undergone a severe contraction, and, therefore, only 5%
of mouse light chains are of
the type.
J region diversity is notably
higher in TCR -chain genes
than in other TCR or Ig genes
“Light-chain” segment use
and variable regions
encoded by V and J segments
and variable regions
encoded by V and J segments
“Heavy-chain” segment use
VH regions encoded by V, D
and J segments
and variable regions
encoded by V, D and J segments
Absolute dependence on
RAG1/2 expression
Yes
Yes
Junctional diversity: P nucleotides and N-nucleotide
addition
Yes
Yes
Many fewer N nucleotides
found in Ig light chains
because of developmental
regulation of TdT
Multiple D regions per
recombined chain
Not in Ig heavy chains
Present in TCR , but not TCR
chains
The presence of two D segments allows an additional
site for N nucleotide addition.
Allelic exclusion of receptor
gene expression
Absolute
Allelic exclusion of TCR
genes is not absolute.
On activation, secretes product with the same binding
site as the receptor
Yes
TCR is not found in secreted
form.
Nature of constant region
determines function
Yes; constant region of
secreted antibody product
determines its function.
No secreted product. Constant region of membrane
receptor anchors receptor in
membrane and connects
with signal transduction
complex.
Constant region of membrane receptor anchors
receptor in membrane and
connects with signal transduction complex
Receptor genes undergo
somatic hypermutation following antigen stimulation
Yes
No
The Organization and Expression of Lymphocyte Receptor Genes
|
CHAPTER 7
255
BOX 7-3
CLINICAL FOCUS
Some Immunodeficiencies Result from Impaired Receptor
Gene Recombination
Since the RAG1/2 enzymes are
responsible for both BCR and TCR gene
rearrangements, mutations in the genes
encoding RAG1 or RAG2 have catastrophic consequences for the immune
system, resulting in patients with severe
combined immunodeficiency, or SCID.
The RAG1/2 genes are located on human
chromosome 11 (i.e., an autosome), so
such mutations are inherited in an autosomally recessive manner.
Babies born with nonfunctional RAG1
or RAG2 genes have essentially no circulating B or T cells, although they do
express normal levels of natural killer (NK)
cells and their myeloid and erythroid cells
are normal in number and function.
Because infants receive antibodies passively from the maternal circulation, the
first manifestation of this disease is the
complete absence of T-cell function, and
hence such infants suffer from severe,
recurrent infections with fungi and viruses
that would normally be combated by T
cells in a healthy neonate. SCID used to be
inevitably fatal within the first few months
of life, unless babies were delivered
directly into a sterile environment. The life
span of a SCID patient can be prolonged
by preventing contact with all potentially
harmful microorganisms. Air must be fil-
tered, all food must be sterilized, and
there must be no direct contact with
other people. Such isolation is feasible
only as a temporary measure, pending
treatment.
Nowadays, babies diagnosed sufficiently early with RAG1/2 deficiency can
be successfully treated with bone marrow
transplantation. If a suitable donor can be
found, and transplantation is performed
early in the patient’s first year of life, the
chances of the patient surviving to live a
normal life are 97% or greater.
Patients who carry mutations resulting
in partially active or impaired RAG1 or
RAG2 genes are diagnosed with Omenn
syndrome. Omenn syndrome patients
have no circulating B cells and abnormal
lymph node architecture with deficiencies
in the B-cell zones of the lymph nodes.
Although some T cells are present, they
are oligoclonal (derived from a very few
precursors and hence have very few different receptors), and these T cells tend to
be inappropriately activated. They do not
respond normally to mitogens or to antigens in vitro. This condition, like RAG1/2
deficiency, is invariably fatal unless corrected by bone marrow transplantation.
If deficiency in RAG1/2 genes causes
such a catastrophic loss of immune func-
tion, it makes sense that loss of any of the
other genes implicated in V(D)J recombination would also result in a SCID phenotype. In 1998, it was determined that
deficiency of the Artemis DNA repair
enzyme also results in the loss of T- and
B-cell function, in the face of normal NK
cell activity. Loss of Artemis results in a
SCID syndrome referred to as Athabascan
SCID. Since Artemis is necessary for normal
DNA repair, as well as V(D)J recombination, patients with Athabascan SCID also
suffer from increased radiation sensitivity
of skin fibroblasts and bone marrow cells.
Similarly, human patients suffering from a
deficiency in DNA ligase IV present with
an immunodeficiency that affects T, B, and
NK cells. Since DNA ligase IV is implicated
in DNA repair functions outside the
immune system, such patients also suffer
from generalized chromosomal instability,
radiosensitivity, and developmental and
growth retardation.
A SCID defect in mice, first noted and
characterized by Prof. Melvin Bosma, results
from a nonsense mutation in the gene
encoding the DNA-PKcs. Although a similar
recessive mutation has been found in Arabian horses, few, if any, human cases of
SCID have been reported carrying this
mutation.
S U M M A R Y
■
■
■
The antigen receptor on the surface of a B cell is an Ig with
the same binding site as the antibody the B cell’s progeny
will eventually secrete.
The number of different antibody molecules that can be made
by a single individual mouse or human is vast: close to 1013–14.
This number is achieved by combinatorial association of multiple gene segments termed V (variable) and J (joining),
which together encode the variable region of the light chain
of the Ig molecule and V, D (diversity), and J, which encode
the variable region of the heavy chain of the Ig molecule.
V(D)J recombination occurs at the level of the DNA and
is catalyzed by the lymphoid-specific enzymes RAG1 and
RAG2.
■
■
RAG1 and RAG2 recognize Recombination Signal
Sequences (RSSs) that are contiguous with the coding
regions. RSSs consist of a conserved heptamer and a conserved nonamer sequence, separated by a spacer region
conserved in length, but not completely in sequence.
Spacer regions have a length corresponding to one turn of
the double helix (12-bp spacer) or two turns of the double
helix (23-bp spacer). RAG1/2 catalyzes recombination
between segments bordered by different spacers.
At the sites of recombination, hairpin cleavage can be
asymmetric, resulting in P nucleotides, and in the heavychain genes, non-templated nucleotide addition can result
in N nucleotides. Exonuclease nibbling can reduce the
256
■
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
number of N and P nucleotides in the final receptor gene
products.
The variable regions of TCRs are generated by the same
general mechanism. Most T cells have receptors of the
type. The variable regions of chains are made up of V and
J segments, whereas those of chains have V, D, and J
regions. Other T cells have receptors. The variable
regions of chains, like those of chains, have V and J
segments. chains are encoded within the TCR locus,
and their variable regions are made up of V, D, and J segments. but not chains can express more than one D
region per chain.
R E F E R E N C E S
Alder, M. N., et al. 2008. Antibody responses of variable lymphocyte receptors in the lamprey. Nature Immunology 9:319–327.
Calame, K., et al. 1980. Mouse Cmu heavy chain immunoglobulin gene segment contains three intervening sequences separating domains. Nature 284:452–455.
Capone, M., et al. 1993. TCR beta and TCR alpha gene enhancers confer tissue- and stage-specificity on V(D)J recombination events. EMBO Journal 12:4335–4346.
Cesari, I. M., and M. Weigert. 1973. Mouse lambda-chain
sequences. Proceedings of the National Academy of Sciences of
the United States of America 70:2112–2116.
Chien, Y., et al. 1984. A third type of murine T-cell receptor
gene. Nature 312:31–35.
Chien, Y. H., N. R. Gascoigne, J. Kavaler, N. E. Lee, and M. M.
Davis. 1984. Somatic recombination in a murine T-cell receptor gene. Nature 309:322–326.
Chun, J. J., D. G. Schatz, M. A. Oettinger, R. Jaenisch, R., and D.
Baltimore. 1991. The recombination activating gene-1 (RAG1) transcript is present in the murine central nervous system.
Cell 64:189–200.
Clatworthy, A. E., M. a. Valencia-Burton, J. E. Haber, and M. A.
Oettinger. 2005. The MRE11-RAD50-XRS2 complex, in addition to other non-homologous end-joining factors, is required
for V(D)J joining in yeast. Journal of Biological Chemistry
280:20,247–20,252.
Cooper, M. D., and M. N. Alder. 2006. The evolution of adaptive
immune systems. Cell 124:815–822.
Davis, M. M., et al. 1980. An immunoglobulin heavy-chain gene
is formed by at least two recombinational events. Nature
283:733–739.
Davis, M. M., Y. H. Chien, N. R. Gascoigne, and S. M. Hedrick.
1984. A murine T cell receptor gene complex: Isolation, structure and rearrangement. Immunological Reviews 81:235–258.
Early, P., H. Huang, M. Davis, K. Calame, and L. Hood. 1980.
An immunoglobulin heavy chain variable region gene is generated from three segments of DNA: VH, D and JH. Cell
19:981–992.
Early, P., et al. 1980. Two mRNAs can be produced from a single
immunoglobulin mu gene by alternative RNA processing
pathways. Cell 20:313–319.
Early, P.W., M. M. Davis, D. B. Kaback, N. Davidson, and L.
Hood. 1979. Immunoglobulin heavy chain gene organization in
mice: Analysis of a myeloma genomic clone containing variable
and alpha constant regions. Proceedings of the National Academy of Sciences of the United States of America 76:857–861.
Ernst, P., and S. T. Smale. 1995. Combinatorial regulation of
transcription II: The immunoglobulin mu heavy chain gene.
Immunity 2:427–438.
Flajnik, M. F., and M. Kasahara. 2004. Origin and evolution of
the adaptive immune system: Genetic events and selective
pressures. Nature Reviews Genetics 11:47–59.
Gascoigne, N. R., Y. Chien, D. M. Becker, J. Kavaler, and M. M.
Davis. 1984. Genomic organization and sequence of T-cell
receptor beta-chain constant- and joining-region genes.
Nature 310:387–391.
Gearhart, P. J. 2004. The birth of molecular immunology. Journal of Immunology 173:4259.
Gilfillan, S., A. Dierich, M. Lemeur, C. Benoist, and D. Mathis.
1993. Mice lacking TdT: Mature animals with an immature
lymphocyte repertoire. Science 261:1175–1178.
Hedrick, S. M., D. I. Cohen, E. A. Nielsen, and M. M. Davis.
1984. Isolation of cDNA clones encoding T cell-specific
membrane-associated proteins. Nature 308:149–153.
Hedrick, S. M., E. A. Nielsen, J. Kavaler, D. I. Cohen, and M.
M. Davis. 1984. Sequence relationships between putative
T-cell receptor polypeptides and immunoglobulins. Nature
308:153–158.
Herrin, B. R., and M. D. Cooper. 2010. Alternative adaptive
immunity in jawless vertebrates. Journal of Immunology
185:1367.
Hood, L., et al. 1981. Two types of DNA rearrangements in
immunoglobulin genes. Cold Spring Harbor Symposia on
Quantitative Biology 45(Pt 2):887–898.
Hozumi, N., and S. Tonegawa. 1976. Evidence for somatic rearrangement of immunoglobulin genes coding for variable and
constant regions. Proceedings of the National Academy of Sciences of the United States of America 73:3628–3632.
Ju, Z., et al. 2007. Evidence for physical interaction between the
immunoglobulin heavy chain variable region and the 3 regulatory region. Journal of Biological Chemistry 282:35,169–
35,178.
Kasahara, M., T. Suzuki, and L. D. Pasquier. 2004. On the origins of the adaptive immune system: Novel insights from
invertebrates and cold-blooded vertebrates. Trends in Immunology 25:105–111.
The Organization and Expression of Lymphocyte Receptor Genes
Kavaler, J., M. M. Davis, and Y. Chien. 1984. Localization of a
T-cell receptor diversity-region element. Nature 310:421–423.
Koralov, S. B., T. I. Novobrantseva, J. Konigsmann, A. Ehlich,
and K. Rajewsky. 2006. Antibody repertoires generated by VH
replacement and direct VH to JH joining. Immunity 25:43–53.
Landau, N. R., D. G. Schatz, M. Rosa, and D. Baltimore. 1987.
Increased frequency of N-region insertion in a murine pre-B-cell
line infected with a terminal deoxynucleotidyl transferase retroviral expression vector. Molecular and Cellular Biology 7:3237–3243.
Mansilla-Soto, J., and P. Cortes. 2003. VDJ recombination: Artemis and its in vivo role in hairpin opening. Journal of Experimental Medicine 197:543–547.
Oestreich, K. J., et al. 2006. Regulation of TCR beta gene assembly
by a promoter/enhancer holocomplex. Immunity 24:381–391.
Oettinger, M. A., D. G. Schatz, C. Gorka, and D. Baltimore.
1990. RAG-1 and RAG-2, adjacent genes that synergistically
activate V(D)J recombination. Science 248:1517–1523.
Pancer, Z., and M. D. Cooper. 2006. The evolution of adaptive
immunity. Annual Review of Immunology 24:497–518.
Rivera-Munoz, P., et al. 2007. DNA repair and the immune
system: From V(D)J recombination to aging lymphocytes.
European Journal of Immunology 37(Suppl 1):S71–82.
Rooney, S., et al. 2003. Defective DNA repair and increased
genomic instability in Artemis-deficient murine cells. Journal
of Experimental Medicine 197:553–565.
Saito, H., et al. 1984. Complete primary structure of a heterodimeric T-cell receptor deduced from cDNA sequences. Nature
309:757–762.
Saito, H., et al. 1984. A third rearranged and expressed gene in
a clone of cytotoxic T lymphocytes. Nature 312:36–40.
|
CHAPTER 7
257
Wei, X. C., et al. 2005. Characterization of the proximal
enhancer element and transcriptional regulatory factors for
murine recombination activating gene-2. European Journal of
Immunology 35:612–621.
Weigert, M. G., I. M. Cesari, S. J. Yonkovich, and M. Cohn.
1970. Variability in the lambda light chain sequences of mouse
antibody. Nature 228:1045–1047.
Zha, S., F. W. Alt, H. L. Cheng, J. W. Brush, and G. Li. 2007.
Defective DNA repair and increased genomic instability in
Cernunnos-XLF-deficient murine ES cells. Proceedings of the
National Academy of Sciences of the United States of America
104:4518–4523.
Zhang, Z. 2007. VH replacement in mice and humans. Trends in
Immunology 28:132–137.
Useful Web Sites
http://imgt.org The creators of this website monitor the
literature for information on the numbers of genes encoding Ig and TCRs in a variety of species, and regularly
update information regarding the numbers of gene segments that have been sequenced. Also includes information on other gene families related to the immune system,
such as the MHC.
http://cellular-immunity.blogspot.com/2007/12/vdjrecombination.html A clear and brief explanation of
V(D)J recombination, with clickable links.
http://highered.mcgraw-hill.com/sites/
0072556781/student_view0/chapter32/animation_
quiz_2.html A clear, even if rather simplistic, animation
Schatz, D. G., and Y. Ji. 2011 Recombination centres and the
orchestration of V(D)J recombination. Nature Reviews Immunology 11:251–263.
of V(D)J recombination, which offers a “big picture” of the
process. May be useful to view before launching into more
detailed reading. Note that this animation speaks of there
being five constant regions, rather than the eight referred to
in this chapter; this is because two of the heavy-chain
classes— and —can be divided into subclasses, which are
ignored in this animation.
Schatz, D. G., M. A. Oettinger, and D. Baltimore. 1989. The V(D)
J recombination activating gene, RAG-1. Cell 59:1035–1048.
http://users.rcn.com/jk imball.ma.ultranet/
BiologyPages/A/AgReceptorDiversity.html A useful
Schatz, D. G., and D. Baltimore. 1988. Stable expression of
immunoglobulin gene V(D)J recombinase activity by gene
transfer into 3T3 fibroblasts. Cell 53:107–115.
Tonegawa, S., C. Brack, N. Hozumi, G. Matthyssens, and R.
Schuller. 1977. Dynamics of immunoglobulin genes. Immunological Reviews 36:73–94.
Tonegawa, S., C. Brack, N. Hozumi, and V. Pirrotta. 1978. Organization of immunoglobulin genes. Cold Spring Harbor Symposia in Quantitative Biology 42(Pt 2):921–931.
Tonegawa, S., C. Brack, N. Hozumi, and R. Schuller. 1977. Cloning of an immunoglobulin variable region gene from mouse
embryo. Proceedings of the National Academy of Sciences of the
United States of America 74:3518–3522.
Tonegawa, S., N. Hozumi, G. Matthyssens, and R. Schuller.
1977. Somatic changes in the content and context of immunoglobulin genes. Cold Spring Harbor Symposia in Quantitative
Biology 41(Pt 2):877–889.
review of antigen receptor diversity with clickable links. Do
not be put off by minor differences between this website and
this chapter in the numbers of gene segments identified as
belonging to particular families—these numbers are constantly being updated, and they vary among different individuals and inbred strains of animals.
http://biophilessurf.info/immuno.html A useful collection of databases pertaining to immunological topics,
including RAG1 mutations that give rise to immunodeficiencies.
www.ncbi.nlm.nih.gov The National Center for Biotechnology Information (NCBI) site offers library tools as well as
numerous sequence analysis and protein structure resources
(under the “Resources” tab) pertinent to immunology.
PA R T I I I
258
S T U D Y
|
Adaptive Immunity: Antigen Receptors and MHC
Q U E S T I O N S
1. Indicate whether each of the following statements is true or
false. If you think a statement is false, explain why.
a. In generating a B-cell receptor gene, V segments some-
times join to C segments.
b. In generating a T-cell receptor gene, V segments some-
times join to C .
c. The switch in constant region use from to IgM to IgD is
mediated by DNA rearrangements.
d. Although each B cell carries two alleles encoding the
immunoglobulin heavy and light chains, only one allele
is expressed.
e. Like the variable regions of the heavy chain of the B-cell
Ig receptor, TCR variable genes are all encoded in three
segments.
2. Explain why a VH segment cannot join directly with a JH
segment in heavy-chain gene rearrangement.
3. For each incomplete statement below, select the phrase(s)
that correctly completes the statement. More than one
choice may be correct.
a. Recombination of Ig gene segments serves to:
(1) Promote Ig diversification
(2) Assemble a complete Ig coding sequence
(3) Allow changes in coding information during B-cell
maturation
(4) Increase the affinity of Ig for antibody
(5) All of the above
b. Kappa and lambda light-chain genes:
(1) Are located on the same chromosome
(2) Associate with only one type of heavy chain
(3) Can be expressed by the same B cell
(4) All of the above
(5) None of the above
c. Generation of combinatorial diversity within the vari-
able regions of Ig involves
mRNA splicing
DNA rearrangement
Recombination signal sequences
One-turn/two-turn joining rule
Switch sites
d. The mechanism that permits Ig to be synthesized in
either a membrane-bound or secreted form is:
(1) Allelic exclusion
(2) Codominant expression
(3) Class switch recombination
(4) The one-turn/two-turn joining rule
(5) Differential RNA processing
e. During Ig VH recombination, the processes that contribute to additional diversity at the third complementarity-determining region of Ig variable regions include:
(1) Introduction of the DH gene segments into the
heavy-chain V gene
(2) mRNA splicing of the membrane form of the C
region of the constant chain
(3) Exonuclease cleavage of the ends of the gene segments
(4) P nucleotide addition
(5) N nucleotide addition
(1)
(2)
(3)
(4)
(5)
4. You have been given a cloned mouse myeloma cell line that
secretes IgG with the molecular formula 22. Both the
heavy and light chains in this cell line are encoded by genes
derived from allele 1 (i.e., the first of the two homologous
alleles encoding each type of chain). Indicate the form(s) in
which each of the genes listed below would occur in this cell
line using the following symbols: G germ-line form; P
productively rearranged form; NP nonproductively rearranged form. State the reason for your choice in each case.
(a) Heavy-chain allele 1
(b) Heavy-chain allele 2
(c) -chain allele 1
(d) -chain allele 2
(e) -chain allele 1
(f) -chain allele 2
5. You have identified a B-cell lymphoma that has made non-
productive rearrangements for both heavy-chain alleles.
What is the arrangement of its light-chain DNA? Why?
6. The random addition of nucleotides by TdT is a wasteful
and potentially risky evolutionary strategy. State why it
may be disadvantageous to the organism and why, therefore, you think it is sufficiently useful to have been retained
during vertebrate evolution.
7. Are there any differences in the genetic strategies used to
generate complete V genes in T and BCRs?
8. What known features of the TCR did Hedrick, Davis, and
colleagues use in their quest to isolate the TCR genes?
9. The following figure describes the end of a V region
sequence and the beginning of the D region sequence to
which is it about to be joined. Arrows mark the cleavage
points where the RAG1/RAG2 complex will make the cut
and recombination will be targeted.
V
5'
3'
AGCATC
TATCGA
TCGTAG
ATAGCT
D
3'
5'
a. Is this a heavy-chain or a light-chain sequence? How do
you know?
b. What DNA sequence structure would you find just
downstream of the AGCATC sequence immediately
adjacent to the 3 end of the V segment?
c. Here is one possible VD joint structure formed after
recombination between these two gene segments:
5'
3'
V
AGCATCGACGCCGTATCGA
D
TCGTAGCTGCGGCATAGCT
(1) Which residue(s) MAY be P-region nucleotide(s)?
(2) Can we know for certain that this residue is a
P-region nucleotide?
(3) Which residues must have been added by TdT, and
therefore must be N-region nucleotides?
(4) Can we know for certain whether a residue is an
N-region nucleotide?
The Organization and Expression of Lymphocyte Receptor Genes
10. Challenge data question. In Classic Experiment Box 7.1,
Figure 1, we see the pattern of bands that Hozumi and
Tonegawa’s experiment would have revealed, using the particular cell line that they used. Below, see a pair of gels that
represent the results of a hypothetical experiment performed using the same general protocol. In this hypothetical
experiment, our probes correspond to either the V region or
the C region. Furthermore, the investigators used a different
tumor cell line and a different restriction endonuclease.
Germline DNA
DNA from antibody producing cells
C-region
probe
V-region
probe
C-region
probe
V-region
probe
1
2
3
4
|
CHAPTER 7
259
a. Why are there two C-region bands in the germ-line blot?
b. Why are these two bands still present in the myeloma
blot, and why are there two bands recognized by the
V-region probe on the myeloma blot?
c. From this plot, would you hypothesize that the cell had
achieved a successful arrangement at the first allele?
Why or why not?
d. How would you prove your answer to part c?
This page lelt intentionally blank.
8
The Major
Histocompatibility
Complex and Antigen
Presentation
A
lthough both T and B cells use surface
molecules to recognize antigen, they
accomplish this in very different ways. In
contrast to antibodies or B-cell receptors,
which can recognize an antigen alone, T-cell receptors
only recognize pieces of antigen that are positioned on
the surface of other cells. These antigen pieces are held
within the binding groove of a cell surface protein called the
major histocompatibility complex (MHC) molecule,
encoded by a cluster of genes collectively called the MHC
locus. These fragments are generated inside the cell
following antigen digestion, and the complex of the
antigenic peptide plus MHC molecule then appears on
the cell surface. MHC molecules thus act as a cell surface
vessel for holding and displaying fragments of antigen so
that approaching T cells can engage with this molecular
complex via their T-cell receptors.
The MHC got its name from the fact that the genes in
this region encode proteins that determine whether a
tissue transplanted between two individuals will be
accepted or rejected. The pioneering work of Benacerraf,
Dausset, and Snell helped to characterize the functions
controlled by the MHC, specficially, organ transplant fate
and the immune responses to antigen, resulting in the
1980 Nobel Prize in Medicine and Physiology for the
trio (see Table 1-2). Follow-up studies by Rolf
Zinkernagel and Peter Doherty illustrated that the
proteins encoded by these genes play a seminal role in
adaptive immunity by showing that T cells recognize
MHC proteins as well as antigen. Structural studies done
by Don Wiley and others showed that different MHC
proteins bind and present different antigen fragments.
There are many alleles of most MHC genes, and the
specific alleles one inherits play a significant role in
susceptibility to disease, including the development of
autoimmunity. The mechanisms by which this family of
molecules exerts such a strong influence on the
development of immunity to nearly all types of antigens
has become a major theme in immunology, and has taken
Ribbon diagram of a human MHC class I
molecule (blue) with a space-filled peptide
(orange) held in the binding groove.
■
The Structure and Function of MHC Molecules
■
General Organization and Inheritance of the MHC
■
The Role of the MHC and Expression Patterns
■
■
The Endogenous Pathway of Antigen Processing
and Presentation
The Exogenous Pathway of Antigen Processing and
Presentation
■
Cross-Presentation of Exogenous Antigens
■
Presentation of Nonpeptide Antigens
the study of the MHC far beyond its origins in the field of
transplantation biology.
There are two main classes of MHC molecules: class I
and class II. These two molecules are very similar in their
final quaternary structure, although they differ in how
they create these shapes via primary through quaternary
protein arrangements. Class I and class II MHC molecules
also differ in terms of which cells express them and in the
source of the antigens they present to T cells. Class I
molecules are present on all nucleated cells in the body
and specialize in presenting antigens that originate from
261
262
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
the cytosol, such as viral proteins. These are presented to
CD8⫹ T cells, which recognize and kill cells expressing
such intracellular antigens. In contrast, class II MHC
molecules are expressed almost exclusively on a subset of
leukocytes called antigen-presenting cells (APCs) and
specialize in presenting antigens from extracellular spaces
that have been engulfed by these cells, such as fungi and
extracellular bacteria. Once expressed on the cell surface,
the MHC class II molecule presents the antigenic peptide
to CD4⫹ T cells, which then become activated and go on
to stimulate immunity directed primarily toward
destroying extracellular invaders.
This chapter will begin by discussing the structure of
MHC molecules, followed by the genetic organization of
the DNA region that encodes these proteins and
inheritance patterns. We then turn to the function of the
MHC in regulating immunity, including reference to
several seminal studies in these areas. In this section we
also discuss regulation of MHC expression. This is
followed by a detailed discussion of the cellular pathways
that lead to antigen degradation and association with each
type of MHC molecule (antigen processing) and the
appearance of these MHC-peptide complexes on the cell
surface for recognition by T cells (antigen presentation).
The role of particular APCs in these processes will also be
presented (no pun intended!). The chapter concludes with
a discussion of unique processing and presentation
pathways, such as cross-presentation and the handling of
nonpeptide antigens by the immune system.
The Structure and Function
of MHC Molecules
Class I and class II MHC molecules are membrane-bound
glycoproteins that are closely related in both structure and
function. Both classes of MHC molecule have been isolated
and purified, and the three-dimensional structures of their
extracellular domains have been resolved by x-ray crystallography. These membrane glycoproteins function as highly
specialized antigen-presenting molecules with grooves that
form unusually stable complexes with peptide ligands, displaying them on the cell surface for recognition by T cells via
T-cell receptor (TCR) engagement. In contrast, class III
MHC molecules are a group of unrelated proteins that do
not share structural similarity or function with class I and II
molecules, although many of them do participate in other
aspects of the immune response.
Class I Molecules Have a Glycoprotein Heavy
Chain and a Small Protein Light Chain
Two polypeptides assemble to form a single class I MHC
molecule: a 45-kilodalton (kDa) chain and a 12-kDa
2-microglobulin molecule (Figure 8-1, left). The chain is
organized into three external domains (1, 2, and 3), each
approximately 90 amino acids long; a transmembrane domain
of about 25 hydrophobic amino acids followed by a short
stretch of charged (hydrophilic) amino acids; and a cytoplasmic anchor segment of 30 amino acids. Its companion,
2-microglobulin, is similar in size and organization to the
3 domain. 2-microglobulin does not contain a transmembrane region and is noncovalently bound to the MHC class I
chain. Sequence data reveal strong homology between the
3 domain of MHC class I, 2-microglobulin, and the constant-region domains found in immunoglobulins.
The 1 and 2 domains interact to form a platform of
eight antiparallel strands spanned by two long -helical
regions (Figure 8-2a). The structure forms a deep groove, or
cleft, with the long helices as sides and the strands of the
sheet as the bottom (Figure 8-2b). This peptide-binding
groove is located on the top surface of the class I MHC molecule, and it is large enough to bind a peptide of 8 to 10 amino
acids. During the x-ray crystallographic analysis of class I
molecules, small noncovalently associated peptides that had
co-crystallized with the protein were found in the groove.
The big surprise came when these peptides were later discovered to be processed self-proteins bound to this deep groove
and not the foreign antigens that were expected.
The 3 domain and 2-microglobulin are organized into
two pleated sheets each formed by antiparallel strands of
amino acids. As described in Chapter 3, this structure, known
as the immunoglobulin fold, is characteristic of immunoglobulin domains. Because of this structural similarity, which
is not surprising given the considerable sequence similarity
with the immunoglobulin constant regions, class I MHC molecules and 2-microglobulin are classified as members of the
immunoglobulin superfamily (see Figure 3-19). The 3
domain appears to be highly conserved among class I MHC
molecules and contains a sequence that interacts strongly with
the CD8 cell surface molecule found on TC cells.
All three molecules (class I chain, 2-microglobulin,
and a peptide) are essential to the proper folding and expression of the MHC-peptide complex on the cell surface. This is
demonstrated in vitro using Daudi tumor cells, which are
unable to synthesize 2-microglobulin. These tumor cells
produce class I MHC chains but do not express them on
the cell membrane. However, if Daudi cells are transfected
with a functional gene encoding 2-microglobulin, class I
molecules will appear on the cell surface.
Class II Molecules Have Two Non-Identical
Glycoprotein Chains
Class II MHC molecules contain two different polypeptide
chains, a 33-kDa chain and a 28-kDa chain, which associate by noncovalent interactions (see Figure 8-1, right). Like
class I chains, class II MHC molecules are membranebound glycoproteins that contain external domains, a transmembrane segment, and a cytoplasmic anchor segment. Each
The Major Histocompatibility Complex and Antigen Presentation
|
CHAPTER 8
Class I molecule
α2
Class II molecule
α1
Peptide-binding
groove
α1
Membrane - distal
domains
S
S
S
α3
S
β1
S S
S
Membrane -proximal
domains
(Ig-fold structure)
263
S
S
β2-microglobulin
α2
S
S
S
β2
Transmembrane segment
Cytoplasmic tail
FIGURE 8-1 Schematic diagrams of class I and class II MHC molecules showing the external domains, transmembrane segments, and cytoplasmic tails. The peptide-binding groove is formed by the membrane-distal domains in both class I and class II molecules
(blue). The membrane-proximal domains (green and red) possess the basic immunoglobulin-fold structure; thus, both class I and class II MHC
molecules are classified as members of the immunoglobulin superfamily.
chain in a class II molecule contains two external domains: 1
and 2 domains in one chain and 1 and 2 domains in the
other (Figure 8-2c). The membrane-proximal 2 and 2
domains, like the membrane-proximal 3/2-microglobulin
domains of class I MHC molecules, bear sequence similarity
to the immunoglobulin-fold structure. For this reason, class II
MHC molecules are also classified in the immunoglobulin
superfamily. The 1 and 1 domains form the peptide-binding
groove for processed antigen (Figure 8-2d). Although similar
to the peptide-binding groove of MHC class I, this groove is
thus formed by the association of two separate chains.
X-ray crystallographic analysis reveals the similarity
between these two classes of molecule, strikingly apparent when
class I and class II molecules are superimposed (Figure 8-3).
Interestingly, despite the fact that these two structures are
encoded quite differentially (one chain versus two), the final
quaternary structure is similar and retains the same overall
function: the ability to bind antigen and present it to T cells.
The peptide-binding groove of class II molecules, like that found
in class I molecules, is composed of a floor of eight antiparallel
strands and sides of antiparallel helices, where peptides
typically ranging from 13 to 18 amino acids can bind. The class
II molecule lacks the conserved residues in the class I molecule
that bind to the terminal amino acids of short peptides, and
therefore forms more of an open pocket. In this way, class I
presents more of a socket-like opening, whereas class II pos-
sesses an open-ended groove. The functional consequences of
these differences in fine structure will be explored below.
Class I and II Molecules Exhibit Polymorphism
in the Region That Binds to Peptides
Several hundred different allelic variants of class I and II
MHC molecules have been identified in humans. Any one
individual, however, expresses only a small number of these
molecules—up to six different class I molecules and 12 or
more different class II molecules. Yet this limited number of
MHC molecules must be able to present an enormous array
of different antigenic peptides to T cells, permitting the
immune system to respond specifically to a wide variety of
antigenic challenges. Thus, peptide binding by class I and II
molecules does not exhibit the fine specificity characteristic
of antigen binding by antibodies and T-cell receptors. Instead,
a given MHC molecule can bind numerous different peptides,
and some peptides can bind to several different MHC molecules. Because of this broad specificity, the binding between a
peptide and an MHC molecule is often referred to as “promiscuous.” This promiscuity of peptide binding allows many
different peptides to match up with the MHC binding groove
and for exchange of peptides to happen on occasion, unlike
the relatively stable, high-affinity cognate interactions of antibodies and T-cell receptors with their specific ligands.
264
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
(a) MHC class I
(b) MHC class I peptide-binding groove
Peptide-binding
groove
α1 domain
α1 domain
α helix
α2 domain
β sheets
α2 domain
β2-microglobulin
α3 domain
(c) MHC class II
(d) MHC class II peptide-binding groove
Peptide-binding
groove
α1 domain
α helix
β1 domain
α1 domain
β sheets
β1 domain
β2 domain
α2 domain
FIGURE 8-2 Representations of the three-dimensional
structure of the external domains of human MHC class I and
class II molecules based on x-ray crystallographic analysis.
(a, c) Side views of class I and II, respectively, in which the
strands are depicted as thick arrows and the helices as spiral
ribbons. Disulfide bonds are shown as two interconnected
spheres. The 1 and 2 domains of class I and the 1 and 1
domains of class II interact to form the peptide-binding groove. Note
the immunoglobulin-fold structure of each membrane proximal
domain, including the 2-microglobulin molecule. (b, d) The 1
(dark blue) and 2 domains (light blue) of class I and the 1 (dark
blue) and 1 (light blue) domains of class II as viewed from the
top, showing the peptide-binding grooves, consisting of a base of
antiparallel strands and sides of helices for each molecule.
This groove can accommodate peptides of 8 to 10 residues for
class I and 13 to 18 residues for class II.
The Major Histocompatibility Complex and Antigen Presentation
|
CHAPTER 8
265
no such requirement for class II peptide binding. The association of peptide and MHC molecule is very stable under
physiologic conditions (Kd values range from ~ 106 to
1010). Thus, most of the MHC molecules expressed on the
membrane of a cell will be associated with a peptide.
FIGURE 8-3 The peptide-binding groove of a human class II
MHC molecule (blue), superimposed over the corresponding
regions of a human class I MHC molecule (red). Overlapping
regions are shown in white. These two molecules create very similar
binding grooves such that most of the structural differences (long
stretches, kinks, bends, etc.) lie outside the peptide-binding regions.
[From J. H. Brown et al., 1993, Nature 364:33.]
Given the similarities in the structures of the peptidebinding grooves in class I and II MHC molecules, it is not
surprising that they exhibit some common peptide-binding
features (Table 8-1). In both types of MHC molecules, peptide ligands are held in a largely extended conformation that
runs the length of the groove. The peptide-binding groove in
class I molecules is blocked at both ends, whereas the ends
of the groove are open in class II molecules (Figure 8-4). As
a result of this difference, class I molecules bind peptides that
typically contain 8 to 10 amino acid residues, whereas the
open groove of class II molecules accommodates slightly
longer peptides of 13 to 18 amino acids. Another difference
is that class I binding requires that the peptide contain specific amino acid residues near the N and C termini; there is
TABLE 8-1
Class I MHC-Peptide Interaction
Class I MHC molecules bind peptides and present them to
CD8⫹ T cells. These peptides are often derived from endogenous intracellular proteins that are digested in the cytosol. The
peptides are then transported from the cytosol into the endoplasmic reticulum (ER), where they interact with class I MHC
molecules. This process, known as the cytosolic or endogenous
processing pathway, is discussed in detail later in this chapter.
Each human or mouse cell can express several unique
class I MHC molecules, each with slightly different rules for
peptide binding. Because a single nucleated cell expresses
about 105 copies of each of these unique class I molecules,
each with its own peptide promiscuity rules, many different
peptides will be expressed simultaneously on the surface of a
nucleated cell. This means that although many of the peptide
fragments of a given foreign antigen will be presented to
CD8⫹ T cells, the set of MHC class I alleles inherited by each
individual will determine which specific peptide fragments
from a larger protein get presented.
The bound peptides isolated from different class I molecules have two distinguishing features: they are 8 to 10
amino acids in length, most commonly 9, and they contain
specific amino acid residues at key locations in the
sequence. The ability of an individual class I MHC molecule to bind to a diverse spectrum of peptides is due to the
presence of the same or similar amino acid residues at
these key positions along the peptides (Figure 8-5). Because
these amino acid residues anchor the peptide into the
groove of the MHC molecule, they are called anchor residues. The side chains of the anchor residues in the peptide
are complementary with surface features of the binding
groove of the class I MHC molecule. The amino acid residues lining the binding sites vary among different class I
allelic variants, and this is what determines the chemical
Peptide binding by class I and class II MHC molecules
Class I molecules
Class II molecules
Peptide-binding domain
1/2
1/1
Nature of peptide-binding groove
Closed at both ends
Open at both ends
General size of bound peptides
8–10 amino acids
13–18 amino acids
Peptide motifs involved in
binding to MHC molecule
Anchor residues at both ends of peptide;
generally hydrophobic carboxyl-terminal
anchor
Conserved residues distributed along the
length of the peptide
Nature of bound peptide
Extended structure in which both ends
interact with MHC groove but middle
arches up away from MHC molecule
Extended structure that is held at a constant
elevation above the floor of the MHC groove
266
(a)
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
α1 domain
Eluted
from
MHC
protein 1
Eluted
from
MHC
protein 2
1
2
3
4
5
6
7
8
9
H3 N
V
G
P
Q
K
N
E
N
L
COO−
H3 N
S
G
P
R
K
A
I
A
L
COO−
H3N
V
G
P
S
G
K
Y
F
I
COO−
H3N
S
G
P
E
R
I
L
S
L
COO−
H3 N
S
Y
F
P
E
I
T
H
I
COO−
H3N
T
Y
Q
R
T
R
A
L
V
COO−
H3N
S
Y
I
G
S
I
N
N
I
COO−
A = alanine
E = glutamic acid
F = phenylalanine
G = glycine
H = histidine
I = isoleucine
α2 domain
(b)
α1 domain
K = lysine
L = leucine
N = asparagine
P = proline
Q = glutamine
R = arginine
S = serine
T = threonine
V = valine
Y = tyrosine
FIGURE 8-5 Examples of anchor residues (blue) in nonameric peptides eluted from two different class I MHC molecules. Anchor residues, at positions 2/3 and 9, that interact with the
class I MHC molecule tend to be hydrophobic amino acids. The two
MHC proteins bind peptides with different anchor residues. [Data
from V. H. Engelhard, 1994, Current Opinion in Immunology 6:13.]
β1 domain
FIGURE 8-4 MHC class I and class II molecules with bound
peptides. (a) Ribbon model of human class I molecule HLA-A2 (1
in dark blue, 2 in light blue) with a peptide from the HIV-1 envelope
protein GP120 (space-filling, orange) in the binding groove. (b) Ribbon model of human class II molecules HLA-DR1 with the DR chain
shown in dark blue and the DR chain in light blue. The peptide
(space-filling, orange) in the binding groove is from influenza hemagglutinin (amino acid residues 306–318). [Part (a) PDB ID 1HHG, part (b)
PDB ID 1DLH.]
identity of the anchor residues that can interact with a
given class I molecule.
In a critical study of peptide binding by MHC molecules,
peptides bound by two different class I MHC proteins were
released chemically and analyzed by high-performance liquid chromatography (HPLC) mass spectrometry. More than
2000 distinct peptides were found among the peptide ligands
released from these two class I MHC molecules. Since there
are approximately 105 copies of each class I protein per cell,
it is estimated that each of the 2000 distinct peptides is presented with a frequency of between 100 and 4000 copies per
cell. Evidence suggests that even a single MHC-peptide complex may be sufficient to target a cell for recognition and lysis
by a cytotoxic T lymphocyte with a receptor specific for that
target structure.
All peptides examined to date that bind to class I molecules contain a carboxyl-terminal anchor (e.g., see position
9 in Figure 8-5). These anchors are generally hydrophobic
residues (e.g., leucine, isoleucine), although a few charged
amino acids have been reported. Besides the anchor residue
found at the carboxyl terminus, another anchor is often
found at the second, or second and third, positions at the
amino-terminal end of the peptide. In general, any peptide
of the correct length that contains the same or chemically
similar anchor residues will bind to the same class I MHC
molecule. Knowledge of these key positions and the chemical restrictions for amino acids at these positions may allow
for more nuanced clinical design studies, such as future vaccines targeted at eliciting protective immunity to particular
pathogens.
X-ray crystallographic analyses of class I MHC-peptide
complexes have revealed how the peptide-binding groove in
a given MHC molecule can interact stably with a broad spectrum of different peptides. The anchor residues at both ends
of the peptide are buried within the binding groove, thereby
holding the peptide firmly in place (see Figure 8-4a). Nonameric peptides are bound preferentially. The main contacts
between class I MHC molecules and the peptides they bind
The Major Histocompatibility Complex and Antigen Presentation
Bulge
4
5
6
N 1
2
3
7
8
9
C
Hydrogen bonds with
MHC molecule
FIGURE 8-6 Conformation of peptides bound to class I
MHC molecules. Schematic diagram depicting, in a side view, the
conformational variance in MHC class I-bound peptides of different
lengths. Longer peptides bulge in the middle, presumably interacting more with TCR, whereas shorter peptides lay flat in the groove.
Contact with the MHC molecule is by hydrogen bonds to anchor
residues 2 and/or 3 and 9. [Adapted from P. Parham, 1992, Nature 360:300,
|
CHAPTER 8
267
but the binding characteristics are determined by the central
13 residues. The peptides that bind to a particular class II
molecule often have internal conserved sequence motifs, but
unlike class I–binding peptides, they appear to lack conserved anchor residues (see Table 8-1). Instead, hydrogen
bonds between the backbone of the peptide and the class II
molecule are distributed throughout the binding site rather
than being clustered predominantly at the ends of the site as
is seen in class I–bound peptides. Peptides that bind to class
II MHC molecules contain an internal sequence of 7 to 10
amino acids that provide the major contact points. Generally, this sequence has an aromatic or hydrophobic residue at
the amino terminus and three additional hydrophobic residues in the middle portion and carboxyl-terminal end of the
peptide. In addition, over 30% of the peptides eluted from
class II molecules contain a proline residue at position 2 and
another cluster of prolines at the carboxyl-terminal end. This
relative flexibility contributes to peptide binding promiscuity.
© 1992 Macmillan Magazines Limited]
involve residue 2 at the amino-terminal end and residue 9 at
the carboxyl terminus of the peptide. Between the anchors,
the peptide arches away from the floor of the groove in the
middle (Figure 8-6), allowing peptides that are slightly longer or shorter to be accommodated. Amino acids in the
center of the peptide that arch away from the MHC molecule
are more exposed and presumably can interact more directly
with the T-cell receptor.
Class II MHC-Peptide Interaction
Class II MHC molecules bind and present peptides to CD4⫹
T cells. Like class I molecules, MHC class II molecules can
bind a variety of peptides. These peptides are typically
derived from exogenous proteins (either self or nonself) that
are degraded within the exogenous processing pathway,
described later in this chapter. Most of the peptides associated with class II MHC molecules are derived from self
membrane-bound proteins or foreign proteins internalized
by phagocytosis or by receptor-mediated endocytosis and
then processed through the exogenous pathway.
Peptides recovered from class II MHC-peptide complexes
generally contain 13 to 18 amino acid residues, somewhat
longer than the nonameric peptides that most commonly
bind to class I molecules. The peptide-binding groove in
class II molecules is open at both ends (see Figure 8-4b),
allowing longer peptides to extend beyond the ends, like an
extra-long hot dog in a bun. Peptides bound to class II MHC
molecules maintain a roughly constant elevation on the floor
of the binding groove, another feature that distinguishes
peptide binding to class I and class II molecules.
Peptide-binding studies and structural data for class II
molecules indicate that a central core of 13 amino acids
determines the ability of a peptide to bind class II. Longer
peptides may be accommodated within the class II groove,
General Organization and
Inheritance of the MHC
Every vertebrate species studied to date possesses the tightly
linked cluster of genes that constitute the MHC. As we have
just discussed, MHC molecules have the important job of
deciding which fragments of a foreign antigen will be “seen”
by the host T cells. In general terms, MHC molecules face a
similar ligand binding challenge to that faced, collectively, by
B-cell and T-cell receptors: they must be able to bind a wide
variety of antigens, and they must do so with relatively
strong affinity. However, these immunologically relevant
molecules meet this challenge using very different strategies.
Although B- and T-cell receptor diversity is generated
through genomic rearrangement and gene editing (described
in Chapter 7), MHC molecules have opted for a combination
of peptide binding promiscuity (discussed above) and the
expression of several different MHC molecules on every cell
(described below). Using this clever combined strategy, the
immune system has evolved a way of maximizing the
chances that many different regions, or epitopes, of an antigen will be recognized.
Studies of the MHC gene cluster originated when it was
found that the rejection of foreign tissue transplanted between
individuals in a species was the result of an immune response
mounted against cell surface molecules, now called histocompatibility antigens. In the mid-1930s, Peter Gorer, who was
using inbred strains of mice to identify blood-group antigens,
identified four groups of genes that encode blood-cell antigens.
He designated these I through IV. Work carried out in the 1940s
and 1950s by Gorer and George Snell established that antigens
encoded by the genes in the group designated as II took part in
the rejection of transplanted tumors and other tissue. Snell
called these histocompatibility genes; their current designation
as histocompatibility-2 (H-2, or MHC) genes in the mouse was
in reference to Gorer’s original group II blood-cell antigens.
268
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
Mouse H-2 complex
Complex
H–2
II
MHC class
I
Region
K
IA
IE
H–2K
IA
αβ
IE
αβ
Gene
products
III
I
S
D
TNF-α
Lymphotoxin-α
C′ proteins
H–2D
H–2L*
* Not present in all haplotypes
Human HLA complex
HLA
Complex
II
MHC class
I
III
Region
DP
DQ
DR
Gene
products
DP
αβ
DQ
αβ
DR
αβ
C4, C2, BF
C′ proteins
TNF-α
Lymphotoxin-α
B
C
A
HLA-B
HLA-C
HLA-A
FIGURE 8-7 Comparison of the organization of the major
histocompatibility complex (MHC) in mouse and human. The
MHC is referred to as the H-2 complex in mice and as the HLA complex in humans. In both species, the MHC is organized into a number
of regions encoding class I (pink), class II (blue), and class III (green)
gene products. The class I and II gene products shown in this figure
are considered to be the classical MHC molecules. The class III gene
products include other immune function–related compounds such
as the complement proteins (C’) and tumor necrosis factors (TNF-
and Lymphotoxin ).
The MHC Locus Encodes Three Major
Classes of Molecules
human, this region in mouse is noncontinuous, interrupted
by the class II and III regions (Figure 8-8). Recall that there
are two chains to the MHC class I molecule: the more
variable and antigen-binding chain and the common
-2-microglobulin chain. The chain molecules are encoded
by the K and D regions in mice, with an additional L region
found in some strains, and by the A, B, and C loci in humans.
2-microglobulin is encoded by a gene outside the MHC.
Collectively, these are referred to as classical class I molecules;
all posses the functional capability of presenting protein
fragments of antigen to T cells.
Additional genes or groups of genes within the class I
region of both mouse and human encode nonclassical class I
molecules that are expressed only in specific cell types and
have more specialized functions. Some appear to play a role in
self/nonself discrimination. One example is the class I HLA-G
molecule. These are present on fetal cells at the maternal-fetal
interface and are credited with inhibiting rejection by maternal CD8⫹ T cells by protecting the fetus from identification as
foreign, which may occur when paternally derived antigens
begin to appear on the developing fetus.
Class II MHC molecules are encoded by the IA and IE
regions in mice and by the DP, DQ, and DR regions in
humans (see Figures 8-7 and 8-8). The terminology is somewhat confusing, since the D region in mice encodes class I
MHC molecules, whereas DP, DQ, and DR in humans refers
to class II genes and molecules. Recall that class II molecules
are composed of two chains, both of which interact with
The major histocompatibility complex is a collection of
genes arrayed within a long continuous stretch of DNA on
chromosome 6 in humans and on chromosome 17 in mice.
The MHC is referred to as the human leukocyte antigen
(HLA) complex in humans and as the H-2 complex in mice,
the two species in which these regions have been most studied. Although the arrangement of genes is somewhat different in the two species, in both cases the MHC genes are
organized into regions encoding three classes of molecules
(Figure 8-7):
• Class I MHC genes encode glycoproteins expressed on
the surface of nearly all nucleated cells; the major function of the class I gene products is presentation of
endogenous peptide antigens to CD8⫹ T cells.
• Class II MHC genes encode glycoproteins expressed
predominantly on APCs (macrophages, dendritic cells,
and B cells), where they primarily present exogenous
antigenic peptides to CD4⫹ T cells.
• Class III MHC genes encode several different proteins,
some with immune functions, including components of
the complement system and molecules involved in
inflammation.
Class I MHC molecules were the first discovered and are
expressed in the widest range of cell types. Unlike in the
The Major Histocompatibility Complex and Antigen Presentation
|
CHAPTER 8
269
MOUSE CHROMOSOME 17
H-2
Complex
1500 kb
D
L
Telomere
C′
IE α
LMP2
TAP2
LMP7
TAP1
Oβ
IAβ
IA α
IEβ
I
TNF- α
Lymphotoxin-α
III
II
Oα
Mα
Mβ
Loci
I
K
Class
Centromere
HUMAN CHROMOSOME 6
HLA
4000 kb
KEY
Gene
LMP2, LMP7
TAP1, TAP2
TNF- α, Lymphotoxin- α
HLA-G
HLA-E
HLA-B
HLA-C
HLA-F
Telomere
Centromere
I
TNF- α
Lymphotoxin-α
C′
DRα
III
DRβ
DP β
DP α
Loci
DO α
DM α
DMβ
LMP2
TAP1
LMP7
TAP2
DOβ
DQβ
DQα
II
Class
HLA-A
Complex
Encoded protein
Proteasome-like subunits
Peptide-transporter subunits
Tumor necrosis factor α and lymphotoxin β
FIGURE 8-8 Simplified map of the mouse and human MHC loci. The MHC class I genes are colored red, MHC class II genes are colored blue, and genes in MHC III are colored green. Classical class I genes are labeled in red, classical class II genes are labeled in blue, and the
nonclassical MHC genes are labeled in black. The concept of classical and nonclassical does not apply to class III. Only some of the non-classical
class I genes are shown; the functions of only some of their proteins are known.
antigen. The class II region of the MHC locus encodes both
the chain and the chain of a particular class II MHC
molecule, and in some cases multiple genes are present for
either or both chains. For example, individuals can inherit
up to four functional DR -chain genes, and all of these are
expressed simultaneously in the cell. This allows any DR
-chain gene product to pair with any DR chain product.
Since the antigen-binding groove of class II is formed by a
combination of the and chains, this creates several
unique antigen-presenting DR molecules on the cell.
As with the class I loci, additional nonclassical class II
molecules with specialized immune functions are encoded
within this region. For instance, human nonclassical class II
genes designated DM and DO have been identified. The DM
genes encode a class II–like molecule (HLA-DM) that
facilitates the loading of antigenic peptides into class II
MHC molecules. Class II DO molecules, which are expressed
only in the thymus and on mature B cells, have been shown
to serve as regulators of class II antigen processing.
Class I and II MHC molecules have common structural
features, and both have roles in antigen processing and presentation. By contrast, the class III MHC region encodes
several molecules that are critical to immune function but
have little in common with class I or II molecules. Class III
products include the complement components C4, C2, and
factor B (see Chapter 6), as well as several inflammatory
cytokines, including the two tumor necrosis factor proteins
(TNF- and Lymphotoxin- [TNF-]). Allelic variants of
some of these class III MHC gene products have been linked
to certain diseases. For example, polymorphisms within the
TNF- gene, which encodes a cytokine involved in many
immune processes (see Chapter 4), have been linked to susceptibility to certain infectious diseases and some forms of
autoimmunity, including Crohn’s disease and rheumatic
arthritis. Despite its differences from class I and class II
genes, the linked class III gene cluster is conserved in all species with an MHC region, suggesting similar evolutionary
pressures have come to bear on this cluster of genes.
PA R T I I I
270
|
Adaptive Immunity: Antigen Receptors and MHC
The Exon/Intron Arrangement of Class I and II
Genes Reflects Their Domain Structure
Separate exons encode each region of the class I and II proteins. Each of the mouse and human class I genes has a 5
leader exon encoding a short signal peptide followed by five
or six exons encoding the chain of the class I molecule
(Figure 8-9a). The signal peptide serves to facilitate insertion
of the chain into the ER and is removed by proteolytic
enzymes after translation is complete. The next three exons
encode the extracellular 1, 2, and 3 domains, and the
following downstream exon encodes the transmembrane
(Tm) region. Finally, one or two 3-terminal exons encode
the cytoplasmic domains (C).
Like class I MHC genes, the class II genes are organized
into a series of exons and introns mirroring the domain
(a)
α1
L
α2
α3
Tm
C
3′
L α1
α2
α3
Allelic Forms of MHC Genes Are Inherited
in Linked Groups Called Haplotypes
The genes that reside within the MHC region are highly
polymorphic; that is, many alternative forms of each gene,
or alleles, exist within the population. The individual genes of
the MHC loci (class I, II, and III) lie so close together that their
inheritance is linked. Crossover, or recombination between
genes, is more likely when genes are far apart. For instance, the
recombination frequency within the H-2 complex (i.e., the
(b)
C
DNA 5′
structure of the and chains (Figure 8-9b). Both the
and the genes encoding mouse and human class II MHC
molecules have a leader exon, an 1 or 1 exon, an 2 or
2 exon, a transmembrane exon, and one or more cytoplasmic exons.
L
β1
L β1
(A)n
β2
Tm+C C C
(A) n
mRNA
Class II MHC
molecule
α3
S
C
β chain
α2
S
C
3′
α chain
Class I MHC
molecule
Tm+C
DNA 5′
Tm C C
mRNA
β2
S
S
S
S
COOH
β1
S
H2N
H2N
α1
β2
S
α1
S
S
COOH
S
S
COOH
α2
H2N
α chain
β2 - microglobulin
(A) n
mRNA
L α1
α2 Tm+C C
α1
α2
DNA 5′
3′
L
FIGURE 8-9 Schematic diagram of (a) class I and (b) class II
MHC genes, mRNA transcripts, and protein molecules. There is
strong correspondence between exons and the domains in the gene
products of MHC molecules. Note that the mRNA transcripts are
spliced to remove the intron sequences. Each exon, with the excep-
Tm+C
C
tion of the leader (L) exon, encodes a separate domain of an MHC
molecule. The leader peptide is removed in a post-translational reaction before the molecule is expressed on the cell surface. The gene
encoding 2-microglobulin is located on a different chromosome in
both human and mouse. Tm transmembrane; C cytoplasmic.
The Major Histocompatibility Complex and Antigen Presentation
frequency of chromosome crossover events during meiosis,
indicative of the distance between given genes) is only 0.5%.
Thus, crossover occurs only once in every 200 meiotic cycles.
For this reason, most individuals inherit all the alleles
encoded by these genes as a set (known as linkage disequilibrium). This set of linked alleles is referred to as a haplotype. An individual inherits one haplotype from the mother
and one haplotype from the father, or two sets of alleles.
In outbred populations, such as humans, the offspring are
generally heterozygous at the MHC locus, with different
alleles contributed by each of the parents. If, however, mice
are inbred, each H-2 locus becomes homozygous because
the maternal and paternal haplotypes are identical, and all
offspring begin to express identical MHC molecules. Certain
mouse strains have been intentionally inbred in this manner
and are employed as prototype strains. The MHC haplotype
expressed by each of these strains is designated by an arbitrary italic superscript (e.g., H-2a, H-2b). These designations
refer to the entire set of inherited H-2 alleles within a strain
without having to list the specific allele at each locus individually (Table 8-2). Different inbred strains may share the
same set of alleles, or MHC haplotype, with another strain
(i.e., CBA, AKR, and C3H) but will differ in genes outside
the H-2 complex.
Detailed analysis of the H-2 complex in mice has been
made possible by the development of congenic H-2 strains
that differ only at the MHC locus. Inbred mouse strains are
said to be syngeneic, or identical at all genetic loci. Two
strains are congenic if they are genetically identical except at
a single genetic region. Any phenotypic differences that can
be detected between congenic strains is therefore related to
the genetic region that distinguishes the two strains. Congenic strains that are identical with each other except at the
TABLE 8-2
|
CHAPTER 8
271
MHC can be produced by a series of crosses, backcrosses,
and selections between two inbred strains that differ at the
MHC. A frequently used congenic strain, designated B10.A,
is derived from B10 mice (which is H-2b) genetically manipulated to possess the H-2a haplotype at the MHC locus.
Recombination within the H-2 region of congenic mouse
strains then allows the study of individual MHC genes and
their products. Examples of these are included in the list in
Table 8-2. For example, the B10.A (2R) strain has all the
MHC genes from the a haplotype except for the D region,
which is derived from the H-2b parent.
MHC Molecules Are Codominantly Expressed
The genes within the MHC locus exhibit a codominant form
of expression, meaning that both maternal and paternal gene
products (from both haplotypes) are expressed at the same
time and in the same cells. Therefore, if two mice from inbred
strains possessing different MHC haplotypes are mated, the
F1 generation inherits both parental haplotypes and will
express all these MHC alleles. For example, if an H-2b strain
is crossed with an H-2k strain, then the F1 generation inherits
both parental sets of alleles and is said to be H-2b/k (Figure 8-10a). Because such an F1 generation expresses the
MHC proteins of both parental strains on its cells, it is said
to be histocompatible with both parental strains. This means
offspring are able to accept grafts from either parental
source, each of which expresses MHC alleles viewed as
“self ” (Figure 8-10b). However, neither of the inbred parental strains can accept a graft from its F1 offspring because half
of the MHC molecules (those coming from the other parent)
will be viewed as “nonself,” or foreign, and thus subject to
recognition and rejection by the immune system.
H-2 haplotypes of some mouse strains
H-2 ALLELES
Prototype strain
Other strains with the same haplotype
Haplotype
K
IA
IE
S
D
CBA
AKR, C3H, B10.BR, C57BR
k
k
k
k
k
k
DBA/2
BALB/c, NZB, SEA, YBR
d
d
d
d
d
d
C57BL/10 (B10)
C57BL/6, C57L, C3H.SW, LP, 129
b
b
b
b
b
b
A
B10.A(2R)*
B10.A(3R)
B10A.(4R)
A/He, A/Sn, A/Wy, B10.A
a
h2
i3
h4
k
k
b
k
k
k
b
k
k
k
k
b
d
d
d
b
d
b
d
b
A.SW
B10.S, SJL
s
s
s
s
s
s
t1
s
k
k
k
d
q
q
q
q
q
q
A.TL
DBA/1
STOLI, B10.Q, BDP
*The R designates a recombinant haplotype, in this case between the H-2a and H-2b types. Gene contribution from the a strain is shown in yellow and from the b strain
in red.
FIGURE 8-10 Illustration of inheri-
(a) Mating of inbred mouse strains with different MHC haplotypes
Homologous chromosomes with MHC loci
H-2b parent
H-2k parent
b/b
b/b
k/k
k/k
F1 progeny (H-2b/k)
b/k
b/k
(b) Skin transplantation between inbred mouse strains with same or
different MHC haplotypes
Parental recipient
Skin graft donor
Progeny recipient
b/b
Parent
b/k
k/k
Parent
b/k
b/k
Progeny
b/k
b/b
k/k
b/b
k/k
b/b
k/k
(c) Inheritance of HLA haplotypes in a typical human family
Parents
A/B
C/D
Progeny
A/C
A/D
B/R
B/C
B/D
272
tance of MHC haplotypes in inbred
mouse strains. (a) The letters b/b designate a
mouse homozygous for the H-2b MHC haplotype, k/k homozygous for the H-2k haplotype,
and b/k a heterozygote. Because the MHC
genes are closely linked and inherited as a set,
the MHC haplotype of F1 progeny from the
mating of two different inbred strains can be
predicted easily. (b) Acceptance or rejection
(X) of skin grafts is controlled by the MHC type
of the inbred mice. The progeny of the cross
between two inbred strains with different
MHC haplotypes (H-2b and H-2k) will express
both haplotypes (H-2b/k) and will accept grafts
from either parent and from one another.
However, neither parent strain will accept
grafts from the offspring. (c) Inheritance of
HLA haplotypes in a hypothetical human family. For ease, the human paternal HLA haplotypes are arbitrarily designated A and B,
maternal C and D. Note that a new haplotype,
R (recombination), can arise from rare recombination of a parental haplotype (maternal
shown here).
The Major Histocompatibility Complex and Antigen Presentation
In an outbred population such as humans, each individual is generally heterozygous at each locus. The human HLA
complex is highly polymorphic, and multiple alleles of each
class I and class II gene exist. However, as with mice, the
human MHC genes are closely linked and usually inherited
as a haplotype. When the father and mother have different
haplotypes, as in the example shown in Figure 8-10c, there is
a one-in-four chance that siblings will inherit the same
paternal and maternal haplotypes and therefore will be histocompatible (i.e., genetically identical at their MHC loci)
with each other; none of the offspring will be fully histocompatible with the parents.
Although the rate of recombination by crossover is low
within the HLA complex, it still contributes significantly to
the diversity of the loci in human populations. Genetic
recombination can generate new allelic combinations, or
haplotypes (see haplotype R in Figure 8-10c), and the high
number of intervening generations since the appearance of
humans as a species has allowed extensive recombination.
As a result of recombination and other mechanisms for generating mutations, it is rare for any two unrelated individuals
to have identical sets of HLA genes. This makes transplantation between individuals who are not identical twins quite
challenging! To address this, clinicians begin by looking for
family members who will be at least partially histocompatible with the patient, or they rely on donor databases to look
for an MHC match. Even with partial matches, physicians
still need to administer heavy doses of immunosuppressive
drugs to inhibit the strong rejection responses that typically
follow tissue transplantation due to differences in the MHC
proteins (see Chapter 16).
Class I and Class II Molecules Exhibit Diversity
at Both the Individual and Species Levels
As noted earlier, any particular MHC molecule can bind
many different peptides (called promiscuity), which gives the
host an advantage in responding to pathogens. Rather than
relying on just one gene for this task, the MHC region has
evolved to include multiple genetic loci encoding proteins
with the same function. In humans, HLA-A, B, or C molecules can all present peptides to CD8 T cells and HLA-DP,
DQ, or DR molecules can present to CD4 T cells. The MHC
region is thus said to be polygenic because it contains multiple genes with the same function but with slightly different
structures. Since the MHC alleles are also codominantly
expressed, heterozygous individuals will express the gene
products encoded by both alleles at each MHC gene locus. In
a fully heterozygous individual this amounts to 6 unique classical class I molecules on each nucleated cell. An F1 mouse,
for example, expresses the K, D, and L class I molecules from
each parent (six different class I MHC molecules) on the
surface of each of its nucleated cells (Figure 8-11). The
expression of so many individual class I MHC molecules
allows each cell to display a large number of different peptides in the peptide-binding grooves of its MHC molecules.
|
CHAPTER 8
Dk
Class I
molecules
Dd
Kd
273
Lk
Kk
Ld
Maternal MHC
Kk IAα k β k IEα k β k D k Lk
I
II
I
K d IAα d β d IEα d β d D d Ld
IE αk β k
IAα k β k
Paternal MHC
IE αd β d
IAα d β d
IE αk β d
Class II
molecules
IE αd β k
IAα k β d
IAα d β k
FIGURE 8-11 Diagram illustrating the various MHC molecules expressed on antigen-presenting cells of a heterozygous H-2k/d mouse. Both the maternal and paternal MHC genes are
expressed (codominant expression). Because the class II molecules are
heterodimers, new molecules containing one maternal-derived and
one paternal-derived chain are also produced, increasing the diversity
of MHC class II molecules on the cell surface. The 2-microglobulin
component of class I molecules (pink) is encoded by a nonpolymorphic gene on a separate chromosome and may be derived from
either parent.
MHC class II molecules have even greater potential for
diversity. Each of the classical class II MHC molecules is
composed of two different polypeptide chains encoded by
different loci. Therefore, a heterozygous individual can
express - combinations that originate from the same
chromosome (maternal only or paternal only) as well as class
II molecules arising from unique chain pairing derived from
separate chromosomes (new maternal-paternal - combinations). For example, an H-2k mouse expresses IAk and IEk
class II molecules, whereas an H-2d mouse expresses IAd and
IEd molecules. The F1 progeny resulting from crosses of these
two strains express four parental class II molecules (identical
to their parents) and also four new molecules that are mixtures from their parents, containing one parent’s chain and
the other parent’s chain (as shown in Figure 8-11). Since
the human MHC contains three classical class II genes (DP,
DQ, and DR), a heterozygous individual expresses six class II
molecules identical to the parents and six molecules containing new and chain combinations. The number of
different class II molecules expressed by an individual can be
further increased by the occasional presence of multiple
-chain genes in mice and humans, and in humans by the
presence of multiple -chain genes. The diversity generated
by these new MHC molecules likely increases the number of
different antigenic peptides that can be presented and is
therefore advantageous to the organism in fighting infection.
274
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
The variety of peptides displayed by MHC molecules
echoes the diversity of antigens bound by antibodies and
T-cell receptors. This evolutionary pressure to diversify comes
from the fact that both need to be able to interact with antigen
fragments they have never before seen, or that may not yet
have evolved. However, the strategy for generating diversity
within MHC molecules and the antigen receptors on T and
B cells is not the same. Antibodies and T-cell receptors are
generated by several somatic processes, including gene rearrangement and the somatic mutation of rearranged genes (see
Chapters 7 and 12). Thus, the generation of T- and B-cell
receptors is dynamic, changing over time within an individual. By contrast, the MHC molecules expressed by an individual are fixed. However, promiscuity of antigen binding
ensures that even “new” proteins are likely to contain some
fragments that can associate with any given MHC molecule.
TABLE 8-3
Genetic diversity of MHC loci
in the human population
MHC CLASS I
HLA locus
Number of allotypes (proteins)
A
1448
B
1988
C
1119
E
3
F
4
G
16
MHC CLASS II
HLA locus
Number of allotypes (proteins)
DMA
4
DMB
7
DOA
3
DOB
5
DPA1
17
DPB1
134
DQA1
47
DQB1
126
DRA
2
DRB1
860
DRB3
46
DRB4
8
DRB5
17
Source: Data obtained from http://hla.alleles.org, a Web site maintained by the
HLA Informatics Group based at the Anthony Nolan Trust in the United Kingdom,
with up-to-date information on the numbers of HLA alleles and proteins.
Collectively, this builds in enormous flexibility within the host
for responding to unexpected environmental changes that
might arise in the future—an elegant evolutionary strategy.
Countering this limitation on the range of peptides that can
be presented by any one individual is the vast array of peptides
that can be presented at the species level, thanks to the diversity
of the MHC in any outbred population. The MHC possesses an
extraordinarily large number of different alleles at each locus
and is one of the most polymorphic genetic complexes known
in higher vertebrates. These alleles differ in their DNA
sequences from one individual to another by 5% to 10%. The
number of amino acid differences between MHC alleles can be
quite significant, with up to 20 amino acid residues contributing to the unique structural nature of each allele. Analysis of
human HLA class I genes as of July 2012 reveals approximately
2013 A alleles, 2605 B alleles, and 1551 C alleles (Table 8-3
shows the number of protein products; not all alleles encode
expressed proteins). In mice, the polymorphism is similarly
enormous. The human class II genes are also highly polymorphic, and in some cases, different individuals can even inherit
different numbers of genes. The number of HLA-DR -chain
genes (DRB) may vary from 2 to 9 in different haplotypes, and
over 1200 alleles of DRB genes alone have been reported. Interestingly, the DRA gene is highly conserved, with only seven
different alleles and two proteins identified. Current estimates
of actual polymorphism in the human MHC are likely on the
low side because the earliest and most detailed data have primarily concentrated on populations of European descent.
This enormous polymorphism results in a tremendous
diversity of MHC molecules within a species. Using the numbers given above for the allelic forms of human HLA-A, -B,
and -C, we can calculate a theoretical number of combinations
that can exist, which is upward of 1.7 billion different class I
haplotypes possible in the population. If class II loci are considered, the numbers are even more staggering, with over 1015
different class II combinations. Because each haplotype contains
both class I and class II genes, multiplication of the numbers
gives a total of more than 1.7 1024 possible combinations of
these class I and II alleles within the entire human population!
Some evidence suggests that a reduction in MHC polymorphism within a species may predispose that species to infectious disease (see Evolution Box 8-1). In one example, cheetahs
and certain other wild cats, such as Florida panthers, that have
been shown to be highly susceptible to viral disease also have
very limited MHC polymorphism. It is postulated that the
present cheetah population arose from a limited breeding
population, or genetic bottleneck, causing a loss of MHC diversity. This increased susceptibility of cheetahs to various viruses
may result from a reduction in the number of different MHC
molecules available to the species as a whole and a corresponding limitation on the range of processed antigens with which
these MHC molecules can interact. As a corollary, this suggests
that the high level of MHC polymorphism that has been
observed in many outbred species, including humans, may
provide a survival advantage by supplying a broad range of
MHC molecules and thus a broad range of presentable antigens.
The Major Histocompatibility Complex and Antigen Presentation
|
CHAPTER 8
275
BOX 8-1
EVOLUTION
The Sweet Smell of Diversity
As early as
the mid-1970s, mate
choice in mice was shown to be influenced by genes at the MHC (H-2) locus.
This was followed by investigations in
other rodents, fish, and birds, all with similar conclusions. Thanks to results from
several studies conducted in the last 15
years, it looks like humans could be added
to this list. However, questions remain as
to the precise evolutionary pressures, the
mechanism, and the magnitude of this
effect, among others, in influencing mate
choice in humans.
In terms of evolutionary pressure, local
pathogens play a significant role in maintaining MHC diversity in the population
and in selecting for specific alleles. As we
now appreciate, this is because the MHC
influences immune responsiveness. Due
to its role in selecting the peptide fragments that will be presented, the inheritance of specific alleles at particular loci
can predispose individuals to either
enhanced susceptibility or resistance to
specific infectious agents and immune
disorders (see Clinical Focus Box 8-2). Over
long time periods, endemic local pathogens exert evolutionary pressures that
drive higher- or lower-than-expected
rates of certain MHC alleles in a population, as well as overrepresentation of
other resistance-associated genes.
The degree of diversity at the MHC
locus clearly influences susceptibility to
disease in populations; witness the
enhanced viral susceptibility seen in
cheetahs (see page 274) and some devastating human stories. For instance, the
European introduction of the smallpox
virus to New World populations is credited with wiping out large Native American groups. This may be due to a lack of
past evolutionary pressure for conservation of resistance-associated MHC alleles,
which would therefore be rare or nonexistent in this population, as well as to a lack
of any individuals with immunity from
prior infection.
But how does an individual evaluate
how well a potential mate will contribute
“new” MHC alleles to one’s offspring, leading to greater diversity and the potential
for enhanced fitness? The primary candidate is odor: the MHC is known to influence odor in many vertebrate species. For
example, the urine of mice from distinct
MHC congenic lines can be distinguished
by both humans and rodents. Mice show
a distinct preference for mating with animals that carry MHC alleles that are dissimilar to their own. In terms of
maximizing the range of peptides that
can be presented, this makes clear evolutionary sense, as increased diversity at the
MHC should increase the number of different pathogenic peptides that can be
“seen” by the immune system, increasing
the likelihood of effective anti-pathogen
responses. To back this up, the advantage
of overall MHC diversity has been shown
experimentally in mice, where most simulated epidemic experiments have found
a survival advantage for H-2 heterozygous
animals over their homozygous counterparts. In humans, research in HIV-infected
individuals has shown that extended survival and a slower progression to AIDS are
correlated with full heterozygosity at the
HLA class I locus, as well as absence of
certain AIDS-associated HLA-B and -C
alleles. This specific link to class I is not
surprising in light of the key role of CD8
T cells in combating viral infections.
Human studies of attraction and mating also point to preferences for individuals with MHC dissimilar alleles. In one key
study involving what is commonly known
as the “sweaty T-shirt test,” college-age
volunteers were asked to rate their preference or sexual attraction to the odor of
T-shirts worn by individuals of the opposite sex. In general, both males and
females preferred the odor of T-shirts
worn by individuals with dissimilar HLA
types. The one key exception was seen in
women who were concurrently using an
estrogen-based birth control pill; they
instead showed a preference for the odors
of MHC-similar individuals, suggesting
that this hormone not only interferes with
this response but that it potentially shifts
the outcome.
But how? you ask. Soluble forms of
MHC molecules have been found in
many bodily fluids, including urine, saliva,
sweat, and plasma. However, these molecules are unlikely to be small or volatile
enough to account for direct olfactory
detection. Other hypotheses for how
MHC might influence odor include via
olfactory recognition of natural ligands
or specific volatile peptides carried by
MHC molecules, or by MHC-driven differences in natural flora, which are also
known to impact body odor. The recent
finding that some polymorphic olfactory
receptor genes are closely linked to the
MHC may also help explain this apparent
association with mating preference. To
date, the jury is still out on which
mechanism(s) likely account(s) for MHCspecific odor differences.
As one might imagine, large methodological challenges are inherent in asking
questions related to mate choice in
humans, where differing levels of outbreeding as well as social and cultural
factors influence the outcome. Nonetheless, it appears that like most primate species tested so far, we humans may also
have the capacity to use smell as an evolutionary strategy for promoting a robust
immune response in our offspring.
J. Havlicek, and S. C. Roberts. 2009. MHC-correlated
mate choice in humans: A review. Psychoneuroendocrinology 34:497–512.
C. Wedekind et al. 1995. MHC-dependent mate
preferences in humans. Proceedings. Biological Science 260:245–249.
C. Wedekind et al. 1997. Body odor preferences
in men and women: Do they aim for specific
MHC combinations or simply heterozygosity? Proceedings. Biological Science 264:1471–
1479.
276
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
Although some individuals within a species may not be able to
develop an immune response to a given pathogen and therefore will be susceptible to infection, extreme polymorphism
ensures that at least some members of a species will be resistant
to that disease. In this way, MHC diversity at the population
level may protect the species as a whole from extinction via a
wide range of infectious diseases.
MHC Polymorphism Has Functional Relevance
Although the sequence divergence among alleles of the MHC
within a species is very high, this variation is not randomly
distributed along the entire polypeptide chain. Instead, polymorphism in the MHC is clustered in short stretches, largely
within the membrane-distal 1 and 2 domains of class I
molecules (Figure 8-12a). Similar patterns of diversity are
observed in the 1 and 1 domains of class II molecules.
(a)
α2
α3
Variability
α1
Structural comparisons have located the polymorphic residues within the three-dimensional structure of the membranedistal domains in class I and class II MHC molecules and
have related allelic differences to functional differences (Figure 8-12b). For example, of 17 amino acids previously shown
to display significant polymorphism among HLA-A molecules, 15 were shown by x-ray crystallographic analysis to be in
the peptide-binding groove of this molecule. The location of so
many polymorphic amino acids within the binding site for
processed antigen strongly suggests that allelic differences contribute to the observed differences in the ability of MHC molecules to interact with a given peptide ligand. Polymorphisms
that lie outside these regions and might affect basic domain
folding are rare. This clustering of polymorphisms around
regions that make contact with antigen also suggest possible
reasons why certain MHC genes or haplotypes can become
associated with certain diseases (see Clinical Focus Box 8-2).
20
40
60
80
100
120
140
160
Residue number
180
200
220
240
260
(b)
12
45
62
63
66
70
74
95
9
97 116
114
156
N
107
105
FIGURE 8-12 Variability in the amino acid sequences of allelic class I HLA molecules. (a) In the external domains, most of the
variable residues are in the membrane-distal 1 and 2 domains. (b) Location of polymorphic amino acid residues (red) in the 1/2 domain
of a human class I MHC molecule. [Part (a) adapted from R. Sodoyer et al., 1984, EMBO Journal 3:879, reprinted by permission of Oxford University Press; part
(b) adapted, with permission, from P. Parham, 1989, Nature 342:617, © 1989 Macmillan Magazines Limited.]
The Major Histocompatibility Complex and Antigen Presentation
|
CHAPTER 8
277
BOX 8-2
CLINICAL FOCUS
MHC Alleles and Susceptibility to Certain Diseases
Some HLA alleles
occur at a
much higher frequency in people suffering from certain diseases than in the general population. The diseases associated
with particular MHC alleles include autoimmune disorders, certain viral diseases,
disorders of the complement system,
some neurologic disorders, and several
different allergies. In humans, the association between an HLA allele and a given
disease may be quantified by determining
the frequency of that allele expressed by
individuals afflicted with the disease, then
comparing these data with the frequency
of the same allele in the general population. Such a comparison allows calculation
of an individual’s relative risk (RR).
RR
frequency of disease in the allele group
frequency of disease in the allele group
(Eq. 8.1)
A RR value of 1 means that the HLA
allele is expressed with the same frequency
in disease-afflicted and general populations, indicating that this allele confers no
increased risk for the disease. A RR value
substantially above 1 indicates an association between the HLA allele and the disease. For example, individuals with the
HLA-B27 allele are 90 times more likely
(RR 90) to develop the autoimmune
disease ankylosing spondylitis, an inflammatory disease of vertebral joints characterized by destruction of cartilage, than are
individuals who lack this HLA-B allele. Other
disease associations with significantly high
RR include HLA-DQB1 and narcolepsy
(RR 130) and HLA-DQ2 with celiac disease (RR50), an allergy to gluten. A few
HLA alleles have also been linked to relative
protection from disease or clinical progres-
sion. This is seen in the case of individuals
inheriting HLA-B57, which is associated
with greater viral control and a slower progression to AIDS in HIV-infected individuals.
When the associations between MHC
alleles and disease are weak, reflected by
low RR values, it is possible that multiple
genes influence susceptibility, of which only
one lies within the MHC locus. The genetic
origins of several autoimmune diseases,
such as multiple sclerosis (associated with
DR2; RR = 5) and rheumatoid arthritis (associated with DR4; RR = 10) have been studied in depth. The observation that these
diseases are not inherited by simple Mendelian segregation of MHC alleles can be
seen clearly in identical twins, where both
inherit the same MHC risk factor, but frequently only one develops the disease. This
finding suggests that multiple genetic factors
plus one or more environmental factors are at
play in development of this disease. As we
highlight in Chapter 16, this combined role
for genes and the environment in the development of autoimmunity is not uncommon.
The existence of an association
between an MHC allele and a disease
should not be interpreted to imply that
the expression of the allele has caused the
disease. The relationship between MHC
alleles and development of disease is
more complex, partly thanks to the
genetic phenomenon of linkage disequilibrium. The fact that some of the class I
MHC alleles are in linkage disequilibrium
with the class II and class III alleles can
make their contribution to disease susceptibility appear more pronounced than
it actually is. If, for example, DR4 contrib-
The preceding discussion points to additional parallels
between MHC molecules and lymphocyte antigen receptors.
The somatic hypermutations seen in B-cell receptor genes is
also not randomly arrayed within the molecule, but instead
is clustered in the regions most likely to interact directly with
peptide (see Chapter 12), providing yet another example of
how the immune system has solved a similar functional
dilemma using a very different strategy.
utes to risk of a disease and it also occurs
frequently in combination with HLA-A3
because of linkage disequilibrium, then
A3 would incorrectly appear to be associated with the disease. Improved genomic
mapping techniques make it possible to
analyze the fine linkage between genes
within the MHC and various diseases
more fully and to assess the contributions
from other loci. In the case of ankylosing
spondylitis, for example, it has been suggested that alleles of TNF-␣ and Lymphotoxin-␣ may produce protein variants that
are involved in destruction of cartilage,
and these alleles happen to be linked to
certain HLA-B alleles. In the case of HLAB57 and AIDS progression, the allele has
been linked more directly to disease. It is
believed that this class I allele is particularly efficient at presenting important
components of the virus to circulating
Tc cells, leading to increased destruction
of virally infected cells and delayed disease progression.
Other hypotheses have also been
offered to account for a direct role of particular MHC alleles in disease susceptibility. In some rare cases, certain allelic forms
of MHC genes can encode molecules that
are recognized as receptors by viruses or
bacterial toxins, leading to increased susceptibility in the individuals who inherit
these alleles. As will be explored further in
Chapters 15 through 19, in many cases
complex interactions between multiple
genes, frequently including the MHC, and
particular environmental factors are
required to create a bias toward the development of certain diseases.
The Role of the MHC
and Expression Patterns
As we have just discussed, several genetic features help ensure
a diversity of MHC molecules in outbred populations, including polygeny, polymorphism, and codominant expression. All
this attention paid to maximizing the number of different
278
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
binding grooves suggests that variety within the MHC plays
an important role in survival (see Clinical Focus Box 8-2). In
fact, in addition to fighting infection, MHC expression
throughout the body plays a key role in maintaining homeostasis and health even when no foreign antigen is present.
Although the presentation of MHC molecules complexed
with foreign antigen to T cells garners much attention (and
space in this book!), most MHC molecules spend their lives
presenting other things, and often to other cells. There are
several reasons why an MHC molecule on the surface of a
cell is important. In general, these include the following:
• To display self class I to demonstrate that the cell is
healthy
• To display foreign peptide in class I to show that the cell
is infected and to engage with TC cells
• To display a self-peptide in class I and II to test developing T cells for autoreactivity (primary lymphoid organs)
• To display a self-peptide in class I and II to maintain tolerance to self-proteins (secondary lymphoid organs)
• To display a foreign peptide in class II to show the body
is infected and activate TH cells
It is worth noting that although some of these instances lead
to immune activation, some do not. It is also important to
note that the cell type, tissue location, and timing of expression vary for each of these situations. For instance, developing T cells first encounter MHC class I and II molecules
presenting self peptides in the thymus, where these signals
are designed to inhibit the ability of T cells to attack selfstructures later (see Chapter 9). Others occur only after
certain types of exposure, such as during an immune
response to extracellular pathogens or when cells in the body
are infected with viral invaders, and is designed to activate T
cells to act against the pathogen (see Chapter 11).
To help set the stage for this discussion, we turn now to
the where, when, and why of MHC expression. This is followed by a description of the how, detailing the pathways
that lead to peptide placement in the binding groove of each
class of molecule. As we will see in greater detail in the following sections, the type of MHC molecule(s) expressed by
a given cell is linked to the role of the cell and the function
of each class of molecule.
MHC Molecules Present Both Intracellular
and Extracellular Antigens
In very general terms, the job of MHC class I molecules is to
collect and present antigens that come from intracellular locations. This is a form of ongoing surveillance of the internal
happenings of the cell—in essence, a window for displaying
on the surface of the cell snippets of what is occurring inside.
Basically, all cells in the body need this form of check and balance, and this shows up in the fairly ubiquitous nature of
MHC class I expression in the body (there are a few notable
exceptions, which we will touch on later). Often, nothing
other than normal cellular processes are occurring in the cytosol, and in these instances our cells present self-peptides in the
groove of MHC class I molecules. The expresssion of self-MHC
class I (with self peptides) signals that a cell is healthy; absence
of self-MHC class I (as can occur in virus-infected and tumor
cells) targets that cell for killing by NK cells (see Chapter 13).
When foreign proteins are present in the cytosol and begin to
appear in the groove of MHC class I on the cell surface, this alerts
CD8 T cells to the presence of this unwelcome visitor, targeting the cell for destruction. In this case, the cell is called a target cell because it becomes a target for lysis by cytotoxic T cells.
Conversely, MHC class II molecules primarily display
peptides that have come from the extracellular spaces of the
body. Since this sampling of extracellular contents is not a
form of policing that all cells need to perform, only specialized leukocytes posses this ability. These cells are collectively
referred to as antigen-presenting cells (APCs), because their
job is to present extracellular antigen to T cells, charging
them with the ultimate job of coordinating the elimination
of this extracellular invader. While all cells in the body
express MHC class I proteins and can present peptides from
foreign intracellular antigens, the term APC is usually
reserved for MHC class II-expressing cells.
MHC Class I Expression Is Found
Throughout the Body
In general, classical class I MHC molecules are expressed
constitutively on almost all nucleated cells of the body. However, the level of expression differs among different cell types,
with the highest levels of class I molecules found on the surface of lymphocytes. These molecules may constitute approximately 1% of the total plasma membrane proteins, or some
5 105 MHC class I molecules per lymphocyte. In contrast,
some other cells, such as fibroblasts, muscle cells, liver hepatocytes, and some neural cells, express very low to undetectable levels of class I MHC molecules. This low-level expression
on liver cells may contribute to the relative success of liver
transplants by reducing the likelihood of graft rejection by Tc
cells of the recipient. A few cell types (e.g., subsets of neurons
and sperm cells at certain stages of differentiation) appear to
lack class I MHC molecules altogether. Nucleated cells without class I expression are, however, very rare. Non-nucleated
cells, such as red blood cells in mammals, do not generally
express any MHC molecules.
In normal, healthy cells, class I molecules on the surface
of the cell will display self-peptides resulting from normal
turnover of self-proteins inside the cell. In cells infected by a
virus, viral peptides as well as self-peptides will be displayed.
Therefore, a single virus-infected cell can be envisioned as
having various class I molecules on its membrane, some
displaying a subset of viral peptides derived from the viral
proteins being manufactured within. Because of individual
allelic differences in the peptide-binding grooves of the class I
MHC molecules, different individuals within a species will
have the ability to bind and present different sets of viral
The Major Histocompatibility Complex and Antigen Presentation
peptides. In addition to virally infected cells, altered self cells
such as cancer cells, aging body cells, or cells from an allogeneic graft (i.e., from a genetically-different individual), also
can serve as target cells due to their expression of foreign
MHC proteins and can be lysed by TC cells. The importance
of constitutive expression of class I is highlighted by the
response of natural killer (NK) cells to somatic cells that lack
MHC class I, as can occur during some viral infections. NK
cells can kill a cell that has stopped expressing MHC class I
on the surface, presumably because this suggests that the cell
is no longer healthy or has been altered by the presence of an
intracellular invader.
Expression of MHC Class II Molecules Is Primarily
Restricted to Antigen-Presenting Cells
MHC class II molecules are found on a much more restricted
set of cells than class I, and sometimes only after an inducing
event. As mentioned above, cells that display peptides associated with class II MHC molecules to CD4 TH cells are
called antigen-presenting cells (APCs), and these cells are
primarily certain types of leukocytes. APCs are specialized
for their ability to alert the immune system to the presence
of an invader and drive the activation of T cell responses.
Among the various APCs, marked differences in the level
of MHC class II expression have been observed. In some
cases, class II expression depends on the cell’s differentiation
stage or level of activation (such as in macrophages; see
below). APC activation usually occurs following interaction
with a pathogen and/or via cytokine signaling, which then
induces significant increases in MHC class II expression.
A variety of cells can function as bona fide APCs. Their
distinguishing feature is their ability to express class II MHC
molecules and to deliver a costimulatory, or second activating signal, to T cells. Three cell types are known to have these
characteristics and are thus often referred to as professional
antigen-presenting cells (pAPCs): dendritic cells, macrophages, and B lymphocytes. These cells differ from one
another in their mechanisms of antigen uptake, in whether
they constitutively express class II MHC molecules, and in
their costimulatory activity, as follows:
• Dendritic cells are generally viewed as the most effective
of the APCs. Because these cells constitutively express a
TABLE 8-4
Antigen-presenting cells
Professional antigenpresenting cells
Nonprofessional antigenpresenting cells
Dendritic cells
(several types)
Fibroblasts (skin)
Thymic epithelial
cells
Macrophages
Glial cells (brain)
Thyroid epithelial
cells
B cells
Pancreatic beta
cells
Vascular endothelial
cells
|
CHAPTER 8
279
high levels of class II MHC molecules and have inherent
costimulatory activity, they can activate naïve TH cells.
• Macrophages must be activated (e.g., via TLR signaling)
before they express class II MHC molecules or costimulatory membrane molecules such as CD80/86.
• B cells constitutively express class II MHC molecules
and posses antigen-specific surface receptors, making
them particularly efficient at capturing and presenting
their cognate antigen. However, they must be activated by,
for example, antigen, cytokines, or pathogen-associated
molecular patterns (PAMPs), before they express the
costimulatory molecules required for activating naïve
TH cells.
Several other cell types, classified as nonprofessional
APCs, can be induced to express class II MHC molecules
and/or a costimulatory signal under certain conditions
(Table 8-4). These cells can be deputized for professional
antigen presentation for short periods and in particular situations, such as during a sustained inflammatory response.
MHC Expression Can Change with
Changing Conditions
As noted above, MHC class I is constitutively expressed by
most cells in the body, whereas class II is only expressed
under certain conditions and in a very limited number of
cell types. The different roles of the two molecules help
explain this, and suggest that in certain instances specific
changes in MHC expression may prove advantageous. The
MHC locus can respond to both positive and negative regulatory pressures. For instance, MHC class I production can
be disrupted or depressed by some pathogens. MHC class II
expression on APCs is already quite variable, and the microenvironment surrounding an APC can further modulate
expression, usually enhancing the expression of these molecules. APCs in particular are conditioned to respond to local
cues leading to their activation and heightened MHC
expression, enabling them to arm other cells in the body for
battle. As one might imagine, this activation and arming of
APCs must be carefully regulated, lest these cells orchestrate
unwanted aggressive maneuvers against self components or
benign companions, as occurs in clinical conditions such as
autoimmunity or allergy, respectively. The mechanisms driving these changes in expression are described below.
Genetic Regulatory Components
The presence of internal or external triggers, such as intracellular invaders or cytokines (see below), can induce a signal
transduction cascade that leads to changes in MHC gene
expression. Research aimed at understanding the mechanism
of control of MHC expression has been advanced by the now
complete sequence of the mouse genome. Both class I and
class II MHC genes are flanked by 5 promoter sequences that
bind sequence-specific transcription factors. The promoter
280
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
motifs and the transcription factors that bind to these motifs
have been identified for a number of MHC genes, with
examples of regulation mediated by both positive and negative elements. For example, a class II MHC transcriptional
activator called CIITA (also known as class II, major histocompatibility complex, transactivator) and another transcription factor called RFX, have both been shown to activate
the promoter of class II MHC genes. Defects in these transcription factors cause one form of bare lymphocyte syndrome. Patients with this disorder lack class II MHC molecules
on their cells and suffer from severe immunodeficiency, highlighting the central role of class II molecules in T-cell maturation and activation.
Viral Interference
One clear example of negative regulation of MHC comes
from viruses that interfere with MHC class I expression and
thus avoid easy detection by CD8 T cells. These viruses
include human cytomegalovirus (CMV), hepatitis B virus
(HBV), and adenovirus 12 (Ad12). In some cases, reduced
expression of class I MHC molecules is due to decreased
levels of a component needed for peptide transport or MHC
class I assembly rather than decreased transcription. For
example, in the case of cytomegalovirus infection, a viral
protein binds to 2-microglobulin, preventing assembly of
class I MHC molecules and their transport to the plasma
membrane. Adenovirus 12 infection causes a pronounced
decrease in transcription of the transporter genes (TAP1 and
TAP2). As described in the following section on antigen
processing, the TAP gene products play an important role in
peptide transport from the cytoplasm into the rough endoplasmic reticulum (RER). Blocking of TAP gene expression
inhibits peptide transport; as a result, class I MHC molecules
cannot assemble with 2-microglobulin or be transported to
the cell membrane. These observations are especially important because decreased expression of class I MHC molecules,
by whatever mechanism, is likely to help viruses evade the
immune response. Lower levels of class I decrease the likelihood that virus-infected cells can display viral peptide complexes on their surface and become targets for CTL-mediated
destruction.
Cytokine-Mediated Signaling
The expression of MHC molecules is externally regulated by
various cytokines. Leading among these are the interferons
(, , and ) and the tumor necrosis factors (TNF- and
Lymphotoxin-), each of which have been shown to increase
expression of class I MHC molecules on cells. Typically,
phagocytic cells that are involved in innate responses, or
locally infected cells, are the first to produce these MHCregulating cytokines. In particular, IFN- (produced by a
cell following viral or bacterial infection) and TNF-
(secreted by APCs after activation) are frequently the first
cytokines to kick off an MHC class I up-regulation event. In
the later stages, interferon gamma (IFN- ), secreted by activated TH cells as well as other cell types, also contributes to
increased MHC expression.
Binding of these cytokines to their respective receptors
induces intracellular signaling cascades that activate transcription factors and alter expression patterns. These factors
bind to their target promoter sequences and coordinate
increased transcription of the genes encoding the class I
chain, 2-microglobulin, and other proteins involved in
antigen processing and presentation. IFN- has been shown
to induce expression of the class II transcriptional activator
(CIITA), thereby indirectly increasing expression of class II
MHC molecules on a variety of cells, including non-APCs
(e.g., skin keratinocytes, intestinal epithelial cells, vascular
endothelium, placental cells, and pancreatic beta cells).
Other cytokines influence MHC expression only in certain cell types. For example, IL-4 increases expression of
class II molecules in resting B cells, turning them into more
efficient APCs. Conversely, expression of class II molecules
by B cells is down-regulated by IFN- . Corticosteroids and
prostaglandins can also decrease expression of class II MHC
molecules. These naturally occurring, membrane-permeable
compounds bind to intracellular receptors and are some of
the most potent suppressors of adaptive immunity, primarily
based on this ability to inhibit MHC expression. This property is exploited in many clinical settings, where these compounds are used as treatments to suppress overly zealous
immune events, such as in allergic responses or during
transplant rejection (see Chapters 15 and 16).
Class II MHC Alleles Play a Critical Role
in Immune Responsiveness
MHC haplotype plays a strong role in the outcome of an
immune response, as these alleles determine which fragments
of protein will be presented. Recall that class II MHC molecules present foreign antigen to CD4 TH cells, which go on to
activate B cells to produce serum antibodies. Early studies by
Benacerraf in which guinea pigs were immunized with simple
synthetic antigens were the first to show that the ability of an
animal to mount an immune response, as measured by the
production of serum antibodies, is determined by its MHC
haplotype. Later experiments by H. McDevitt, M. Sela, and
their colleagues used congenic (matched at the MHC locus
but not elsewhere) and recombinant congenic mouse strains
to specifically map the control of immune responsiveness to
class II MHC genes. In early reports, the genes responsible for
this phenotype were designated Ir or immune response genes;
retaining the initial I, mouse class II products are now called
IA and IE. We now know that the dependence of immune
responsiveness on the genes within the class II MHC reflects
the central role of these molecules in determining which fragments of foreign proteins are presented as antigen to TH cells.
Two explanations have been proposed to account for this
variability in immune responsiveness observed among different haplotypes. According to the determinant-selection
model, different class II MHC molecules differ in their ability
to bind particular processed antigens. In the end, some peptides may be more crucial to eliminating the pathogen than
others. A separate hypothesis, termed holes-in-the-repertoire
model, postulates that T cells bearing receptors that recognize
|
The Major Histocompatibility Complex and Antigen Presentation
certain foreign antigens which happen to closely resemble
self-antigens may be eliminated during T-cell development,
leaving the organism without these cells/receptors for future
responses to foreign molecules. Since the T-cell response to
an antigen involves a trimolecular complex of the T-cell
receptor, an antigenic peptide, and an MHC molecule (discussed in detail in Chapter 11), both models appear correct.
That is, the absence of an MHC molecule that can bind and
present a particular peptide, or the absence of T-cell receptors
that can recognize a given MHC-peptide molecule complex,
both result in decreased immune responsiveness to a given
foreign substance, and can account for the observed relationship between MHC haplotype and the ability to respond to
particular exogenous antigens.
CHAPTER 8
Antigen
Strain 2 or 13
or (2 × 13) F1
Strain 2 or 13
or (2 × 13) F1
7 days
Lymph node cells
(source of T cells)
Peritoneal exudate cells
(source of macrophages)
Adherence
column
(retains
macrophages)
Adherent cells
T Cells Are Restricted to Recognizing Peptides
Presented in the Context of Self-MHC Alleles
In the 1970s a series of experiments were carried out to further explore the relationship between the MHC and immune
response. These investigations contributed two very crucial
discoveries: (1) that both CD4 and CD8 T cells can recognize antigen only when it is presented in the groove of an
MHC molecule, and (2) that the MHC haplotype of the APC
and the T cell must match. This happens naturally in the
host, where T cells develop alongside host APCs, both
expressing only that individual’s MHC molecules (see Chapter 9). This constraint is referred to as self-MHC restriction,
which refers to the dual specificity of T cells for self MHC as
well as for foreign antigen.
A. Rosenthal and E. Shevach showed that antigen-specific
proliferation of TH cells occurs only in response to antigen
presented by macrophages of the same MHC haplotype as
the T cells recognizing the antigen. In their experimental
system, guinea pig macrophages from strain 2 were initially
incubated with an antigen (Figure 8-13). After the “antigenpulsed” macrophages had processed the antigen and presented it on their surface, they were mixed with T cells from
the same strain (strain 2), a different strain (strain 13), or
their F1 progeny (2 X 13), and the magnitude of T-cell proliferation in response to the antigen-pulsed macrophages
was measured. The results of these experiments showed that
strain-2 antigen-pulsed macrophages activated T cells from
strain-2 and F1 mice but not T cells from strain-13 animals.
Similarly, strain-13 antigen-pulsed macrophages activated
strain-13 and F1 T cells but not strain-2 T cells. Subsequently, congenic and recombinant congenic strains of mice,
which differed from each other only in selected regions of
the MHC were used as the source of macrophages and T
cells. These experiments confirmed that the CD4 TH cell is
activated and proliferates only in the presence of antigenpulsed macrophages that share class II MHC alleles. Thus,
antigen recognition by the CD4 TH cell is said to be class II
MHC restricted. In 1974, Zinkernagel and Doherty similarly
demonstrated the self-MHC class I restriction of CD8 T cells
(see Classic Experiment Box 8-3). Over two decades later,
the pair was awarded the Nobel Prize in Medicine for their
281
Peritoneal macrophages
Antigen
Antigen-primed
T cells
Antigen-pulsed
macrophages
Measure T-cell
proliferation
Antigen-primed
T cell
Strain 2
Strain 13
(2 × 13) F1
Antigen-pulsed macrophages
Strain 2
Strain 13
(2 × 13) F1
+
−
+
−
+
+
+
+
+
FIGURE 8-13 Experimental demonstration of self-MHC
restriction of TH cells. Peritoneal exudate cells from strain 2, strain
13, or (2 X 13) F1 guinea pigs were incubated in plastic petri dishes,
allowing enrichment of macrophages, which are adherent cells. The
peritoneal macrophages were then incubated with antigen. These
“antigen-pulsed” macrophages were incubated in vitro with T cells
from strain 2, strain 13, or (2 X 13) F1 guinea pigs, and the degree of
T-cell proliferation was assessed ( vs ). The results indicated that
TH cells could proliferate only in response to antigen presented by
macrophages that shared MHC alleles. [Adapted from A. Rosenthal and E.
Shevach, 1974, Journal of Experimental Medicine 138:1194, by copyright permission of the Rockefeller University Press.]
seminal studies (see Table 1-2 for a list of Nobel Prizes
related to Immunology). As we will see in Chapter 9, selfMHC restriction is the result of T-cell development in the
thymus, where only T cells that recognize self-MHC are
selected for survival.
282
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
BOX 8-3
CLASSIC EXPERIMENT
Demonstration of the Self-MHC Restriction of CD8 T Cells
In 1974, R. M. Zinkernagel and P. C.
Doherty demonstrated that CD8 T cells
are restricted to recognizing antigen in
the context of self-MHC molecules. For
this seminal study, they took advantage of
recently derived inbred mice strains, a
virus that causes widespread neurological
damage in infected animals, and new
assays that allowed quantification of cytotoxic T cell responses.
In their experiments, mice were first
immunized with lymphocytic choriomeningitis virus (LCMV), a pathogen that
results in central nervous system inflammation and neurological damage. Several days later, the animals’ spleen cells,
which included TC cells specific for the
virus, were isolated and incubated with
LCMV-infected target cells of the same
or different haplotype (Figure 1). The
assay relied on measuring the release of
a radioisotope (51Cr) from labeled target
cells (called a chromium release assay;
see Figure 13-18). They found that the TC
cells killed syngeneic virus-infected target cells (or cells with matched MHC
LCMV
H–2k
Spleen cells
(containing Tc cells)
51Cr
H–2k target cells
–51Cr release
(no lysis)
H–2k LCMV-infected
target cells
H–2b LCMV-infected
target cells
+51Cr release
(lysis)
–51Cr release
(no lysis)
FIGURE 1
Classic experiment of Zinkernagel and Doherty demonstrating that antigen recognition by TC cells exhibits MHC restriction. H-2k mice were primed with the lymphocytic
choriomeningitis virus (LCMV) to induce cytotoxic T lymphocytes (CTLs) specific for the virus. Spleen
cells from this LCMV-primed mouse were then added to target cells of different H-2 haplotypes that
were intracellularly labeled with 51Cr (black dots) and either infected or not with the LCMV. CTL-mediated
killing of the target cells, as measured by the release of 51Cr into the culture supernatant, occurred only
if the target cells were infected with LCMV and had the same MHC haplotype as the CTLs. [Adapted from
P. C. Doherty and R. M. Zinkernagel, 1975, Journal of Experimental Medicine 141:502.]
alleles) but not uninfected cells or
infected cells from a donor that shared
no MHC alleles with these cytotoxic
cells.
Later studies with congenic and
recombinant congenic strains showed
that the TC cell and the virus-infected target cell must specifically share class I
molecules encoded by the K or D regions
of the MHC. Thus, antigen recognition by
CD8 cytotoxic T cells is class I MHC
restricted. In 1996, Doherty and Zinkernagel were awarded the Nobel Prize in
Medicine for their major contribution to
understanding the role of the MHC in cellmediated immunity.
Before this discovery, histocompatibility antigens were mainly blamed for transplantation rejection, but their importance
in the everyday process of cellular immunity was not appreciated. Zinkernagel and
Doherty’s discovery is all the more revolutionary when one takes into consideration
that little was known about cellular immunity and no T-cell receptor had yet been
discovered!
This milestone in immunologic understanding set the stage for the development of two key models of how T cells
respond to foreign antigen: altered self and
dual recognition. The altered self hypothesis posited that histocompatibility molecules that associate with foreign particles,
such as viruses, may appear to be altered
forms of self-proteins. The dual recognition
model proposed that T cells must be capable of simultaneous recognition of both
the foreign substance and these selfhistocompatibility molecules. We now
appreciate the validity of both of these
models, and the role of MHC restriction in
the immune response to both microorganisms and allogeneic transplants.
Doherty, P. C., and R. M. Zinkernagel. 1975. H-2
compatibility is required for T-cell mediated
lysis of target cells infected with lymphocytic choriomeningitis virus. Journal of
Experimental Medicine 141:502.
The Major Histocompatibility Complex and Antigen Presentation
Processing of Antigen Is Required
for Recognition by T Cells
In the 1980s, K. Ziegler and E. R. Unanue showed that an
intracellular processing step by APCs was required to activate T cells. These researchers observed that TH-cell activation by bacterial protein antigens was prevented by treating
the APCs with paraformaldehyde prior to antigen exposure
(in essence, killing the cells and immobilizing the antigens in
the membrane; see Figure 8-14a). However, if the APCs were
first allowed to ingest the antigen and were fixed with paraformaldehyde 1 to 3 hours later, TH-cell activation still
occurred (Figure 8-14b). During that interval of 1 to 3 hours,
the APCs had taken up the antigen, processed it into peptide
fragments, and displayed these fragments on the cell membrane in a form capable of activating T cells.
Subsequent experiments by R. P. Shimonkevitz showed
that internalization and processing could be bypassed if
APCs were exposed to already digested peptide fragments
instead of the native antigen (Figure 8-14c). In these experi-
|
CHAPTER 8
283
ments, APCs were treated with glutaraldehyde (this chemical, like paraformaldehyde, fixes the cell, rendering it
metabolically inactive) and then incubated with native ovalbumin or with ovalbumin that had been subjected to partial
enzymatic digestion. The digested ovalbumin was able to
interact with the glutaraldehyde-fixed APCs, thereby activating ovalbumin-specific TH cells. However, the native ovalbumin failed to do so. These results suggest that antigen
processing requires the digestion of the protein into peptides
that can be recognized by the ovalbumin-specific TH cells.
At about the same time, several investigators, including W.
Gerhard, A. Townsend, and their colleagues, began to identify the proteins of influenza virus that were recognized by TC
cells. Contrary to their expectations, they found that internal
proteins of the virus, such as polymerase and nucleocapsid
proteins, were often recognized by TC cells better than the
more exposed envelope proteins found on the surface of the
virus. Moreover, Townsend’s work revealed that TC cells recognized short linear peptide sequences of influenza proteins.
EXPERIMENTAL CONDITIONS
T-CELL
ACTIVATION
(a)
Antigen
–
1h
APC
APC
Fixation
(b)
Antigen
Fixation
+
1h
TH cell
APC
APC
Fixation
APC
(c)
+
Antigen
peptides
APC
FIGURE 8-14 Experimental demonstration that antigen
processing is necessary for TH-cell activation. (a) When antigenpresenting cells (APCs) are fixed before exposure to antigen, they are
unable to activate TH cells. (b) In contrast, APCs fixed at least 1 hour
after antigen exposure can activate TH cells. (This simplified figure
TH cell
APC
does not show costimulatory molecules needed for T-cell activation.)
(c) When APCs are fixed before antigen exposure and incubated with
peptide digests of the antigen (rather than native antigen), they also
can activate TH cells. TH-cell activation is determined by measuring a
specific TH-cell response (e.g., cytokine secretion).
284
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
In fact, when noninfected target cells were incubated in vitro
with synthetic peptides corresponding to only short sequences
of internal influenza proteins, these cells could be recognized
by TC cells and subsequently lysed just as well as target cells
that had been infected with whole, live influenza virus. These
findings, along with those presented below, suggest that antigen processing is a metabolic process that digests proteins into
peptides, which can then be displayed on the cell surface
together with a class I or class II MHC molecule.
Endogenous pathway
(class I MHC)
Endogenous
antigen
Proteasome
Evidence Suggests Different Antigen
Processing and Presentation Pathways
The immune system typically uses different pathways to
eliminate intracellular and extracellular antigens. As a general rule, endogenous antigens (those generated within the
cell) are processed in the cytosolic or endogenous pathway
and presented on the membrane with class I MHC molecules. Exogenous antigens (those taken up from the extracellular environment by endocytosis) are typically processed in
the exogenous pathway and presented on the membrane
with class II MHC molecules (Figure 8-15).
Experiments carried out by L. A. Morrison and T. J. Braciale provided an excellent demonstration that the antigenic
peptides presented by class I and class II MHC molecules
follow different routes during antigen processing. To do this,
they used a set of T-cell clones specific for an influenza virus
antigen; some recognized the antigen presented by MHC
class I, and others recognized the same antigen presented by
MHC class II. Examining the T-cell responses, they derived
the following general principles about the two pathways:
• Class I presentation requires internal synthesis of virus
protein, as shown by the requirement that the target cell
be infected by live virus, and by the inhibition of class I
presentation observed when protein synthesis was
blocked by the inhibitor emetine.
• Class II presentation can occur with either live or replication-incompetent virus; protein synthesis inhibitors
had no effect, indicating that new protein synthesis is
not a necessary condition of class II presentation.
• Class II, but not class I, presentation is inhibited by
treatment of the cells with an agent that blocks endocytic processing within the cell (e.g., chloroquine).
These studies support the distinction between the processing of exogenous and endogenous antigens. They suggest
a preferential, but not absolute, association of exogenous
antigens with class II MHC molecules and of endogenous
antigens with class I molecules. What they do not show is
that both the APC of choice and the antigen of choice can
influence the outcome, as will be seen later when these rules
are subverted during cross-presentation. In the experiments
described above, association of viral antigen with class I
MHC molecules required replication of the influenza virus
and viral protein synthesis within the target cells, but asso-
Exogenous pathway
(class II MHC)
Peptide
Rough endoplasmic
reticulum (RER)
Class I
MHC
Class II
MHC
Golgi complex
Exogenous
antigen
Class I
MHC
Class II
MHC
FIGURE 8-15 Overview of endogenous and exogenous
pathways for processing antigen. In the endogenous pathway
(left), antigens are degraded by the proteasome, converting proteins
into smaller peptides. In the exogenous pathway (right), extracellular
antigens are engulfed into endocytic compartments where they are
degraded by acidic pH-dependent endosomal and lysosomal
enzymes. The antigenic peptides from proteasome cleavage and
those from endocytic compartments associate with class I or class II
MHC molecules respectively, and the MHC-peptide complexes are
then transported to the cell membrane. It should be noted that the
ultimate fate of most peptides in the cell is neither of these pathways
but rather to be degraded completely into amino acids.
ciation with class II did not. These findings suggested that
the peptides presented by class I and class II MHC molecules
are trafficked through separate intracellular compartments;
class I MHC molecules interact with peptides derived from
cytosolic degradation of endogenously synthesized proteins,
class II molecules with peptides derived from endocytic degradation of exogenous antigens. The next two sections examine these two pathways in detail.
The Major Histocompatibility Complex and Antigen Presentation
The Endogenous Pathway of Antigen
Processing and Presentation
In eukaryotic cells, protein levels are carefully regulated.
Every protein is subject to continuous turnover and is
degraded at a rate that is generally expressed in terms of its
half-life. Some proteins (e.g., transcription factors, cyclins,
and key metabolic enzymes) have very short half-lives.
Denatured, misfolded, or otherwise abnormal proteins also
are degraded rapidly. Defective ribosomal products are polypeptides that are synthesized with imperfections and constitute a large part of the products that are rapidly degraded.
The average half-life for cellular proteins is about 2 days, but
many are degraded within 10 minutes. The consequence of
steady turnover of both normal and defective proteins is a
constant deluge of degradation products within a cell. Most
will be degraded to their constituent amino acids and recycled, but some persist in the cytosol as peptides. The cell
samples these peptides and presents some on the plasma
membrane in association with class I MHC molecules,
where cells of the immune system can sample these peptides
to survey for foreign proteins. The pathway by which these
endogenous peptides are generated for presentation with
class I MHC molecules utilizes mechanisms similar to those
involved in the normal turnover of intracellular proteins, but
exactly how particular proteins are selected for degradation
and peptide presentation still remains unclear.
Peptides Are Generated by Protease
Complexes Called Proteasomes
Intracellular proteins are degraded into short peptides by a
cytosolic proteolytic system present in all cells, called the
proteasome (Figure 8-16a). The large (20S) proteasome is
composed of 14 subunits arrayed in a barrel-like structure of
symmetrical rings. Many proteins are targeted for proteolysis when a small protein called ubiquitin is attached to
them. These ubiquitin-protein conjugates enter the proteasome complex, consisting of the 20S base and an attached
19S regulatory component, through a narrow channel at the
19S end. The proteasome complex cleaves peptide bonds in
an ATP-dependent process. Degradation of ubiquitin-protein
complexes is thought to occur within the central hollow of
the proteasome.
The immune system also utilizes this general pathway of
protein degradation to produce small peptides for presentation by class I MHC molecules. In addition to the standard
20S proteasomes resident in all cells, a distinct proteasome of
the same size can be found in pAPCs and the cells of infected
tissues. This distinct proteasome, called the immunoproteasome, has some unique components that can be induced by
exposure to interferon- or TNF- (Figure 8-16b). LMP2
and LMP7, genes that are located within the class I region
(see Figure 8-8) and are responsive to these cytokines,
encode replacement catalytic protein subunits that convert
|
CHAPTER 8
(a)
285
(b)
Protein
β1
Ubiquitin
conjugation
β5
β2
Constitutive proteasome
Ub
20S
proteasome
19S
regulator
β1i
β5i
β2i
Immunoproteasome
FIGURE 8-16 Cytosolic proteolytic system for degradation of intracellular proteins. (a) Endogenous proteins may be
targeted for degradation by ubiquitin conjugation. These proteins are
degraded by the 26S proteasome complex, which includes the 20S
constitutive proteasome and a 19S regulator. (b) In activated APCs,
several proteins in the constitutive proteasome (1, 2, and 5) are
replaced by proteins encoded by the LMP genes and specific to the
immunoproteasome (1i, 2i, and 5i). This immunoproteasome has
increased proteolytic efficiency for creating peptides that can assemble with MHC class I molecules. [Adapted from M. Groettrup et al. 2010.
Proteasomes in immune cells: More than peptide producers. Nature Reviews.
Immunology 10:73–78. doi:10.1038/nri2687]
standard proteasomes into immunoproteasomes, increasing
the production of peptides that bind efficiently to MHC class
I proteins. The immunoproteasome turns over more rapidly
than a standard proteasome, possibly because the increased
level of protein degradation in its presence may have consequences beyond the targeting of infected cells. It is possible
that in some cases autoimmunity results from increased
processing of self-proteins in cells with high levels of immunoproteasomes.
Peptides Are Transported from
the Cytosol to the RER
Insight into the cytosolic processing pathway came from
studies of cell lines with defects in peptide presentation by
class I MHC molecules. One such mutant cell line, called
RMA-S, expresses about 5% of the normal levels of class I
MHC molecules on its membrane. Although RMA-S cells
synthesize normal levels of class I chain and 2-microglobulin,
few class I MHC complexes appear on the membrane. A clue
to the mutation in the RMA-S cell line was the discovery by
Townsend and his colleagues that “feeding” these cells peptides restored their level of membrane-associated class I
MHC molecules to normal. These investigators suggested
that peptides might be required to stabilize the interaction
between the class I chain and 2-microglobulin. The ability
PA R T I I I
286
|
Adaptive Immunity: Antigen Receptors and MHC
to restore expression of class I MHC molecules on the membrane by feeding the cells predigested peptides suggested that
the RMA-S cell line might have a defect in peptide transport.
Subsequent experiments showed that the defect in the
RMA-S cell line occurs in the protein that transports peptides from the cytoplasm into the RER, where class I molecules are synthesized. When RMA-S cells were transfected
with a functional gene encoding the transporter protein, the
cells began to express class I molecules on the membrane.
The transporter protein, designated TAP (for transporter
associated with antigen processing), is a membrane-spanning
heterodimer consisting of two proteins: TAP1 and TAP2
(Figure 8-17a). In addition to their multiple transmembrane
TAP1
(a)
TAP2
ATP
ATP
Cytosol
RER membrane
RER
lumen
Cytosol
(b)
Amino acids
Protein
Peptides
Immunoproteasome
ATP
ADP + Pi
TAP
RER lumen
Peptide ready to be
loaded onto class I MHC
molecule
FIGURE 8-17 TAP (transporter associated with antigen
processing). (a) Schematic diagram of TAP, a heterodimer anchored
in the membrane of the rough endoplasmic reticulum (RER). The two
chains are encoded by TAP1 and TAP2. The cytosolic domain in each
TAP subunit contains an ATP-binding site, and peptide transport
depends on the hydrolysis of ATP. (b) In the cytosol, association of 1i,
2i, and 5i (colored spheres) with a proteasome changes its catalytic specificity to favor production of peptides that bind to class I
MHC molecules. These peptides are translocated by TAP into the RER
lumen, where, in a process mediated by several other proteins, they
will associate with class I MHC molecules.
segments, the TAP1 and TAP2 proteins each have a domain
projecting into the lumen of the RER and an ATP-binding
domain that projects into the cytosol. Both TAP1 and TAP2
belong to the family of ATP-binding cassette proteins found
in the membranes of many cells, including bacteria. These
proteins mediate ATP-dependent transport of amino acids,
sugars, ions, and peptides.
Peptides generated in the cytosol by the proteasome are
translocated by TAP into the RER by a process that requires
the hydrolysis of ATP (Figure 8-17b). TAP has affinity for
peptides containing 8 to 16 amino acids. The optimal peptide length for class I MHC binding is around 9 amino
acids, and longer peptides are trimmed by enzymes present
in the ER, such as ERAP (endoplasmic reticulum aminopeptidase). In addition, TAP appears to favor peptides with
hydrophobic or basic carboxyl-terminal amino acids, the
preferred anchor residues for class I MHC molecules. Thus,
TAP is optimized to transport peptides that are most likely
to interact with class I MHC molecules. The TAP1 and
TAP2 genes map within the class II MHC region, adjacent
to the LMP2 and LMP7 genes, and different allelic forms of
these genes exist within the population. TAP deficiencies
can lead to a disease syndrome that has aspects of both
immunodeficiency and autoimmunity (see Clinical Focus
Box 8-4).
Chaperones Aid Peptide Assembly
with MHC Class I Molecules
Like other proteins destined for the plasma membrane, the
chain and 2-microglobulin components of the class I
MHC molecule are synthesized on ribosomes on the RER.
Assembly of these components into a stable class I MHC
molecular complex that can exit the RER requires the presence of a peptide in the binding groove of the class I molecule. The assembly process involves several steps and includes
the participation of molecular chaperones that facilitate the
folding of polypeptides.
The first molecular chaperone involved in class I MHC
assembly is calnexin, a resident membrane protein of the
ER. ERp57, a protein with enzymatic activity, and calnexin
associate with the free class I chain and promote its folding
(Figure 8-18). When 2-microglobulin binds to the chain,
calnexin is released and the class I molecule associates with
the chaperone calreticulin and with tapasin. Tapasin (TAPassociated protein) brings the TAP transporter into proximity with the class I molecule and allows it to acquire an
antigenic peptide. The TAP protein promotes peptide capture by the class I molecule before the peptides are exposed
to the luminal environment of the RER.
Exoproteases in the ER will act on peptides not associated with class I MHC molecules. One ER aminopeptidase,
ERAP1, removes the amino-terminal residue from peptides
to achieve optimum class I binding size (see Figure 8-18).
ERAP1 has little affinity for peptides shorter than eight
amino acids in length. As a consequence of productive
The Major Histocompatibility Complex and Antigen Presentation
|
CHAPTER 8
287
BOX 8-4
CLINICAL FOCUS
Deficiencies in TAP Can Lead to Bare Lymphocyte Syndrome
A relatively rare condition known
as bare lymphocyte syndrome (BLS) has
been recognized for more than 20 years.
The lymphocytes in BLS patients express
MHC molecules at below-normal levels
and, in some cases, not at all. Type 1 BLS
is caused by a deficiency in MHC class I
molecules; in type 2 BLS, expression of
class II molecules is impaired. The pathogenesis of type 1 BLS underscores the
importance of the class I family of MHC
molecules in their dual roles of preventing autoimmunity as well as defending
against pathogens.
One study identified a group of
patients with type 1 BLS caused by defects
in TAP1 or TAP2 genes. As described in this
chapter, TAP proteins are necessary for the
loading of peptides onto class I molecules,
a step that is essential for expression of
class I MHC molecules on the cell surface.
Lymphocytes in individuals with TAP deficiency express levels of class I molecules
significantly lower than those of normal
controls. Other cellular abnormalities
include increased numbers of NK and T
cells and decreased levels of CD8 T
cells. As we will see, the disease manifestations are reasonably well explained by
these deviations in the levels of certain
cells involved in immune function.
In early life, the TAP-deficient individual suffers frequent bacterial infections of
the upper respiratory tract and in the second decade begins to experience chronic
infection of the lungs. It is thought that a
postnasal-drip syndrome common in
younger patients promotes the bacterial
lung infections in later life. Noteworthy is
the absence of susceptibility to severe
viral infection, which is common in immunodeficiencies with T-cell involvement
(see Chapter 18). Bronchiectasis (dilation
of the bronchial tubes) often occurs, and
recurring infections can lead to lung damage that may be fatal. The most characteristic mark of the deficiency is the
occurrence of necrotizing skin lesions on
the extremities and the midface (Figure 1).
These lesions ulcerate and may cause
disfigurement. The skin lesions are probably due to activated NK cells and T cells;
NK cells were isolated from biopsied skin
from several patients, supporting this possibility. Normally, the activity of NK cells is
limited through the action of killer-cellinhibitory receptors (KIRs), which deliver a
negative signal to the NK cell following
interaction with class I molecules (see
Chapter 13). The lack of class I molecules
in BLS patients with TAP deficiency explains
the excessive activity of the NK cells.
Activation of NK cells further explains the
absence of severe viral infections, which
are limited by NK and T cells.
The best treatment for the characteristic lung infections appears to be a combination of antibiotics and intravenous
immunoglobulin. Attempts to limit the
skin disease by immunosuppressive regimens, such as steroid treatment or cytotoxic agents, can lead to exacerbation of
lesions and are therefore contraindicated.
Mutations in the promoter region of TAP
that prevent expression of the gene were
found for several patients, suggesting the
possibility of gene therapy, but the cellular distribution of class I is so widespread
that it is not clear what cells would need
to be corrected to alleviate all symptoms.
FIGURE 1
Necrotizing granulomatous lesions in the midface of a patient with TAPdeficiency syndrome. TAP deficiency leads to a condition with symptoms characteristic of
autoimmunity, such as the skin lesions that appear on the extremities and the midface, as well as
immunodeficiency that causes chronic sinusitis, leading to recurrent lung infection. [From S. D. Gadola
et al., 1999, Lancet 354:1598, and 2000, Clinical and Experimental Immunology 121:173.]
PA R T I I I
288
|
Adaptive Immunity: Antigen Receptors and MHC
RER lumen
ERp57
Tapasin
ERAP
Calreticulin-tapasin–
associated class I
MHC molecule
ERp57
β2-microglobulin
Calreticulin
Calreticulin
Peptide
Exit RER
+
Class I MHC
α chain
Tapasin
Calnexin
TAP
Peptides
Class I MHC
molecule
Proteasome
Cytosol
FIGURE 8-18 Assembly and stabilization of class I MHC
molecules. Within the rough endoplasmic reticulum (RER) membrane, a newly synthesized class I chain associates with calnexin, a
molecular chaperone, and ERp57 until 2-microglobulin binds to
the chain. The binding of 2-microglobulin releases calnexin and
allows binding to calreticulin and to tapasin, which is associated
peptide binding, the class I molecule displays increased stability and can dissociate from the complex with calreticulin,
tapasin, and ERp57. The class I molecule can then exit from
the RER and proceed to the cell surface via the Golgi complex.
The Exogenous Pathway of Antigen
Processing and Presentation
APCs can internalize particulate material by simple phagocytosis (also called “cell eating”), where material is engulfed by
pseudopods of the cell membrane, or by receptor-mediated
endocytosis, where the material first binds to specific surface
receptors. Macrophages and dendritic cells internalize antigen
by both processes. Most other APCs, whether professional or
not, demonstrate little or no phagocytic activity and therefore
typically internalize exogenous antigen only by endocytosis
(either receptor-mediated endocytosis or by pinocytosis, “cell
drinking”). B cells, for example, internalize antigen very effectively by receptor-mediated endocytosis using their antigenspecific membrane immunoglobulin as the receptor.
Peptides Are Generated from Internalized
Antigens in Endocytic Vesicles
Once an antigen is internalized, it is degraded into peptides
within compartments of the endocytic processing pathway.
with the peptide transporter TAP. This association promotes binding
of an antigenic peptide. Antigens in the ER can be further processed
via exopeptidases such as ERAP1, producing fragments ideally
suited for binding to class I. Peptide association stabilizes the class I
molecule-peptide complex, allowing it to be transported from the RER
to the plasma membrane.
As the experiment shown in Figure 8-14 demonstrated,
internalized antigen takes 1 to 3 hours to traverse the endocytic pathway and appear at the cell surface in the form of
class II MHC-peptide complexes. The endocytic antigen
processing pathway appears to involve several increasingly
acidic compartments, including early endosomes (pH 6.0–
6.5); late endosomes, or endolysosomes (pH 4.5–5.0); and
lysosomes (pH 4.5). Internalized antigen progresses
through these compartments, encountering hydrolytic
enzymes and a lower pH in each compartment (Figure 8-19).
Antigen-presenting cells have a unique form of late endosome, the MHC class II-containing compartment (MIIC),
in which final protein degradation and peptide loading
into MHC class II proteins occurs. Within the compartments of the endocytic pathway, antigen is degraded into
oligopeptides of about 13 to 18 residues that meet up with
and bind to class II MHC molecules in late endosomes.
Because the hydrolytic enzymes are optimally active under
acidic conditions (low pH), antigen processing can be
inhibited by chemical agents that increase the pH of the compartments (e.g., chloroquine) as well as by protease inhibitors (e.g., leupeptin).
The mechanism by which internalized antigen moves
from one endocytic compartment to the next has not been
conclusively demonstrated. It has been suggested that early
endosomes from the periphery move inward to become late
endosomes and eventually lysosomes. Alternatively, small
The Major Histocompatibility Complex and Antigen Presentation
CHAPTER 8
289
the ER and do not transit past the cis-Golgi. However, in
cells transfected with both the class II MHC genes and
the Ii gene, the class II molecules were localized in the
cytoplasmic vesicular structures of the endocytic pathway. The invariant chain contains sorting signals in its
cytoplasmic tail that direct the transport of the class II
MHC complex from the trans-Golgi network to the endocytic compartments.
Antigen
Recycling
of receptors
Clathrincoated
vesicle
|
Early endosome
pH 6.0–6.5
Endoplasmic reticulum
Golgi complex
MHC class II-Ii
MIIC Late endosome
pH 4.5-5.0
FIGURE 8-19 Generation of antigenic peptides in the
exogenous processing pathway. Internalized exogenous antigen moves through several acidic compartments ending in specialized MIIC late endosomes, where it is degraded into peptide
fragments which associate with class II MHC molecules transported
in vesicles from the Golgi complex. The cell shown here is a B cell,
which internalizes antigen by receptor-mediated endocytosis, with
the membrane-bound antibody functioning as an antigen-specific
receptor.
transport vesicles may carry antigens from one compartment to the next. Eventually the endocytic compartments,
or portions of them, return to the cell periphery, where
they fuse with the plasma membrane. In this way, the surface receptors are recycled.
The Invariant Chain Guides Transport of Class II
MHC Molecules to Endocytic Vesicles
Since APCs express both class I and class II MHC molecules,
some mechanism must exist to prevent class II MHC molecules from binding to the antigenic peptides destined for the
class I molecules. When class II MHC molecules are synthesized within the RER, these class II chains associate with
a protein called the invariant chain (Ii, CD74). This conserved, non-MHC encoded protein interacts with the class II
peptide-binding groove preventing any endogenously
derived peptides from binding while the class II molecule is
within the RER (Figure 8-20a). The invariant chain also appears
to be involved in the folding of the class II and chains, their
exit from the RER, and the subsequent routing of class II
molecules to the endocytic processing pathway from the
trans-Golgi network.
The role of the invariant chain in the routing of class
II molecules has been demonstrated in transfection
experiments with cells that lack both the genes encoding
class II MHC molecules and the invariant chain. Immunofluorescent labeling of these cells transfected only with
class II MHC genes revealed that, in the absence of
invariant chain, class II molecules remain primarily in
Peptides Assemble with Class II MHC
Molecules by Displacing CLIP
Recent experiments indicate that most class II MHCinvariant chain complexes are transported from the RER,
where they are formed, through the Golgi complex and
trans-Golgi network, and then through the endocytic pathway, moving from early endosomes to the MIIC late endosomal compartment in some cases. As the proteolytic activity
increases in each successive compartment, the invariant
chain is gradually degraded. However, a short fragment of the
invariant chain termed CLIP (for class II–associated invariant chain peptide) remains bound to the class II molecule
after the majority of the invariant chain has been cleaved
within the endosomal compartment. Like antigenic peptide,
CLIP physically occupies the peptide-binding groove of the
class II MHC molecule, preventing any premature binding of
antigen-derived peptide (Figure 8-20b).
A nonclassical class II MHC molecule called HLA-DM
is required to catalyze the exchange of CLIP with antigenic
peptides (see Figure 8-20a). The DM and DM genes are
located near the TAP and LMP genes in the MHC complex
of humans, with similar genes in mice (see Figure 8-8).
Like other class II MHC molecules, HLA-DM is a heterodimer of and chains. However, unlike other class II
molecules it is relatively nonpolymorphic and is not normally expressed at the cell membrane but is found predominantly within the endosomal compartment. HLA-DM
has been found to associate with the class II MHC chain
and to function in removing or “editing” peptides, including CLIP, that associate transiently with the binding
groove of classical class II molecules. Peptides that make
especially strong molecular interactions with class II
MHC, creating long-lived complexes, are harder for HLADM to displace, and thus become the repertoire of peptides that ultimately make it to the cell surface as
MHC-peptide complexes.
As with class I MHC molecules, peptide binding is
required to maintain the structure and stability of class II
MHC molecules. Once a peptide has bound, the class II
MHC-peptide complex is transported to the plasma membrane, where the neutral pH appears to enable the complex
to assume a compact, stable form. Peptide is bound so
strongly in this compact form that it is difficult to replace a
class II–bound peptide on the membrane with another peptide at physiologic conditions.
PA R T I I I
290
(a)
|
Adaptive Immunity: Antigen Receptors and MHC
FIGURE 8-20 Assembly of class II MHC molecules. (a) Within
Digested invariant chain
Peptides
CLIP
αβ
+
MHC
Class II
Released CLIP
–
+
Invariant
chain
HLA-DM
HLA-DO
(b)
α1
C
N
β1
the rough endoplasmic reticulum, a newly synthesized class II MHC
molecule binds an invariant chain. The bound invariant chain prevents premature binding of peptides to the class II molecule and
helps to direct the complex to endocytic compartments containing
peptides derived from exogenous antigens. Digestion of the invariant chain leaves CLIP, a small fragment remaining in the binding
groove of the class II MHC molecule. HLA-DM, a nonclassical MHC
class II molecule present within the MIIC compartment, mediates
exchange of antigenic peptides for CLIP. The nonclassical class II
molecule HLA-DO may act as a negative regulator of class II antigen
processing by binding to HLA-DM and inhibiting its role in the dissociation of CLIP from class II molecules. (b) Comparison of threedimensional structures showing the binding groove of HLA class II
molecules (1, 1) containing different antigenic peptides or CLIP.
The red lines show HLA-DR4 complexed with collagen II peptide,
yellow lines are HLA-DR1 with influenza hemagglutinin peptide,
and dark blue lines are HLA-DR3 associated with CLIP. (N indicates
the amino-terminus and C the carboxyl-terminus of the peptides.)
No major differences in the structures of the class II molecules or in
the conformation of the bound peptides are seen. This comparison
shows that CLIP binds the class II molecule in a manner identical to
that of antigenic peptides. [Part b from A. Dessen et al., 1997, Immunity
7:473–481; courtesy of Don Wiley, Harvard University.]
One additional nonclassical member of the class II MHC
family, HLA-DO, like DM, is another relatively nonpolymorphic class II molecule. Unlike other class II molecules,
its expression is not induced by IFN- . HLA-DO has been
found to act as a negative regulator of antigen binding,
modulating the function of HLA-DM and changing the
repertoire of peptides that preferentially bind to classical
class II molecules. In cells that express both DO and DM,
these two molecules strongly associate in the ER and maintain this interaction all the way to the endosomal compartments. Although this interaction has been recognized for
many years, the function of this negative regulator and the
impact of this changed peptide repertoire is only now being
resolved.
Originally observed only in B cells and the thymus, the
cellular expression profile of HLA-DO has recently been
expanded to dendritic cells (DCs), where it is thought to
play a role in the maintenance of self-tolerance (discussed
further in Chapter 16). This phenomenon was studied by L.
K. Denzin and colleagues using diabetes-prone mice engineered to express human HLA-DO. DCs are known to be
essential in the establishment of self-tolerance, as well as in
the presentation of autoantigens to self-reactive T cells. Selfreactive T cells develop in these non-obese diabetic animals,
and these cells ultimately destroy pancreatic cells, causing
diabetes. In this study, the development of diabetes was
blocked by the presence of the HLA-DO transgene in DCs.
In addition, using specific monoclonal antibodies, they
showed that the repertoire of peptides being presented to
the autoreactive T cells was significantly altered, resulting in
a reduced efficiency in presenting key self-peptides. This
and other work suggest that normal DO expression may
play an important role in modulating DM behavior to
ensure the presentation of a self-peptide repertoire that
encourages tolerance to self-antigens. Interestingly, in wildtype mice and humans, DO expression is down-regulated
following DC activation by antigen, releasing HLA-DM to
carry out its normal function of encouraging the presentation of a diverse array of peptides, many of which will presumably be derived from the foreign proteins that stimulated
these APCs.
Figure 8-21 illustrates the endogenous pathway (left side)
and compares it with the separate exogenous pathway (right),
described above. Whether an antigenic peptide associates
with class I or with class II molecules is partially dictated by
the mode of entry into the cell, either exogenous or endogenous, by the site of processing, and by the cell type. However,
in the next section, we will see that these assignments are not
absolute and that exogenous antigens may, in some APCs, be
presented by class I antigens via a unique pathway.
|
The Major Histocompatibility Complex and Antigen Presentation
Endogenous pathway
(class I MHC)
CHAPTER 8
291
Exogenous pathway
(class II MHC)
Endogenous
antigen
1
Endogenous antigen is
degraded by proteasome.
Proteasome
2
Rough endoplasmic
reticulum (RER)
1
Peptide is transported to
RER via TAP.
TAP
ERp57
Calreticulin
Invariant
chain
Class II MHC ␣ and 
bind invariant chain,
blocking binding of
endogenous antigen.
Tapasin
β2- microglobulin
Class I
MHC
Calnexin
Class II
MHC
2
MHC complex is routed
through Golgi to
endocytic pathway
compartments.
Peptide
3
Class I MHC ␣ chain
binds calnexin and ERp57,
then β2 microglobulin.
Calnexin dissociates.
Calreticulin and tapasin,
bind. MHC captures peptide,
chaperones dissociate.
3
Golgi complex
CLIP
Digested
invariant
chain
4
Invariant chain is
degraded, leaving CLIP
fragment.
Exogenous antigen is
taken up, degraded,
routed to endocytic
pathway compartments.
4
Class I MHC–peptide is
transported from RER to
Golgi complex to plasma
membrane.
Exogenous
antigen
5
HLA-DM mediates
exchange of CLIP for
antigenic peptide.
6
Class I
MHC
Class II
MHC
Class II MHC–peptide is
transported to plasma
membrane.
FIGURE 8-21 Separate antigen-presenting pathways are utilized for endogenous (green) and exogenous (red) antigens.
The mode of antigen entry into cells and the site of antigen processing determine whether antigenic peptides associate with class I MHC molecules in the rough endoplasmic reticulum or with class II molecules in endocytic compartments.
Cross-Presentation of
Exogenous Antigens
We have seen that after pAPCs internalize extracellular
antigen they can process and present these antigens via
the exogenous pathway. Because pAPCs also express
costimulatory molecules, engagement of CD4 T cells
with these class II MHC-peptide complexes can lead to
activation of T helper responses. On the other hand,
infected cells will generally process and present cytosolic
peptides via the endogenous pathway, leading to class I
MHC-peptide complexes on their surface. However,
unless these infected cells are pAPCs, they will not express
the costimulatory molecules necessary to activate naïve
CD8 T cells. This leaves us with a dilemma: how does the
immune system activate naïve CD8 T cells to eliminate
292
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
intracellular microbes unless a professional antigenpresenting cell happens to become infected? And if a
pAPC is not harboring an intracellular infection, how
does this cell present pathogens, such as viruses that have
been engulfed from extracellular sources in ways that will
activate the needed CTL responses (usually mediated by
the endogenous pathway)?
The answer to this dilemma is a process called crosspresentation. In some instances, APCs will divert antigen
obtained by endocytosis (exogenous antigen) to a pathway
that leads to class I MHC loading and peptide presentation
to CTLs (like in the endogenous pathway)—in other words,
crossing the two pathways. First reported by Michael Bevan
and later described in detail by Peter Cresswell and colleagues, the phenomenon of cross-presentation requires that
internalized antigens that would normally be handled by the
exogenous pathway leading to class II MHC presentation
somehow become redirected to a class I peptide loading pathway. When this form of antigen presentation leads to the
activation of CTL responses, it is referred to as cross-priming;
when it leads to the induction of tolerance in these CD8
T cells it is called cross-tolerance.
Dendritic Cells Appear to Be the Primary
Cross-Presenting Cell Type
The concept of cross-presentation was first recognized as
early as 1976, but the relevant cell types and mechanisms of
action remained a mystery until more recently. New studies,
focused on subsets of DCs that are now recognized to be the
most efficient of the cross-presenters, are just beginning to
shed light on this process. For example, in vivo depletion or
inactivation of DCs compromises cross-presentation. The
most potent of the cross-presenting DCs appear to reside in
secondary lymphoid organs, where they are believed to
receive antigen from extracellular sources by handoff from
migrating APCs or from dying infected cells. A few other
cells types, such as B cells, macrophages, neutrophils, and
mast cells, have been found capable of cross-presentation in
vitro, although evidence that they cross-present in vivo or
that they have the ability to prime CTL responses (i.e., activate naïve CD8 T cells) is still lacking.
Mechanisms and Functions
of Cross-Presentation
Two possible models have been proposed for how DCs
accomplish cross-presentation. The first hypothesizes that
cross-presenting cells possess special antigen-processing
machinery that allows loading of exogenously derived
peptides onto class I MHC molecules. The second theory
postulates specialized endocytosis machinery that can
send internalized antigen directly to an organelle (such as
a phagosome or early endosome), where peptides from
those antigens are then loaded onto class I MHC mole-
cules using conventional machinery. These two proposed
mechanisms are not mutually exclusive, and evidence for
both exists.
The means by which antigen achieves the crossover
from its exogenous or extracellular origins to the endogenous pathway is unresolved. However, in many instances
cross-presented antigen has been found to enter the cytoplasm. In the early 1990s, it was shown that bead-conjugated
antigens captured by cross-presenting DCs could reach the
cytosol of these cells. Later, M. L. Lin and colleagues showed
that exogenously added cytochrome c, which causes programmed cell death when present in the cytosol, could
induce cross-presenting DCs to undergo apoptosis. It is
now proposed that retro-translocation, or movement of
endocytosed proteins out of endocytic compartments and
into the cytosol, can occur via TAP molecules present in
these endocytic membranes—a sort of backflow out of
endocytic vesicles. However, TAP-independent cross-presentation of some antigens has also been observed, suggesting multiple mechanisms may exist for targeting externally
derived peptides to class I molecules.
The ability of some DCs to cross-present antigens has
great advantage for the host. It allows these APCs to capture
virus from the extracellular environment or from dying cells,
process these viral antigens, and activate CTLs that can
attack virus-infected cells, inhibiting further spread of the
infection. One outstanding question remains: if this pathway
is so important, why don’t all APCs cross-present? The
answer may lie in the fact that cross-presenting cells could
quickly become targets of lysis themselves. Another dilemma
concerns how cross-presenting DCs handle self-peptides
presented in MHC class I. Shouldn’t these cells break tolerance by presenting extracellularly-derived self-peptides to
CD8 T cells, activating anti-self CTL responses?
To help explain how cross-presentation is regulated and
tolerance is maintained, scientists have proposed that DCs
might first need to be “licensed” before they can crosspresent. The cell type postulated to supply this licensing role
is activated CD4 T cells. The way this is believed to work
is that, first, the classical exogenous pathway of antigen
processing in DCs leads to presentation of antigen to CD4
T cells via class II, leading to activation of these cells (Figure 8-22a). These activated helper cells might then return
the favor by inducing costimulatory molecules in the DC
and by cytokine secretion (e.g., IL-2), supplying a “second
opinion” that, respectively, licenses the DC to present internalized antigens via MHC class I and helps activate naïve
CD8 T cells (Figure 8-22b). This requirement for licensing
by a TH cell could help avoid accidental induction of CTLs to
nonpathogenic antigens or self-proteins.
If this is the case, any unlicensed DCs that cross-present
antigens may serve the opposite and equally important purpose. Since they have not received the go-ahead from the TH
cell to activate CTL responses, they may instead induce tolerance in the CD8 T cells they encounter, helping to
dampen reactivity to these antigens.
The Major Histocompatibility Complex and Antigen Presentation
(a) DC licensing by TH cell
Dendritic cell
Exogenous
antigen
Exogenous
pathway
Class II MHC
CD4
CD3
TCR
TH cell
CD40
CD40L
(b) DC cross-presentation and activation of CTL
Cross-presenting
dendritic cell
Exogenous
antigen
TLR
Crossover
pathway
Class I MHC
CD8
CD3
CD80/CD86
CD28
IL-2
Naïve
TC cell
FIGURE 8-22 Activation of naïve Tc cells by exogenous
antigen requires DC licensing and cross-presentation. (a) Dendritic cells (DCs) first internalize and process antigen through the
exogenous pathway, presenting to CD4 TH cells via MHC class II
molecules and activating these cells through, among other things,
CD40-CD40L engagement. (b) These activated TH cells can then serve
as a bridge to help activate CTL responses; they provide local IL-2 and
they in turn license the DC to cross-present internalized antigen in
MHC class I, up-regulate costimulatory molecules, and down-regulate
their inhibitory counterparts. DC licensing creates an ideal situation
for the stimulation of antigen-specific CD8 T cell responses. When
the TLRs on these DCs are engaged, this further activates these cells,
providing added encouragement for cross-presentation. Dashed
arrows indicate antigen directed for cross-presentation. [Adapted from
Kurts et al., 2010, Nature Reviews Immunology 10:403.]
Presentation of Nonpeptide Antigens
To this point, the discussion of the presentation of antigens
has been limited to protein antigens and their presentation
by classical class I and II MHC molecules. It is well known
|
CHAPTER 8
293
that some nonprotein antigens are also recognized by T cells,
and in the 1980s T-cell proliferation was detected in the presence of nonprotein antigens derived from infectious agents.
Many reports now indicate that various types of T cells
(expressing
as well as T–cell receptors) can react
against lipid antigens, such as mycolic acid, derived from
well-known pathogens, such as Mycobacterium tuberculosis.
These antigens are presented by members of the CD1 family
of nonclassical class I molecules.
CD1 molecules share structural similarity to classical
MHC class I but overlap functionally with MHC class II.
Five human CD1 genes are known; all are encoded on a
chromosome separate from the classical class I molecules
and display very limited polymorphism. Much like classical
MHC class I, CD1 proteins are formed by a transmembrane
heavy chain, composed of three extracellular domains,
which associates noncovalently with 2-microglobulin. In
terms of trafficking and expression profile, however, most
CD1 molecules resemble MHC class II proteins, moving
intracellularly to endosomal compartments, where they
associate with exogenous antigen. Like MHC class II molecules, CD1 proteins are expressed by many immune cell
types, including thymocytes, B cells, and DCs, although
some members of the family have also been found on hepatocytes and epithelial cells.
Many different lipid or lipid-linked structures have been
found to associate non-covalently with CD1 molecules. In
general, most of these are glycolipid or lipoprotein antigens,
where the lipid moiety fits into deep pockets within the CD1
binding groove and the hydrophilic head group remains
exposed, allowing recognition by T cells. Crystal structures
have demonstrated that CD1 contains a binding groove that
is both deeper and narrower than classical MHC molecules.
These grooves are lined with nonpolar amino acids, which
can easily accommodate lipid structures. Some of these
pockets correspond roughly with the antigen-binding pockets of classical MHC class I and allow antigen access to a
deep groove through a narrow opening—more like a foot
sliding into a shoe (Figure 8-23a). Figure 8-23b illustrates a
comparison of the binding groove of CD1b holding a lipid
antigen with a classical class I molecule complexed with
peptide antigen. The relatively nonspecific manner of antigen association with the CD1 binding groove, which relies
primarily on many hydrophobic interactions, probably
accounts for the diversity of self- and foreign antigens that
can be presented by these nonclassical class I molecules. A
variety of T cells are known to bind these nonclassical MHC
molecules.
It is now hypothesized that short-chain self-lipids with
relatively low affinity are loaded onto CD1 molecules in the
ER, shortly after translation, and allow proper CD1 protein
folding. These self-antigen loaded CD1 molecules then travel
to the cell surface, where in some cases exogenous lipids may
be exchanged with these low-affinity self-antigens. Like with
MHC class II molecules, after endocytosis and movement to
lower pH environments, CD1 proteins are believed to
294
PA R T I I I
|
Adaptive Immunity: Antigen Receptors and MHC
(a)
A
C
F
T
β2-microglobulin
(b)
FIGURE 8-23 Lipid antigen binding to the CD1 molecule. (a) Cartoon of the deep binding pockets (A, C, and F) and the deep binding
tunnel (T) that accommodate lipid antigens in a foot-in-shoe fashion in the members of the CD1 family of nonclassical class I molecules. (b)
Ribbon diagram of the binding groove or pocket of human CD1b complexed with lipid (left) compared with a classical class I molecule binding
peptide antigen (right). [Part b PDB IDs 1GZQ (left) and 1HHG (right).]
exchange their self peptide, low-affinity binding partners for
exogenously derived long-chain, high-affinity lipid antigens
resident in phagolysosomes or late endosomes. These newly
loaded CD1 molecules then return to the cell surface for
recognition by CD1-restricted T cells. Recently constructed
transgenic mice that express human CD1 molecules will likely
lead the way to new murine models aimed at better defining
the role of CD1 restricted T cells in protection from infection.
The Major Histocompatibility Complex and Antigen Presentation
|
CHAPTER 8
295
S U M M A R Y
■
■
■
■
■
■
■
The major histocompatibility complex (MHC) encodes class
I and II molecules, which function in antigen presentation to
T cells, and class III molecules, which have diverse functions.
Class I MHC molecules consist of a large glycoprotein
chain encoded by an MHC gene, and 2-microglobulin, a
protein with a single domain that is encoded elsewhere.
Class II MHC molecules are composed of two noncovalently associated glycoproteins, the and chains, encoded
by separate MHC genes.
MHC genes are tightly linked and generally inherited as a
unit from parents; these linked units are called haplotypes.
MHC genes are polymorphic (many alleles exist for each
gene in the population), polygenic (several different MHC
genes exist in an individual), and codominantly expressed
(both maternal and paternal copies).
MHC alleles influence the fragments of antigen that are
presented to the immune system, thereby influencing susceptibility to a number of diseases.
Class I molecules are expressed on most nucleated cells;
class II molecules are restricted to B cells, macrophages,
and dendritic cells (pAPCs).
■
■
■
■
■
■
In most cases, class I molecules present processed endogenous antigen to CD8 TC cells and class II molecules
present processed exogenous antigen to CD4 TH cells.
Endogenous antigens are degraded into peptides within
the cytosol by proteasomes, assemble with class I molecules in the RER, and are presented on the membrane to
CD8 TC cells. This is the endogenous processing and
presentation pathway.
Exogenous antigens are internalized and degraded within the
acidic endocytic compartments and subsequently combine
with class II molecules for presentation to CD4 TH cells.
This is the exogenous processing and presentation pathway.
Peptide binding to class II molecules involves replacing a
fragment of invariant chain in the binding groove by a process catalyzed by nonclassical MHC molecule HLA-DM.
In some cases, exogenous antigens in certain cell types
(mainly DCs) can gain access to class I presentation pathways in a process called cross-presentation.
Presentation of nonpeptide (lipid and lipid-linked) antigens
derived from pathogens involves the nonclassical class I–like
CD1 molecules.
R E F E R E N C E S
Amigorena, S., and A. Savina. 2010. Intracellular mechanisms
of antigen cross presentation in dendritic cells. Current Opinion in Immunology 22:109.
Margulies, D. 1999. The major histocompatibility complex. In
Fundamental Immunology, 4th ed. W. E. Paul, ed. LippincottRaven, Philadelphia.
Brown, J. H., et al. 1993. Three-dimensional structure of the human
class II histocompatibility antigen HLA-DR1. Nature 364:33.
Moody, D. B., D. M. Zajonc, and I. A. Wilson. 2005. Anatomy of
CD1-lipid antigen complexes. Nature Reviews in Immunology
5:387.
Doherty, P. C., and R. M. Zinkernagel. 1975. H-2 compatibility
is required for T-cell mediated lysis of target cells infected
with lymphocytic choriomeningitis virus. Journal of Experimental Medicine 141:502.
Fahrer, A. M., et al. 2001. A genomic view of immunology.
Nature 409:836.
Gadola, S. D., et al. 2000. TAP deficiency syndrome. Clinical
and Experimental Immunology 121:173.
Horton, R., et al. 2004. Gene map of the extended human MHC.
Nature Reviews Genetics 5:1038.
International Human Genome Sequencing Consortium. 2001.
Initial sequencing and analysis of the human genome. Nature
409:860.
Rock, K. L., et al. 2004. Post-proteosomal antigen processing for
major histocompatibility complex class I presentation. Nature
Immunology 5:670.
Rothenberg, B. E., and J. R. Voland. 1996. Beta 2 knockout mice
develop parenchymal iron overload: A putative role for class I
genes of the major histocompatibility complex in iron metabolism.
Proceedings of the National Academy of Science U.S.A. 93:1529.
Rouas-Freiss, N., et al. 1997. Direct evidence to support the role
of HLA-G in protecting the fetus from maternal uterine natural killer cytolysis. Proceedings of the National Academy of
Science USA. 94:11520.
Strominger, J. L. 2010. An alternative path for antigen presentation: Group 1 CD1 proteins. Journal of Immunology 184(7):3303
Kelley, J., et al. 2005. Comparative genomics of major histocompatibility complexes. Immunogenetics 56:683.
Wearsch, P. A., and P. Cresswell. 2008. The quality control of MHC
class I peptide loading. Current Opinion in Cell Biology 20(6):624.
Kurts, C., B. W. S. Robinson, and P. A. Knolle. 2010. Cross-priming in health and disease. Nature Reviews in Immunology 10:403.
Yaneva, R., et al. 2010. Peptide binding to MHC class I and II
proteins: New avenues from new methods. Molecular Immunology 47:649.
Li, X. C., and M. Raghavan. 2010. Structure and function of
major histocompatibility complex class I antigens. Current
Opinions in Organ Transplantation 15:499.
Madden, D. R. 1995. The three-dimensional structure of peptide-MHC complexes. Annual Review of Immunology 13:587.
Yi, W., et al. 2010. Targeted regulation of self-peptide presentation prevents type I diabetes in mice without disrupting general immunocompetence. Journal of Clinical Investigation
120(4):1324.
PA R T I I I
296
|
Adaptive Immunity: Antigen Receptors and MHC
map of the human MHC from the Horton et al. reference
cited above.
Useful Web Sites
www.bioscience.org This Frontiers in Bioscience site
includes in the Research Tools section a database that lists
gene knockouts. Information is available on studies of the
consequences of targeted disruption of MHC molecules and
other component molecules, including 2-microglobulin
and the class II invariant chain.
www.bshi.org.uk The British Society for Histocompatibility and Immunogenetics home page contains information
on tissue typing and transplantation and links to worldwide
sites related to the MHC.
www.ebi.ac.uk/imgt/hla The International ImMunoGeneTics (IMGT) database section contains links to sites
with information about HLA gene structure and genetics
and up-to-date listings and sequences for all HLA alleles
officially recognized by the World Health Organization HLA
nomenclature committee.
www.hla.alleles.org A Web site maintained by the HLA
Informatics Group based at the Anthony Nolan Trust in the
United Kingdom, with up-to-date information on the numbers of HLA alleles and proteins.
www.nature.com/nrg/journal/v5/n12/poster/
MHCmap/index.html A poster with an extended gene
S T U D Y
Q U E S T I O N S
CLINICAL FOCUS QUESTION Patients with TAP deficiency have
partial immunodeficiency as well as autoimmune manifestations. How do the profiles for patients’ immune cells explain
the partial immunodeficiency? Why is it difficult to design a
gene therapy treatment for this disease, despite the fact that a
single gene defect is implicated?
MHC molecules expressed on the
membrane of the transfected L cells
Transfected
gene
Dk
Db
Kk
Kb
IAk
IAb
None
1. Indicate whether each of the following statements is true or
false. If you think a statement is false, explain why.
a. A monoclonal antibody specific for 2-microglobulin
b.
c.
d.
e.
f.
g.
can be used to detect both class I MHC K and D molecules on the surface of cells.
Antigen-presenting cells express both class I and class II
MHC molecules on their membranes.
Class III MHC genes encode membrane-bound
proteins.
In outbred populations, an individual is more likely to
be histocompatible with one of its parents than with
its siblings.
Class II MHC molecules typically bind to longer peptides than do class I molecules.
All nucleated cells express class I MHC molecules.
The majority of the peptides displayed by class I and
class II MHC molecules on cells are derived from selfproteins.
d
k
2. You cross a BALB/c (H-2 ) mouse with a CBA (H-2 )
mouse. What MHC molecules will the F1 progeny express
on (a) its liver cells and (b) its macrophages?
3. To carry out studies on the structure and function of the
class I MHC molecule Kb and the class II MHC molecule
IAb, you decide to transfect the genes encoding these
proteins into a mouse fibroblast cell line (L cell) derived
from the C3H strain (H-2k). L cells do not normally
function as antigen-presenting cells. In the following
table, indicate which of the listed MHC molecules will
() or will not () be expressed on the membrane of the
transfected L cells.
Kb
IA␣b
IAb
IA␣b and IAb
s
4. The SJL mouse strain, which has the H-2 haplotype, has a
deletion of the IE locus.
a. List the classical MHC molecules that are expressed on
the membrane of macrophages from SJL mice.
k
b. If the class II IE and IE genes from an H-2 strain are
transfected into SJL macrophages, what additional classical MHC molecules would be expressed on the transfected macrophages?
5. Draw diagrams illustrating the general structure, including
the domains, of class I MHC molecules, class II MHC molecules, and membrane-bound antibody on B cells. Label each
chain and the domains within it, the antigen-binding regions,
and regions that have the immunoglobulin-fold structure.
6. One of the characteristic features of the MHC is the large
number of different alleles at each locus.
a. Where are most of the polymorphic amino acid resi-
dues located in MHC molecules? What is the significance of this location?
b. How is MHC polymorphism thought to be generated?
7. As a student in an immunology laboratory class, you have
been given spleen cells from a mouse immunized with the
LCM virus (LCMV). You determine the antigen-specific
The Major Histocompatibility Complex and Antigen Presentation
functional activity of these cells with two different assays.
In assay 1, the spleen cells are incubated with macrophages
that have been briefly exposed to LCMV; the production of
interleukin 2 (IL-2) is a positive response. In assay 2, the
spleen cells are incubated with LCMV-infected target cells;
lysis of the target cells represents a positive response in this
assay. The results of the assays using macrophages and
target cells of different haplotypes are presented in the table
below. Note that the experiment has been set up in a way to
exclude alloreactive responses (reactions against nonself
MHC molecules).
a. The activity of which cell population is detected in each
b.
c.
d.
e.
of the two assays?
The functional activity of which MHC molecules is
detected in each of the two assays?
From the results of this experiment, which MHC molecules are required, in addition to the LCM virus, for specific reactivity of the spleen cells in each of the two assays?
What additional experiments could you perform to
unambiguously confirm the MHC molecules required
for antigen-specific reactivity of the spleen cells?
Which of the mouse strains listed in the table could
have been the source of the immunized spleen cells
tested in the functional assays? Give your reasons.
MHC haplotype
of macrophages
and virus-infected
target cells
Response of
spleen cells
CHAPTER 8
297
10. The hypothetical allelic combination HLA-A99 and HLA-
B276 carries a relative risk of 200 for a rare, and yet
unnamed, disease that is fatal to preadolescent children.
a. Will every individual with A99/B276 contract the disease?
b. Will everyone with the disease have the A99/B276
combination?
c. How frequently will the A99/B276 allelic combination be
observed in the general population? Do you think that this
combination will be more or less common than predicted
by the frequency of the two individual alleles? Why?
11. Explain the difference between the terms antigen-presenting
cell and target cell, as they are commonly used in immunology.
12. Define the following terms:
a.
b.
c.
d.
e.
f.
Self-MHC restriction
Antigen processing
Endogenous antigen
Exogenous antigen
Anchor residues
Immunoproteasome
13. Ignoring cross-presentation, indicate whether each of the
cell components or processes listed is involved in the processing and presentation of exogenous antigens (EX), endogenous antigens (EN), or both (B). Briefly explain the
function of each item.
a.
b.
c.
d.
e.
f.
______ Class I MHC molecules
______ Class II MHC molecules
______ Invariant (Ii) chains
______ Lysosomal hydrolases
______ TAP1 and TAP2 proteins
______ Transport of vesicles from the RER to the Golgi
complex
______ Proteasomes
______ Phagocytosis or endocytosis
______ Calnexin
______ CLIP
______ Tapasin
Mouse
strain
used as
source of
macrophages
and target
cells
K
IA
IE
D
C3H
k
k
k
k
BALB/C
d
d
d
d
d/k
d/k
d/k
d/k
A.TL
s
k
k
d
a. If chloroquine is added to the incubation mixture, pre-
B10.A (3R)
b
b
b
d
B10.A (4R)
k
k
—
b
sentation of the native protein is inhibited, but the
peptide continues to induce TH-cell activation. Explain
why this occurs.
b. If chloroquine addition is delayed for 3 hours, presentation of the native protein is not inhibited. Explain why
this occurs.
(BALB/C X
B10.A)F1
IL-2
production
in response
Lysis of
to LCMVLCMVpulsed
infected
macrophages
cells
(assay 1)
(assay 2)
|
8. A TC-cell clone recognizes a particular measles virus pep-
tide when it is presented by H-2Db. Another MHC molecule has a peptide-binding groove identical to the one in
H-2Db but differs from H-2Db at several other amino acids
in the 1 domain. Predict whether the second MHC molecule could present this measles virus peptide to the TC-cell
clone. Briefly explain your answer.
9. Human red blood cells are not nucleated and do not
express any MHC molecules. Why is this property fortuitous for blood transfusions?
g.
h.
i.
j.
k.
14. Antigen-presenting cells have been shown to present lyso-
zyme peptide 46–61 together with the class II IAk molecule.
When CD4 TH cells are incubated with APCs and native
lysozyme or the synthetic lysozyme peptide 46–61, TH-cell
activation occurs.
15. Cells that can present antigen to TH cells have been classified
into two groups: professional and nonprofessional APCs.
a. Name the three types of professional APCs. For each
type, indicate whether it expresses class II MHC molecules and a costimulatory signal constitutively or must
be activated before doing so.
b. Give three examples of nonprofessional APCs. When are
these cells most likely to function in antigen presentation?
PA R T I I I
298
|
Adaptive Immunity: Antigen Receptors and MHC
16. Predict whether TH-cell proliferation or CTL-mediated
21. Define the following terms and give examples using the
cytolysis of target cells will occur with the following mixtures of cells. The CD4 TH cells are from lysozymeprimed mice, and the CD8 CTLs are from
influenza-infected mice. Use R to indicate a response and
NR to indicate no response.
human HLA locus: polygeny, polymorphism, and codominant expression. How exactly does each contribute to
ensuring that a diversity of antigens can be presented by
each individual?
k
b.
c.
d.
e.
f.
phages
______ H-2k TH cells lysozyme-pulsed H-2b/k macrophages
______ H-2k TH cells lysozyme-primed H-2d macrophages
______ H-2k CTLs influenza-infected H-2k macrophages
______ H-2k CTLs influenza-infected H-2d macrophages
______ H-2d CTLs influenza-infected H-2d/k macrophages
17. Molecules of the CD1 family were recently shown to pres-
ent nonpeptide antigens.
a. What is a major source of nonpeptide antigens?
b. Why are CD1 molecules not classified as members of
the MHC family even though they associate with
2-microglobulin?
c. What evidence suggests that the CD1 pathway is different
from that utilized by classical class I MHC molecules?
18. A slide of macrophages was stained by immunofluores-
cence using a monoclonal antibody for the TAP1/TAP2
complex. Which of the following intracellular compartments would exhibit positive staining with this antibody?
a.
b.
c.
d.
Cell surface
Endoplasmic reticulum
Golgi apparatus
The cells would not be positive because the TAP1/TAP2
complex is not expressed in macrophages.
19. HLA determinants are used not only for tissue typing of
transplant organs, but also as one set of markers in paternity
testing. Given the following phenotypes, which of the potential fathers is most likely the actual biological father? Indicate
why each could or could not be the biological father.
22. What is the cellular phenotype that results from inactivation
or mutation in both copies of the gene for the invariant chain?
23. In general terms, describe the process of cross-presenta-
tion. Which cell types are most likely to be involved in
cross-presentation, and what unique role does this process
play in the activation of naïve CD8 T cells?
ANALYZE THE DATA Smith and colleagues (2002, Journal of
Immunology 169:3105) examined the ability of two peptides
to bind two different allotypes—Ld and Lq—of the mouse class
I MHC molecule. They looked at (a) the binding of a murine
cytomegalovirus peptide (MCMV; amino acid sequence YPHFMPTNL) and (b) a synthesized peptide, tum P91A 14–22
(TQNHRALDL). Before and after pulsing the cells with the
target peptides, the investigators measured the amount of
peptide-free MHC molecules on the surface of cells expressing either Ld, Lq, or a mutant Ld in which tryptophan had been
mutated to arginine at amino acid 97 (Ld W97R). The ability of
the peptide to decrease the relative number of open forms on
the cell surface reflects increased peptide binding to the class I
MHC molecules. Answer the following questions based on the
data and what you have learned from reading this book.
(a) 80
Relative amount of open forms
on cell surface
k
a. ______ H-2 TH cells lysozyme-pulsed H-2 macro-
(b) 60
70
50
No exogenous
peptide
+ MCMV
peptide
60
50
No exogenous
peptide
+ tum_ P91A 14–22
peptide
40
40
30
0.77
30
20
20
10
0
10
0.04
Ld
0.08
Lq
0.11
LdW97R
0.71
0.02
Ld
Lq
LdW97R
a. Are there allotypic differences in the binding of peptide
HLA-A
HLA-B
HLA-C
Offspring
A3, 43
B54, 59
C5, 8
Biological Mother
A3, 11
B59, 78
C8, 8
Potential Father 1
A3, 33
B54, 27
C5, 5
Potential Father 2
A11, 43
B54, 27
C5, 6
Potential Father 3
A11, 33
B59, 26
C6, 8
20. Assuming the greatest degree of heterozygosity, what is the
maximum number of different classical HLA class I molecules
that one individual can employ to present antigens to CD8
T cells? Assuming no extra beta chain genes, what is the maximum number of different class II molecules (isoforms) that
one individual can use to present antigen to CD4 T cells?
by native class I MHC molecules on the cell surface?
b. Is there a difference in the binding of MCMV peptide to
Ld after a tryptophan (W) to arginine (R) mutation in
the Ld molecule at position 97? Explain your answer.
2
c. Is there a difference in the binding of tum P91A 14–22
d
peptide to L after a tryptophan (W) to arginine (R) mutation in the Ld molecule at position 97? Explain your answer.
d. If you wanted to successfully induce a CD8 T-cell response
2
against tum P91A 14–22 peptide, would you inject a
mouse that expresses Lq or Ld? Explain your answer.
e. The T cells generated against the MCMV peptide have a
different specificity than the T cells generated against
the tum2 P91A 14–22 peptide when Ld is the restricting MHC molecule. Explain how different peptides can
bind the same Ld molecule yet restrict/present peptide
to T cells with different antigen specificities.
9
T-Cell Development
T
cells initiate the adaptive immune response by
interacting, via their T-cell receptors, with
MHC/peptide complexes on antigen-presenting
cells that have been exposed to pathogens. The
range of pathogens to which we are exposed is
considerable, and, as described in Chapter 7, vertebrates
have evolved a remarkable mechanism to generate a
comparable range of T-cell (and B-cell) receptor
specificities. Each of the several million T cells
circulating in the body expresses a distinct T-cell
receptor. The generation of this diverse population, with
its diverse receptor repertoire, takes place in the thymus,
an organ both required for and dedicated to the
development of T cells. Immature cells entering the
thymus from the bone marrow express no mature
lymphocyte features and no antigen receptors; those
leaving the thymus are mature T cells that are diverse in
their receptor specificities, and are both tolerant to self
and restricted to self-MHC.
How do we generate such a diverse self-restricted
and self-tolerant group of T cells? This question has
intrigued immunologists for decades and has inspired
and required much experimental ingenuity. The
innovation has paid off and, although our
understanding is not comprehensive, we now have a
fundamental appreciation of the remarkable strategies
taken by the thymus, the nursery of immature T cells or
thymocytes, to produce a functional, safe, and useful
T-cell repertoire.
In this chapter we describe these strategies and
divide T-cell development into two clusters of events:
early thymocyte development, during which a
dizzyingly diverse TCR⫹ population of immature
T cells is generated, and selection events that depend
on TCR interactions to shape this population so that
only those cells that are self-restricted and selftolerant will leave to populate the periphery
(Overview Figure 9-1). Early thymocyte development
is T-cell receptor independent and brings cells
through uncommitted CD4⫺CD8⫺ (double negative,
DN) stages to the T-cell receptor-expressing,
Thymocytes in the cortex of the thymus.
■
Early Thymocyte Development
■
Positive and Negative Selection
■
Lineage Commitment
■
Exit from the Thymus and Final Maturation
■
Other Mechanisms That Maintain Self-Tolerance
■
Apoptosis
CD4⫹CD8⫹ (double positive, DP) stage. The specific
events in this early stage include:
1.
commitment of hematopoietic precursors to the T cell
lineage,
2.
the initiation of antigen receptor gene rearrangements, and
3.
the selection and expansion of cells that have successfully rearranged one of their T-cell receptor genes
(-selection).
The second phase of T-cell development is largely
dependent on T-cell receptor interactions and brings cells
to maturity from the CD4⫹CD8⫹ stage to the CD4⫹ or
CD8⫹ single positive (SP) stage. The events in this last
phase of development include:
1.
positive selection, selection for those cells whose T-cell
receptors respond to self-MHC,
2.
negative selection, selection against those cells whose
T-cell receptors react strongly to self-peptide/MHC
combinations, and
3.
lineage commitment, commitment of thymocytes to
effector cell lineages, including CD4⫹ helper or CD8⫹
cytotoxic populations.
299
9-1
OVERVIEW FIGURE
Development of T cells in the mouse
Hematopoietic
stem cell (HSC)
Marrow
Hematopoietic
precursor
Blood
migration
T-cell
precursor
Dendritic cell,
B cell, NK
DN1
TCR locus rearrangement,
proliferation
DN2
Thymic
cortex
Early
thymocyte
development
CD4⫺CD8⫺ (DN)
T-cell commitment
DN3
-selection, proliferation
DN4
TCR ␣ locus rearrangement
Double positive
CD4⫹CD8⫹ (DP)
Death by neglect
Positive and
negative
selection
CD8⫹
CD4⫹
Single positive
(SP)
Thymic
medulla
Blood
Negative
selection
CD8⫹
␥␦ T cell; few
further changes
in surface
phenotype
Other lineages,
including
TREGs, TH17,
IEL, NK1.1
CD4⫹
migration
␥␦ T cell
Peripheral
tissues
CD8⫹
T-cell precursors from the bone marrow travel to the thymus via
the bloodstream, undergo development to mature T cells, and are
exported to the periphery, where they can undergo antigeninduced activation and differentiation into effector cells and
memory cells. Each stage of development occurs in a specific
microenvironment and is characterized by specific intracellular
events and distinctive cell-surface markers. The most immature
thymocytes are CD4⫺CD8⫺ (double negative, DN) and pass
through several stages (DN1-DN4) during which they commit to
the T-cell lineage and begin to rearrange their T-cell receptor (TCR)
gene loci. Those that successfully rearrange their TCR chain proliferate, initiate rearrangement of their TCR␣ chains, and become
CD4⫹CD8⫹ (double positive, DP) thymocytes, which dominate the
300
CD4⫹
thymus. DP thymocytes undergo negative and positive selection
in the thymic cortex (see text and Overview Figure 9-5 for details).
Positively selected thymocytes continue to mature and migrate to
the medulla, where they are subject to another round of negative
selection to self-antigens that include tissue-specific proteins.
Mature T cells express either CD4 or CD8 (single positive, SP) and
leave the thymus with the potential to initiate an immune
response. Although most thymocytes develop into conventional
TCR␣ CD4⫹ or CD8⫹ T cells, some DN and DP thymocyte cells
develop into other cell lineages, including lymphoid dendritic
cells, TCR␥␦ T cells, natural killer T cells (NKT), regulatory T cells
(TREG), and intraepithelial lymphocytes (IELs), each of which has a
distinct function (see text).
T-Cell Development
Understanding T-cell development has been a
challenge for students and immunologists alike, but it
helps to keep in mind the central purpose of the process:
to generate a large population of T cells that express a
diverse array of receptor specificities that can interact with
self-MHC, but do not respond to self proteins.
|
CHAPTER 9
301
Bone marrow stem cells
Early Thymocyte Development
T-cell development occurs in the thymus and begins with
the arrival of small numbers of lymphoid precursors
migrating from the bone marrow and blood into the thymus, where they proliferate, differentiate, and undergo
selection processes that result in the development of
mature T cells.
The precise identity of the bone marrow cell precursor
that gives rise to T cells is still debated. However, it is clear
that this precursor, which is directed to the thymus via chemokine receptors, retains the potential to give rise to more
than one type of cell, including natural killer (NK) cells,
dendritic cells (DC), B cells, and even myeloid cells. This
precursor only becomes fully committed to the T-cell lineage
in the late DN2 stage of T-cell development (see Overview
Figure 9-1).
Studies revealed that commitment to the T-cell lineage
was dependent on a receptor, Notch, which had been classically associated with embryonic cell development. Notch, in
fact, regulates the decision of a lymphoid precursor to
become a T versus a B lymphocyte. When a constitutively
active version of Notch1, one of four versions of Notch, is
overexpressed in hematopoietic cells, T cells rather than B
cells develop in the bone marrow. Reciprocally, when the
Notch1 gene is knocked out among hematopoietic precursors, B cells rather than T cells develop in the thymus (!).
The importance of Notch in T-cell commitment was
underscored by an in vitro system for studying T-cell
development. Until recently, we thought that development
of T cells from their hematopoietic precursors required the
intact architecture of the thymus. Whereas B-cell development could be achieved in vitro using a preparation of bone
marrow stem cells grown on stromal cells in the presence of
appropriate cytokines, in vitro studies of T-cell development
once could only be performed using intact fetal thymic
organ culture systems. In 2002, J. C. Zuniga-Pfluker and colleagues demonstrated that T cells could be induced to
develop when bone marrow stem cells were cultured on a
stromal cell line that expressed a ligand for Notch. As shown
in Figure 9-2, growth of hematopoietic stem cells (HSCs) on
stromal cells that express Notch ligand drives the development of these multipotent stem cells to the T-cell lineage.
This assay system has been invaluable in defining interactions that control early T-cell development and have helped
investigators reveal, for instance, that the transcription factor GATA-3 is a critical participant in Notch-mediated T-cell
commitment.
Stromal cells
Notch1⫹
Notch1⫺
Growth factors
IL-7, FLT3L
1-2 weeks
Harvest T-lineage cells
Harvest B-lineage cells
FIGURE 9-2 Development of T cells from hematopoietic
stem cells on bone marrow stromal cells expressing the
Notch ligand. Investigators can induce lymphoid development
from hematopoietic stem cells in vitro using a combination of stromal cell lines and soluble cytokines and growth factors, as indicated.
Investigators discovered that Notch signaling was the key to inducing development to the T- rather than B-lymphocyte lineage. After
transfecting the stromal cell line with a gene encoding the Notch
ligand, lymphoid precursors would adopt the T-cell lineage. Otherwise, they would develop into B cells. [Adapted from J. C. Zuniga-Pflucker,
2002, Nature Reviews Immunology 4:67–72.]
Thymocytes Progress through Four DoubleNegative Stages
T-cell development is elegantly organized, spatially and temporally. Different stages of development take place in distinct
microenvironments that provide membrane-bound and soluble signals that regulate maturation. After arriving in the thymus from the bone marrow via blood vessels at the
cortico-medullary boundary, T-cell precursors encounter
Notch ligands, which are abundantly expressed by the thymic
epithelium. Recall that T-cell precursors first travel to the
outer cortex where they slowly proliferate, then they pass
through the thymic medulla before exiting at the corticomedullary junction (see Figure 2-6). During the time it takes
cells to develop in the thymus (1 to 3 weeks), thymocytes pass
through a series of stages defined by changes in their cell surface phenotype (see Overview Figure 9-1). The earliest T cells
lack detectable CD4 and CD8 and are therefore referred to as
double-negative (DN) cells. DN T cells can be subdivided
302
TABLE 9-1
PA R T I V
|
Adaptive Immunity: Development
Double-negative thymocyte development
Genotype
DN1
c-kit (CD117)⫹⫹, CD44⫹, CD25⫺
⫹⫹
⫹
⫹
Location
Description
Bone marrow to thymus
Migration to thymus
TCR␥, ␦, and  chain rearrangement;
T-cell lineage commitment
DN2
c-kit (CD117)
, CD44 , CD25
Subcapsular cortex
DN3
c-kit (CD117)⫹, CD44⫺, CD25⫹
Subcapsular cortex
Expression of pre-TCR;  selection
DN4
c-kit (CD117)low/⫺, CD44⫺, CD25⫺
Subcapsular cortex to cortex
Proliferation, allelic exclusion of -chain
locus; ␣-chain locus rearrangement
begins; becomes DP thymocyte
into four subsets (DN1-4) based on the presence or absence of
other cell surface molecules, including c-kit (CD117), the
receptor for stem cell growth factor; CD44, an adhesion molecule; and CD25, the ␣ chain of the IL-2 receptor (Table 9-1).
DN1 thymocytes are the first to enter the thymus and
are still capable of giving rise to multiple cell types. They
express only c-kit and CD44 (c-kit⫹⫹CD44⫹CD25⫺), but
once they encounter the thymic environment and become
resident in the cortex, they proliferate and express CD25,
becoming DN2 thymocytes (c-kit⫹⫹CD44 ⫹CD25 ⫹).
During this critical stage of development, the genes for
the TCR ␥, ␦, and  chains begin to rearrange; however, the
TCR ␣ locus does not rearrange, presumably because the
region of DNA encoding TCR␣ genes is not yet accessible to
the recombinase machinery. At the late DN2 stage, T-cell
precursors fully commit to the T-cell lineage and reduce
expression of both c-kit and CD44. Cells in transition
from the DN2 to DN3 (c-kit⫹CD44⫺CD25⫹) stages continue rearrangement of the TCR␥, TCR␦, and TCR
chains and make the first major decision in T-cell development: whether to join the TCR␥␦ or TCR␣ T-cell lineage. Those DN3 T cells that successfully rearrange their
 chain and therefore commit to the TCR␣ T-cell lineage
lose expression of CD25, halt proliferation, and enter the
final phase of their DN stage of development, DN4
(c-kit low/⫺CD44⫺CD25⫺), which mature directly into
CD4⫹CD8⫹ DP thymocytes.
Thymocytes Can Express Either TCR␣
or TCR␥␦ Receptors
Vertebrates generate two broad categories of T cells: those
that express TCR␣ and  receptor chains and those that
express TCR␥ and ␦ receptor chains. TCR␣ cells are the
dominant participants in the adaptive immune response in
secondary lymphoid organs; however, TCR␥␦ cells also play
an important role, particularly in protecting our mucosal
tissues from outside infection. Both types of cells are generated in the thymus, but how does a cell make the decision to
become one or the other? To a large extent, the choice to
become a ␥␦ or ␣ T cell is dictated by when and how fast
the genes that code for each of the four receptor chains successfully rearrange.
Recall from Chapter 7 that TCR genes are generated by
the shuffling (rearrangement) of V and J (and sometimes D)
segments, an event responsible for the vast diversity of
receptor specificities. Rearrangement of the , ␥, and ␦ loci
begins during the DN2 stage. To become an ␣ T cell, a cell
must generate a TCR chain—an event that depends on a
single in-frame VDJ rearrangement event. To become a ␥␦
cell, however, a thymocyte must generate two functional
proteins that depend on two separate in-frame rearrangement events. Probability favors the former fate and, in fact,
T cells are at least three times as likely to become TCR␣
cells than TCR␥␦ cells.
TCR␥␦ T-cell generation is also regulated developmentally. They are the first T cells that arise during fetal development, and provide a very important protective function
perhaps even prior to birth. Studies show, for instance,
that ␥␦ T cells are required to protect very young mice
against the protozoal pathogen that causes coccidiosis.
However, production of ␥␦ T cells declines after birth, and
the TCR␥␦ T-cell population represents only 0.5% of all
mature thymocytes in the periphery of an adult animal
(Figure 9-3).
Most TCR␥␦ T cells are quite distinct in phenotype and
function from conventional TCR␣ T cells. Most do not
go through the DP stage of thymocyte development and
leave the thymus as mature DN T cells. Many emerge from
the thymus with the ability to secrete cytokines, a capacity
gained by most TCR␣ cells only after they encounter
antigen in secondary lymphoid tissues. The TCR␥␦ T-cell
population also expresses receptors that are not as diverse
as TCR␣ T cells, and many appear to recognize unconventional antigens, including lipids associated with unconventional MHC molecules. Many take up long-term
residence in mucosal tissues and skin and join innate
immune cells in providing a first line of attack against
invading microbes, as well as the response to cellular
stress.
T-Cell Development
100
γδ Thymocytes
CD3+ cells, %
75
αβ Thymocytes
2.
3.
25
4.
0
14
15
16
17 18 19 Birth
Days of gestation
Adult
FIGURE 9-3 Time course of appearance of ␥␦ thymocytes
and ␣ thymocytes during mouse fetal development. The
graph shows the percentage of cells in the thymus that are double
negative (CD4⫺CD8⫺) and bear the ␥␦ T-cell receptor (black) or are
double positive (CD4⫹CD8⫹) and bear the ␣ T-cell receptor (blue).
Fetal animals generate more ␥␦ T cells than ␣ T cells, but the proportion of ␥␦ T cells generated drops off dramatically after birth. This
early dominance of TCR␥␦ cells may have adaptive value: a large portion of these cells express nondiverse TCR specificities for common
pathogen proteins and can mount a quick defense before the more
traditional adaptive immune system has fully developed.
For the rest of this chapter, we will focus on the development of TCR␣ T cells, which make up the vast majority of
T lymphocytes in the body and are continually being generated, even after puberty, when the thymus shrinks considerably in size (involutes).
DN Thymocytes Undergo -Selection, Which
Results in Proliferation and Differentiation
Double-negative (DN) thymocytes that have successfully
rearranged their TCR chains are valuable, and are identified and expanded via a process known as -selection (see
Overview Figure 9-1). This process involves a protein that is
uniquely expressed at this stage of development, a 33-kDa
invariant glycoprotein known as the pre–T␣ chain. Pre-T␣
acts as a surrogate for the real TCR␣ chain, which has yet to
rearrange, and assembles with a successfully rearranged and
translated  chain, as well as CD3 complex proteins. This
precursor TCR/CD3 complex is known as the pre-TCR
(Figure 9-4a) and acts as a sensor by initiating a signal transduction pathway. The signaling that the pre-TCR complex
initiates is dependent on many of the same T-cell specific
kinases used by a mature TCR (see Chapter 3), but does not
appear to be dependent on ligand binding. In fact, little if
any of the complex is expressed on the cell surface; rather,
successful assembly of the complex may be sufficient to acti-
CHAPTER 9
303
vate the signaling events. As we will learn in Chapter 10, B
cells also express an immature receptor complex (the preBCR) that is coded by distinct genes, but is fully analogous
to the pre-TCR.
Pre-TCR signaling results in the following cascade of
events:
1.
50
|
5.
6.
Maturation to the DN4 stage (c-kitlow/⫺CD44⫺CD25⫺)
Rapid proliferation in the subcapsular cortex
Suppression of further rearrangement of TCR -chain
genes, resulting in allelic exclusion of the -chain
locus
Development to the CD4⫹CD8⫹ double-positive (DP)
stage
Cessation of proliferation
Initiation of TCR␣ chain rearrangement
It is important to note that the proliferative phase prior to
␣ chain rearrangement enhances receptor diversity considerably by generating clones of cells with the same TCR
-chain rearrangement. Each of the cells within a clone can
then rearrange a different ␣-chain gene, thereby generating
an even more diverse population than if the original cell had
undergone rearrangement at both the - and ␣-chain loci
prior to proliferation. TCR ␣-chain gene rearrangement
does not begin until double-positive thymocytes stop proliferating.
Most T cells fully rearrange and express a TCR chain
from only one of their two TCR alleles, a phenomenon
known as allelic exclusion. Allelic exclusion is the result of
inhibition of further rearrangement at the other TCR
allele (which must be fully rearranged to be expressed).
This can be accomplished by reducing RAG expression so
no more rearrangement can occur, as well as by making the
locus inaccessible to further RAG interaction via more
permanent changes in chromatin packaging. The details of
the mechanisms responsible for this shutdown are still
being investigated. However, negative feedback signals
from a successfully assembled pre-TCR/pre-T␣ complex
during -selection clearly have a significant influence.
Other events, including the proliferative burst that follows
-selection, which dilutes RAG protein levels, can also play
a role.
RAG levels continue to change after -selection. They are
restored after the proliferative burst and allow TCR␣ rearrangement to occur. They decrease once again after expression of a successfully assembled TCR␣ dimer.*
*
The mechanisms that turn off rearrangement at this (DP) stage of
development are not as efficient as those that turn off additional TCR chain
rearrangement. In fact, both TCR␣ alleles successfully rearrange more frequently than we originally supposed. This means that many T cells actually
express two TCR specificities: an interesting reality check that complicates a
fundamentalist interpretation of the clonal selection hypothesis.
304
PA R T I V
|
Adaptive Immunity: Development
(a) Pre-TCR (DN3, DN4, DP)
(b) Mature αβTCR, immature signaling pathways (DP)
TCR β
Pre-Tα
δ ε
TCR α
γ ε
ζ ζ
δ ε
Signals
γ ε
ζ ζ
Signals
Cell becomes
permissive for
TCR α-chain locus
rearrangement
Stops additional
TCR β-chain locus
rearrangements
(allelic exclusion)
Stimulates
expression
of CD4 and
CD8 coreceptors
TCR β
Stimulates
proliferation
Positive selection
(intermediate affinity)
and differentiation
Negative selection
(high-affinity and
costimulatory signals)
FIGURE 9-4 Changes in the structure and activity of the
(c) Mature αβTCR, mature signaling pathways (SP)
TCR α
δ ε
TCR β
γ ε
ζ ζ
Signals
Proliferation
Effector function
(cytokine secretion
and/or cytotoxicity)
Survival
Once a young double-positive (DP) thymocyte successfully rearranges and expresses a TCR␣ chain, this chain will
associate with the already produced TCR chain, taking the
place of the surrogate pre-TCR␣ chain, which is no longer
actively expressed (Figure 9-4b). At this point, several
“goals” have been accomplished by the thymus: hematopoietic cell precursors have expanded in the subcapsular cortex,
committed to the T-cell lineage, and rearranged a set of TCR
genes. They have also “chosen” to become a TCR␣ or
TCR␥␦ T cell. This TCR␣ population now expresses both
CD4 and CD8, and is ready for the second stage of T-cell
development: selection.
T-cell receptor through T-cell development. (a) The pre-TCR.
The pre-TCR is assembled during the DN stage of development when
a successfully rearranged TCR chain dimerizes with the nonvariant
pre-T␣ chain. Like the mature ␣TCR dimer, the pre-TCR is noncovalently associated with the CD3 complex. Assembly of this complex
results in intracellular signals that induce a variety of processes, including the maturation to the DP stage. (b) Mature ␣TCR expressed at the
DP stage of development. Once a DP thymocyte has successfully rearranged a TCR␣ chain, it will dimerize with the TCR, replacing the preT␣ chain. This mature ␣TCR generates the signals that lead to either
positive or negative selection (differentiation or death, respectively),
depending on the affinity of the interaction. Note that the TCR␣ chain
has a shorter intracellular region than the pre-T␣ chain and cannot
generate intracellular signals independently. (c) Mature ␣TCR
expressed at the SP stage. Although the ␣TCR/CD3 complex
expressed by mature SP T cells is structurally the same as that expressed
by DP thymocytes, the signals it generates are distinct. It responds to
high-affinity engagement not by dying, but by initiating cell proliferation, activation, and the expression of effector functions. Low-affinity
signals generate survival signals. The basis for the differences in signals
generated by DP and SP TCR complexes is still unknown.
Positive and Negative Selection
CD4⫹CD8⫹ (DP) thymocytes, small, nonproliferating cells
that reside in the thymic cortex, are the most abundant subpopulation in the thymus, comprising more than 80% of
cells. Most important, they are the first subpopulation of
thymocytes that express a fully mature surface TCR␣/CD3
complex and are therefore the primary targets of thymic
selection (see Figure 9-1). Thymic selection shapes the TCR
repertoire of DP thymocytes based on the affinity of their
T-cell receptors for the MHC/peptides they encounter as
they browse the thymic cortex.
T-Cell Development
Why is thymic selection necessary? The most distinctive property of the mature T cells is that they recognize
only foreign antigen combined with self-MHC molecules.
However, randomly generated TCRs will certainly have
no inherent affinity for “foreign antigen plus self-MHC
molecules.” They could just as well recognize foreign
MHC/peptide combinations, which would not be useful,
or self-MHC/self-peptide combinations, which could be
dangerous.
Two distinct selection processes are required:
• Positive selection, which selects for those thymocytes
bearing receptors capable of binding self-MHC molecules, resulting in MHC restriction
• Negative selection, which selects against thymocytes
bearing high-affinity receptors for self-MHC/peptide
complexes, resulting in self-tolerance.
Because only self-peptides are presented in the thymus,
and only in association with self-MHC molecules, these two
selection processes ensure that surviving thymocytes
express TCRs that have low affinity for self-peptides in selfMHC. On the other hand, the processes do not guarantee
that the T cells generated will bear receptors with high affinity for any specific self-MHC/foreign peptide combination.
The ability of the immune system to respond to a foreign
antigen depends on the probability that one of the millions
of T cells that survives selection will bind one of the many
MHC/peptide combinations expressed by an antigenpresenting cell that has processed pathogenic proteins outside the thymus.
The vast majority of DP thymocytes (~98%) never
meet the selection criteria and die by apoptosis within the
thymus (Overview Figure 9-5). The bulk of DP thymocyte
death (~95%) occurs among thymocytes that fail positive
selection because their receptors do not specifically recognize self-MHC molecules. These cells do not receive
survival signals through their TCRs, and die by a process
known as death by neglect. A small percentage of cells
(2%–5%) are eliminated by negative selection. Only 2% to
5% of DP thymocytes actually exit the thymus as mature
T cells.
Below, we briefly describe the experimental evidence
for thymic involvement in MHC restriction. We also
describe what is currently known about the positive and
negative selection processes, as well as the models that
have been advanced to explain what is known as “the thymic paradox.”
Thymocytes “Learn” MHC Restriction
in the Thymus
Early evidence for the role of the thymus in selection of the
T-cell repertoire came from chimeric mouse experiments
performed by R. M. Zinkernagel and his colleagues. Recall
that Zinkernagel won the Nobel Prize for showing that
mature T cells are MHC restricted (i.e., they need to recog-
|
CHAPTER 9
305
nize antigen in the context of the MHC of their host to
respond; see Classic Experiment in Chapter 8). However, he
and his colleagues were curious to know how T cells became
so “restricted.” They considered two possibilities: Either the
T cell and APC simply had to have matching MHC types
(i.e., an “A” strain T cell had to see an “A” strain antigenpresenting cell) or T cells, regardless of their own MHC type,
“learned” the MHC type of their host sometime during
development. Zinkernagel and colleagues thought that such
learning could take place in the thymus, the T-cell nursery.
To determine if T cells could be “taught” to recognize the
host MHC, they removed the thymus (thymectomized) and
irradiated (A B) F1 mice so they had no functional
immune system (Figure 9-6). They then reconstituted the
hematopoietic cells with an intravenous infusion of F1 bone
marrow cells, but replaced the thymus with one from a
B-type mouse. (To be certain that the thymus graft did not
contain any mature T cells, they irradiated it before transplantation.)
In this experimental system, T-cell progenitors from the
(A B) F1 bone marrow would mature within a thymus that
expresses only B-haplotype MHC molecules on its stromal
cells. Would these (A B) F1 T cells now be MHC restricted
to the B haplotype of the thymus in which they developed?
Or, because they expressed both A and B MHC, would they
be able to recognize both A and B MHC haplotypes? To
answer this question, the investigators infected the chimeric
mice with lymphocytic choriomeningitis virus (LCMV, the
antigen) and removed the immunized, mature splenic T cells
to see which LCMV-infected target cells (APCs) they could
kill. They tested them against infected APCs from strain A,
strain B, and strain A B mice. As shown in Figure 9-6,
T cells from the chimeric mice could only lyse LCMVinfected target cells from strain B mice. Thus, the MHC
haplotype of the thymus in which T cells develop determines
their MHC restriction. T cells “learned” which MHC haplotype they are restricted to during their early days in the
thymus. Although once we referred to this process as “T-cell
education,” we now know that it is the consequence of a
brutal selection process.
T Cells Undergo Positive and
Negative Selection
In the thymus, thymocytes come into contact with thymic
epithelial cells that express high levels of class I and class II
MHC molecules on their surface. These self-MHC molecules present self-peptides, which are typically derived from
intracellular or extracellular proteins that are degraded in
the normal course of cellular metabolism. DP thymocytes
undergo positive and negative selection, depending on the
signals they receive when they encounter self-MHC/selfpeptide combinations with their TCRs.
While working to understand negative and positive
selection, it helps to picture an early model that was
advanced by several seminal immunologists (see Overview
9-5
OVERVIEW FIGURE
Positive and Negative Selection of Thymocytes in the Thymus
Cortex
CD8
CD3
T-cell receptor
DP thymocyte
2%–5%
high-affinity
interaction
Negative
selection
cTEC
CD4
90–96%
no interaction
2%–5%
low/int-affinity
interaction
Death by
neglect
Positive
selection
Class I and/or
Class II MHC
molecules
CD4+
TH cell
CD8+
TC cell
Medulla
Negative
selection
AIRE+
CD4+
CD8+
mTEC
TH cell
TC cell
Dendritic cell
Mature CD4+ or CD8+ T lymphocytes
emigrate to periphery
Thymic selection involves multiple interactions of DP and SP thymocytes with both the cortical and medullary thymic stromal cells, as
well as dendritic cells and macrophages. Selection results in a mature
T-cell population that is both self-MHC restricted and self-tolerant. DP
thymocytes that express new TCR␣ dimers browse the MHC/
peptide complexes expressed by the cortical thymic epithelial cells
(cTECs). The large majority of DP thymocytes die in the cortex by
neglect because of their failure to bind MHC/peptide combinations
with sufficient affinity. The small percentage whose TCRs bind MHC/
306
peptide with high affinity die by clonal deletion (negative selection).
Those DP thymocytes whose receptors bind to MHC/peptide with
intermediate affinity are positively selected and mature to singlepositive (CD4⫹ or CD8⫹) T lymphocytes. These migrate to the
medulla, where they are exposed to AIRE⫹ medullary thymic epithelial cells (mTECs), which express tissue-specific antigens and can
mediate negative selection. Medullary dendritic cells can acquire
mTEC antigens by engulfing mTECs, and mediate negative selection
(particularly of MHC Class II restricted CD4⫹ thymocytes).
T-Cell Development
CONTROL
Infect with LCMV
(A × B)F1
Spleen cells
LCMV-infected
strain A cells
Killing
LCMV-infected
strain B cells
Killing
EXPERIMENT
(A × B) F1 (H–2a/b)
1 Thymectomy
2 Lethal x-irradiation
Strain B thymus graft (H–2b)
(A × B) F1 hematopoietic stem
cells (H–2a/b)
Infect with LCMV
Spleen cells
LCMV-infected
strain A cells
No killing
LCMV-infected
strain B cells
Killing
FIGURE 9-6 Experimental demonstration that the thymus selects for maturation only those T cells whose T-cell
receptors recognize antigen presented on target cells with
the haplotype of the thymus. Control (A ⫻ B) F1 mice infected
with LCMV produce mature T cells that are able to kill both
infected-strain A cells and infected-strain B cells, demonstrating
that these T cells are restricted to MHC molecules expressed by
both strain A (H-2a) and strain B (H-2b). In order to determine the
involvement of the thymus in the restriction specificity of T cells,
investigators grafted thymectomized and lethally irradiated (A ⫻ B)
F1 (H-2a/b) mice with a strain-B (H-2b) thymus and reconstituted it
with (A ⫻ B) F1 bone marrow cells. After infection with LCMV, the
CD8⫹ cytotoxic (CTL) cells from this reconstituted mouse were
assayed for their ability to kill 51Cr-labeled strain A or strain B target
cells infected with LCMV. Only strain B target cells were lysed, suggesting that the H-2b grafted thymus had selected for maturation
only those T cells that could recognize antigen combined with
H-2b MHC molecules.
|
CHAPTER 9
307
Figure 9-5). Although we now know that some aspects of
this model are too simplistic, its core principles remain relevant and provide a very useful framework for understanding thymic selection. Briefly, DP thymocytes can be
considered to have one of three fates, depending on the
affinity of their new T-cell receptors for the MHC/selfpeptide combinations that they encounter: If their newly
generated TCRs do not bind to any of the MHC/self-peptides
they encounter on stromal cells, they will die by neglect. If
they bind too strongly to MHC/self-peptide complexes they
encounter, they will be negatively selected. If they bind with
a low, “just right” affinity to MHC/self-peptide complexes,
they will be positively selected and mature to the singlepositive stage of development. (We will also see that cells
that are positively selected in the cortex undergo a second,
very important round of negative selection in the medulla,
following lineage commitment to CD4⫹ or CD8⫹ cells.) We
review the origins of the model and discuss some of the
modifications that have been suggested by recent data
below. See Classic Experiment Box 9-1 for experimental
evidence in support of this model from the earliest TCR
transgenic mouse.
Thymic stromal cells, including epithelial cells, macrophages, and dendritic cells, play essential roles in positive
and negative selection and should be part of our visualization of thymic selection events. Young DP thymocytes are in
intimate contact with these cells and “browse” the MHC/
self-peptides displayed on their surfaces. Each of these cell
types has the capacity to express high levels of class I and
class II MHC proteins, and typically have extended processes that contact many developing thymocytes. Some also
express costimulatory ligands, including CD80 (B7-1) and
CD86 (B7-2). As the thymocytes migrate through the thymus, they encounter multiple different stromal cell surfaces,
and have the opportunity to bind many different MHC/
peptide combinations.
Positive Selection Ensures MHC Restriction
If a CD4⫹CD8⫹ thymocyte recognizes a self-MHC/peptide
complex on the cortical epithelial cells that they browse, it
will undergo positive selection, a process that induces both
survival and differentiation of DP thymocytes. Remarkably,
the majority of newly generated thymocytes do not successfully engage the MHC/peptides they encounter with their
TCRs. Either they have not generated a functional TCR␣
combination or the combination does not bind MHC/peptide
complexes with sufficient affinity. These cells have “failed”
the positive selection test and die by apoptosis within 3 to
4 days. TCR/MHC/self-peptide interactions that initiate
positive selection are thought to be at least three times lower
in intensity than interactions that initiate negative selecting
signals (Figure 9-7).
The signal cascade generated by positive selecting TCR
interactions has not yet been fully characterized. We do
know that many of the signaling molecules required for
CLASSIC EXPERIMENT
Insights about Thymic Selection from the First TCR Transgenic
Mouse Have Stood the Test of Time
In the late 1980s
Harald von
Boehmer and colleagues published a series
of seminal experiments that provided direct
and compelling evidence for the influence
of T-cell receptor interactions on positive
and negative selection. They recognized
that in order to understand how positive
and negative selection worked, they would
need to be able to trace and compare the
fate of thymocytes with defined T-cell
receptor specificities. But how could one do
that if every one of the hundreds of millions
of normal thymocytes expresses a different
receptor? The researchers decided to
develop a system in which all thymocytes
expressed a T-cell receptor of known specificity, and thereby generated one of the first
TCR transgenic mice.
Transgenic animals are made by injecting a gene under the control of a defined
promoter into a fertilized egg (zygote)
(see Chapter 20). The gene—in fact, often
many copies of the gene—will be incorporated randomly into the genome of the
zygote. Therefore, the gene will be present in all cells of the mouse that develop
from that zygote; however, only those
cells that can activate the promoter will
express the gene.
The team of von Boehmer et al. developed the first TCR transgenic animals by
isolating fully rearranged TCR␣ and TCR
genes of a mature CD8⫹ T-cell line that
was known to be specific for an antigen
expressed only in male mice (the “H-Y”
antigen) in the context of H-2Db Class I
MHC molecules (Figure 1). They generated a genetic construct that included
both rearranged genes under the control
of a T-cell specific promoter and injected
this into a mouse zygote. Since the receptor transgenes were already rearranged,
other TCR gene rearrangements were
suppressed in the transgenic mice (the
phenomenon of allelic exclusion; see text
and Chapter 7); therefore, a very high percentage of the thymocytes in the transgenic mice expressed the TCR␣
combination specific for the H-Y/H-2Db.
(It turns out that not all thymocytes
expressed this TCR combination. Why not?
308
Recall that allelic exclusion does not operate well for the TCR␣ locus in particular.)
The team’s choice of specificities was
very clever. Because all thymocytes
expressed an anti-male antigen TCR specificity, the researchers could directly compare the phenotype of autoreactive
thymocytes and non-autoreactive thymocytes simply by comparing male and
female mice in the same litter. Because the
MHC restriction of the TCR was known, they
could also observe the influence of MHC
haplotype on thymocyte development
simply by breeding the mice to other
inbred strains.
The results of a comparison of CD4⫹
versus CD8⫹ phenotype among thymocytes in H-2Db male versus female mice is
shown in Figure 1. The primary data from
one of the first publications are shown in
CTL
H-Y specific
H-2Db restricted
×
H-Y
peptide
Clone TCR
α and β genes
Male cell (H-2Db )
α
Female cell (H-2Db )
β
Use to make α H-Y TCR transgenic mice
Male H-2Db
Female H-2Db
Female H-2Dd
+
−
−
CD4− 8−
++
+
+
CD4+ 8+
+
++
++
CD4+
+
+
−
CD8+
−
++
−
H-Y expression
Thymocytes
FIGURE 1
Experimental demonstration that negative selection of thymocytes requires both
self-antigen and self-MHC, and positive selection requires self-MHC. In this experiment, H-2Db male and female transgenics were generated by injecting TCR transgenes specific for H-Y
antigen plus the Db MHC molecule into zygotes. The H-Y antigen is expressed only in males. H-2Dd mice
were also generated by backcrossing these transgenics onto an H-2d strain (e.g., BALB/c). FACS analysis
of thymocytes from the transgenics showed that mature CD8⫹ T cells expressing the transgene were
absent in the male H-2Db mice but present in the female H-2Db mice, suggesting that thymocytes that
bind with high affinity to a self antigen (in this case, H-Y antigen in the male mice) are deleted during
thymic development. Results also show that DP but not CD8⫹ thymocytes develop in female H-2Dd
mice, which do not express the proper self-MHC. These findings indicate that positive selection and
maturation require self-MHC interactions. [Adapted from H. von Boehmer and P. Kisielow, 1990, Science 248:1369.]
BOX 9-1
Figure 2 and depict flow cytometric data
of CD4 and CD8 expression of thymocytes.
The flow cytometric profile (see Chapter 20)
of a typical wild-type thymus is shown in
Figure 2a and can be used for comparison.
In this profile, the CD8-expression status of
a cell is shown on the X-axis, the CD4expression status on the Y-axis. The data
are quantified in the upper right corner of
the profile. As you can see, more than 80%
of normal thymocytes are DP, 10% to 15%
are CD4 single positive, 3% to 5% are CD8
single-positive cells, and only a small percentage of cells are DN. How do the CD4
versus CD8 phenotypes of male and
female H-Y TCR transgenic mice (Figure 2b)
differ, and what does this tell you?
Thymocytes in female mice are in all
stages of development: DN, DP, and SP. In
fact, they seem to have an abundance of
mature CD8⫹ SP thymocytes. We’ll come
back to this below. The male mice, however,
have few if any anti-H-Y TCR transgenic DP
thymocytes. These data show that in an
environment where DP thymocytes are
exposed to the antigen for which they are
specific (in this case the male H-Y antigen),
they are eliminated. This result was fully
consistent with the concept of negative
selection and showed that self-reactive
cells were removed from the developing
repertoire, perhaps at the DP stage. However, the mice offered insights into many
more aspects of thymic selection.
Evidence for the role of TCR interactions
in positive selection, for instance, came
(a) Wild-type thymus
108
cells
(b)
from a different experiment in which the
investigators asked what would happen if
the “correct” restricting element, H-2Db,
was not present in the mouse. To do this
they performed backcrosses of their TCR
transgenic mice to mice that expressed a
different H-2 haplotype, an H-2Dd mouse.
(For those who want to brush up on their
genetics, determine what crosses you
would make to do this, and include a plan
for assessing the genotype of your mice.)
Figure 2c shows the CD4 versus CD8 phenotype of H-Y TCR transgenic thymocytes
from female H-2Db/d mice and H-2Dd mice.
What did they find?
H-2Dd females had no mature T cells,
indicating that in the absence of the MHC
haplotype to which the TCR was restricted,
thymocytes could not mature beyond the
DP stage. These data provided direct evidence for the necessity of a TCR/self-MHC
interaction for thymocyte maturation
(positive selection) to occur.
These “simple” experiments also revealed
a third feature of thymic development and
initiated a controversy that continues to this
day. Return to Figure 2b and take a look at
the female mouse data. Unlike wild-type
thymocytes, which typically include 3%- to
-5% CD8⫹ T cells, 46% of developing thymocytes in the anti-H-Y TCR transgenic were
mature CD8⫹ cells. What did this mean?
Recall that the T-cell line that the TCR genes
came from originally was a CD8⫹ T-cell line,
restricted to Class I MHC (H-2Db). These data
showed that the MHC restriction specificity
Male H-2Db
Female H-2Db
(c)
of the TCR (in this case, a restriction for
Class I) influenced its CD4 versus CD8 lineage choice. In other words, the results suggested that thymocytes with new TCRs that
preferentially bound Class II would become
CD4⫹ SP mature T cells, and thymocytes
with TCRs that preferentially bound Class I
would develop into CD8⫹ SP mature T cells.
The fundamental conclusions of this
now classic set of experiments have stood
the test of time and have been supported
by many different, equally clever experiments: (1) Negative selection of autoreactive
cells results in their elimination at the DP
stage and beyond. (It is important to note
that it later became clear that the promoter
driving these first H-Y TCR transgenes permitted earlier-than-normal expression of
the TCR, which resulted in elimination of
autoreactive cells at an earlier-than-usual
point in development. Experiments now
indicate that negative selection by clonal
deletion typically occurs after positive
selection as cells transit between the DP
and SP stages of development.) (2) Positive
selection involves a low affinity interaction
between thymocyte TCR and self-MHC/
peptide interactions. (3) Commitment to the
CD4 versus CD8 lineage is determined by the
preference of the TCR for MHC Class I versus
Class II. How this last event happens remains
a topic of intense controversy (see text).
von Boehmer, H., H. Teh, and P. Kisielow. 1989.
The thymus selects the useful, neglects the
useless and destroys the harmful. Immunology
Today 10:57–61.
Female H-2Db/d
Female H-2Dd
11 83
3 3
⫹
CD4
⫺
⫺
CD8
⫹
FIGURE 2
Primary data from experiments summarized in Figure 1. The relative staining by an anti-CD8 antibody is shown on the X-axis, and the relative
staining by an anti-CD4 antibody appears on the Y-axis. DP, DN, CD4 SP, and CD8 SP phenotypes are divided into quadrants, and percentages are given by
quadrant. (a) CD4 versus CD8 phenotype of a wild-type thymus. (b) CD4 versus CD8 phenotype of thymocytes from H-Y TCR transgenic H-2b male and female.
(c) CD4 versus CD8 phenotype of thymocytes from H-Y TCR transgenic H-2b/d and H-2d/d females. [Parts (b) and (c) from H. von Boehmer and P. Kisielow, 1990, Science
248:1369.]
309
PA R T I V
310
|
Adaptive Immunity: Development
Cell number
Negative Selection (Central Tolerance)
Ensures Self-Tolerance
TCR affinity for self
Affinity too low
Intermediate
affinity
Affinity too high
Fail to be
positively selected
Positively
selected
Negatively
selected
“Death by neglect”
Survive
Deleted
90%–96%
2%–5%
2%–5%
FIGURE 9-7 Relationship between TCR affinity and selection. This figure schematically illustrates the association between
thymocytes’ fate and the affinity of their TCR for self-MHC/peptide
complexes that they encounter in the thymus. Fewer than 5% of
thymocytes produce TCRs that bind to MHC/peptide complexes with
high affinity. Most of these will be deleted by negative selection
(some will become regulatory T cells and other specialized cell types).
More than 90% generate TCRs that either do not bind to MHC/
peptide complexes or bind them with very low affinity. These die by
neglect. Fewer than 5% generate TCRs that bind with just the right
intermediate affinity to self-MHC/peptide complexes. These will survive and mature.
positive selection are also involved in T-cell receptor mediated activation of mature T cells (see Chapter 3), an event,
however, that requires higher-affinity TCR interactions. We
still do not fully understand how low-affinity positive selecting signals initiate maturation to the single-positive stage, or
how they differ from pro-apoptotic negative selecting signals, which are also TCR mediated.
What is the adaptive value of positive selection? Why
didn’t we evolve a system that negatively selects but leaves the
T-cell receptor repertoire otherwise intact? Some have suggested that the paring down of the repertoire is important to
increasing the efficiency of negative selection, as well as to
increasing the efficiency of peripheral T-cell responses. Without positive selection, the system would be cluttered with a
great many cells that are unlikely to recognize anything, and
reduce the probability that a T cell will find “its” antigen in a
reasonable period of time. This is a reasonable speculation;
however, positive selection may also offer more subtle advantages that have often inspired discussion in the field.
Autoreactive CD4⫹CD8⫹ thymocytes with high-affinity
receptors for self-MHC/self-peptide combinations are
potentially dangerous to an organism, and many are killed
by negative selection in the thymus. In fact, errors in the
negative selection process are responsible for a host of autoimmune disorders, including Type 1 diabetes (see Clinical
Focus). Negative selection is defined broadly as any process
that rids a repertoire of autoreactive clones and is responsible for central tolerance. It is likely that most negative
selection occurs via a process known as clonal deletion,
where high-affinity TCR interactions directly induce apoptotic signals.
Clonal deletion of DP thymocytes appears to be optimally
mediated by the same cells (APCs) and same interactions
(high-affinity TCR engagement coupled with costimulatory
signals) that activate mature T cells. Why strong TCR signals
result in death of immature T cells, but proliferation and differentiation of mature T cells (see Figure 9-4c) remains an
active area of investigation.
Thymic dendritic cells and macrophages, which are
found in multiple areas of the thymus, clearly have the ideal
features to mediate negative selection, but interestingly, so
do medullary epithelial cells, which express high levels of
the costimulatory ligands CD80 and CD86 as well as a
unique transcription factor that allows them to present
tissue-specific antigens (see below). Therefore, both the
cortex and the medulla have the potential to induce negative selection.
Do We Delete Thymocytes Reactive
to Tissue-Specific Antigens?
We have seen that MHC molecules can present peptides
derived from both endogenous and exogenous proteins on
the surface of antigen-presenting cells (see Chapter 8).
However, only a fraction of cell types—thymocytes, stromal
cells, macrophages, and other antigen-presenting cells—
reside in the thymus, so we would expect these cells to
produce only a subset of the proteins encoded in the
genome. How, then, can the thymus possibly get rid of
developing T cells that are autoreactive to tissue-specific
antigens—for example, proteins specific to the brain, the
liver, the kidney, or the pancreas?
This question bothered immunologists for a long time. For
a bit, we assumed that other mechanisms of tolerance in the
periphery took care of autoreactivity to tissue-specific proteins,
but investigations in the late 1990s surprised us all and revealed
that the thymus had an extraordinary capacity to express and
present proteins from all over the body. Bruno Kyewski and
colleagues showed that this capacity was a unique feature of
thymic medullary epithelial cells. Diane Mathis, Cristoph Benoist, and their colleagues went on to show that some medullary
epithelial cells express a unique protein, AIRE, that allows cells
T-Cell Development
|
CHAPTER 9
311
BOX 9-2
CLINICAL FOCUS
How Do T Cells That Cause Type 1 Diabetes Escape
Negative Selection?
We have an extensive appreciation
of the pain caused by autoimmune disease and its clinical progression, but
although we have also gained a deeper
understanding of the mechanisms
behind immune tolerance, we still know
little about the precise origins of autoimmune disorders. Indisputably, the primary
cause of autoimmune disease is the
escape of self-reactive lymphocytes—B
cells, T cells, or both—from negative
selection. Surprisingly little is known
about the reasons for this escape, or the
specificity and phenotype of the escapees that cause even the best-characterized
disorders, including Type I diabetes
(T1D), multiple sclerosis (MS), rheumatoid arthritis (RA), and systemic lupus
erythematosus (SLE).
Recent work on a mouse strain, the
non-obese diabetic or NOD mouse, which
is markedly susceptible to Type 1 diabetes, has shed light on the features of the
self-reactive T cells that cause the disease,
and revealed some interesting reasons for
their escape from central tolerance. T1D is
a T-cell mediated autoimmune disease
caused by T cells that kill insulin-producing
beta islet cells in the pancreas. Many of
the autoreactive T cells actually recognize a specific peptide from the insulin
protein itself.
When Emil Unanue and colleagues
examined the fine specificity of T cells
that had infiltrated the islet cells in diabetesprone mice, they found two types of cells:
T cells that could respond to the insulin
peptide that was generated via intracellular MHC Class II processing pathways
by dendritic cells (“Type A” T cells), and an
unexpected population of T cells that
responded to the insulin peptide only if it
were added exogenously to the MHC
Class II molecules on dendritic cells (“Type
B” T cells). The investigators speculate
that the endogenously processed MHC/
insulin peptide combination and the
exogenously formed MHC/insulin peptide complex differ in conformation and
therefore activate different T-cell receptor
specificities.
Perhaps more important, these observations suggest an elegant and precise
reason for the escape of autoreactive T
cells from the thymus. Dendritic cells in
the pancreas, but not other tissues,
appear uniquely capable of forming MHC
Class II/insulin peptide from exogenous
sources, presumably because they are in
an environment replete with extracellular
insulin secreted by the beta islet cells.
Thymic medullary epithelial cells or dendritic cells would not have this capacity,
and therefore “Type B” T cells would sneak
to express, process, and present proteins that are ordinarily
only found in specific organs. This group discovered AIRE
while studying the molecular basis for autoimmunity by examining human disease. They traced a mutation that caused
autoimmune polyendocrinopathy syndrome 1 (APS1, otherwise known as APECED, see Chapter 18) to this gene, and
called it AIRE for “autoimmune regulator.” The title of the
paper that announced their discovery, “Projection of an Immunological Self Shadow within the Thymus by the AIRE Protein,” aptly described AIRE’s powerful function.
AIRE’s mechanism of action is still under investigation. It
has classic features of a transcription factor and may be part
through the negative selection process in
the thymus.
Why “Type A” T cells also escape is less
clear. They are autoreactive to more conventionally processed insulin/MHC Class
II complexes, which are clearly present
on thymic medullary cells. Perhaps Type
A cells expressed too low an affinity for
insulin/MHC Class II peptides in the thymus, but were inspired by the high levels
of insulin and inflammatory signals in a
diabetic islet? Perhaps they escaped negative selection because the level of insulin peptide expression on medullary
epithelium was too low in the NOD
mouse strain? In fact, recent data from
both mice and humans suggest that the
level of insulin expression significantly
influences the progression of disease
and the efficiency of negative selection
in the thymus.
Understandably, most current therapies for autoimmune disease focus on
inhibiting the secondary but most proximal cause of autoimmune disease: the
peripheral activation of autoreactive T and
B cells that escaped negative selection.
However, by defining the molecular reasons for self-reactive lymphocyte escape,
we may ultimately find a way to develop
therapeutic approaches to correct negative selection defects, too.
of a transcriptional complex that facilitates expression of
tissue-specific genes by regulating not only translation, but
also chromatin packing. Thus, it allows medullary epithelial
cells to express proteins not ordinarily found in the thymus,
process them, and present them in MHC molecules. This is
particularly useful for the presentation of MHC Class I peptides and negative selection of CD8⫹ thymocytes. However,
neighboring dendritic cells and macrophages are also
thought to phagocytose medullary epithelial cells and so can
then present their protein contents in MHC Class II molecules,
and mediate negative selection of CD4⫹ thymocytes (see
Overview Figure 9-5).
312
PA R T I V
|
Adaptive Immunity: Development
Other Mechanisms of Central Tolerance
Other mechanisms of thymic negative selection (central
tolerance) that do not necessarily involve cell death have
been proposed. They include clonal arrest, where thymocytes that express autoreactive T-cell receptors are prevented from maturation; clonal anergy, where autoreactive
cells are inactivated, rather than deleted; and clonal editing,
where autoreactive cells are given a second or third chance
to rearrange a TCR␣ gene. Each of these mechanisms has
some experimental support, but clonal deletion is probably
the most common mechanism responsible for thymic negative selection. The generation of regulatory T cells from
autoreactive thymocytes can certainly be considered of
importance among central tolerance mechanisms and will
be discussed below.
Superantigens
Superantigens, viral or bacterial proteins that bind simultaneously to the V domain of a T-cell receptor and to the ␣
chain of a Class II MHC molecule (outside the peptide binding groove), can induce activation of all T cells that express
that particular family of V chains (see Figure 11-5). Superantigens are also expressed in the thymus of mice and
humans and influence thymocyte maturation. Because they
mimic high-affinity TCR interactions, superantigens will
induce the negative selection of all DP thymocytes whose
receptors possess V domains targeted by the superantigen.
However, because we continually generate T cells with a
wide range of T-cell receptor specificities, this loss does not
appear to have major clinical consequences.
The Selection Paradox: Why Don’t We Delete
All Cells We Positively Select?
Full understanding of the process of positive and negative
selection requires an appreciation of the following paradox:
If positive selection allows only thymocytes reactive with
self-MHC molecules to survive, and negative selection
eliminates self-MHC-reactive thymocytes, then why are any
T cells allowed to mature? Other factors must operate to
prevent these two MHC-dependent processes from eliminating the entire repertoire of MHC-restricted T cells.
The most straightforward model advanced to explain the
paradox of MHC-dependent positive and negative selection
is the affinity hypothesis,* which asserts that differences in
the strength of TCR-mediated signals received by thymo-
*
Note that a similar version of this hypothesis is also sometimes referred to
as the avidity hypothesis. Avidity and affinity do have distinct meanings,
which can differ for different investigators. For the sake of simplicity, we
consider affinity to mean the strength of interaction between the TCR and
its MHC/peptide ligand. For a more nuanced discussion of the differences
between affinity and avidity and their influence on thymic development, see
the review by Kyewski listed in the reference section of this chapter.
cytes undergoing positive and negative selection determine
the outcome of the interaction. Double-positive thymocytes
that receive low-affinity signals would be positively selected,
those that received high-affinity (autoreactive) signals
would be negatively selected, and those that received no
signal at all would die by neglect (see Figure 9-7 and Overview Figure 9-5).
The OT-I TCR transgenic mouse, developed by Kristin
Hogquist and colleagues, is a superb model for the study of
thymic selection using TCRs and MHC/peptides with
known affinities. Not only is its peptide/MHC specificity
known precisely (it binds a chicken ovalbumin peptide in
the context of the H-2Kb MHC Class I molecule), but its
affinity for a range of peptides that vary in sequence has also
been defined. To determine the influence of affinity on T-cell
development, investigators also took advantage of two other
immunological tools: (1) the fetal thymic organ culture
(FTOC) system, where thymic development can be followed
in vitro (Figure 9-8a), and (2) mice defective in MHC Class
I antigen processing. Several different mutations that lead to
defective MHC Class I presentation exist. An experimental
scheme using the TAP-1 knock out (TAP-1⫺), a mutant
mouse where newly generated peptides cannot access newly
generated MHC Class I molecules (see Chapter 8), is
depicted in Figure 9-8b. These mice cannot load their MHC
Class I with endogenous peptides; they are expressed on the
cell surface but are “empty” and do not hold their shape.
However, they can be loaded with exogenous peptides,
which stabilize their conformation. Therefore, by incubating
fetal thymic organs from these mutant mice with peptides of
choice, investigators were able to control the type of peptide
presented by the MHC Class I.
What did investigators find? As expected, OT-I⫹/TAP-1⫹
fetal thymic organ cultures produced more CD8⫹ cells than
wild-type FTOC. This is because virtually all of their thymocytes expressed a receptor that bound to the variety of
MHC Class I (H-2Kb)/self-peptide complexes expressed by
the normal stroma with intermediate affinity, generating a
signal that results in positive selection (see Classic Experiment Box 9-1). Also, as expected, CD8⫹ mature T cells
failed to develop in OT-I⫹/TAP-1⫺ thymic organ cultures
because there was no normal MHC Class I and, therefore,
no TCR signal—a failure of positive selection, resulting in
death by neglect (see Figure 9-8b, top row). However, when
a low-affinity peptide was added to the OT-I⫹/TAP-1⫺ culture, MHC Class I was able to load itself with peptide and
assume a normal conformation (see Figure 9-8b, middle
row). Positive selection occurred, and CD8⫹ T cells developed successfully. Addition of high-affinity peptides
induced a strong TCR signal and negative selection, resulting in deletion of DP cells in OT-I⫹/TAP-1⫺ thymic organ
cultures (see Figure 9-8b, bottom row). Interestingly, low
concentrations of high-affinity peptides also induced positive selection, although the cells that developed were not
fully functional. These results and many others provided
clear support for the affinity model, showing that the
T-Cell Development
|
CHAPTER 9
313
(a) Experimental procedure—fetal thymic organ culture (FTOC)
Remove
thymus
Place in
FTOC
Porous membrane
Growth medium
(b)
Peptide added
Signal generated
CD8+ T cell selection
None
cTEC
None
None
Empty MHC
class I
Intermediate
Low affinity
cTEC
Positive
MHC class I
filled with lowaffinity peptide
Strong
High affinity
cTEC
Negative
MHC class I
filled with highaffinity peptide
FIGURE 9-8 Role of TCR affinity for peptide in thymic
selection. Fetal thymocyte populations have not yet undergone
positive and negative selection and can be easily manipulated to
study the development and selection of single-positive (CD4⫹CD8⫺
and CD4⫺CD8⫹) T cells. (a) Outline of the experimental procedure for
in vitro fetal thymic organ culture (FTOC). Fetal thymi are cultured on
a filter disc at the interface between medium and air. Reagents added
to the medium are absorbed by the thymi. In this experiment, peptide is added to the medium of FTOC from TAP-1 knockout (TAP-1⫺)
mice, which are unable to form peptide-MHC Class I complexes
unless peptide is added exogenously to the culture medium. (b) The
development and selection of CD8⫹CD4⫺ class I–restricted T cells
strength of TCR/MHC/peptide interaction does, indeed,
influence cell fate.
The OT-I system has also been used to estimate the range
of TCR affinities that define positive versus negative selection outcomes. It appears as if negative selection occurs at
affinities that are threefold higher than those that induce
positive selection.
depends on TCR peptide–MHC class I interactions. The mice used in
this study were transgenic for OT-I TCR ␣ and chains, which recognize an ovalbumin peptide when associated with MHC Class I; the
proportion of CD8⫹ T cells that develop in these thymi is higher than
normal because all thymocytes in the mouse express the MHC Class
I restricted specificity. Different peptide/MHC complexes interact
with the TCR with different affinities. Varying the peptide added to
FTOC from OT-I⫹/TAP-1⫺ mice revealed that peptides that interact
with low affinities (middle row) resulted in positive selection, and
levels of CD4⫺CD8⫹ cells that approached that seen in the OT-I⫹/
TAP-1⫹ mice. Peptides that interact with high affinities (bottom row)
cause negative selection, and fewer CD4⫺CD8⫹ T cells appeared.
An Alternative Model Can Explain the
Thymic Selection Paradox
Philippa Marrack, John Kappler, and their colleagues
thoughtfully offered an alternative possibility to explain the
“thymic paradox”—that is, why we don’t negatively select all
cells that we positively select. These investigators advanced
314
PA R T I V
|
Adaptive Immunity: Development
the altered peptide model, a suggestion that cortical epithelium, which induces positive selection, makes peptides that
are unique and distinct from peptides made by all other
cells, including the thymic cells that induce negative selection. Thus, those thymocytes selected on these unique peptide/
MHC complexes would not be negatively selected when they
browse the medulla and other negatively selecting cells.
Initial experiments did not support this model; however,
advances in our ability to analyze peptides presented on cell
surfaces in detail suggests the model has merit. Cortical epithelial cells may, indeed, process peptides differently, and are
likely to present a different array of peptides to developing
T cells. Recall that peptides are processed intracellularly by
several mechanisms, including the activity of the proteosome (see Figure 8-16). It appears that components of the
proteosome expressed by cortical epithelial cells (the “thymoproteasome”) are unique. The thymic epithelium may
also express unique versions of proteases (e.g., cathepsins).
The altered peptide and affinity models are by no means
mutually exclusive. Investigations firmly establish that affinity plays an important role in distinguishing positive from
negative selection. That the cortex may also generate novel
peptides simply underscores the possibility that the thymus
evolved multiple ways to ensure that we would be able to
generate a sufficiently large pool of T cells with diverse
specificities.
Do Positive and Negative Selection Occur
at the Same Stage of Development,
or in Sequence?
Strictly interpreted, the affinity model assumes that positive
and negative selection operate on exactly the same target
population (DP thymocytes) and in the same microenvironment of the thymus (the thymic cortex). Although it is clear
that thymocytes undergo positive selection in the cortex, as
we have seen above, the medulla is the site of negative selection to tissue-specific antigens. This compartmentalization
of function suggests some thymocytes are positively selected
and initiate a maturation program prior to negative selection. That this can happen is supported by observations
showing that “semi-mature” thymocytes—those that have
been positively selected and are in transition from the DP to
the SP stages—are excellent targets of negative selection.
Most likely, thymocytes can be negatively selected at
more than one point of development (illustrated in Overview Figure 9-5). Those thymocytes that bind with high
affinity as they browse MHC/peptide complexes in the cortex are negatively selected, and those that receive positively
selecting signals are given permission not only to mature,
but to migrate to the cortical medullary boundary and ultimately to the medulla. In the medulla, semi-mature thymocytes can again be subject to negative selection if they
interact with high affinity to tissue-specific antigens.
Migration of positively selected thymocytes to the
medulla is now known to be dependent on expression of the
CCR7 chemokine receptor: CCR7 deficient thymocytes fail
to enter the medulla. Interestingly, they still mature and are
exported to the periphery, but there they cause autoimmunity—
an observation that, again, underscores the central importance of the medulla in removing autoreactive cells from the
T-cell repertoire.
Lineage Commitment
As thymocytes are being screened on the basis of their TCR
affinity for self-antigens, they are also being guided in their
lineage decisions. Specifically, a positively selected doublepositive thymocyte must decide whether to join the CD8⫹
cytotoxic T-cell lineage or the CD4⫹ helper T-cell lineage.
Lineage commitment requires changes in genomic organization and gene expression that result in (1) silencing of one
coreceptor gene (CD4 or CD8) as well as (2) expression of
genes associated with a specific lineage function. Immunologists continue to debate about the cellular and molecular
mechanisms responsible for lineage commitment. We will
review the most current perspectives.
Several Models Have Been Proposed to
Explain Lineage Commitment
When early studies with TCR transgenic mice revealed that
affinity for MHC Class I versus Class II preference dictated
the CD8⫹ versus CD4⫹ fate of developing thymocytes (see
Classic Experiment Box 9-1, investigators advanced two
simple testable models to explain how developing T cells
“matched” their TCR preference with their coreceptor
expression.
In the instructive model, TCR/CD4 and TCR/CD8
coengagement generates unique signals that directly initiate
distinct developmental programs (Figure 9-9a). For example, if a thymocyte randomly generated a TCR with an affinity for MHC Class I, the TCR and CD8 would bind MHC
Class I together, and generate a signal that specifically initiated a program that silenced CD4 expression and induced
expression of genes specific for cytotoxic T-cell lineage function. Likewise, TCR/CD4 coengagement would generate a
unique signal that initiated CD8 silencing and the helper
T-cell developmental program.
In the stochastic model, a positively selected thymocyte
randomly down-regulates CD4 or CD8 (Figure 9-9b). Only
those cells that express the “correct” coreceptor—the ones
that can coengage MHC with the TCR—generate a TCR
signal strong enough to survive to mature. In this model,
TCR/CD4 and TCR/CD8 coengagement does not necessarily generate distinct signals. Unfortunately, studies that followed the consequences of such mismatches confounded
researchers by providing evidence in support of both models! Clearly they were too simplistic.
Further experiments challenged some of the assumptions
on which these early efforts were based and indicated that
T-Cell Development
(a) Instructive model
CD4+ 8+
|
CHAPTER 9
315
CD8
engagement signal
CD4lo 8hi
CD4− 8+ T cell
CD4
engagement signal
CD4+ 8+
CD4hi 8lo
CD4+ 8− T cell
(b) Stochastic model
Able to bind
Ag + class I MHC
CD4+ 8+
Random
CD4
CD4− 8 + T cell
Not able to bind
Ag + class I MHC
CD4lo 8hi
Apoptosis
Able to bind
Ag + class II MHC
CD4+ 8+
Random
CD8
CD4+ 8 − T cell
Not able to bind
Ag + class II MHC
CD4hi 8lo
Apoptosis
(c) Kinetic signaling model
Disrupted TCR
signal, IL–7
CD4–8+ T cell
Engagement signal
CD8
CD4+8+
CD4+8lo
Continuous
TCR signal
CD4+8– T cell
FIGURE 9-9 Proposed models of lineage commitment, the
decision of double-positive thymocytes to become helper
CD4ⴙ or cytotoxic CD8ⴙ T cells. (a) According to the instructive
model, interaction of a coreceptor with the MHC molecule for which
it is specific results in down-regulation of the other coreceptor.
(b) According to the stochastic model, down-regulation of CD4 or CD8
is a random process. (c) According to the kinetic signaling model,
the decision to commit to the CD4 or CD8 lineage is based on the
continuity of the TCR signal that a thymocyte receives. Positive selection results in down-regulation of CD8 on all thymocytes. This will not
alter the intensity of a TCR/CD4/MHC Class II signal, and cells receiving
this signal will continue development to the CD4 SP lineage. However,
down-regulation of CD8 diminishes (interrupts) a TCR/CD8/MHC Class
I signal, an experience that sends a cell toward the CD8 lineage. IL-7
signals are required to “seal” CD8 lineage commitment.
the strength and duration of the T-cell receptor/coreceptor
signal experienced by a thymocyte play a more important
role in dictating cell fate than its specificity for MHC Class I
or Class II. In fact, thymocytes whose TCR had a known
preference of MHC Class II could be coaxed into the CD8⫹
T-cell fate simply by weakening the CD4/Class II interaction
(e.g., by mutating kinases downstream of TCR signaling so
that they signaled more or less effectively). By inhibiting
TCR signaling, investigators could coax cells that normally
commit to the CD4 lineage to become CD8⫹ cells. By
enhancing TCR signaling, investigators could coax MHC
Class I restricted cells to become CD4⫹. These data suggested that stronger positive selecting TCR signals resulted
in CD4 lineage commitment, and weaker positive selecting
TCR signals resulted in CD8 lineage commitment. They
were consistent with the observation that the intracellular
tail of CD4 interacts more effectively with the tyrosine
kinase lck than the intracellular tail of CD8. Therefore, a
TCR/CD4 coengagement is likely to generate stronger signals than TCR/CD8 and result in CD4⫹ T-cell commitment.
316
PA R T I V
|
Adaptive Immunity: Development
The strength of signal model represented an advance in our
understanding, but it, too, is likely to be too simplistic. Alfred
Singer and colleagues have proposed the kinetic signaling
model, which incorporates historical data, our improved
understanding of changes in CD8 expression during positive
selection, as well as recent advances in the understanding of
the complexity of coreceptor gene silencing (Figure 9-9c).
Briefly, they propose that thymocytes commit to the CD4⫹
T-cell lineage if they receive a continuous signal in response to
TCR/coreceptor engagement, but commit to the CD8 lineage
if the TCR signal is interrupted. We now know that all
CD4⫹CD8⫹ thymocytes down-regulate surface levels of CD8
in response to positive selection. Given this response, only
MHC Class II restricted T cells will maintain continuous
TCR/CD4/MHC Class II interaction, and therefore develop to
the CD4 lineage. However, with the loss of CD8 expression,
MHC Class I restricted T cells will lose the ability to maintain
TCR/CD8/MHC Class I interactions. Singer et al. provide
evidence that these thymocytes, interrupted from their TCR/
coreceptor engagement, are subsequently rescued by IL-7,
which facilitates their commitment to the CD8⫹ lineage.
Transcription factors that specifically regulate development to the CD4 and CD8 lineage have recently been identified. At present, ThPok and Runx3 are taking center stage as
transcriptional factors required for CD4 and CD8 commitment, respectively, although others also play a role. Both
ThPok and Runx3 act, at least in part, by suppressing genes
involved in differentiation to the other lineage. ThPok inhibits expression of genes that regulate CD8 differentiation,
including Runx3. Runx3 inhibits expression of CD4, itself, as
well as ThPok. This is just the beginning of our understanding of the transcriptional networks that regulate CD4 and
CD8 developmental programs. Likely to play a role, too, are
miRNAs; some have already been implicated in CD8 singlepositive T-cell differentiation. There is clearly more to come.
Double-Positive Thymocytes May Commit
to Other Types of Lymphocytes
Small populations of DP thymocytes can also commit to
other T-cell types, including the NK T cell, regulatory T cell,
and intraepithelial lymphocyte (IEL) lineages. NKT cells
(which include mature cells that express only CD4 and cells
that have lost both CD4 and CD8) play a role in innate
immunity (see Chapter 5) and express a TCR that includes
an invariant TCR␣ chain (V␣14). For this reason, NKT cells
are sometimes referred to as iNKT cells for invariant. Their
invariant TCR interacts not with classical MHC, but with the
related molecule CD1, which presents glycolipids, not peptides (see Chapter 8). Intraepithelial lymphocytes (IEL),
most of which are CD8⫹, also have features of innate
immune cells and patrol mucosal surfaces. Regulatory T cells
(TREG), another CD4⫹ subset discussed in more detail below,
quench adaptive immune reactions. All three subpopulations are thought to develop from DP thymocytes in
response to autoreactive, high-affinity TCR interactions—
the same interactions, in fact, that mediate negative selection. What determines whether a thymocyte undergoes
negative selection or an alternative developmental pathway
remains a topic of much interest.
Exit from the Thymus
and Final Maturation
Once a thymocyte successfully passes through selection and
makes a lineage decision, it enters a quiescent state and
leaves the thymus. The cellular and molecular basis for thymocyte egress was unknown until we gained a better understanding of the receptors and ligands that regulate cell
migration. As mentioned above, one set of investigators discovered that thymocytes failed to enter the medulla if they
were deficient for the CCR7 chemokine receptor. However,
these cells still were able to leave the thymus, indicating that
other receptors control thymic exit. The identity of this
receptor was discovered when investigators found that few if
any T cells made it out of the thymus in a sphingosine-1phosphate receptor (SIPR) knockout mouse.
Current observations suggest that a cascade of events controls these final stages of maturation: positive selecting TCR
signals up-regulate Foxo1, a transcription factor that controls
expression of several genes related to T-cell function. Foxo1
regulates expression of Klf2, which, in turn, up-regulates
SIPR. Foxo1 also up-regulates both IL-7R, which helps to
maintain mature T-cell survival, and CCR7, the chemokine
receptor that directs mature T-cell traffic to the lymph nodes.
Mature T cells that exit the thymus are referred to as recent
thymic emigrants (RTEs). It is now clear that recent thymic
emigrants are not as functionally mature as most naïve T cells
in the periphery: they do not proliferate or secrete cytokines
as vigorously in response to T-cell receptor stimulation. RTEs
also can be distinguished from the majority of peripheral
naïve T cells because their levels of expression of several surface proteins (e.g., the maturation markers HSA and Qa-2, as
well as the IL-7 receptor) are more similar to their immature
thymocyte ancestors than their fully mature T-cell descendants. Investigators are particularly interested in RTEs because
they are an important source of mature T cells in individuals
who are lymphopenic (i.e., have a reduced pool of functional
lymphocytes), including people who have undergone chemotherapy, newborns, and the aged. Although research is ongoing, studies suggest that the final maturation of these cells is
influenced by their interactions with both MHC and nonMHC ligands in secondary lymphoid organs.
Other Mechanisms That Maintain
Self-Tolerance
As we have seen, negative selection of thymocytes can rid a
developing T-cell repertoire of cells that express a high
affinity for both ubiquitous self-antigens and, thanks to the
|
T-Cell Development
CHAPTER 9
317
and thwart autoimmune reactions. What determines
whether a self-reactive thymocyte dies or differentiates
into a TREG cell is still an open question. Investigators are
currently trying to understand if the choice is made based
on subtle differences in affinity for self or on differences in
their maturation state when they receive a high-affinity
signal. Recent work suggests that TREGs develop in a
unique microenvironmental niche within the thymus, and
that the available space for developing cells in this niche is
limited. These findings suggest that thymocytes that commit to the regulatory T-cell lineage are likely to receive
unique stimulatory signals. These natural TREGs share the
periphery with induced TREGs that can develop from conventional mature T cells that are exposed to TGF- and
IL-10 cytokines (see Chapter 11).
Animal studies show that members of the FoxP3⫹ TREG
population inhibit the development of autoimmune diseases
such as experimentally induced inflammatory bowel disease,
experimental allergic encephalitis, and autoimmune diabetes. Suppression by these regulatory cells is antigen specific
because it depends on activation through the T-cell receptor.
Exactly how TREGs quench responses is still debated, although
they probably do so via a variety of means: directly inhibiting an antigen-presenting cell’s ability to activate T cells,
directly killing T cells, indirectly inhibiting T-cell activity by
secreting inhibitory cytokines IL-10 and TGF-, and/or
depleting the local environment of stimulatory cytokines
such as IL-2 (Figure 9-10).
activity of AIRE, tissue-specific antigens. However, negative
selection in the thymus is not perfect. Autoreactive T cells do
escape, either because they have too low an affinity for self to
induce clonal deletion, or they happen not to have browsed
the “right” tissue-specific antigen/MHC combination. The
body has evolved several other mechanisms to avoid autoimmunity, including what has become a major focus of interest
for immunologists: the development in both the thymus and
the periphery of a fascinating group of cells now known as
regulatory T cells.
TREG Cells Negatively Regulate
Immune Responses
Regulatory T cells (TREG cells) can inhibit the proliferation
of other T-cell populations in vitro, effectively suppressing
autoreactive immune responses. They express on their surface CD4 as well as CD25, the ␣ chain of the IL-2 receptor.
However, TREG cells are more definitively identified by their
expression of a master transcriptional regulator, FoxP3, the
expression of which is necessary and sufficient to induce differentiation to the TREG lineage.
TREG cells can develop in the thymus and appear to
represent an alternative fate for autoreactive T cells. As we
have seen, most thymocytes that express receptors with
high affinity for self-antigen die via negative selection.
However, a small fraction appear to commit to the regulatory T-cell lineage and leave the thymus to patrol the body
2
T cell
Inhibitory
cytokines
T cell
TGF
1
Cytokine
deprivation
IL–2R
FoxP3
3
TCR
APC
Inhibiting antigen
presenting cells
MHC
4
Cytotoxicity
T cell
FIGURE 9-10 How regulatory T cells inactivate traditional T cells. Some possible mechanisms of TREG activity are
illustrated in this schematic. These may all contribute to quelling
immune responses in vivo. (1) Cytokine deprivation. TREGs express
relatively high levels of high-affinity IL-2 receptors and can compete for the cytokines that activated T cells need to survive and
proliferate. (2) Cytokine inhibition. TREGs secrete several cytokines,
T cell
including IL-10 and TGF-, which bind receptors on activated
T cells and reduce signaling activity. (3) Inhibition of antigenpresenting cells. TREGs can interact directly with MHC Class II
expressing antigen-presenting cells and inhibit their maturation,
leaving them less able to activate T cells. (4) Cytotoxicity. TREGs can
also display cytotoxic function and kill cells by secreting perforin
and granzyme.
318
PA R T I V
|
Adaptive Immunity: Development
The existence of regulatory T cells that specifically suppress immune responses has clinical implications. Depletion
or inhibition of TREG cells before immunization may enhance
immune responses to conventional vaccines. Elimination of
T cells that suppress responses to tumor antigens may also
facilitate the development of antitumor immunity. Conversely, increasing the suppressive activity of regulatory T-cell
populations could be useful in the treatment of allergic or
autoimmune diseases. The ability to increase the activity of
regulatory T-cell populations might also be useful in suppressing organ and tissue rejection. Investigators doing future
work on this regulatory cell population will seek deeper
insights into the mechanisms by which members of FoxP3⫹
T-cell populations regulate immune responses. They will also
make determined efforts to discover ways the activities of
these populations can be increased to diminish unwanted
immune responses and decreased to promote desirable ones.
Peripheral Mechanisms of Tolerance also
Protect Against Autoreactive Thymocytes
The body has evolved several other mechanisms to manage the
autoreactive escapee in the periphery. Briefly, many antigens
are “hidden” from autoreactive T cells because only a subset of
cells (professional APCs) express the right costimulatory molecules needed to initiate the immune response. Autoreactive
naïve T cells that can see an MHC/self-peptide combination on
a nonprofessional antigen-presenting cell will not receive the
correct costimulatory signals and will not divide or differentiate. For example, if a thymocyte specific for a peptide made by
a kidney cell escaped from the thymus, it will not be activated
unless that peptide were presented on a professional antigenpresenting cell. Kidney cells do not express the costimulatory
ligands required for activating a CD4⫹ or a CD8⫹ T cell. In
fact, a high-affinity interaction with MHC/peptide combinations on the surface of kidney cell, in the absence of costimulatory ligands, could result in T-cell anergy—another peripheral
tolerance mechanism that is described in more detail in Chapter 11. The clinical consequences of failures of central and
peripheral tolerance will be discussed in Chapter 16.
Apoptosis
As we have seen here, and will also see in Chapter 10, apoptosis features prominently during T- and B-lymphocyte
development. T-cell development is a particularly expensive
process for the host. An estimated 95% to 98% of all thymocytes do not mature—most die by apoptosis within the thymus either because they fail to make a productive TCR gene
rearrangement or because they fail to interact with selfMHC. Some (2%–5%) die, also via apoptosis, because they
are negatively selected.
Not only does apoptosis shape the developing lymphocyte repertoire; it also regulates immune cell homeostasis by
returning myeloid, T-, and B-cell populations to their appro-
priate levels after bursts of infection-inspired proliferation,
and it is the means by which a cell dies after being targeted
by cytotoxic T cells or NK cells. Understanding its fundamental biological features is an important part of an overall
understanding of immune cell function. Because apoptosis
is such a critical part of lymphocyte development, we will
take the opportunity in this chapter to briefly describe the
apoptotic process and how it is regulated.
Apoptosis Allows Cells to Die without
Triggering an Inflammatory Response
Apoptosis, or programmed cell death, is an energy-dependent
process by which a cell brings about its own demise. It is most
often contrasted with necrosis, a form of cell death arising from
injury or toxicity (Figure 9-11a). In necrosis, injured cells swell
and burst, releasing their contents and possibly triggering a
damaging inflammatory response. Apoptotic cells, in contrast,
dismantle their contents without disrupting their membranes
and induce neighboring cells to engulf them before they can
release any inflammatory material. Morphologic changes associated with apoptosis include a pronounced decrease in cell
volume; modification of the cytoskeleton, which results in
membrane blebbing; a condensation of the chromatin; and
degradation of the DNA into fragments (Figure 9-11b).
Different Stimuli Initiate Apoptosis,
but All Activate Caspases
Although a lymphocyte can be signaled to die in several
ways, all apoptotic signals ultimately activate a class of proteases called caspases. Caspase activation is common to
almost all death pathways known in both vertebrates and
invertebrates, demonstrating that apoptosis is an ancient
process. Many of its molecular participants have been conserved through evolution (Table 9-2).
The role of caspases was first revealed by studies of developmentally programmed cell death in the nematode Caenorhabditis elegans, where the death of cells was shown to be
totally dependent on the activity of a gene that encoded a
cysteine protease with specificity for aspartic acid residues.
We now know that mammals have at least 14 cysteine-aspartic
proteases, or caspases, and all apoptotic cell deaths require
the activity of at least a subset of these molecules.
How do caspases trigger cell death? Apoptotic pathways
typically involve two classes of caspases, both of which are
maintained in an inactive state until apoptosis is initiated.
Initiator caspases are activated by a death stimulus and, in
turn, activate effector caspases, which execute the death
program by (1) cleaving critical targets necessary for cell
survival (e.g., cytoskeletal proteins) as well as (2) activating
other enzymes that carry out the dismantling of the cell.
Ultimately, the catalytic cascade initiated by caspases induces
death via an orderly disassembly of intracellular molecules.
It also results in the packaging of cell contents into vesicles
that are ultimately engulfed.
T-Cell Development
|
CHAPTER 9
319
(a)
NECROSIS
APOPTOSIS
Mild convolution
Chromatin compaction
and segregation
Condensation of
cytoplasm
Chromatin clumping
Swollen organelles
Flocculent mitochondria
Nuclear fragmentation
Blebbing
Apoptotic bodies
Disintegration
Release of
intracellular
contents
Phagocytosis
Apoptotic body
Phagocytic
cell
Inflammation
(b)
Normal
Apoptotic
TEM
FIGURE 9-11 Morphological changes that occur during
SEM
apoptosis. (a) Comparison of morphologic changes that occur in
apoptosis versus necrosis. Apoptosis, which results in the programmed cell death of hematopoietic cells, does not induce a local
inflammatory response. In contrast, necrosis, the process that leads to
death of injured cells, results in release of the cells’ contents, which
may induce a local inflammatory response. (b) Microscopic images of
apoptotic cells. Transmission and scanning electron micrographs
(TEM and SEM) of normal and apoptotic thymocytes, as indicated.
[Part (b) from B. A. Osborne and S. Smith, 1997, Journal of NIH Research 9:35;
courtesy of B. A. Osborne, University of Massachusetts at Amherst.]
320
PA R T I V
TABLE 9-2
|
Adaptive Immunity: Development
Proteins involved in apoptosis
Protein
Location in cell
Function
Role in apoptosis
Death receptors
(e.g., Fas, TNFR)
Membrane
Activates caspase cascade
after binding ligand
(e.g., FasL, TNF)
Promotes
Initiator caspase
(e.g., caspase-8, caspase-9)
Cytosol
Protease; cleaves and
activates effector caspases
Promotes
Effector caspase
(e.g., caspase-3)
Cytosol, nucleus
Protease; cleaves and
activates enzymes, cleaves
and disassembles structural
proteins
Promotes
Cytochrome c
Intermembrane space,
mitochondria
Participates in formation
of apoptosome
Promotes
Apaf-1
Cytosol
Participates in formation
of apoptosome
Promotes
Anti-apoptotic Bcl-2
family members
(e.g., Bcl-2, Bcl-xL)
Mitochondria, ER
Regulates cytochrome c
release
Inhibits
Pro-apoptotic Bcl-2
family members
(e.g., Bax, Bak)
Mitochondria
Regulates cytochrome
c release; opposes
Bcl-2, Bcl-xL
Promotes
BH3 proteins
(e.g., Bim, Bid, PUMA)
Cytosol and mitochondria
Opposes activity of antiapoptotic Bcl-2 family
members at mitochondria
Promotes
Stimuli that lead to apoptosis can be divided into two
categories based on how they activate the effector caspases.
Signals such as radiation, stress, and growth factor withdrawal induce caspase activation and apoptosis via the
intrinsic pathway that is mediated by mitochondrial molecules, while membrane-bound or soluble ligands that bind
to membrane receptors stimulate caspase activation and
apoptosis via the extrinsic pathway.
Apoptosis of Peripheral T Cells Is Mediated
by the Extrinsic (Fas) Pathway
Outside of the thymus, most of the TCR-mediated apoptosis
of mature T cells is induced via the extrinsic pathway
through membrane-associated death receptors, including Fas
(CD95). Repeated or persistent stimulation of peripheral T
cells via their T-cell receptors results in the co-expression of
both Fas and Fas ligand (FasL) on T cells, ultimately leading
to Fas/FasL-mediated death. Apoptosis induced by antigen
receptor activation is called activation-induced cell death
(AICD) and is a major homeostatic mechanism, helping to
reduce the pool of activated T cells after antigen is cleared
and helping to remove stray autoreactive T cells that are
stimulated by self-antigens.
The signaling cascade initiated by Fas (and by all death
receptors) leads directly to the activation of the initiator caspase,
caspase-8. When FasL binds Fas, procaspase-8, an inactive form
of caspase-8, is recruited to the intracellular tail of Fas by an
adaptor molecule called FADD (Fas-associated protein with
death domain) (Figure 9-12a). The aggregation of procaspase-8
results in its cleavage and conversion to active caspase-8, which
will then activate effector caspase-3, which in turn initiates the
proteolytic cascade that leads to the death of the cell.
The importance of Fas and FasL in the removal of mature
activated T cells is underscored by abnormalities in lpr/lpr mice,
a naturally occurring loss of function mutation in Fas. When T
cells become activated in these mice, the Fas/FasL pathway is
not operative, and stimulated T cells continue to proliferate.
These mice spontaneously develop strikingly large lymph nodes
that are filled with excessive numbers of lymphocytes. Ultimately, they develop autoimmune disease, clearly demonstrating the consequences of a failure to delete activated T cells. An
additional spontaneous mutant, gld/gld, has the complementary
loss of function. These mice lack functional FasL and display
abnormalities very similar to those found in the lpr/lpr mice.
Recently, humans with defects in Fas have been reported. As
expected, these individuals display characteristics of autoimmune disease (see Clinical Focus Box 9-3).
T-Cell Development
(a)
(b)
T cell
FasL
MHC
Fas
TCR
FADD
Bcl-2
Caspase-8
(active)
Mitochondrion
Bid
Released
cytochrome c
Truncated
Bid
Procaspase-3
(inactive)
Caspase-9
Caspase-3
(active)
Apoptosis
substrates
Apoptosome
Active
apoptotic
effectors
Apaf-1
Procaspase-9
Apoptosis
FIGURE 9-12 Two pathways to apoptosis in T cells.
(a) Activated peripheral T cells are induced to express high levels of
Fas and FasL and stimulate the extrinsic apoptotic pathway. FasL
induces the trimerization of Fas on a neighboring cell. FasL can also
engage Fas on the same cell, resulting in a self-induced death signal. Trimerization of Fas leads to the recruitment of FADD, which
leads in turn to the cleavage of associated molecules of procaspase-8
to form active caspase-8. Caspase-8 cleaves procaspase-3, producing active caspase-3, which results in the death of the cell.
Caspase-8 can also cleave the Bcl-2 family member Bid to a truncated form that can activate the mitochondrial death pathway.
(b) Other signals, such as the engagement of the TCR by peptideMHC complexes on an APC during T-cell development, result in the
activation of the intrinsic, mitochondrial death pathway. A key feature of this pathway is the release of cytochrome c from the inner
mitochondrial membrane into the cytosol, an event that is regulated by Bcl-2 family members. Cytochrome c interacts with Apaf-1
and subsequently with procaspase-9 to form the active apoptosome. The apoptosome initiates the cleavage of procaspase-3,
producing active caspase-3, which initiates the execution phase of
apoptosis by proteolysis of substances whose cleavage commits
the cell to apoptosis. [Adapted in part from S. H. Kaufmann and M. O.
Hengartner, 2001, Trends in Cell Biology 11:526.]
CHAPTER 9
321
Fas and FasL are members of a family of related receptor/
ligands including tumor necrosis factor (TNF) and its
ligand, TNFR (tumor necrosis factor receptor), which can
also induce apoptosis via the activation of caspase-8 followed by the activation of effector caspases such as caspase-3
(see Figure 4-14).
TCR-Mediated Negative Selection in the
Thymus Induces the Intrinsic (MitochondriaMediated) Apoptotic Pathway
Bim
Procaspase-8
(inactive)
|
Fas/FasL interactions do not appear to play a central role in
negative selection in the thymus. Instead, TCR-mediated
negative-selecting signals in the thymus induce a route to
apoptosis in which mitochondria play a central role (Figure 9-12b). In mitochondria-dependent (intrinsic) apoptotic
pathways, cytochrome c, which normally resides between
the inner and outer mitochondrial membranes, leaks into
the cytosol. The release of cytochrome c is regulated by
several different protein families. In thymocytes, Bim, a
Bcl-2 family member, and Nur77, an orphan nuclear receptor, both play a role.
In the cytosol, cytochrome c binds to a protein known as
Apaf-1 (apoptotic protease-activating factor 1), which then
oligomerizes. Binding of oligomeric Apaf-1 to procaspase-9
results in its transformation to active caspase-9, an initiator
caspase. The cytochrome c/Apaf-1/caspase-9 complex, called
the apoptosome, proteolytically cleaves procaspase-3, generating active caspase-3, which initiates the cascade of reactions that kills the cell.
Interestingly, although over 95% of thymocytes die during development, apoptotic thymocytes are very difficult to
find in a normal thymus. Investigators showed that if you
inhibit the activity of thymic macrophages, apoptotic thymocytes are abundantly evident. This experiment dramatically underscored both the importance and the efficiency of
phagocytes in clearing dying cells from the thymus.
Cell death induced by Fas/FasL is swift, with rapid activation of the caspase cascade leading to cell death in 2 to 4
hours. On the other hand, TCR-induced negative selection
appears to involve a more circuitous pathway, requiring the
activation of several processes, including mitochondrial
membrane failure, the release of cytochrome c, and the formation of the apoptosome before caspases become involved.
Consequently, TCR-mediated negative selection can take as
long as 8 to 10 hours.
Bcl-2 Family Members Can Inhibit
or Induce Apoptosis
The Bcl-2 (B-cell lymphoma 2) family of genes, which regulate
cytochrome c release, plays a prominent role in immune cell
physiology. They include both anti-apoptotic and proapoptotic proteins that can insert into the mitochondrial
membrane. Although their precise mechanism of action is still
322
PA R T I V
|
Adaptive Immunity: Development
CLINICAL FOCUS
Failure of Apoptosis Causes Defective
Lymphocyte Homeostasis
The maintenance of appropriate
Normal control
Patient A
20%
CD4–/CD8+
1%
CD4+/CD8+
24%
CD4–/CD8+
1%
CD4+/CD8+
4%
CD4–/CD8–
75%
CD4+/CD8–
43%
CD4–/CD8–
32%
CD4+/CD8–
103
CD8
numbers of various types of lymphocytes
is extremely important to an effective
immune system. One of the most important elements in this regulation is apoptosis mediated by the Fas/FasL system. The
following excerpts from medical histories
show what can happen when this key
regulatory mechanism fails.
Patient A: A woman, now 43, has had
a long history of immunologic imbalances and other medical problems. By
age 2, she was diagnosed with CanaleSmith syndrome (CSS), a severe enlargement of such lymphoid tissues as lymph
nodes (lymphadenopathy) and spleen
(splenomegaly). Biopsy of lymph nodes
showed that, in common with many
other CSS patients, she had greatly
increased numbers of lymphocytes. She
had reduced numbers of platelets
(thrombocytopenia) and, because her red
blood cells were being lysed, she was
anemic (hemolytic anemia). The reduction in numbers of platelets and the lysis
of red blood cells could be traced to the
action of circulating antibodies that
reacted with these host components. At
age 21, she was diagnosed with grossly
enlarged pelvic lymph nodes that had to
be removed. Ten years later, she was
again found to have an enlarged abdominal mass, which on surgical removal
turned out to be a half-pound lymph
node aggregate. She has continued to
have mild lymphadenopathy and, typical
of these patients, the lymphocyte populations of enlarged nodes had elevated
numbers of T cells (87% as opposed to a
104
102
101
100
100
101
102
CD4
103
104
100
101
102
CD4
103
104
FIGURE 1
Flow-cytometric analysis of T cells in the blood of CSS patient A and a control
subject. Mature T cells are either CD4⫹ or CD8⫹. Although almost all of the T cells in the control
subject are CD4⫹ or CD8⫹, the CSS patient shows high numbers of double-negative T cells (43%),
which express neither CD4 nor CD8. The percentage of each category of T cells is indicated in the
quadrants. [Adapted from M. D. Drappa et al., 1996, New England Journal of Medicine 335:1643.]
normal range of 48%–67% T cells). Examination of these cells by flow cytometry
and fluorescent antibody staining
revealed an excess of double-negative T
cells (Figure 1). Also, like many patients
with CSS, she has had cancer: breast cancer at age 22 and skin cancer at ages 22
and 41.
Patient B: A man who was eventually
diagnosed with CSS had severe lymphadenopathy and splenomegaly as an
infant and child. He was treated from age
4 to age 12 with corticosteroids and the
immunosuppressive drug mercaptopurine. These appeared to help, and the
swelling of lymphoid tissues became
milder during adolescence and adulthood. At age 42, he died of liver cancer.
controversial, the balance of pro-apoptotic and anti-apoptotic
family members influences a cell’s response to stress. The
Bcl-2 family also includes an important, third group of proteins, the BH3 family, all of which do not insert in the mitochondrial membrane, but instead inhibit the anti-apoptotic
Bcl-2 family members, and so are ultimately pro-apoptotic.
Intrinsic apoptotic stimuli act, in part, by activating BH3
family members, which are often sequestered from the mito-
Patient C: An 8-year-old boy, the son
of patient B, was also afflicted with CSS
and showed elevated T-cell counts and
severe lymphadenopathy at the age of 7
months. At age 2 his spleen became so
enlarged that it had to be removed. He
also developed hemolytic anemia and
thrombocytopenia. However, although he
continued to have elevated T-cell counts,
the severity of his hemolytic anemia and
thrombocytopenia have so far been controlled by treatment with methotrexate, a
DNA-synthesis-inhibiting drug used for
immunosuppression and cancer chemotherapy.
Recognition of the serious consequences of a failure to regulate the number of lymphocytes, as exemplified by
chondria in a healthy cell. For example, TCR-mediated
negative selection activates Bim, a BH3 family protein that
facilitates cytochrome c release (see Figure 9-12b). However,
even the extrinsic pathway can activate BH3 family members. For instance, caspase-8 generated by the Fas pathway
cleaves the BH3 family member Bid, releasing it from
sequestration and allowing it to activate the mitochondrial
death pathway.
|
T-Cell Development
CHAPTER 9
323
BOX 9-3
60
Percentage of T cells killed
these case histories, emerged from
detailed study of several children whose
enlarged lymphoid tissues attracted medical attention. In each of these cases of
CSS, examination revealed grossly
enlarged lymph nodes that were 1 cm to
2 cm in girth and sometimes large enough
to distort the local anatomy. In four of a
group of five children who were studied
intensively, the spleens were so massive
that they had to be removed.
Even though the clinical picture in
CSS can vary from person to person, with
some individuals suffering severe chronic
affliction and others only sporadic episodes of illness, there is a common feature: a failure of activated lymphocytes to
undergo Fas-mediated apoptosis. Isolation and sequencing of Fas genes from a
number of patients and more than a
hundred controls reveals that CSS patients
are heterozygous (fas⫹/⫺) at the fas locus
and thus carry one copy of a defective
fas gene. A comparison of Fas-mediated
cell death in T cells from normal controls
who do not carry mutant Fas genes with
death induced in T cells from CSS patients
shows a marked defect in Fas-induced
death (Figure 2). Characterization of the
Fas genes so far seen in CSS patients
reveals that they have mutations in or
around the region encoding the deathinducing domain (the “death domain”)
of this protein. Such mutations result in
the production of Fas protein that lacks
biological activity but still competes
with normal Fas molecules for interactions with essential components of the
Fas-mediated death pathway. Other
mutations have been found in the extracellular domain of Fas, often associated
with milder forms of CSS or no disease
at all.
Normal controls
Patient A
40
Patient B
20
16
80
Anti-Fas antibody (ng/ml)
400
FIGURE 2
Fas-mediated killing takes place when Fas is cross-linked by FasL, its normal ligand, or by treatment with
anti-Fas antibody, which artificially cross-links Fas molecules. This experiment shows the percentage of
dead T cells after induction of apoptosis in T cells from patients and controls by cross-linking Fas with
increasing amounts of an anti-Fas monoclonal antibody. T cells from the Canale-Smith patients (A and
B) are resistant to Fas-mediated death. [Adapted from M. D. Drappa et al., 1996, New England Journal of Medicine
335:1643.]
A number of research groups have
conducted detailed clinical studies of CSS
patients, and the following general observations have been made:
•
The cell populations of the blood and
lymphoid tissues of CSS patients show
dramatic elevations (5-fold to as much
as 20-fold) in the numbers of lymphocytes of all sorts, including T cells, B
cells, and NK cells.
•
Most of the patients have elevated levels of one or more classes of immunoglobulin (hyper-gammaglobulinemia).
•
Immune hyperactivity is responsible
for such autoimmune phenomena as
the production of autoantibodies
against red blood cells, resulting in
hemolytic anemia, and a depression in
platelet counts due to the activity of
antiplatelet autoantibodies.
The signature member of this gene family, bcl-2, was discovered near the breakpoint of a chromosomal translocation
in a cancer known as human B-cell lymphoma. The translocation moved the bcl-2 gene into the immunoglobulin
heavy-chain locus, resulting in transcription of bcl-2 along
with the immunoglobulin gene, with consequent overproduction of the encoded Bcl-2 protein by the lymphoma cells.
The resulting high levels of Bcl-2 resulted in the accumula-
These observations establish the
importance of the death-mediated regulation of lymphocyte populations in
lymphocyte homeostasis. Such cell
death is necessary because the immune
response to antigen results in a sudden
and dramatic increase in the populations of responding clones of lymphocytes and temporarily distorts the
representation of these clones in the
repertoire. In the absence of cell death,
the periodic stimulation of lymphocytes
that occurs in the normal course of life
would result in progressively increasing,
and ultimately unsustainable, lymphocyte levels. As CSS demonstrates, without the essential culling of lymphocytes
by apoptosis, severe and life-threatening
disease can result.
tion of long-lived B cells and contributed to their transformation into cancerous lymphoma cells.
Bcl-2 levels play an important role in regulating the
normal life span of various hematopoietic cell lineages,
including lymphocytes. A normal adult has about 5 liters of
blood, with about 2000 lymphocytes/mm3, for a total of
about 1011 circulating lymphocytes. During acute infection, the lymphocyte count increases fourfold or more. The
324
PA R T I V
|
Adaptive Immunity: Development
immune system cannot sustain such a massive increase in
cell numbers for an extended period, so the system must
eliminate unneeded activated lymphocytes once the antigenic threat has passed. Activated lymphocytes have been
found to express lower levels of Bcl-2, and therefore are
more susceptible to the induction of apoptotic death than
are naïve lymphocytes or memory cells. However, if the
lymphocytes continue to be activated by antigen, signals
received during activation block the apoptotic signal. As
antigen levels subside, the levels of the signals that block
apoptosis diminish, and the lymphocytes begin to die by
apoptosis.
S U M M A R Y
■
■
■
■
■
■
■
■
■
■
Uncommitted white blood cell progenitors enter the thymus from the bone marrow. Notch/Notch ligand interactions are required for T-cell commitment.
Immature T cells are called thymocytes, and stages of
development can be defined broadly by the expression of
the coreceptors CD4 and CD8. The most immature thymocytes express neither coreceptor and are referred to as
CD4⫺CD8⫺ or double-negative (DN) cells. These progress
to the CD4⫹CD8⫹ (double positive or DP) stage, which in
turn mature to the CD4⫹CD8⫺ or CD4⫺CD8⫹ (single
positive or SP) stages.
DN thymocytes progress through four stages of development (DN1-DN4) defined by CD44, CD25, and c-Kit
expression. During these stages they proliferate and rearrange the TCR, ␦, and ␥ antigen receptor genes.
Thymocytes that rearrange ␦ and ␥ receptor genes successfully mature to the TCR␥␦ lineage.
Those thymocytes that successfully rearrange the  receptor chain are detected by a process called -selection.
-selection is initiated by assembly of the TCR protein
with a surrogate, invariant TCR␣ chain and CD3 complex
proteins. Assembly of this pre-TCR results in commitment to the TCR␣ lineage, another burst of proliferation,
maturation to the DP stage of development, and initiation
of TCR␣ rearrangement.
DP thymocytes are the most abundant subpopulation in
the thymus and are the first cells to express a mature
TCR␣ receptor. They are the targets of positive and
negative selection, which are responsible for self-tolerance
and self-MHC restriction, respectively.
Positive and negative selection are regulated by the affinity
of a DP thymocyte’s TCR for MHC/self-peptide complexes expressed by the thymic epithelium.
High-affinity TCR/MHC/peptide interactions result in
negative selection, typically by initiating apoptosis (clonal
deletion).
Low/intermediate-affinity signals result in positive selection and initiate a maturation program to the helper CD4⫹
or cytotoxic CD8⫹ single-positive lineages.
The large majority of thymocytes (95%) do not interact
with any MHC/self-peptides expressed by the thymic epithelium and die by neglect.
■
■
■
■
■
■
■
■
■
■
The distinction among TCR signaling cascades that activate positive versus negative selection is not fully understood but may involve differences in MAPK and Ca2⫹
signaling.
Positively selected DP thymocytes have to decide whether
to become helper CD4⫹ or cytotoxic CD8⫹ T cells. This
process, called lineage commitment, appears to depend on
continuity of the signals positively selected DP thymocytes receive through their TCR.
Thymocytes whose TCR preferentially interacts with
MHC Class II generate a continuous signal that initiates a
CD4⫹ helper T-cell developmental program.
Thymocytes whose TCR preferentially interacts with
MHC Class I cannot generate a continuous signal because
CD8 surface levels are down-regulated in response to
positive selection. The interruption in signaling, followed
by further stimulation, initiates a CD8 developmental
program.
Transcription factors Th-Pok and Runx3 play important roles
in CD4⫹ helper cell and CD8⫹ cytotoxic cell commitment.
They work in part by reciprocally regulating each other.
Positively selected thymocytes migrate from the thymic
cortex to the thymic medulla via interactions with the
CCR7 chemokine receptor. These cells, in transition from
the DP to the SP stage of development, are negatively
selected for tissue-specific antigens in the medulla. This is
the second opportunity to remove autoreactive T cells
from the developing repertoire.
The thymic cortex and thymic medulla carry out distinct
functions in the thymus. Medullary epithelial cells, but not
cortical epithelial cells, express the transcription factor
AIRE, which is responsible for their unique capacity to
express tissue-specific antigens.
Fully mature thymocytes exit the thymus via interactions
via the sphingosine-1-phosphate receptor (S1PR) and
undergo final functional maturation in peripheral lymphoid tissues.
A small percentage of cells also develop within the thymus
to other cell lineages, including NK T cells, IELs, and
regulatory T cells.
Regulatory T cells that develop in the thymus are called natural TREGs. They share the periphery with regulatory T cells
T-Cell Development
■
■
that develop from conventional T cells (induced TREGs) and
play an important role in inhibiting autoimmune responses.
Apoptosis, a process of cell death that is internally initiated and highly regulated, has a major influence on the
shape of the T-cell repertoire. Thymocytes that do not pass
-selection (and do not successfully express a TCR␥␦
complex) die by apoptosis; 95% of developing DP thymocytes also die by apoptosis, the vast majority because they
do not receive sufficient TCR signaling (death by neglect)
and a smaller population because they receive too high a
level of TCR signaling (negative selection).
The process by which a cell undergoes apoptosis signaling
is evolutionarily conserved and inspired by many different receptor interactions in most if not all cell types in the
body.
■
■
■
|
CHAPTER 9
325
Apoptotic stimuli either activate death receptors (extrinsic
pathway) or mitochondrial cytochrome c release (intrinsic
pathway). Both pathways ultimately activate effector caspases, which finalize cell death events.
Apoptotic cells are rapidly engulfed by neighboring
phagocytes and, for this reason, are difficult to detect in
living tissues.
Bcl-2 family members include both anti-apoptotic and
pro-apoptotic family members that reside in the mitochondrial membrane and regulate cytochrome c release.
A third set of family members, the pro-apoptotic BH3
proteins, can be activated by apoptotic stimuli, favoring a shift in balance of Bcl-2 family member activity
at the mitochondrial membrane toward cytochrome c
release.
R E F E R E N C E S
Alam, S., et al. 1999. Qualitative and quantitative differences in
T cell receptor binding of agonist and antagonist ligands.
Immunity 10:227–237.
Germain, R. 2008. Special regulatory T-cell review: A rose by
any other name: from suppressor T cells to TREGs, approbation
to unbridled enthusiasm. Immunology 123:20–27.
Anderson, M., et al. 2002. Projection of an immunological self
shadow within the thymus by the aire protein. Science
298:1395–1401.
He, X., et al. 2005. The zinc finger transcription factor Th-POK
regulates CD4 versus CD8 T-cell lineage commitment. Nature
433:826–833.
Baldwin, T., K. Hogquist, and S. Jameson. 2004. The fourth way?
Harnessing aggressive tendencies in the thymus. Journal of
Immunology 173:6515–6520.
Hogquist, K., T. Baldwin, and S. Jameson. 2005. Central tolerance: Learning self-control in the thymus. Nature Reviews.
Immunology 5:772–782.
Baldwin, T., M. Sandau, S. Jameson, and K. Hogquist. 2005. The
timing of TCR alpha expression critically influences T cell
development and selection. Journal of Experimental Medicine
202:111–121.
Hogquist, K., M. Gavin, and M. Bevan. 1993. Positive selection
of CD8⫹ T cells induced by major histocompatibility complex
binding peptides in fetal thymic organ culture. Journal of
Experimental Medicine 177:1469–1473.
Bouillet, P., et al. 2002. BH3-only Bcl-2 family member Bim is
required for apoptosis of autoreactive thymocytes. Nature
415:922–926.
Hogquist, K., S. Jameson, and M. Bevan. 1995. Strong agonist
ligands for the T cell receptor do not mediate positive selection of functional CD8⫹ T cells. Immunity 3:79-86.
Carpenter, A., and R. Bosselut. 2010. Decision checkpoints in
the thymus. Nature Immunology 11:666–673.
Hogquist, K., et al. 1994. T cell receptor antagonist peptides
induce positive selection. Cell 76:17–27.
Caton, A., et al. 2004. CD4(⫹) CD25(⫹) regulatory T cell
selection. Annals of the New York Academy of Sciences
1029:101–114.
Kappes, D., and X. He. 2006. ole of the transcription factor ThPOK in CD4:CD8 lineage commitment. Immunological
Reviews 209:237–252.
Ceredig, R., and T. Rolink. 2002. A positive look at doublenegative thymocytes. Nature Reviews Immunology 2:888–897.
Drennan, M., D. Elewaut, and K. Hogquist. 2009. Thymic emigration: sphingosine-1-phosphate receptor-1-dependent
models and beyond. European Journal of Immunology
39:925–930.
Kisielow, P., H. Blüthmann, U. Staerz, M. Steinmetz, and H. von
Boehmer, H. 1988. Tolerance in T-cell-receptor transgenic
mice involves deletion of nonmature CD4⫹8⫹ thymocytes.
Nature 333:742–746.
Gascoigne, N. 2010. CD8⫹ thymocyte differentiation: T cell
two-step. Nature Immunology 11:189–190.
Klein, L., M. Hinterberger, G. Wirnsberger, and B. Kyewski.
2009. Antigen presentation in the thymus for positive selection and central tolerance induction. Nature Reviews. Immunology 9:833–844.
Germain, R. 2002. T-cell development and the CD4-CD8 lineage decision. Nature Reviews. Immunology 2:309–322.
Kyewski, B., and P. Peterson. 2010. Aire: Master of many trades.
Cell 140:24–26.
326
PA R T I V
|
Adaptive Immunity: Development
Li, R., and D. Page. 2001. Requirement for a complex array of
costimulators in the negative selection of autoreactive thymocytes in vivo. Journal of Immunology 166:6050–6056.
Uematsu, Y., et al. 1988. In transgenic mice the introduced
functional T cell receptor beta gene prevents expression of
endogenous beta genes. Cell 52:831–841.
Marrack, P., L. Ignatowicz, J. Kappler, J. Boymel, and J. Freed.
1993. Comparison of peptides bound to spleen and thymus
class II. Journal of Experimental Medicine 178:2173–2183.
Venanzi, E., C. Benoist, and D. Mathis. 2004. Good riddance:
Thymocyte clonal deletion prevents autoimmunity. Current
Opinion in Immunology 16:197–202.
Marrack, P., J. McCormack, and J. Kappler. 1989. Presentation
of antigen, foreign major histocompatibility complex proteins
and self by thymus cortical epithelium. Nature 338:503–505.
von Boehmer, H., H. Teh, and P. Kisielow. 1989. The thymus
selects the useful, neglects the useless and destroys the harmful. Immunology Today 10:57–61.
Mathis, D., and C. Benoist. 2009. Aire. Annual Review of Immunology 27:287–312.
Wang, L., and R. Bosselut. 2009. CD4-CD8 lineage differentiation: Thpok-ing into the nucleus. Journal of Immunology
183:2903–2910.
McNeil, L., T. Starr, and K. Hogquist. 2005. A requirement for
sustained ERK signaling during thymocyte positive selection
in vivo. Proceedings of the National Academy of Sciences of the
United States of America 102:13,574–13,579.
Mohan, J., et al. 2010. Unique autoreactive T cells recognize
insulin peptides generated within the islets of Langerhans in
autoimmune diabetes. Nature Immunology 11:350–354.
Page, D., L. Kane, J. Allison, and S. Hedrick. 1993. Two signals
are required for negative selection of CD4⫹CD8⫹ thymocytes. Journal of Immunology 151:1868–1880.
Park, J., et al. 2010. Signaling by intrathymic cytokines, not T
cell antigen receptors, specifies CD8 lineage choice and promotes the differentiation of cytotoxic-lineage T cells. Nature
Immunology 11:257–264.
Pui, J., et al. 1999. Notch1 expression in early lymphopoiesis influences B versus T lineage determination. Immunity 11:299–308.
Punt, J., B. Osborne, Y. Takahama, S. Sharrow, and A. Singer.
1994. Negative selection of CD4⫹CD8⫹ thymocytes by T cell
receptor-induced apoptosis requires a costimulatory signal
that can be provided by CD28. Journal of Experimental Medicine 179:709–713.
Wang, L., et al. 2008. Distinct functions for the transcription
factors GATA-3 and ThPOK during intrathymic differentiation of CD4(⫹) T cells. Nature Immunology 9:1122–1130.
Wilson, A., H. MacDonald, and F. Radtke. 2001. Notch
1-deficient common lymphoid precursors adopt a B cell
fate in the thymus. Journal of Experimental Medicine
194:1003–1012.
Zúñiga-Pflücker, J. 2004. T-cell development made simple.
Nature Reviews. Immunology 4:67–72.
Useful Web Sites
http://www.bio.davidson.edu/courses/immunology/
Flash/Main.htm A Dr. Victor Lemas–generated animation of positive and negative selection events in the thymus,
used at Davidson College.
http://www.bio.davidson.edu/courses/movies.
html A full list of animations assembled by and in many
Sambandam, A., et al. 2005. Notch signaling controls the generation and differentiation of early T lineage progenitors.
Nature Immunology 6:663–670.
cases generated by individuals associated with Davidson
College.
Schmitt, T., and J. Zúñiga-Pflücker. 2002. Induction of T cell
development from hematopoietic progenitor cells by deltalike-1 in vitro. Immunity 17:749–756.
http://bmc.med.utoronto.ca/student_video_
gallery.php?titleimages/stories/videos/Janice_
Yau&h360&w640&m4v1 Another animation of
Singer, A. 2010. Molecular and cellular basis of T cell lineage
commitment: An overview. Seminars in Immunology 22:253.
T-cell development generated as part of a master’s project by
Janice Yau. This is from an impressive site that features multiple video projects generated by master’s students in the
University of Toronto at Mississauga’s Biomedical Communications Graduate Program.
Stadinski, B., et al. 2010. Diabetogenic T cells recognize insulin
bound to IAg7 in an unexpected, weakly binding register.
Proceedings of the National Academy of Sciences of the United
States of America 107:10,978–10,983.
Starr, T., S. Jameson, and K. Hogquist. 2003. Positive and negative selection of T cells. Annual Review of Immunology
21:139–176.
http://bio-alive.com/categories/apoptosis.htm
Videos of cells undergoing apoptosis.
www.celldeath.de/encyclo/aporev/aporev.htm
Sun, G., et al. 2005. The zinc finger protein cKrox directs CD4
lineage differentiation during intrathymic T cell positive selection. Nature Immunology 6:373–381.
Detailed online review of apoptosis biology.
Teh, H., et al. 1988. Thymic major histocompatibility complex
antigens and the alpha beta T-cell receptor determine the
CD4/CD8 phenotype of T cells. Nature 335:229–233.
Access to original Kerr, Wyllie, and Currie British Journal
of Cancer 1972 article where the term apoptosis was
coined.
www.ncbi.nlm.nih.gov/pmc/articles/PMC2008650
T-Cell Development
CHAPTER 9
327
Q U E S T I O N S
2. Susceptibility to many autoimmune diseases has been linked
to MHC gene variants. One of the best examples of such a
linkage is provided by multiple sclerosis (MS), a human autoimmune disease caused by autoreactive T cells whose activity
ultimately damages the myelin sheaths around neurons. Susceptibility to MS has been consistently associated with variants in the HLA-DR2 gene. Although this link was first
recognized in 1972, we still don’t fully understand the basis
for this susceptibility. One perspective on the reasons for the
link between MHC variations and autoimmune disease was
offered in a recent review article. The authors state, “The
mechanisms underlying MHC association in autoimmune
disease are not clearly understood. One long-held view suggests a breakdown in immunological tolerance to self-antigens through aberrant class II presentation of self or foreign
peptides to autoreactive T lymphocytes. Thus, it seems likely
that specific MHC class II alleles determine the targeting of
particular autoantigens resulting in disease-specific associations.” (Fernando, M. M. A., et al. 2008 Defining the role of
the MHC in autoimmunity: A review and pooled analysis.
PLoS Genet 4:4: e1000024. doi:10.1371/journal.pgen.1000024.)
a. Paraphrase this perspective using your own words.
What, specifically, might the authors mean by “aberrant
class II presentation . . . to autoreactive T lymphocytes”?
b. (Advanced question.) Although this speculation has some
merit, it does not resolve all questions. Why? Pose one
question that this explanation inspires or does not answer.
c. (Very advanced question.) Offer one addition to the
explanation (or an alternative) that helps resolve the
question you posed above.
3. Over a period of several years, a group of children and ado-
lescents are regularly dosed with compound X, a lifesaving
drug. However, in addition to its beneficial effects, it was
found that this drug interferes with Fas-mediated signaling.
a. What clinical manifestations of this side effect of compound X might be seen in these patients?
b. If white blood cells from an affected patient are stained
with a fluorescein-labeled anti-CD4 and a phycoerythrin-labeled anti-CD8 antibody, what might be seen in
the flow-cytometric analysis of these cells? What pattern would be expected if the same procedure were
performed on white blood cells from a healthy control?
STUDY QUESTIONS
1. Each of the following statements is false. Correct them
(and explain your correction[s]).
Knockout mice lacking class I MHC molecules fail to produce CD4⫹ mature thymocytes.
-selection initiates negative selection.
2. Whereas the majority of T cells in our bodies express an
␣ TCR, up to 5% of T cells express the ␥␦ TCR instead.
Explain the difference in antigen recognition between
these two cell types.
3. You have fluorescein-labeled anti-CD4 and phycoerythrin-
labeled anti-CD8. You use these antibodies to stain thymocytes and lymph-node cells from normal mice and from
RAG-1 knockout mice. In the forms below, draw the FACS
plots that you would expect.
Thymus
Normal mice
RAG-1 knockout mice
CD4
basis and has been linked to many different gene loci. Identify
three possible genes other than Fas and AIRE (whose connection to autoimmunity have been explicitly described in
this chapter) that could be involved in an increased susceptibility to autoimmune disease. Explain your reasoning.
CD8
Normal mice
CD8
Lymph node
RAG-1 knockout mice
CD4
1. The susceptibility to autoimmune diseases often has a genetic
Negative selection to tissue-specific antigens occurs in the
cortex of the thymus.
Most thymocytes die in the thymus because they are autoreactive.
The extrinsic pathway of apoptosis never activates the
intrinsic pathway.
Cytochrome c is an important downstream molecule in the
extrinsic apoptotic pathway.
Bcl-2 enhances the activity of Bax and therefore inhibits
apoptosis.
CD4
CLINICAL FOCUS QUESTIONS
CD4
S T U D Y
|
CD8
CD8
4. What stages of T-cell development (DN1, DN2, DN3, DN4,
DP, CD4 SP, or CD8 SP) would be affected in mice with the
following genetic modifications? Justify your answers.
a. Mice that do not express MHC Class II.
b. Mice that do not express AIRE.
c. Mice that do not express the TCR␣ chain.
5. You stain thymocytes with PE conjugated anti-CD3 and
FITC conjugated anti-TCR. Most cells stain with both.
However, you find a proportion of cells that stain with
neither antibody. You also find a small population that
stain with anti-CD3, but not with anti-TCR. What thymocyte populations might each of these populations
represent?
328
PA R T I V
|
Adaptive Immunity: Development
k
6. You immunize an H-2 mouse with chicken ovalbumin (a
protein against which the mouse will generate an immune
response) and isolate a CD4⫹ mature T cell specific for an
ovalbumin peptide. You clone the ␣ TCR genes from this
cell line and use them to prepare transgenic mice with the
H-2k or H-2d haplotype.
a. What approach can you use to distinguish immature
thymocytes from mature CD4⫹ thymocytes in the
transgenic mice?
b. Would thymocytes from a TCR transgenic mouse of the
H-2k background have a proportion of CD4⫹ thymocytes
that is higher, lower, or the same as a wild-type mouse?
c. Would thymocytes from a TCR transgenic mouse of the
H-2d background have a proportion of CD4⫹ thymocytes
that is higher, lower, or the same as a wild-type mouse?
Speculate and explain your answer.
d. You find a way to “make” the medullary epithelium of
an H-2k TCR transgenic mouse express and present the
ovalbumin peptide for which your T cell is specific.
What would the CD4 versus CD8 profile of a TCR
transgenic thymus look like in these mice?
You also find a way to “make” the cortical epithelium
express this ovalbumin peptide in its MHC Class II dimer.
What might the CD4 versus CD8 profile of this TCR transgenic thymus look like?
7. In his classic chimeric mouse experiments, Zinkernagel
took bone marrow from a mouse of one MHC haplotype
(mouse 1) and a thymus from a mouse of another MHC
haplotype (mouse 2) and transplanted them into a third
mouse, which was thymectomized and lethally irradiated.
He then immunized this reconstituted mouse with the
lymphocytic choriomeningitis virus (LCMV) and examined the activity of the mature T cells isolated from the
spleen and lymph nodes of the mouse.
He was specifically interested to see if the mature CD8⫹ T
cells in these mice could kill target cells infected with LCMV
with the MHC haplotype of mouse 1, 2, or 3. The results of
two such experiments using H-2b strain C57BL/6 mice and
H-2d strain BALB/c mice as bone marrow and thymus
donors, respectively, are shown in the following table.
Lysis of LCMVinfected target
cells
Bone
Thymectomized
marrow Thymus x-irradiated
Experiment donor
donor
recipient
H-2d H-2k H-2b
A
C57BL/6 BALB/
c (H-2d)
(H-2b)
C57BL/6 ⫻
BALB/c
⫹
⫺
⫺
B
BALB/c
(H-2d)
C57BL/6 C57BL/6 ⫻
(H-2b)
BALB/c
⫺
⫺
⫹
b
a. Why were the H-2 target cells not lysed in experiment
A but lysed in experiment B?
k
b. Why were the H-2 target cells not lysed in either
experiment?
⫹
k
8. You have a CD8 CTL clone (from an H-2 mouse) that has
a T-cell receptor specific for the H-Y antigen. You clone the
␣ TCR genes from this cloned cell line and use them to
prepare transgenic mice with the H-2k or H-2d haplotype.
a. How can you distinguish the immature thymocytes from
the mature CD8⫹ thymocytes in the transgenic mice?
b. For each of the following transgenic mice, indicate with
(1) or (2) whether the mouse would have immature
double-positive and mature CD8⫹ thymocytes bearing
the transgenic T-cell receptor: H-2k female, H-2k male,
H-2d female, H-2d male.
k
c. Explain your answers for the H-2 transgenics.
d
d. Explain your answers for the H-2 transgenics.
9. To demonstrate positive thymic selection experimentally,
researchers analyzed the thymocytes from normal H-2b mice,
which have a deletion of the class II IE gene, and from H-2b
mice in which the class II IA gene had been knocked out.
a. What MHC molecules would you find on antigenpresenting cells from the normal H-2b mice?
b. What MHC molecules would you find on antigenpresenting cells from the IA knockout H-2b mice?
⫹
⫹
c. Would you expect to find CD4 T cells, CD8 T cells,
or both in each type of mouse? Why?
10. You wish to determine the percentage of various types of
thymocytes in a sample of cells from mouse thymus using
the indirect immunofluorescence method.
a. You first stain the sample with goat anti-CD3 (primary
antibody) and then with rabbit FITC-labeled antigoat Ig
(secondary antibody), which emits a green color. Analysis
of the stained sample by flow cytometry indicates that 70%
of the cells are stained. Based on this result, how many of
the thymus cells in your sample are expressing antigenbinding receptors on their surface? Would all be expressing the same type of receptor? Explain your answer. What
are the remaining unstained cells likely to be?
⫹
b. You then separate the CD3 cells with the fluorescenceactivated cell sorter (FACS) and restain them. In this
case, the primary antibody is hamster anti-CD4, and the
secondary antibody is rabbit PE-labeled antihamster Ig,
which emits a red color. Analysis of the stained CD3⫹
cells shows that 80% of them are stained. From this
result, can you determine how many TC cells are present
in this sample? If yes, then how many TC cells are there?
If no, what additional experiment would you perform to
determine the number of TC cells that are present?
10
B-Cell Development
M
illions of B lymphocytes are generated in the
bone marrow every day and exported to the
periphery. The rapid and unceasing
generation of new B cells occurs in a carefully
regulated sequence of events. Cell transfer experiments, in
which genetically marked donor hematopoietic stem cells
(HSCs) are injected into an unmarked recipient, have
indicated that B-cell development from HSC to mature B
cell takes from 1 to 2 weeks; donor-derived mature B cells
can be detected in the recipient by 2 weeks following
transfer of HSCs into recipient mice.
B-cell development begins in the bone marrow with the
asymmetric division of an HSC and continues through a
series of progressively more differentiated progenitor stages
to the production of common lymphoid progenitors
(CLPs), which can give rise to either B cells or T cells (see
Overview Figure 10-1). Progenitor cells destined to
become T cells migrate to the thymus where they complete
their maturation (see Chapter 9); the majority of those that
remain in the bone marrow become B cells. As
differentiation proceeds, the developing B cell expresses on
its cell surface a precisely calibrated sequence of cell-surface
receptor and adhesion molecules. Some of the signals
received from these receptors induce the differentiation of
the developing B cell; others trigger its proliferation at
particular stages of development and yet others direct its
movements within the bone marrow environment. These
signals collectively allow differentiation of the CLP through
the early B-cell stages to form the immature B cell that
leaves the marrow to complete its differentiation in the
spleen. For the investigator, the expression of different cellsurface molecules at each stage of B-cell maturation
provides an invaluable experimental tool with which to
recognize and isolate B cells poised at discrete points in
their development.
The primary function of mature B cells is to secrete
antibodies that protect the host against pathogens, and
so one major focus of those studying B-cell
differentiation is the analysis of the timing and order of
rearrangement and expression of immunoglobulin receptor
heavy- and light-chain genes. Recall from Chapter 7 that
immunoglobulin gene rearrangements begin with
B cells at different stages of development
seek contact with stromal cells expressing
CXCL12 (pre-pro-B cells, left) or IL-7 (pro-B
cells, right). [Tokoyoda et al. 2004. Cellular Niches
Controlling B Lymphocyte Behavior within Bone
Marrow during Development. Immunity Vol. 20, Issue
6, 707–718. © 2004 Elsevier Ltd. All rights reserved.]
■
The Site of Hematopoiesis
■
B-Cell Development in the Bone Marrow
■
■
The Development of B-1 and Marginal-Zone
B Cells
Comparison of B- and T-Cell Development
heavy-chain D to JH gene-segment rearrangement,
followed by the stitching together of the heavy-chain
VH and DJH segments. These rearrangements culminate
in the cell-surface expression of the pre-B-cell receptor
during the pre-B-cell stage, in which the rearranged
heavy chain is expressed in combination with the
surrogate light chain. Rearrangement of the light chain
is initiated after several rounds of division of cells
bearing the pre-BCR.
Like T cells, developing B cells must solve the
problem of creating a repertoire of receptors capable of
recognizing an extensive array of antigens, while
ensuring that self-reactive B cells are either eliminated
by apoptosis or rendered functionally unreactive or
anergic. However, unlike T-cell receptors, B-cell
receptors are not constrained by the need to be MHC
restricted. Again, unlike T-cell maturation, B-cell
development is almost complete by the time the B cell
leaves the bone marrow; in mammalian systems there is
no thymic equivalent in which B-cell development is
accomplished. Instead, immature B cells are released to
329
330
PA R T I V
|
Adaptive Immunity: Development
10-1
OVERVIEW FIGURE
B-Cell Development Begins in the Bone Marrow and Is
Completed in the Periphery
Early ProB
ProB or preB
• DJ H chain
• Complete
recombination
VDJ H chain
• Start of VDJ H chain recombination
recombination
• Clonal expansion
• VJ L chain
recombination
Immature B
Negative selection
• Deletion
• Receptor
editing
Endogenous
antigen
Bone
endosteum
Central
sinus
Bone marrow
Transitional-2
Mature B
IgM
IgD
Transitional-1
Spleen
B-cell development begins with a hematopoietic stem cell (HSC)
and passes through progressively more delimited progenitor-cell
stages until it reaches the pro-B cell stage. At this stage, the precursor cell is irreversibly committed to the B-cell lineage and the
the periphery, where they complete their developmental
program in the spleen.
In this chapter, we will follow B-cell development from
its earliest stages in the primary lymphoid organs to the
generation of fully mature B cells in the secondary
lymphoid tissues. As for T cells, multiple B-cell subsets
exist, and we will briefly address how the process of
differentiation of the minority B-1 and marginal-zone
(MZ) B-cell subsets differs from the developmental
program followed by the predominant B-2 B-cell subset.
We will conclude with a brief comparison of the
maturational processes of T and B lymphocytes.
The Site of Hematopoiesis
In adult animals, hematopoiesis, the generation of blood
cells, occurs in the bone marrow; the HSCs in the marrow
are the source of all blood cells of the erythroid, myeloid, and
recombination of the immunoglobulin genes begins. Once the
completed immunoglobulin is expressed on the cell surface, the
immature B cell, now a transitional B cell, leaves the bone marrow
to complete its maturation in the spleen.
lymphoid lineages (Chapter 2). Various non-hematopoietic
cells in the bone marrow express cell-surface molecules
and secrete hormones that guide hematopoietic cell development. Developing lymphocytes move within the bone marrow as they mature, thus interacting with different
populations of cells and signals at various developmental
stages. However, fetal animals face particular challenges to
their developing immune systems; how can they generate
blood cells when their bones are still not yet fully developed?
The Site of B-Cell Generation Changes
during Gestation
Hematopoiesis is a complex process in the adult animal, and
during fetal maturation additional challenges must be met.
Red blood cells must be quickly generated de novo in order to
provide the embryo with sufficient oxygen, and HSCs must
proliferate at a rate sufficient to populate the adult as well as
provide for the hematopoietic needs of the maturing fetus.
Furthermore, since the bone marrow appears relatively late in
B-Cell Development
Mouse
|
CHAPTER 10
331
Human
Placenta
Yolk sac
Fetal liver
AGM
Specification
Mesoderm
Emergence
Maturation
Pre-HSC
Expansion
HSC
Quiescence/
self-renewal
HSC
HSC
Mouse
7.5
10.5
12.5
15.5 days
21
28
40
70 days
Birth
Human
Birth
Placenta
Fetal liver
AGM
Primitive
streak
Bone marrow
Yolk sac
FIGURE 10-2 The anatomy and timing of the earliest
become blood cells emerge first as rapidly proliferating pre-HSCs
and eventually mature into relatively quiescent hematopoietic stem
cells that populate the bone marrow. The colored bars in the timeline illustrate the ages at which the various murine and human
hematopoietic sites are active. Mesoderm (gray); generation of fetal
HSC (yellow); active hematopoietic differentiation (red); emergence
of functional, adult-type HSCs (blue). [Adapted from H. K. Mikkola and S. H.
stages in hematopoiesis. (a) Blood-cell precursors are initially
found in the yolk sac (yellow), then spread to the placenta fetal liver
(pink), and aorta-gonad-mesonephros (AGM) region (green), before
finding their adult home in the bone marrow. The mouse embryo is
shown at 11 days of gestation; the human embryo at the equivalent
5 weeks of gestation. (b) In the embryo, cells within the primitive
streak mesodermal tissue adopt either hematopoietic (blood-cell
forming) or vascular (blood-vessel forming) fates. Those destined to
Orkin, 2006, The journey of developing hematopoietic stem cells, Development
133:3733–3744, Figures 1 and 2. ]
development, the whole process of blood-cell generation must
shift location several times before moving into its final home.
The gestation period for mice is 19 to 21 days. Hematopoiesis begins, in the mouse, around 7 days post fertilization
(Figure 10-2) when precursor cells in the yolk sac begin differentiating to form primitive, nucleated, erythroid cells that
carry the oxygen the embryo needs for early development.
Fetal HSCs capable of generating all blood-cell types can be
detected in the early aorta-gonad-mesonephros (AGM) region
on day 8, when the fetal heart starts beating. On day 10, mature
HSCs capable of completely repopulating the hematopoietic
system of irradiated adult mice can be isolated from the AGM,
and by day 11 they can be found in the yolk sac, placenta, and
fetal liver. Between days 11.5 and 12.5, there is rapid expansion
of the placental HSC pool, at the end of which time the placenta holds more HSCs than either the AGM or the yolk sac.
By day 13.5, the number of HSCs in the placenta begins to
decrease while the HSC pool in the liver continues to expand.
As the mouse embryo completes its development, the
predominant site of HSC generation remains within the fetal
liver, but some hematopoiesis can be detected in the spleen
in the perinatal (around the time of birth) period. The number of fetal liver HSCs in mice reaches a maximum of
approximately 1000 by days 15.5 to 16.5 of embryonic
development, after which it starts to decline. Within the
fetal liver, HSCs differentiate to form progenitor cells. At the
332
PA R T I V
|
Adaptive Immunity: Development
earliest time points, hematopoiesis in the fetal liver is
dominated by erythroid progenitors that give rise to the
true, enucleated mature erythrocytes, in order to ensure a
steady oxygen supply to the growing embryo, and myeloid
and lymphoid progenitors gradually emerge. Pre-B cells
(precursor-B cells), defined as cells that express immunoglobulin in their cytoplasm but not on their surfaces, are
first observed at day 13 of gestation, and surface IgM-positive
B cells are present in detectable numbers by day 17. HSCs
first seed the bone marrow at approximately day 15, and
over a period of a few weeks, the bone marrow takes over as
the main site of B-cell development, remaining so throughout post-natal life.
Hematopoiesis in the Fetal Liver Differs from
That in the Adult Bone Marrow
Developing B cells in the fetal liver differ in important
ways from their counterparts in adult bone marrow. The
liver is the primary site of B-cell generation in the fetus,
and provides the neonatal animal with the cells it needs to
populate its nascent immune system. In order to accomplish this, hematopoietic stem cells and their progeny
must undergo a phase of rapid proliferation, and fetal liver
HSCs, as well as their daughter cells, undergo several
rounds of cell division over a short time. In contrast, HSCs
derived from the bone marrow of a healthy adult animal
are relatively quiescent.
B cells generated from fetal liver precursors are predominantly B-1 B cells, which will be described more fully in
Chapter 12. Briefly, B-1 B cells are primarily located in the
body (specifically the peritoneal and pleural) cavities. They
are therefore well-positioned to protect the gut and the lungs,
which are the major ports of entry of microbes in the fetus
and neonate. Antibodies secreted by B-1 B cells are broadly
cross-reactive; many bind to carbohydrate antigens expressed
by a number of microbial species. Since terminal deoxynucleotidyl transferase (TdT) is minimally expressed at this point
in ontogeny, and the RAG1/2 recombinase proteins appear
not to use the full range of V, D, and J region gene segments
at this stage in embryonic development, the immunoglobulin
receptors of B-1 B cells express minimal receptor diversity. In
expressing an oligoclonal (few, as opposed to many, clones)
repertoire of B-cell receptors that bind to a limited number of
carbohydrate antigens shared among many microbes, B-1 B
cells occupy a functional niche that bridges the innate and
adaptive immune systems. We will describe B-1 B-cell development further at the end of this chapter.
Over a period of 2 to 4 weeks after birth, the process of
hematopoiesis in mice shifts from the fetal liver and spleen to
the bone marrow, where it continues throughout adulthood.
The B-1 B-cell population represents an exception to this general rule, as it is self-renewing in the periphery. This means
that new daughter B-1 B cells are generated continually from
preexisting B-1 B cells in the peritoneal and pleural cavities,
and in those other parts of the body in which B-1 B cells
reside. These daughter cells use the same receptors as their
parents, and no new V(D)J recombinase activity is required.
In humans, the sequence of events is similar to that
described for the mouse, but the time frame is obviously
somewhat elongated. Blood-cell precursors first appear in the
yolk sac in the third week of embryonic development, but
these cells, like their analogues in the mouse provide primarily erythroid progenitors and are not capable of generating all
subsets of blood cells. The first cells capable of entirely repopulating an adult human hematopoietic system arise in the
AGM region of the embryo and/or the yolk sac. By the third
month of pregnancy, these HSCs migrate to the fetal liver,
which then becomes responsible for the majority of hematopoiesis in the fetus. By the fourth month of pregnancy, HSCs
migrate to the bone marrow, which gradually assumes the
hematopoietic role from the fetal liver until, by the time of
birth, it is the primary generative organ for blood cells. Prior
to puberty in humans, most of the bones of the skeleton are
hematopoietically active, but by the age of 18 years only the
vertebrae, ribs, sternum, skull, pelvis, and parts of the
humerus and femur retain hematopoietic potential.
Just as B-cell development in the fetus and neonate differs
from that in the adult, so does B-cell hematopoiesis in the
aging animal. Clinical Focus Box 10-1 describes some
aspects of B-cell development that alter as humans age.
B-Cell Development in the
Bone Marrow
In Chapter 2 (Figure 2-5), we presented the structure of bone
and bone marrow. The bone marrow microenvironment is a
complex, three-dimensional structure with distinctive cellular niches which are specialized to influence the development of the cell populations that mature there. A dense
network of fenestrated (leaky) thin-walled blood vessels—
the bone marrow sinusoids—permeates the marrow, allowing the passage of newly formed blood cells to the periphery
and facilitating blood circulation through the marrow.
In addition to serving as a source of hematopoietic stem
cells, bone marrow also contains stem cells that can differentiate into adipocytes (fat cells), chondrocytes (cartilage
cells), osteocytes (bone cells), myocytes (muscle cells), and
potentially other types of cells as well. Each of these different
classes of stem cells requires specific sets of factors, secreted
by particular bone marrow stromal cells to enable their
proper differentiation.
What are bone marrow stromal cells? The term stroma
derives from the Greek for mattress, and a stromal cell is a
general term that describes a large adherent cell that supports the growth of other cells. During B-cell development,
bone marrow stromal cells fulfill two functions. First, by
interacting with adhesion molecules on the surfaces of HSCs
and progenitor cells, stromal cells retain the developing cell
populations in the specific bone marrow niches where they
can receive the appropriate molecular signals required for
B-Cell Development
|
CHAPTER 10
333
Box 10-1
CLINICAL FOCUS
B-Cell Development in the Aging Individual
People of retirement age and older
represent a greater segment of the population than they used to, and these older
individuals expect to remain active and
productive members of society. However,
physicians and immunologists have long
known that the elderly are more susceptible to infection than are young men and
women, and that vaccinations are less
effective in older individuals. In this feature, we explore the differences in B-cell
development between younger and
older vertebrates, which may account for
some of these immunological disparities
between adult and older individuals.
Aging individuals display deficiencies
in many aspects of B-cell function, including a poor antibody response to vaccination, inefficient generation of memory B
cells, and an increase in the expression of
autoimmune disorders. Does this reflect
defective functioning only in the mature
antigen-responsive B-cell population, or
does it result from problems manifested
during earlier stages in B-cell development? Current research demonstrates
that aging individuals display a range of
shortcomings in developing B cells.
Experiments employing reciprocal
bone marrow chimeras—in which aging
HSCs were transplanted into young recipients or HSCs from young mice were
injected into aging recipients—have
shown that the suboptimal process of
B-cell development in aging individuals
results from deficiencies in both the aging
stem cells and in the supporting stromal
cells. For example, bone marrow stromal
cells from aging mice secrete lower levels
of IL-7 than do stromal cells from younger
animals, suggesting an environmental
defect in the aging bone marrow. However, study of isolated, aging B-cell progenitors reveals that they also respond
less efficiently to IL-7 than do B cells from
younger mice, and so the IL-7 response in
aging individuals is affected at both the
secretory and recipient-cell levels.
Indeed, the problems encountered
by developing B cells from aging indi-
viduals start at the very beginning of
their developmental program. The epigenetic regulation of HSC genes in aging
mice is compromised, resulting in diminished levels of HSC self-renewal. Furthermore, the balance between the
production of myeloid versus lymphoid
progenitors is shifted in older individuals,
with down-regulation of genes associated with lymphoid specification and a
correspondingly enhanced expression of
genes specifying myeloid development.
The net effect of these changes in the
HSC population is a reduction with age
in the numbers of early B-cell progenitors, which is reflected in a decrease in
the numbers of pro- and pre-B-cell precursors at all stages of development.
Detailed studies of the expression of
particular genes important in B-cell development demonstrate that the expression
of important transcription factors, such as
those encoded by the E2A gene, is
reduced in older animals. Furthermore,
the Rag genes, as well as the gene encoding the surrogate light-chain component,
5, are down-regulated in older animals
compared with young adults, resulting in
a reduction in the bone marrow output of
immature B cells.
Multiple mechanisms therefore help
to explain why the numbers of B cells
released from the bone marrow are
smaller in aging than in younger individuals. But is the antigen recognition capacity—
the quality—as well as the quantity of B
cells different between the two populations? In particular, do B cells from aging
mice express a repertoire of receptors
similar to those obtained from younger
animals? The answer to this question has
come from the development of techniques that enable a global assessment of
repertoire diversity. Study of the sizes and
sequences of CDR3 regions from large
numbers of human B cells suggests that
in aging individuals the size of the repertoire (the number of different B-cell
receptors an individual expresses) is drastically diminished, and that this decrease
in repertoire diversity correlates with a
reduction in the health of the aging
patient.
The mechanisms for this age-related
repertoire truncation appear to be complex. A decrease in output of immature B2
cells from the bone marrow could provide
the opportunity for B-1 B cells to increase
their share of the peripheral B-cell niche,
and as is well appreciated, B-1 B cells have
a less diverse receptor repertoire than do
B-2 B cells. A lifetime of generating memory cells may also result in an individual
having less room in B-cell follicles for
newly formed B cells to enter, and a
decreased concentration of the homeostatic regulatory cytokines may make it
more difficult for primary B cells to compete with their more robust memory
counterparts. Clearly, this is an area of
increasing clinical interest as the average
age of the population of the developed
world continues to increase.
REFERENCES
Cancro, M. P., et al. 2009. B cells and aging: Molecules and mechanisms. Trends in Immunology 30:313–318.
Dorshkin, K., E. Montecino-Rodriguez, and R. A.
Signer. 2009. The ageing immune system: Is
it ever too old to become young again?
Nature Reviews Immunology 9:57–62.
Goodnow, C. C. 1992. Transgenic mice and
analysis of B-cell tolerance. Annual Reviews
Immunology 10:489–518.
Labrie, J. E., 3rd, A. P. Sah, D. M. Allman, M. P.
Cancro, and R. M. Gerstein. 2004. Bone marrow microenvironmental changes underlie
reduced RAG-mediated recombination and
B cell generation in aged mice. Journal of
Experimental Medicine 200:411–423.
Nemazee, D. A., and K. Bürki. 1989, February.
Clonal deletion of B lymphocytes in a
transgenic mouse bearing anti-MHC class I
antibody genes. Nature 337:562–566.
doi:10.1038/337562a0
Van der Put, E., E. M. Sherwood, B. B. Blomberg, and R. L. Riley. 2003. Aged mice
exhibit distinct B cell precursor phenotypes differing in activation, proliferation
and apoptosis. Experimental Gerontology
38:1137–1147.
334
PA R T I V
|
Adaptive Immunity: Development
Endothelial
cell
Osteoblast
Bone
IL-7 expressing cell
Pro-B cell
Pre-proB cell
Pre-B cell
Immature
B cell
HSC
Medullary
vascular
sinus
FIGURE 10-3 HSCs and B-cell progenitors make contact
with different sets of bone marrow cells as they progress
through their developmental program. HSCs begin their developmental program close to the osteoblasts (top). An HSC is also
shown entering from the blood (left-hand side), illustrating the fact
that HSCs are capable of recirculation in the adult animal. Progenitor
cells then move to gain contact with CXCL12-expressing stromal
cells, where they mature into pre-pro B cells. By the time differentiatheir further differentiation. Second, diverse populations of
stromal cells express different cytokines. At various points in
their development, progenitor and precursor B cells must
interact with stromal cells secreting particular cytokines,
and thus the developing B cells move in an orderly progression from location to location within the bone marrow. This
progression is guided by chemokines secreted by particular
stromal cell populations. For example, HSCs begin their life
in close contact with osteoblasts located close to the lining of
the endosteal (bone marrow) cavity. Once differentiated to
the pre-pro-B-cell stage, the developing B cells require signals from the chemokine CXCL12, which is secreted by a
specialized set of stromal cells, in order to progress to the
pro-B-cell stage. Pro-B cells then require signaling from the
cytokine IL-7, which is secreted by yet another stromal cell
subset (Figure 10-3). Many of these stromal cell factors serve
to induce the expression of specialized transcription factors
important in B-cell development.
The Stages of Hematopoiesis Are Defined by
Cell-Surface Markers, Transcription-Factor
Expression, and Immunoglobulin Gene
Rearrangements
Full characterization of a developmental pathway requires
that scientists understand the phenotypic and functional
IL-7
CXCL12
Plasma cell
CXCL12
reticular cell
tion has progressed to the pro-B-cell stage, the developing cell has
moved to receive signals from IL-7-producing stromal cells. After
leaving the IL-7-expressing stromal cell, the pre-B cell completes its
differentiation and leaves the bone marrow as an immature B cell.
CXCL-12 is shown in purple; IL-7 in blue. [Adapted from T. Nagasawa,
2006, February, Microenvironmental niches in the bone marrow required for
B-cell development, Nature Reviews Immunology 6:107–116. doi:10.1038/
nri1780]
characteristics of each cell type in that pathway, as well as
the molecular signals and transcription factors that drive
differentiation at each stage. Cells at particular stages of
differentiation can be characterized by their surface molecules, which include cell-surface antigens, adhesion molecules, and receptors for chemokines and cytokines. They
are also defined by the array of active transcription factors
that determine which genes are expressed at each step in
the developmental process. Finally, in the case of B cells,
the developmental stages are defined by the status of the
rearranging heavy- and light-chain immunoglobulin
genes. B-cell development is not yet completely understood; however, most of the important cellular intermediates have been defined, and developmental immunologists
are gradually filling in the gaps in our knowledge.
Investigators delineating the path of B-lymphocyte differentiation employed three general experimental strategies.
First, they generated antibodies against molecules (antigens or
markers) present on the surface of bone marrow cells. They
then determined which of these molecules were present at
the same time as other antigens, and which combinations of
antigens appeared to define unique cell types (Figure 10-4a).
In addition, culturing cells in vitro that bear known cellsurface antigens, followed by flow cytometric analysis of the
daughter cells generated in culture, enabled them to describe
the sequential expression of particular combinations of cellsurface molecules (Figure 10-4b).
B-Cell Development
|
CHAPTER 10
335
(d)
(a)
Knocking out particular transcription factors
(TFs) stops development at particular points.
Characterization of progenitors bearing
different sets of cell surface molecules.
Knocking out TF1 leaves only this population. Therefore
TF1 is required to progress to VHD recombination.
(b)
Determining sequence of marker expression by culturing cells
from each stage. Culturing cells with the red antigen gives rise to
daughter cells of both types. Culturing cells with only the blue
antigen gives rise to the cells bearing both blue and green
antigens, but never cells bearing the red antigens. Therefore, we
can sequence the three cell types in this way.
Knocking out TF2 leaves these two populations. Therefore
TF2 is required for progression to light chain rearrangement.
(e)
(c)
Sequencing of different antigen-bearing cells by analyzing each
population for the stage of V(D)J rearrangements in heavy and
light chains.
D-JH
recombination
only
VHDJH recombination
completed
VH and VL
recombination
completed
Placing GFP under the control of the TF2 promoter reveals that TF2
expression occurs in these two cell populations. Clearly, it is turned
on during the end of the blue stage, and is needed for progression
to the blue and green stage of development.
VHDJH
recombination
completed
VH and VL
recombination
completed
FIGURE 10-4 Experimental approaches to the staging and
characterization of B-cell progenitors. In this figure, the different icons do not represent specific antigens, but are used in order to
illustrate the principle of the experiment. Investigators delineated the
stage of B-cell development by using flow cytometry to characterize
the cell-surface expression of developmental markers and molecular
biology to correlate the expression of specific markers with the stage
of immunoglobulin gene rearrangement. The requirement for transcription factor activity at each step was determined using both
knockout and knockin genetic approaches. (See text for details.)
Second, by sorting cells bearing particular combinations
of cell-surface markers, and analyzing those cell populations
for the occurrence of immunoglobulin gene rearrangements,
scientists were able to confirm the staging of the appearance of particular developmental antigens. For example, an
antigen that appears on a cell in which no variable region
gene rearrangement has taken place is clearly expressed
very early in B-cell differentiation. Similarly, an antigen
present on the surface of a cell that has rearranged heavychain, but not light-chain, genes defines a stage in B-cell
differentiation later than that defined by the marker
described above, whereas an antigen present on a cell that
has undergone both heavy-chain and light-chain rear-
rangement characterizes a very late stage in B-cell development. In this way, cell-surface markers were defined that
could serve as indicators of particular steps in B-cell differentiation (Figure 10-4c).
Third, investigators use the power of knockout genetics
to determine the effects on B-cell development of eliminating the expression of particular genes, such as those encoding particular transcription factors (Figure 10-4d). For
example, knocking out a gene encoding a particular transcription factor eliminates an animal’s ability to complete VH
to DJH recombination while still allowing D to JH rearrangement. This tells us that the particular transcription factor in
question is not necessary for stages of B-cell differentiation
336
PA R T I V
|
Adaptive Immunity: Development
ADVANCES
The Role of miRNAs in the Control of B-Cell Development
Geneticists have long known that
only a small fraction of chromosomal DNA
specifies protein sequences, and early
papers relegated the nonprotein-coding
DNA segments to the somewhat ignominiously described status of “junk DNA.”
In 1993, however, scientists studying the
genome of the nematode C. elegans
described groundbreaking investigations
of some of the nonprotein-coding
sequences that they had identified as having been transcribed but not translated.
They showed that these primary transcripts were processed into small pieces
of RNA, 18 to 30 nucleotides in length,
that were capable of exerting control over
the level of expression of mRNA.
The biosynthesis of these micro-RNAs
follows a similar form in eukaryotes as
diverse as C. elegans and humans (Figure 1).
Fully capped and polyadenylated RNAs
(pri-microRNAs) are synthesized by RNA
polymerase II and are then cleaved into a
hairpin-shaped 70- to 100-nucleotide premicroRNA by the nuclear RNAase Drosha,
which works in tandem with a second
double-stranded RNA-binding protein,
DGCR8. The cleaved pre-micro-RNA is
then exported to the cytoplasm, where a
second RNAase, Dicer, acting in association with two other proteins, processes it
to an 18- to 30-nucleotide miRNA duplex,
consisting of the mature miRNA and its
anti-sense strand. In a final step, the
mature miRNA, now single stranded, associates with a protein complex called the
RNA-induced silencing complex, or RISC.
Nucleus
DNA
Cytoplasm
RNA
polymerase
mRNA-target
pri-microRNA
Drosha/
DGCR8
RISC
pre-microRNA
RISC
Dicer
PACT and TRBP
FIGURE 1
The generation of functioning microRNAs (miRNAs). Just like mRNA, miRNA species are
transcribed as long, capped and polyadenylated RNA species (pri-miRNA) by RNA polymerase ll. They
are then cleaved by a nuclear RNase, Drosha, into a hairpin shaped nucleotide precursor molecule,
termed a pre-miRNA. Drosha works in a protein complex with the protein DGCR8 (DiGeorge
syndrome chromosomal region 8). Pre-miRNAs are then exported to the cytoplasm where a second
ribonuclease, Dicer, in association with the proteins PACT and TRBP processes the pre-miRNA into a
19 to 24 nucleotide miRNA duplex, by removing the terminal loop. Next, a protein complex called RISC
(RNA-Induced Silencing Complex) binds to one of the two strands of the duplex. The strand of miRNA
that binds to RISC is the mature miRNA, and it drives the RISC enzyme to the target mRNA, resulting in
mRNA silencing and/or destruction. [Adapted from Vasilatou et al., 2009. The role of microRNAs in normal and
malignant hematopoiesis. European Journal of Hematology, 84, 1 to 16. Figure 1.]
The mature miRNA operates by complementary binding of a so-called “seed”
region of 6 to 8 nucleotides at its 5 end to
a region on its target mRNA. Once the
miRNA has bound, three things can happen: the target mRNA can be directly tar-
prior to D to JH rearrangement, but it is required for one or
more stages, starting with VH to DJH recombination.
One drawback of the knockout approach, however, is that
it only defines the first stage in differentiation at which the
transcription factor is required. More recent variations have
exploited knockin genetics (Figure 10-4e) to generate animals that express fluorescent markers under the control of
geted for cleavage; the mRNA can be
destabilized; or translation from the mRNA
can be repressed. Furthermore, we now
know that a single miRNA can target the
synthesis of many proteins, and each
mRNA can be the target of more than one
the transcription factor promoters, so that every point at
which the transcription factor is expressed can be delineated. The staging of transcription factor expression is then
correlated with the expression of both cell-surface markers
and immunoglobulin gene rearrangement. The sequence of
B-cell development described in the next few sections has
been elucidated using a combination of these strategies.
B-Cell Development
|
CHAPTER 10
337
BOX 10-2
miRNA, thus adding to both the flexibility
and the complexity of this mode of control over gene expression.
But does regulation by miRNAs operate during B-cell development? Several
lines of evidence suggest that the answer
to this question is an unequivocal “yes.”
From a theoretical standpoint, it is clear
that the developmental changes that
occur as B cells mature require rapid
changes in the concentrations of such
important proteins as transcription factors
and pro- and anti-apoptotic molecules,
among other regulatory proteins. The
need for such rapid alterations in protein
concentrations can be met efficiently by
the type of post-transcriptional control
mechanisms mediated by miRNAs.
Conditional loss of the gene encoding
the Dicer nuclease destroys all capacity to
synthesize mature miRNAs. Ablation of
Dicer in early B-cell progenitors resulted in
a developmental block at the pro- to pre-Bcell transition. In these experiments, the
pro-apoptotic molecule Bim, was expressed
at higher concentrations in Dicer-ablated
than in normal B-cell progenitors.
Sequences in the 3 untranslated region of
the Bim gene were found to be complementary to miRNAs of the 17~92 family,
suggesting that members of this family
normally down-regulate Bim at this stage
of development, enabling B cells to pass
through this transition. The 17~92 family of
miRNAs was also shown to affect expression of TdT and hence N-sequence addition. Other alterations in immunoglobulin
gene expression were also observed in the
absence of 17~92 miRNAs, including an
increase in the expression of sterile transcripts. These data collectively demonstrate that miRNAs are important in the
control of the pro- to pre-B-cell transition and
affect both the expression of pro-apoptotic
molecules and the nature of the Ig repertoire. As one would predict from these
collective results, animals with increased
expression of miR-17~92 family members
express lower levels of Bim and suffer from
a lympho-proliferative disorder and autoimmune disease.
Members of the 17~92 miRNA family
were also found to control the levels of
the Pten protein, which acts as an inhibitor of the pro-survival PI3 kinase and Akt
signaling pathway. An increase in the levels of the miR-17~92 molecules allows for
greater destruction of the Pten mRNA,
resulting in increased cell survival and a
corresponding increase in the number of
lymphocytes available to proliferate.
Other investigators have addressed the
question of which miRNAs are expressed at
different stages of B-cell development. A
combination of genomic (in silico analyses)
and more classical molecular biological
approaches have identified several miRNA
species and/or families that are implicated
in the control of B-cell development.
Perhaps the most well studied miRNA
is miR-150, which is highly expressed in
mature and resting B cells but not in their
progenitors. The miR-150 molecule has
been shown to depress the level of
expression of the transcription factor
c-Myb, known to be essential for the control of B-cell development. As might have
been predicted, the pattern of c-Myb
expression in lymphocyte development is
complementary to that of miR-150, in that
c-Myb is highly expressed in lymphoid
progenitors and is down-regulated upon
their maturation; in addition, transcriptional analysis confirmed that miR-150 is
Recently, attention has begun to focus on small molecular
weight miRNA species, which have profound effects on the
stability of mRNA and hence on the expression of particular
proteins. In Advances Box 10-2, we describe the effects of
some of the miRNAs that have recently been shown to affect
B-cell differentiation. This is currently an extremely active
research area.
an important factor in the regulation of
the levels of the c-Myb transcription factor during B-cell development.
MiR-150 is also implicated in B-1/B-2 lineage specification. B-cell-specific deletion of
the c-myb gene stops B-cell development at
the pro- to pre-B transition and also leads to
the complete disappearance of the B-1 subset of B cells. If miR-150 is responsible for
down-modulating the levels of c-Myb in
vivo, then it might be predicted that a deficiency in miR-150 would result in an opposite phenotype to that expressed by a c-Myb
deficient animal, and such was indeed
found to be the case. Deficiency of miR-150
was found to result in an expansion of the
B-1-B-cell pool, with a resulting increase in
the levels of IgM antibody secretion.
The study of miRNA control of mammalian gene expression and lymphocyte
development is still in its infancy, but the
ability to manipulate and isolate cells at
discrete stages in the developmental
sequence provides a particularly tractable
system in which to analyze the range of
actions of different families of miRNAs. This
field will undoubtedly be one to watch.
REFERENCES
Baltimore et al., 2008. MicroRNAs: new regulators of immune cell development and function. Nature Immunology 9:839–845.
Koralov, S. B., et al. 2008. Dicer ablation affects
antibody diversity and cell survival in the B
lymphocyte lineage. Cell 132:860–874.
Vasilatou, D., S. Papageorgiou, V. Pappa, E. Papageorgious, and J. Dervenoulas. 2009. The
role of microRNAs in normal and malignant
hematopoiesis. European Journal of Haematology 84:1–16.
Xiao, C., and K. Rajewsky. 2009. MicroRNA control in the immune system: Basic principles.
Cell 136:26–36.
The Earliest Steps in Lymphocyte
Differentiation Culminate in the Generation of
a Common Lymphoid Progenitor
In this section, we will describe the process by which an HSC
in the bone marrow develops into a CLP. Unless otherwise
specified, the developmental pathway we describe refers to
338
PA R T I V
|
Erythrocyte
Megakaryocyte
Monocyte
T-cell progenitor
Natural killer
(NK) cell
Dendritic
cell (DC)
Adaptive Immunity: Development
c-Kit
sca-1
flt-3
CD34
IL-7R
Rag1/2
and TDT
+
+
–
–
–
–
Multipotent
progenitor
(MPP)
+
+
–
+
–
–
Lymphoid-primed
multipotent
progenitor cell
(LMPP)
+
+
+
–
–
–
Early lymphoid
progenitor cell
(ELP)
+
+
+
–
+/–
+
low
low
+
–
+
+
Hematopoietic
stem cell
(HSC)
Common
lymphoid
precursor
(CLP)
Pre-pro
B cell
FIGURE 10-5 Expression of cell-surface markers on HSC and lymphoid progenitor cells. The maturation of HSCs into lymphoid
progenitors, and the progressive loss of the ability to differentiate into other blood-cell lineages can be followed by the expression of the cellsurface markers as well as by the acquisition of RAG and TdT activity.
that followed by the predominant B-2-B-cell population. Specific aspects of development that differ among the various
B-cell subsets will be addressed toward the end of this chapter.
HSCs
HSCs are both self-renewing (they can divide to create identical copies of the parent cell) and multipotential (they can
divide to form daughter cells that are more differentiated
than the parent cell and that can develop along distinct
blood-cell lineages), and can give rise to all cells of the blood.
HSCs maintain a relatively large number of genes in a socalled “primed” state, and an individual HSC may possess
primed genes characteristic of multiple cell lineages. Primed
chromatin is associated with lower-than-usual numbers of
nucleosomes, is more accessible to enzyme activity than the
majority of chromatin in the cell, and shows histone methylation and acetylation patterns characteristic of active chromatin. Depending on the environmental stimuli to which
any given HSC is exposed, transcription factors may drive
the cell down a number of possible developmental pathways.
During the differentiation process that follows, primed chromatin regions containing genes that are not needed for the
selected developmental pathway are shut down.
In HSCs bound for a B-cell fate, the transcription factors
Ikaros, Purine box factor 1 (PU.1), and E2A participate in
the earliest stages of B-lineage development. Ikaros recruits
chromatin remodeling complexes to particular regions in
the DNA and ensures the accessibility of genes necessary for
B-cell development. PU.1 presides over a leukocytic “balancing act”; low levels of PU.1 favor lymphoid differentiation,
whereas cells expressing higher levels of PU.1 veer off to a
myeloid fate. The level of PU.1 protein expressed is in turn
regulated by the transcriptional repressor Gfi1, which downregulates the expression of PU.1 to the levels necessary for
progression down the B-cell pathway. E2A expression contributes to the maintenance of the HSC pool by participating
in the regulation of cell cycle control in this population.
As noted in Chapter 9, HSCs express the cell-surface molecule c-Kit (CD117), which is the receptor for stem cell factor
(SCF). SCF is a cytokine that exists in both membrane-bound
and soluble forms and the SCF-c-Kit interaction is critical for
the development, in adult animals, of multipotential progenitor cells (MPPs). Membrane-bound SCF plays a role in
retaining the HSCs and its daughter progenitor cells in the
appropriate environmental niches in the bone marrow. HSCs
also express the stem cell associated antigen-1 (Sca-1). Both
c-Kit and Sca-1 are expressed in parallel on early progenitor
cells, and the levels of their expression drop as the cells commit to a particular cell lineage (Figure 10-5). HSCs are often
described as being Lin, a designation that refers to the fact
that they have no “lineage markers” characteristic of a particular blood-cell subpopulation.
B-Cell Development
MPPs
MPPs generated on receipt of SCF/c-Kit signaling lose the
capacity for extensive self-renewal, but retain the potential to
differentiate into several different hematopoietic lineages.
MPP cells retain the expression of c-Kit and Sca-1 and transiently express the molecule CD34. Indeed, antibodies to
CD34 are used clinically to isolate cells at this stage in hematopoiesis. MPPs also express the chemokine receptor CXCR4,
which enables them to bind the stromal cell-derived chemokine CXCL12. The interaction between CXCL12 and CXCR4
is important in ensuring that the progenitor cell occupies the
correct niche within the bone marrow (see Figure 10-3).
LMPPs
The progenitor cell on its way to becoming a B cell then
begins to express the fms-related tyrosine kinase 3 receptor
(flt-3). Flt-3 binds to the membrane-bound flt-3 ligand on
bone marrow stromal cells and signals the progenitor cell to
begin synthesizing the IL-7 receptor (IL-7R, CD127). Flt-3
is expressed on B-cell progenitors from this point until the
pro-B stage and acts synergistically with IL-7R to promote
the growth of cells bearing flt-3 and IL-7R. The expression of
flt-3 on the surface of the developing cell marks the loss of
the potential of the MPP cell to develop into erythrocytes or
megakaryocytes, and therefore characterizes a new level of
cell commitment; however, this progenitor still retains the
capacity to develop along either the myeloid or the lymphoid
pathways.
These cells, now c-Kit, Sca-1, and flt-3, are termed
lymphoid-primed, multipotential progenitors (LMPPs)
(see Figure 10-5). As they become further committed to the
lymphoid lineage, levels of the stem-cell antigens c-Kit and
Sca-1 fall, and cells destined to become lymphocytes begin
to express RAG1/2 and terminal deoxynucleotidyl transferase (TdT) (Chapter 7). Expression of the genes encoding
RAG1/2, TdT, IL-7R, and the B-cell-specific transcription
factor EBF1 are all up-regulated at the end of this stage.
ELPs
Expression of RAG1/2 defines the cell as an early lymphoid
progenitor cell (ELP). A subset of ELPs migrates out of the
bone marrow to seed the thymus and serve as the T-cell
progenitors (discussed in Chapter 9). The rest of the ELPs
remain in the bone marrow as B-cell progenitors. On these
cells, the levels of the early c-Kit and Sca-1 antigens decrease
as the levels of the IL-7R increase, and the ELP now develops
into a CLP.
CLPs
At the CLP stage, the progenitor on its way to B-cell commitment still retains the potential to mature along the
NK, conventional DC, or T-cell lineages. At this point in
development, signals received through IL-7R promote
cell survival and enhance the production of EBF-1 and
other transcription factors that are required for later steps
|
CHAPTER 10
339
in the B-cell differentiation pathway. Signaling through
the IL-7R occurs via pathways familiar from Chapter 4.
Specifically, an IL-7R-mediated JAK-STAT pathway
induces the up-regulation of the anti-apoptotic molecule
Mcl1. Signaling through IL-7R also results in the upregulation of the C-myc and N-myc genes, which signal
the cell proliferation characteristic of the later, pro-B-cell
stage.
CLPs are c-Kitlow, Sca-1low, and IL-7R and have lost
myeloid potential. However, as a CLP destined to differentiate
along the B cell pathway matures, the chromatin containing
the immunoglobulin locus becomes increasingly accessible
and the developing lymphocyte approaches the point at
which it is irrevocably committed to the B-cell lineage.
The Later Steps of B-Cell Development Result
in Commitment to the B-Cell Phenotype
Figure 10-6 illustrates the expression of cell-surface markers,
and the patterns of rearrangement of immunoglobulin
heavy-chain and light-chain genes starting at the pre-pro
B-cell stage of development. The stages of B-cell differentiation have been defined by more than one group of scientists
and, as a result, two systems of nomenclature are in common
use. The first, and most widely used, is the Basel nomenclature (pre-pro, pro, pre-B, immature B) developed by Melchers
and colleagues. The second (A, B, C, C, D, E) is that defined
by Hardy et al., and the process by which this system of classification of B-cell development was established is described
in detail in Classic Experiment Box 10-3.
Pre-Pro B Cells
With the acquisition of the B-cell lineage-specific marker
B220 (CD45R), and the expression of increasing levels of the
transcription factor EBF1, the developing cell enters the prepro-B-cell stage. EBF1 is an important transcription factor in
lymphoid development, and therefore transcription of the
Ebf1 gene is itself under the control of multiple transcription
factors (Figure 10-7). These each bind at distinct promoter
regions, and hence the level of transcription of the Ebf1 gene
can vary considerably depending on the combination of
controlling factors present at any particular developmental
stage. At the pre-pro-B-cell stage, EBF-1, along with E2A,
binds to the immunoglobulin gene, promoting accessibility
of the D-JH locus and preparing the cells for the first step of
Ig gene recombination. EBF-1 is also essential to the full
expression of many B-cell proteins, including Ig,Ig
(CD79,), and the genes encoding the pre-B-cell receptor,
which will be expressed when heavy-chain VDJ recombination is complete.
Pre-pro B cells remain in contact with CXCL-12-secreting stromal cells in the bone marrow. However, the onset of
D to JH gene recombination classifies the cell as an early
pro-B cell, and at this stage the developing cell moves within
the bone marrow, seeking contact with IL-7 secreting stromal cells.
PA R T I V
340
|
Adaptive Immunity: Development
Status
of Ig
genes
Surface Ig
B220
receptor
c-Kit IL-7R CD25 CD19
(CD45R)
expression
Pre-pro
(Fr. A)1
GL2
None
–
lo
–
–
+
Early pro
(Fr. B)
DJH
None
lo
lo/+
–
+
+
Late pro
(Fr. C)
Some
VHDJH
None
lo
+
+/–
+
+
Large Pre
(Fr. C')
VHDJH
Pre-BCR
–
+
+
+
+
Small Pre
(Fr. D)
VHDJH
VL JL rearrangement
begins
Decreasing
levels
of Pre-BCR
–
+
+
+
+
IgM
–
+
+
+
+
Cell stage
1st checkpoint
Immature B VHDJH
(Fr. E)
VL JL
FIGURE
10-6 Immunoglobulin gene
rearrangements and expression of marker
proteins during B-cell development. The
expression of selected marker proteins is correlated with the extent of Ig gene rearrangement
during B-cell development from the pre-pro B cell
to the immature T1 B-cell stage. (See text for
details.) [Adapted from K. Samitas, J. Lötvall, and A. Bossios, B cells: From early development to regulating allergic
diseases, Archivum Immunologiae et Therapiae Experimentalis 58:209–225, Figure 1.]
2nd checkpoint
Pre-BCR
BCR
1 Labeled fractions refer to the “Hardy nomenclature,” described in the
Classic Experiment Box 10-1.
2 GL = germ line arrangement of heavy and/or light chain V
region segments
IL-7
IL-7Rγ
IL-7Rα
FIGURE 10-7 The interplay of transcripCytoplasm
STAT5
P
P
STAT5
• Transcription of B-cell
associated genes survival
• B-cell specification and
commitment survival
E2A
N-myc
C-myc
EBF1
B-cell proliferation
PAX5
Nucleus
D-JH recombination
VHDJH recombination
Notch1
tion factors during early B-cell development. Dimerization and activation of the
transcription factor STAT5 is stimulated by IL-7
binding to its receptor. STAT5 stimulates B-cell
proliferation by activating the proliferative control
proteins N-myc and C-myc. STAT5 collaborates
with E2A proteins to promote the expression of
early B-cell factor 1 (EBF1). EBF1 in turn promotes
the expression of PAX5, and together the E2A
proteins, EBF1, and PAX5 activate many genes
leading to B-cell lineage specification and commitment. PAX5 and EBF1 both participate in positive feedback loops that enhance the levels of
both EBF1 and PAX5 transcription. [Adapted from B.
L. Kee, 2009, E and ID proteins branch out, Nature Reviews
Immunology 9:175–184.]
B-Cell Development
|
CHAPTER 10
341
BOX 10-3
CLASSIC EXPERIMENT
The Stages of B-Cell Development: Characterization
of the Hardy Fractions
Richard Hardy’s
laboratory was
one of the first to combine flow cytometry and molecular biology in experiments
designed to analyze lymphocyte maturation. In this feature, we describe what
those researchers did and how they generated a model of the sequencing of the
stages of B-cell development from their
data.
When Hardy and colleagues began
their characterization of B-cell lineage
development in the early 1990s, prior work
using molecular analysis of long term bone
marrow cell lines had already established
the sequential rearrangement of heavychain and light-chain immunoglobulin
genes. In addition, the expression of a
number of cell-surface markers on bone
marrow cells had been measured, and
several of these antigens had been shown
to be co-expressed with B220 (CD45R),
which had already been established as a
B-cell differentiation antigen. Hardy’s
approach was to characterize the sequence
of expression of those antigenic markers
that were found on the same cells as B220.
The hypothesis was that some of these
markers may be expressed on early B-cell
progenitors and might therefore help to
generate a scheme of B-cell development.
In order to place cells expressing different
combinations of markers into a developmental lineage, Hardy then sorted cells
bearing each combination of his selected
markers, and placed them into co-cultures
with a bone marrow stromal cell line. After
defined times in culture, he harvested the
hematopoietic cells and re-characterized
their surface marker expression.
The markers used in these experiments included B220 (CD45R) and CD43
(leukosialin), which had previously been
shown to be expressed on granulocytes
and all T cells, but was not present on
mature B cells, with the exception of
plasma cells. In addition, their experiments employed antibodies directed
against Heat Stable Antigen, or HSA
(CD24) and BP-1, an antigen on bone
Start with bone marrow cells
Sort for B220+CD43+
Analyzed B220+
CD43+ cells for
HSA and BP-1
expression
HSABP-1Fraction A
Pre-pro-B cells
HSA+
BP-1Fraction B
Early pro-B cells
HSA+
BP-1+
Fraction C
Lower levels of HSA
Fraction C
Late pro-B cells
Higher levels of HSA
Fraction C'
Large (early)
pre-B cells
FIGURE 1
The isolation of Hardy’s fractions. A, B, C, C. Bone marrow cells were sorted for cells bearing
B220 and CD43 and then analyzed for their expression of the cell-surface markers HSA and BP-1.
marrow cells. Both HSA and BP-1 had
been previously shown to be differentially expressed at varying stages of
lymphoid differentiation.
The first set of experiments analyzed
those cells bearing both B220 and CD43 for
the levels of their expression of HSA and
BP-1 (Figures 1 and 2). Flow cytometry
plots demonstrated that the B220CD43
cells neatly resolved into three discrete
subpopulations. The first, labeled A in Figure 1, expressed neither HSA, nor BP-1. The
second, labeled B, expressed HSA, but not
BP-1, and the third expressed both of these
antigens. Analysis of Ig gene rearrangements in these populations revealed that
no gene rearrangements occurred in
fraction A but that D to JH gene segment
rearrangements had begun in fraction B.
Subsequent work has shown that VH to DJH
Gated B220+ CD43–
FIGURE 2
Flow cytometric characterization of
the early developmental stages of B
cells. (See text for details.) [Hardy et al., J. Exp. Med.
173, 1213–1225. May 1991. By permission of Rockefeller
University Press]
(continued)
342
PA R T I V
|
CLASSIC EXPERIMENT
Adaptive Immunity: Development
(continued)
Start with bone marrow cells
Sort for B220+, CD43-
mIgMlo
H chain rearranged
L chain rearrangement
beginning
mIgMhi
H chain rearranged
L chains rearranged
mIgMhi, mIgDhi
H chain rearranged
L chains rearranged
Fraction D
Small pre-B cells
Fraction E
Immature B cells
Fraction F
Mature B cells
FIGURE 3
The isolation of Hardy’s fractions D, E, and F. Bone marrow cells were sorted for cells
bearing B220, but not CD43 (recognized by monoclonal antibody S7) and then analyzed for cell-surface
expression of IgM and IgD.
Pro-B Cells
In the early pro-B-cell stage, D to JH recombination is completed and the cell begins to prepare for V to DJH joining.
However, this final recombination event awaits the expression of the quintessential B-cell transcription factor, PAX5.
The Pax5 gene is among EBF-1’s transcriptional targets
(see Figure 10-7) and transcription of genes controlled by the
PAX5 transcription factor denotes passage to the pro-B-cell
stage of development, at which point the expression of non-Blineage genes is permanently blocked. PAX5 can act as a transcriptional repressor, as well as an activator, and blocks
Notch-1 gene expression, thereby terminating any residual
potential of the pro-B cell to develop along the T-cell lineage.
Many important B-cell genes are turned on at this stage, under
the control of PAX5 and other transcription factors. Among
these is the gene encoding CD19, which we first encountered in
rearrangements occur in fraction C
(although at the time, the method of analysis that Hardy and colleagues used failed to
reveal this second type of rearrangement).
So at this point, we know that three
distinct types of B-cell precursors express
both B220 and CD43, and can be discriminated on the basis of their levels of the
further two antigens, HSA and BP-1. Fraction A corresponds to what we now know
as pre-pro B cells, fraction B to early pro-B
cells, and fraction C to late pro-B cells.
Culture of fraction C cells yielded cells
that expressed membrane (m) IgM; similarly, culture of fraction B cells also yielded
daughter cells expressing mIgM, but at a
lower frequency than fraction C cells, suggesting that cells in fraction C were further along the differentiation pathway to
mIgM B cells. Furthermore, the three different fractions displayed differential
dependence on the need to adhere to the
stromal cell layer. Cells from fraction A
required stromal cell contact for survival.
Fraction B cells survived best in contact
with the stromal cells, but were able to
survive in a culture in which they were
separated from the stromal cells by a
semipermeable membrane. Under these
conditions, they could still receive soluble
factors generated by the stromal cells, but
were prevented from generating adhesive
interactions with stromal cell-surfacebound growth factors. Fraction C cells
Chapter 3, as one of the components of the B-cell co-receptor.
CD19 is considered a quintessential B-cell marker and is often
used as such in flow cytometry experiments.
Once the PAX5 protein is expressed, a mutual reinforcement occurs between PAX5 and EBF-1 expression, as illustrated in Figure 10-7, with each transcription factor serving
to enhance the expression of the other. The PAX5 protein
continues to be expressed in mature B cells until the B cell
commits to a plasma-cell fate following antigenic stimulation (see Chapter 12). The higher levels of EBF1 expression
induced by PAX5 also allow for an increase in the level of
IL-7R expression.
PAX5 promotes VH to D recombination by contracting the
IgH locus, thus bringing the distant VH gene segments closer
to the D-JH region. B cells deficient in PAX5 permit D to JH
Ig gene rearrangement, but do not allow recombination of
B-Cell Development
|
CHAPTER 10
343
BOX 10-3
survived and proliferated in the absence
of stromal cell contact. Analysis of the factors secreted by the stromal cells that
were necessary for survival and proliferation of the fraction B and C cells revealed
one of them to be interleukin 7.
Hence, using the criteria of Ig gene
rearrangements and phenotypic analysis
of cultured cell populations, Hardy and
colleagues were able to place the three
fractions in sequence; cells of fraction A
gave rise to cells of fraction B, which in
turn mature into cells of fraction C.
Careful analysis of the contour graph of
fraction C reveals that it, in turn, can be
subdivided on the basis of the levels of
expression of HSA. That population of cells
bearing higher levels of HSA, as well as
BP-1, is now defined as fraction C, which
corresponds to early or large pre-B cells.
Hardy and colleagues next turned
their attention to those cells that
expressed B220, but had lost CD43, and
measured their cell-surface expression of
mIgM (Figure 3). Three populations of
cells were again evident, which they
labeled D, E, and F (Figure 4). Cells belonging to fraction D expressed zero to low
levels of mIgM, showed complete heavychain rearrangement, and some lightchain rearrangement and correspond to
small pre-B cells. Cells belonging to fraction E displayed high levels of mIgM as
well as of B220, complete heavy-chain
rearrangement, and most of the cells in
that fraction also displayed light-chain
gene rearrangement. Fraction E cells are
thus immature B cells ready to leave the
bone marrow. Subsequent further characterization of fraction F cells showed that,
in addition to surface IgM, these cells also
bear surface IgD and therefore represent
fully mature B cells, presumably recirculating through the bone marrow.
Thus, Hardy’s experiments revealed
that the pool of progenitor and precursor
B cells in the bone marrow represents a
complex mixture of cells at different
stages of development, with varying
requirements for stromal cell contact and
interleukin support.
These elegant experiments still had
one more story to tell that did not appear
in the original paper, but which emerged
in later publications. Single-cell PCR analysis of fraction C cells showed that many of
them had nonproductive rearrangements
on both heavy-chain chromosomes (see
Figure 2). In contrast, all the cells from the
C fraction demonstrated productive rearrangements on one of the heavy-chain
chromosomes. Fraction C cells therefore
represent B cells that have been unable to
productively rearrange one of their heavychain genes and that will therefore eventually die by apoptosis. The C fraction also
included the highest proportion of cells in
cycle of any of the B220 B-cell stages in
the VH to the D Ig gene segment, indicating that PAX5 is
essential to the second step of Ig gene rearrangement.
Expression of the signaling components of the B-cell receptor, Ig and Ig, also begins at the pro-B-cell stage and the
Ig,Ig signaling complex is briefly placed on the cell surface
in complex with the chaperone protein calnexin. Although this
Ig,Ig complex has been referred to as a “pro-BCR,” no ligand
has yet been established for it, nor do we yet understand the
importance of any signaling that may emanate from it.
Also during the pro-B-cell stage, c-Kit is once more
turned on briefly, enabling the cell to receive signals from
stem cell factor. By the beginning of the pre-B-cell stage of
development, expression of c-Kit is irreversibly turned back
off. By the late pro-B-cell stage, most cells have initiated VH
to DJH Ig gene segment recombination, which is completed
by the onset of the early pre-B-cell stage.
Gated B220+ CD43–
FIGURE 4
Flow cytometric characterization of
the later developmental stages of B
cells. (See text for details.) [Hardy et al., J. Exp. Med.
173, 1213–1225. May 1991. By permission of Rockefeller
University Press]
the marrow. This is consistent with the
notion that the new heavy chain is associating with the surrogate light chain at the
C stage and the pre-B-cell receptor complex is expressed on the cell surface, triggering the period of clonal expansion of B
cells described in this chapter.
REFERENCES
Hardy, R. R., et al. 1991. Resolution and characterization of pro-B and pre-pro-B cell stages
in normal mouse bone marrow. Journal of
Experimental Medicine 173:1213–1225.
Hardy, R. R., P. W. Kincade, and K. Dorshkind.
2007. The protean nature of cells in the B
lymphocyte lineage. Immunity 26:703–714.
Pre-B Cells
During the pre-B-cell stage, the cell expresses a pre-B-cell
receptor composed of the rearranged heavy chain, complexed with the VpreB and 5 components of the surrogate
light chain (see Figure 7-10). The appearance of this pre-Bcell receptor signals the entry of the developing B cell into
the large, or early pre-B-cell phase. As we learned in Chapter 7, the expression of the heavy chain at the cell surface is
necessary for the termination of further heavy-chain rearrangement and ensures allelic exclusion of the Ig heavychain genes. Animals deficient in the expression of either the
pre-B-cell receptor, or of the signaling components Ig,Ig,
fail to progress to the pre-B-cell stage.
Signaling through the pre-B-cell receptor induces a few
rounds of proliferation in the pre-B cell. This proliferative
344
PA R T I V
|
Adaptive Immunity: Development
phase correlates in time with the expression on the pre-Bcell surface of CD25, the chain of the high-affinity IL-2
receptor chain (see Figure 4-8), which first appears on B
cells at the pro-B-cell stage. Since the pre-B-cell proliferative
process appears mainly to be driven by IL-7, the functional
significance of CD25 at this point is unclear. However, its
appearance is frequently used as a marker of the late pro-Bcell to early pre-B-cell stage of development.
VH gene recombination is energetically expensive to the
organism, as not all developing B cells are successful in
undergoing productive VH gene rearrangements, and those
that fail to do so are lost by apoptosis. It therefore seems
logical that those B cells that have achieved productive heavychain expression should be allowed to proliferate. Each individual daughter cell derived from this proliferative process is
then free to participate in a different light-chain rearrangement event. Most individuals will therefore have multiple B
cells expressing precisely the same heavy-chain rearrangement but each with a different light chain, and many different
receptor specificities can thereby be generated from each successful heavy-chain rearrangement. Recall from Chapter 9
that an analogous process of chain rearrangement, followed
by proliferation prior to rearrangement, occurs in T cells.
If the pre-B-cell receptor cannot be displayed on the cell
surface because of non-productive VHDJH gene rearrangements, B-cell development is halted and the cell is lost to
apoptosis. This stage in B-cell development is therefore
referred to as the pre-B-cell (1st) checkpoint (see Figure 10-6).
Progress through this checkpoint depends on some type of
signaling event through the pre-B-cell receptor, and recent
evidence suggests that this is mediated via interactions
between arginine-rich regions in the non-immunoglobulin
portion of the 5 component of the surrogate light chain, or
in the CDR3 regions of some heavy chains, with negatively
charged molecules on the surface of the stromal cells.
Pre-B-cell receptor signaling induces the transient downregulation of RAG1/2 and the loss of TdT activity. Together,
these events ensure that, as soon as one heavy-chain gene has
successfully rearranged, no further heavy-chain recombination is possible. This results in the phenomenon of allelic
exclusion, whereby the genes of only one of the two heavychain alleles can be expressed in a single B cell. As a result of
this pre-B-cell receptor signaling, the chromatin at the unused
heavy-chain locus undergoes a number of physical changes
that render it incapable of participating in further rearrangement events. Recall that IL-7 provided one of the signals that
brought the VH, D, and JH loci into close apposition with one
another at the beginning of VHDJH recombination. A reduction in IL-7 signaling at the pre-B-cell stage now reverses that
initial locus contraction, resulting in the physical separation of
the VH, D, and JH gene segments in the unrearranged heavychain locus. This decontraction is then followed by deacetylation events that deactivate the unused heavy-chain locus and
return it to a heterochromatic (inactive, closed) configuration.
Surrogate light-chain expression is also terminated by a
negative feedback round of signaling through the pre-B-cell
receptor. At the end of pre-BCR-signaled cell proliferation, the
pre-B-cell receptor is lost from the surface, and this signals
entry into the small or late pre-B-cell stage. At this point,
light-chain rearrangement is initiated with the re-expression
of the Rag1/2 genes. Very little TdT activity remains at this
stage, and therefore N region addition occurs less frequently
in light chains than in heavy chains. In the mouse, light-chain
rearrangement begins on one of the chain chromosomes,
followed by the other. If neither chain rearrangement is successful, rearrangement is then successively attempted on each
of the chain chromosomes. In humans, rearrangement is
initiated randomly at either the or the loci.
Once a light-chain gene rearrangement has been successfully completed, the IgM receptor is expressed on the cell
surface, signaling entry into the immature B-cell stage. If the
attempts at light-chain immunoglobulin gene rearrangement
are not successful, the nascent cell is eventually lost at the
immature B-cell (2nd) checkpoint (see Figure 10-6). However,
given the availability of four separate chromosomes on which
to attempt rearrangement, and the opportunity for light-chain
editing in the case of unproductive rearrangement, most
pre-B cells that have successfully rearranged their heavy
chains will progress to the formation of an immature B cell.
Immature B Cells in the Bone Marrow Are
Exquisitely Sensitive to Tolerance Induction
Immature B cells bear a functional receptor in the form of
membrane IgM, but have not yet begun to express any other
class of immunoglobulin. They continue to express B220,
CD25, IL-7R, and CD19.
Once the functional BCR is assembled on the B-cell
membrane, the receptor must be tested for its ability to bind
to self antigens, in order to ensure that as few as possible
autoreactive B cells emerge from the bone marrow. Those
immature B cells that are found to bear autoreactive receptors undergo one of three fates; some are lost from the repertoire prior to leaving the bone marrow, by the BCR-mediated
apoptotic process of clonal deletion. The loss of B cells bearing self-reactive receptors within the bone marrow is referred
to as central tolerance. Other autoreactive B cells reactivate
their RAG genes to initiate the process of light-chain receptor editing (see Chapter 7). Some autoreactive B cells that
recognize soluble self antigens within the bone marrow may
survive to escape the bone marrow environment, but become
anergic, or unresponsive, to any further antigenic stimuli.
The concept of negative selection of lymphocytes bearing
autoreactive receptors should be familiar from the discussion
of T-cell tolerance in Chapter 9. However, functional differences between T cells and B cells mean that the selection
processes against autoreactive B cells are different from those
that protect against the emergence of autoreactive T cells, and
indeed can be somewhat less stringent. Since stimulation of
B-2 B cells requires T-cell help, an autoreactive B cell cannot
respond to antigen with antibody production unless there is
also an autoreactive T cell that can provide the necessary
B-Cell Development
cytokines and costimulation (Chapter 12). Thus, it is quite
possible that most individuals carry significant numbers of
autoreactive B cells within their mature B-cell repertoires that
are never activated.
There are also mechanistic differences between the modes
of negative selection among B and T cells. At this point, no
equivalent of the AIRE protein has been shown to exist for
B-cell selection, and so B-cell negative selection within the
bone marrow is more limited with respect to the available
tolerogenic antigenic specificities than is T-cell negative
selection in the thymus.
Many, but Not All, Self-Reactive B Cells Are
Deleted within the Bone Marrow
Our understanding of how the immune system eliminates or
neutralizes autoreactive threats has been facilitated by the
development of transgenic animals which express both
deliberately introduced auto-antigens and the receptors that
recognize them. It has long been established that crosslinking the IgM receptors of immature B cells in vitro (performed experimentally by treating the cells with antibodies
against the receptor chain) results in death by apoptosis. In
contrast, performing the same experiments with mature B
cells, bearing both IgM and IgD receptors, results in activation. David Nemazee and colleagues set out to test whether
the apoptotic response of immature B cells in vitro reflected
what happens in the bone marrow in vivo when an immature B cell meets a self antigen.
Nemazee et al.’s approach was conceptually simple,
although experimentally complex, particularly for the time
period in which the work was done (1989). They generated
mice transgenic for both a heavy and a light chain specific
for the MHC molecule H-2Kk. All the B cells in this mouse
therefore made only anti-H-2Kk antibodies. If immature B
cells undergo selection to prevent autoimmunity, these cells
would be selected against in a mouse that expresses the
H-2Kk gene for MHC. By appropriate breeding, they introduced the immunoglobulin H-2Kk-specific transgenes into
mice bearing two different MHC genotypes.
In the first group of mice (Figure 10-8a), which bore
H-2Kd but no H-2Kk antigens, they were able to detect the
transgenic antibody at high frequency on the surface of B
cells and at high concentration in the serum (Table 10-1).
This makes sense, as the transgenic antibody would be
unable to bind the H-2Kd molecules and so the B cells that
produce it would not be negatively selected. However, when
they bred these animals with mice of the H-2Kk type (Figure
10-8b), no membrane-bound or secreted anti-H-2Kk antibodies could be detected, suggesting that all immature B
cells bearing the potentially autoimmune receptor antibodies had been deleted in the bone marrow. This deletion
occurred via induction of apoptosis in the autoimmune cells.
Interestingly, in the H-2Kk/d mice, not all B cells bearing
the autoimmune transgenes were deleted, even though all B
cells in this mouse should bear the anti-H-2Kk receptor
|
CHAPTER 10
345
(Figure 10-8c). Closer examination revealed that some of the
residual transgene-expressing B cells in the bone marrow
had undergone light-chain receptor editing (see Chapter 7),
changing their antigen specificity so they no longer bound
the H-2Kk antigen. Recent experiments suggest that in vivo,
a significant fraction of potentially autoimmune B cells
undergo receptor editing (or even VH gene replacement
(Chapter 7), and successfully generate acceptable BCRs prior
to release from the bone marrow.
In normal animals, not all potentially autoimmune B cells
are lost to clonal deletion or altered via receptor editing or
VH gene replacement within the bone marrow, however;
some are released to the periphery and subject to further
rounds of selection.
B Cells Exported from the Bone Marrow Are
Still Functionally Immature
Once the B cell expresses IgM on its membrane (mIgM), it is
referred to as an immature B cell. This B cell is ready for
export to the spleen, where it completes its developmental
program. Immature B cells have a short half-life, in part as a
result of expressing low levels of the anti-apoptotic molecules Bcl-2 and Bcl-xl. They also express high levels of the
cell-surface molecule, Fas, which is capable of transmitting a
death signal when bound by its ligand (see Chapters 4 and
9). Immature B cells are exquisitely susceptible to tolerance
induction, and if they encounter a self antigen at this stage of
development, the B cells will re-express the RAG1 and RAG2
genes and edit their light-chain genes. If receptor editing
fails to yield a suitable receptor, the cell undergoes apoptosis.
The study of B-cell development in the periphery, like that
in the bone marrow, has benefited significantly from the ingenious application of flow cytometry, which has enabled the
classification of immature B cells into two subpopulations of
transitional B cells (T1, T2). These transitional B cells act
sequentially as the precursors to the fully mature B cell.
T1 and T2 Transitional B Cells
T1 and T2 transitional B cells were characterized initially on
the basis of their cell-surface expression of immunoglobulin
receptors and membrane markers (Table 10-2). T1 cells are
mIgMhi, mIgD/lo, CD21, CD23, CD24, and CD93. T2
cells differ from T1 cells in having higher levels of mIgD and
in expressing CD21 (the complement receptor and B-cell
co-receptor; see Figure 3-7) and CD23. T2 cells also express
BAFF-R, the receptor for the B-cell survival factor BAFF,
whose expression is dependent on signals received through
the BCR. As B cells differentiate from the transitional T2
state to full maturity, they raise their levels of mIgD still further, while reducing the expression of mIgM. They also cease
to express CD24 and CD93.
T1 cells that have been labeled and transferred into recipient mice develop into T2 cells. Similarly, both transitional
B-cell subpopulations have been demonstrated to have the
346
PA R T I V
|
Adaptive Immunity: Development
(a) H-2d transgenics
Mature B cells express
anti -Kk
Kd
(b) H-2d/k transgenics
Immature B cells
Anti-Kk
Kd
No mature B cells express
anti-Kk
Kk
Bone marrow
stromal cell
(c) H-2d/k transgenics
Light-chain editing
A few mature B cells with new
light chains no longer bind Kk
FIGURE 10-8 Experimental evidence for negative selection
(clonal deletion) of self-reactive B cells during maturation in
the bone marrow. The presence or absence of mature peripheral B
cells expressing a transgene-encoded IgM against the H-2 class I molecule Kk was determined in H-2d mice (a) and H-2d/k mice (b) and (c).
(a) In the H-2d transgenics, the immature B cells did not bind to a self
antigen and consequently went on to mature, so that splenic B cells
expressed the transgene-encoded anti-Kk as membrane Ig. (b) In the
TABLE 10-1
H-2d/k transgenics, many of the immature B cells that recognized the
self antigen Kk were deleted by negative selection. (c) More detailed
analysis of the H-2d/k transgenics revealed a few peripheral cells that
expressed the transgene-encoded chain but a different light chain.
Apparently, a few immature B cells underwent light-chain editing, so
they no longer bound the Kk molecule and consequently escaped
negative selection. [Adapted from D. A. Nemazee and K. Burki, 1989, Nature
337:562; S. L. Tiegs et al., 1993, Journal of Experimental Medicine 177:1009.]
Expression of transgene encoding IgM antibody to H-2k class I MHC molecules
Expression of transgene
Experimental animal
Number of animals tested
As membrance Ab
As secreted Ab (g/ml)
Nontransgenics
13
()
0.3
H-2d transgenics
7
()
93.0
H-2d/k transgenics
6
()
0.3
[Adapted from D. A. Nemazee and K. Bürki, 1989, February, Clonal deletion of B lymphocytes in a transgenic mouse bearing anti-MHC class I antibody genes, Nature 337:562–566.
doi:10.1038/337562a0.]
B-Cell Development
|
CHAPTER 10
347
Bone marrow
sinusoids, bloodstream
Endosternum
T cell zone
Pro and
pre B
Immature B
Follicle
Central
arteriole
Bone marrow
T2
IgM
IgD
T1
Marginal sinus
Marginal zone
FIGURE 10-9 T2, but not T1, transitional immature cells
can enter the B-cell follicles and recirculate. Immature B cells
leave the bone marrow as T1 transitional immature B cells. They enter
the spleen from the bloodstream through the marginal sinuses, percolating into the T-cell zones, and differentiating into T2 transitional B
cells, which gain the ability to enter the B-cell follicles and recirculate.
There, the T2 cells complete their differentiation into mature follicular
B-2 cells. Marginal zone cells have also been shown to derive from T2
cells. [Adapted from J. B. Chung, M. Silverman, and J. G. Monroe, 2003, Transi-
capacity to differentiate into mature B cells. These experiments
have therefore proven that the order of the developmental
sequence progresses from T1 to T2 to mature B cell. The time
in transit of a T1 cell to a mature B cell has been measured to
be approximately 3 to 4 days. Most T1 B cells differentiate to
T2 cells within the spleen, but a minority of about 25% of transitional B cells emerge from the bone marrow already in the
T2 state. The increased level of maturity of T2 cells correlates
with changes in the expression of chemokine and cytokine
receptors, such that T2 cells, but not T1 cells, are capable of
recirculating among the blood, lymph nodes, and spleen; and
T2 cells, but not T1 cells, can enter B-cell follicles.
Figure 10-9 shows the path of the developing transitional
B cell as it leaves the bone marrow and enters the spleen
through the central arteriole, which deposits it in the marginal sinuses, just inside the outer marginal zone. (In
humans, the anatomy of the spleen is slightly different, and
the cells arrive in the spleen in a peri-follicular zone.) From
there, the T1 B cell percolates through to the T-cell zone,
where some fraction of T1 cells will mature into the T2 state.
T2 B cells are then able to enter the follicles or the marginal
zone where they complete their developmental program into
fully mature, recirculating B lymphocytes.
In Chapter 7, we learned that mature B cells bear on their
surfaces two classes of membrane-bound immunoglobulins—
IgM and IgD—and that the expression of mIgD along with
mIgM requires carefully regulated mRNA splicing events. It is
at the point of transit between the T1 and T2 stages of development that we observe the onset of these splicing capabilities.
Mature B cells bear almost 10 times more mIgD than mIgM,
and so mIgD expression results in significant up-regulation in
the number of B-cell immunoglobulin receptors.
The effect of strong BCR engagement with a multivalent, or
membrane-bound, antigen depends on the maturational status
of the transitional B cell (Figure 10-10 and Table 10-3). Selfreactive T1 B cells are eliminated by apoptosis in response to a
strong antigenic signal, in a process reminiscent of thymocyte
negative selection, leading to peripheral tolerance; recent experiments have suggested that in healthy adults, fully 55% to 75%
of immature B cells are lost by this process. In contrast, once the
B cell has matured into a T2 transitional B cell, it becomes resistant to antigen-induced apoptosis, reminiscent of thymocytes
that have reached the single-positive stage of development. This
TABLE 10-2
Surface marker expression on
transitional T1 and T2 and
mature B-2 B cells
T1
T2
Mature
B-2 cells
mIgM
High
High
Intermediate
mIgD
Low
Intermediate
High
CD24
CD93
CD21
CD23
/
Marker
BAFF receptor
(BAFF-R)
Note: CD93 is defined by the AA4 monoclonal antibody. CD24 is otherwise
known as the Heat Stable Antigen (HSA). CD23 is a low-affinity receptor for
IgE. CD21 is a receptor for complement and part of the B-cell co-receptor.
After D. Allman and S. Pillai. 2008. Peripheral B cell subsets. Current Opinion
in Immunology 20:149–157, and others.
tional B cells: Stem by step towards immune competence, Trends in Immunology
24:343–348, Figure 1; and R. E. Mebius, and G. Kraal, 2005, Structure and function
of the spleen. Nature Reviews Immunology 5:606–616, Figure 1.]
PA R T I V
348
PALS
|
Adaptive Immunity: Development
Follicle
T1
T2
Negative
selection
Positive
selection
Death
Death
B-2
FIGURE 10-10 Transitional B cells bound for a follicular
fate undergo positive and negative selection in the spleen. T1
transitional B cells, which recognize antigen with high affinity in the
spleen, are eliminated by negative selection, and never reach the
splenic follicles. Those T1 cells that escape negative selection enter the
follicles and differentiate into T2 B cells. In the follicles, their BCRs interact with an unknown molecule(s) that deliver(s) a stimulatory survival
signal. Transitional cells that have received this survival signal up-regulate their BAFF receptors (positive selection). Those T2 B cells that fail
to receive the stimulatory signal (or that fail to receive a BAFF survival
signal) die in the spleen. Selecting antigens are shown as violet shapes;
T1 and T2 cells are green. White cells represent dead cells that were
either negatively selected or failed positive selection. [Adapted from
T. T. Su et al., 2004, Signaling in transitional type 2 B cells is critical for peripheral
B-cell development, Immunological Reviews 197:161–178, Figure 3.]
resistance to receptor-induced cell death results in part from
the fact that T2 B cells have increased their expression of the
anti-apoptotic molecule Bcl-xl.
Using conditional knockout genetic techniques (see
Chapter 20), animals can be generated that lack Ig receptor
TABLE 10-3
Responses to strong BCR
signaling in T1, T2, and
mature B-2 B cells
Nature of Response
T1
T2
Mature B-2
B cells
Formation of lipid rafts
/
Increase in cytoplasmic
Ca2 ion concentrations
Increase in diacylglycerol
concentrations
Induction of Bcl-xl
Induction of apoptosis
expression at different stages of B-cell development. Animals
that fail to express a BCR during the immature B-cell stage
lose the capacity to make any mature B cells at all. This indicates that some low level of tonic signaling through the BCR,
analogous to the positive selection signal needed by developing thymocytes, is required for continued generation and
survival of immature B cells. B cells unable to receive this
signal die at the T2 stage (see Figure 10-10). Those T2 B cells
able to receive the follicular signal up-regulate the expression of the receptor for the B-cell survival factor BAFF.
Given the different outcomes of signaling through the
BCR for T1 versus T2 B cells, it is clear that there must be
differences in the signaling pathways downstream of the BCR
in the two transitional B-cell types. Specifically, BCR-mediated
signaling of T1 B cells results in calcium release without
significant production of diacylglycerol, and provides an
apoptotic signal. In contrast, receipt of BCR signals by T2 B
cells induces both an increase in the concentration of intracytoplasmic calcium and in diacylglycerol production. This
combination of intracellular second messengers delivers
both maturational and survival signals to the cell and suggests the involvement of a diacylglycerol-activated protein
kinase in survival signaling (see Chapter 3).
But what causes this difference in the signal transduction
pathways between T1 and T2 cells? A partial answer to this
question appears to lie in differences in the composition of
the lipid membranes of the two types of cells. Immature T1
B cells contain approximately half as much cholesterol as
their more mature counterparts, and this reduction in cholesterol levels appears to prevent efficient clustering of the
B-cell receptor into lipid rafts upon BCR stimulation. This
may cause a reduction of the strength of BCR signaling in T1
versus T2 immature B cells.
The development of B cells through the transitional phase is
absolutely dependent on signaling through the BAFF receptor
(BAFF-R). BAFF-R expression is first detected in T1 B cells
and increases steadily thereafter. BAFF is then required constitutively throughout the life of mature B cells. Signaling through
the BAFF/BAFF-R axis promotes survival of transitional B cells
by inducing the synthesis of anti-apoptotic factors such as Bcl2, Bcl-xl, and Mcl-1, as well as by interfering with the function
of the pro-apoptotic molecule Bim (see Figure 9-12).
The discovery of a B-cell survival signal mediated by
BAFF/BAFF-R interactions, and distinct from survival signals
emanating from the BCR, extends our thinking about how B
cells are selected for survival in the periphery. In the presence
of high levels of BAFF, B cells that may not otherwise receive
sufficient quantities of survival signals via the BCR may survive a selection process that would otherwise eliminate them.
In this way, BAFF can provide plasticity and flexibility in the
process of B-cell deletion. However, this may be accomplished
at the cost of maintaining potentially autoreactive cells.
T3 B Cells Are Primarily Self-Reactive and Anergic
Transitional T3 B cells were first characterized in the blood
and lymphoid organs by flow cytometry and were described
B-Cell Development
as being CD93mIgMlowCD23. The function of CD93 is so
far unknown; CD23 is a low-affinity receptor for some
classes of immunoglobulin, and the two markers were used
in these experiments only to identify the cell populations,
and not because of any particular reasons relevant to their
functionality. Recent experiments have suggested that the T3
population may represent B cells that have been rendered
anergic by contact with soluble self antigen but have not yet
been eliminated from the B-cell repertoire.
A transgenic system developed by Goodnow and colleagues first placed the concept of B-cell anergy, or unresponsiveness, onto a firm experimental footing. Anergic
lymphocytes clearly recognize their antigens, as shown by the
identification of low levels of molecular signals generated
within the cells after binding to antigen. However, rather than
being activated by antigen contact, anergic B cells fail to
divide, differentiate, or secrete antibody after stimulation,
and many die a short time after receipt of the antigenic signal.
Goodnow et al. developed the two groups of transgenic
mice illustrated in Figure 10-11a. One group of mice carried
a hen egg-white lysozyme (HEL) transgene linked to a
metallothionine promoter, which placed transcription of the
HEL gene under the control of zinc levels in the animals’ diet.
This allowed the investigators to alter the levels of soluble
HEL expressed in the experimental animals by changing the
concentration of zinc in their food. Under these experimental
conditions, HEL was expressed in the periphery of the animal, but not in the bone marrow. The other group of transgenic mice carried rearranged immunoglobulin heavy-chain
and light-chain transgenes encoding anti-HEL antibody; in
these transgenic mice, the rearranged anti-HEL transgene is
expressed by 60% to 90% of the mature peripheral B cells.
Goodnow then mated the two groups of transgenics to produce “double-transgenic” offspring carrying both the HEL
and anti-HEL transgenes (Figure 10-11b) and asked what
effect peripheral HEL expression would have on antibody
expression by B cells bearing the anti-HEL antibodies. They
found that the double-transgenic mice continued to generate
mature, peripheral B cells bearing anti-HEL membrane
immunoglobulin of both the IgM and IgD classes, indicating
that the B cells had fully matured. However, these B cells were
functionally nonresponsive, or anergic.
Flow-cytometric analysis of B cells from the doubletransgenic mice showed that, although large numbers of
anergic anti-HEL cells were present, they expressed membrane
IgM at levels about 20-fold lower than anti-HEL single transgenics (Figure 10-11b). When these mice were given an immunizing dose of HEL, few anti-HEL plasma cells were induced
and the serum anti-HEL titer was very low (Table 10-4). Furthermore, when antigen was presented to these anergic B cells
in the presence of T-cell help, many of the anergic B cells
responded by undergoing apoptosis. Additional analysis of
the anergic B cells demonstrated that they had a shorter halflife than normal B cells and appeared to be excluded from
the B-cell follicles in the lymph nodes and spleen. These
properties were dependent on the continuing presence of
|
CHAPTER 10
349
antigen, as the B cell half-lives were restored to normal
lengths upon adoptive transfer of the transgene-bearing B cells
to an animal that was not expressing HEL. The anergic response
appears to be generated in vivo when immature B cells meet
a soluble self antigen.
More recent experiments have focused on defining the differences between the signal transduction events leading to
anergy versus activation. Anergic B cells show much less antigen-induced tyrosine phosphorylation of signaling molecules,
when compared with their nonanergic counterparts and antigen-stimulated calcium release from storage vesicles into the
cytoplasm of the anergic B cells was also dramatically reduced.
Anergic B cells also require higher levels of the cytokine BAFF
for continued survival, and it is likely that their reduced halflives result from unsuccessful competition with normal B cells
for limiting amounts of this survival molecule. One of the outcomes of BAFF signaling is a reduction in the cytoplasmic
levels of the pro-apoptotic molecule BIM; as might be expected,
anergic B cells show higher-than-normal levels of BIM and a
correspondingly increased susceptibility to apoptosis.
The conclusion from these experiments is that mechanisms exist, even after the B cells have exited the bone marrow and entered the periphery, that minimize the risk that B
cells make antibodies to soluble self proteins expressed outside the bone marrow. B cells reactive to such proteins
respond to receptor stimulation in the absence of appropriate T-cell help by anergy and eventual apoptosis.
What might be the function of these anergic cells? One possibility is that they serve to absorb excess self antigens that
might otherwise be able to deliver activating signals to highaffinity B cells and thus lead to autoimmune reactions. Another
is that they represent cells destined for apoptosis that do not yet
display the characteristic microanatomy of apoptotic cells. Yet
another is that these cells will eventually develop into B regulatory cells (see Chapter 12). As is so often the case in the
immune system, it is more than possible that all of these functions are subsumed within this intriguing cell population,
which remains the subject of intensive current investigation.
Mature, Primary B-2 B Cells Migrate to the
Lymphoid Follicles
Fully mature B cells express high levels of IgD and intermediate levels of IgM on their cell surfaces (see Table 10-2).
Mature B cells recirculate between the blood and the lymphoid organs, entering the B-cell follicles in the lymph nodes
and spleen, and responding to antigen encounter in the presence of T-cell help with antibody production (Chapter 12).
Approximately 10 million to 20 million B cells are produced
in the bone marrow of the mouse each day, but only about
10% of this number ever take up residence in the periphery
and only 1% to 3% will ever enter the recirculating follicular
B-2 B-cell pool. Some of these cells are lost to the process of
clonal deletion, but others are perfectly harmless B cells that
nonetheless fail to thrive. Experimental depletion of the
mature B-cell population, either chemically or by irradiation,
350
PA R T I V
|
Adaptive Immunity: Development
(a)
×
Transgenic
(HEL)
Transgenic
(anti–HEL)
Double transgenic
(carrying both HEL and anti–HEL transgenes)
HEL-binding B cells
(b)
Nontransgenic
Anti–HEL transgenic
Anti–HEL/HEL
double transgenic
100
10
1
1
1
10
100
1
10
10
100
IgM expression on membrane (arbitrary fluorescence units)
100
onstrating clonal anergy in mature peripheral B cells. (a)
Production of double-transgenic mice carrying transgenes encoding HEL (hen egg-white lysozyme) and anti-HEL antibody. (b) Flowcytometric analysis of peripheral B cells that bind HEL compared
with membrane IgM levels. The number of B cells binding HEL was
measured by determining how many cells bound fluorescently
labeled HEL. Levels of membrane IgM were determined by incubating the cells with anti-mouse IgM antibody labeled with a fluorescent label different from that used to label HEL. Measurement of the
fluorescence emitted from this label indicated the level of membrane IgM expressed by the B cells. The nontransgenics (left) had
many B cells that expressed high levels of surface IgM but almost no
B cells that bound HEL above the background level of 1. Both antiHEL transgenics (middle) and anti-HEL/HEL double transgenics
(right) had large numbers of B cells that bound HEL (blue), although
the level of membrane IgM was about 20-fold lower in the double
transgenics. The data in Table 10-4 indicate that the B cells expressing anti-HEL in the double transgenics cannot mount a humoral
response to HEL.
followed by in vivo reconstitution, results in rapid replenishment of the B-cell follicular pool. This suggests that the follicular B-cell niches have a designated capacity and that once
full they turn away additional B cells. Most probably, the
mechanism for this homeostatic control of B-cell numbers
relies on competition for survival factors, particularly BAFF
and its related proteins.
Experiments using conditional RAG2 knockout animals, in which all new B-cell development was prevented
in otherwise healthy adult animals, indicate that follicular
B-2 B cells have a half-life of approximately 4.5 months. In
contrast, since B-1 B cells can self-renew in the periphery,
their numbers are unaffected in this experimental knockout animal.
FIGURE 10-11 Goodnow’s experimental system for dem-
TABLE 10-4
Expression of anti-HEL transgene by mature peripheral B cells in single- and
double-transgenic mice
Experimental Group
Anti-HEL single transgenics
Anti-HEL/HEL double transgenics
HEL level
Membrane anti-HEL Ig
Anti-HEL PFC/spleen*
Anti-HEL serum titer
None
High
High
9
Low
Low
10
M
* Experimental animals were immunized with hen egg-white lysozyme (HEL). Several days later hemolytic plaque assays for the number of plasma cells secreting antiHEL antibody were performed and the serum anti-HEL titers were determined. PFC plaque-forming cells.
Adapted from Goodnow, C. C., 1992, Annual Review of Immunology 10:489.
|
B-Cell Development
The Development of B-1 and
Marginal-Zone B Cells
This chapter has so far focused on the development of those
B cells that belong to the best characterized B-cell subpopulation, B-2 B cells (or follicular B cells). Mature B-2 B cells
recirculate between the blood and the lymphoid organs, and
can be found in large numbers in the B-cell follicles of the
lymph nodes and spleen. However, other subsets of B cells
have been recognized that perform distinct functions,
occupy distinct anatomical locations, and pursue different
developmental programs. This section of the chapter will
therefore address the development and function of B-1 B
cells and of marginal-zone B cells (see Figure 10-12 for a
comparison of the properties of these three cell types).
As described more fully in Chapter 12, B-1 B cells generate antibodies against antigens shared by many bacterial
species and may do so even in the absence of antigenic
stimulation. They are the source of the so-called natural
antibodies: serum IgM antibodies that provide a first line of
protection against invasion by many types of microorganisms. Marginal-zone B cells take their name from their location in the outer zones of the white pulp of the spleen. They
are the first B cells encountered by blood-borne antigens
351
entering the spleen and, like B-1 B cells, mainly (although
not exclusively) produce broadly cross-reactive antibodies of
the IgM class. Both B-1 and marginal-zone B cells can generate antibodies in the absence of T-cell help, although the
addition of helper T cells enhances antibody secretion and
allows for some degree of heavy-chain class switching.
B-1 B Cells Are Derived from a Separate
Developmental Lineage
B-1 B cells are phenotypically and functionally distinct from
B-2 B cells in a number of important ways. They occupy different anatomical niches from B-2 B cells, constituting 30% to
50% of the B cells in the pleural and peritoneal cavities of
mice, and representing about 1 million cells in each space. A
similar number of B-1 B cells can also be found in the spleen,
but there they represent a much smaller fraction (around 2%)
of the splenic B-cell population. B-1 B cells have only a relatively limited receptor repertoire, and their receptors tend to be
directed toward the recognition of commonly expressed
microbial carbohydrate antigens. These broadly cross-reactive,
low-affinity antigen receptors expressing minimal repertoire
diversity are reminiscent of the pathogen-associated molecular pattern (PAMP) receptors of the innate immune system
Marginal zone B cells
B-1 B cells
Follicular (B-2) B cells
CD19/
CD21
CHAPTER 10
IgM
CD19/
CD21
IgM
IgM
IgD
CD1
CD5
CD23
(B-1a cells only)
Attribute
Follicular (B-2) B cells
B-1 B cells
Marginal zone B cells
Major sites
Secondary lymphoid organs
Peritoneal and pleural cavities
Marginal zones of spleen
Source of new B cells
From precursors in bone
marrow
Self-renewing (division of existing
B-1 cells)
Long-lived
May be self-renewing
V-region diversity
Highly diverse
Restricted diversity
Somewhat restricted
Somatic hypermutation
Yes
No
Unclear
Requirements for T-cell help
Yes
No
Variable
Isotypes produced
High levels of IgG
High levels of IgM
Primarily IgM; some IgG
Response to carbohydrate
antigens
Possibly
Yes
Yes
Response to protein antigens
Yes
Possibly
Yes
Memory
Yes
Very little or none
Unknown
Surface IgD on mature B cells
Present on naïve B cells
Little or none
Little or none
FIGURE 10-12 The three major populations of mature B cells in the periphery. The cell-surface properties and functions of B-2,
B-1, and marginal-zone (MZ) cells are shown. Conventional B-2 cells were so named because they develop after B-1 B cells.
352
PA R T I V
|
Adaptive Immunity: Development
(Chapter 5), and B-1 B cells are thus considered to play a role
that bridges those of the innate and adaptive immune systems.
In contrast with the T-1 transitional B-2 B cells, which
undergo apoptosis upon antigen challenge, transitional B-1
cells undergo apoptosis unless they interact with self antigens.
Work in a number of different transgenic systems has suggested
that relatively strong BCR engagement by self antigens provides a positive selection rather than a negative selection survival
signal for B-1 B cells. In contrast with B-2 cells and marginalzone cells (see below), B-1 B cells do not require interaction
with BAFF during the transitional stage of development.
For many years, the preeminent issue debated by those
interested in the development of B-1 B cells was whether they
constituted a separate developmental lineage, or whether they
derived from the same progenitors as B-2 B cells. This controversy has since been resolved in favor of the assertion that B-1
and B-2 B cells derive from distinct lineages of progenitor
cells. Several lines of evidence support this conjecture:
• B-1 B cells appear before B-2 B cells during ontogenic
development. B cells generated from the AGM region and
the liver in the fetus have a cell-surface phenotype characteristic of B-1 B cells and secrete natural IgM antibodies
without the need for deliberate immunization. B-1 B cells
may be generated early in development in order to protect
the fetus from commonly encountered bacterial pathogens.
• B-1 B cells display much more limited V region diversity
than B-2 B cells. Generated at a point in ontogeny before
TdT can be efficiently activated, many B-1 cells lack any
evidence of N region addition.
• B-1 B cells populate different anatomical niches in the
mouse than the later-arising B-2 B cells, in much the
same manner that has been described for the fetally
derived T cells (Chapter 9).
• B-cell progenitors of the CD19CD45Rlow/ phenotype
transferred into an immunodeficient mouse were able to
repopulate the B-1, but not the B-2 B-cell compartments.
Conversely, CD19CD45Rhi B-cell progenitors gave rise to
B-2, but not to B-1, daughter cells in an immunodeficient
recipient mouse, supporting the notion that the two subclasses derive from different lineages of progenitor cells.
• Whereas B-2 B cells must be constantly replenished by
the emergence of newly generated cells from the bone
marrow, B-1 B cells are constantly regenerated in the
periphery of the animal. Bone marrow ablation therefore
leaves the mouse with a depleted B-2 pool, but with a
fully functional B-1 population.
There are very few absolutes in biology, and so it should
be clearly stated that it is probable that not all B-1 B cells in
an adult animal are derived from fetal liver precursors and
that some replenishment of B-1 precursors occurs in the
adult bone marrow. Not all B-1 B cells are restricted to IgM
production, and the antibody production of some B-1 B cells
does benefit from the provision of T-cell help. But notwithstanding these variations on the general themes we have
elaborated, it remains the case that evidence clearly supports
the notion that B-1 B cells are derived from a different progenitor cell lineage from B-2 B cells.
Marginal-Zone Cells Share Phenotypic and
Functional Characteristics with B-1 B Cells and
Arise at the T2 Stage
The term marginal zone refers to the fact that marginal zone
(MZ) B cells are located in the outer regions of the white pulp
of the spleen (Figure 10-13). The spreading out of the fastmoving arterial flow into the marginal sinuses results in a
decrease in the rate of blood flow and allows blood-borne
antigens to interact with cells resident within the marginal
zone. Indeed, MZ B cells appear to be specialized for recognizing blood-borne antigens. They are capable of responding
to both protein and carbohydrate antigens, and evidence
suggests that some, but maybe not all, MZ B cells can do so
without the need for help from T cells.
MZ cells are characterized by relatively high levels of
membrane IgM, and the complement receptor/B-cell coreceptor CD21 (see Chapters 3 and 6), but low levels of
membrane IgD and the Fc receptor CD23. They also display
phospholipid receptors and adhesion molecules that enable
them to make adhesive interactions with other cells within
the marginal zone that hold them in place. MZ cells are long
lived and may self-renew in the periphery.
What drives an immature B cell down the MZ pathway?
Phenotypic characterization of bone marrow and peripheral
immature B-cell populations, combined with labeling studies
that define precursor/daughter-cell relationships have suggested that MZ and follicular B-2 B cells both derive from the
T2 transitional population (see Figure 10-9). MZ cells, like
B-2 B cells, are also reliant on the B-cell survival factor BAFF,
which binds to the MZ B cells through the receptor BR3. Like
developing B-1 B cells, developing MZ cells appear to require
relatively strong signaling through the BCR in order to survive. Surprisingly, unlike any other B-cell subset so far
described, the differentiation of MZ B cells also requires signaling through ligands of the Notch pathway (see Chapter 9).
Loss of the Notch 2 receptor or deletion of the Notch 2 ligand,
Delta-like 1 (Dl-1), results in the selective deletion of MZ B
cells. Both MZ and B-1 B cell populations are enriched in cells
that express self antigen-specific receptors, and so relatively
strong signaling of T2 cells by binding of self antigens through
the BCR is also necessary for MZ B-cell differentiation.
Comparison of B- and T-Cell
Development
In closing this chapter, it is instructive to consider the many
points of comparison between the development of the two
arms of the adaptive immune system, B cells and T cells
(Table 10-5). Both cell lineages have their beginnings in the
fetus and the neonate. In the neonate, T cells and B-1 B
cells are dispatched to their own particular peripheral niches,
Central arteriole
T cell zone
Follicle
B cell zone
Marginal sinus
Marginal zone
FIGURE 10-13 The relative locations of the marginal zone and the follicles in the spleen. This figure shows a cross-section of the
spleen, displaying the anatomical relationships between the central arteriole, the T-cell and B-cell zones, and the marginal zone.
TABLE 10-5 Comparison between T-cell and B-cell development
Structure or process
Development begins in the bone marrow.
B cells
T cells
Development continues in the thymus.
Ig heavy chain or TCR chain many gene rearrangement begins
with D-J and continues with V-DJ recombination.
The H chain (BCR) or chain (TCR) is expressed with a
surrogate form of the light chain on the cell surface. Signaling
from this pre-BCR or pre-TCR is necessary for development
to continue.
Signaling through the pre-B- or pre-TCRs results in proliferation.
The / (BCR) or (TCR) chain bears only V and J segments.
Signaling from the completed receptor is necessary for survival
(positive selection).
/ True for the minority B-1
B-cell subset. Ligand for B-2
and MZ subsets unclear.
Low-affinity binding of self
MHC and self peptide in the
thymus is necessary for positive selection.
Receptor editing of / (BCR) or (TCR) chain modifies autoreactive specificities.
Has been shown to occur, but
rarely used as a mechanism
of escaping autoreactivity.
Immature cells bearing high affinity autoreactive receptors are
eliminated by apoptosis (negative selection).
Negative selection involves ectopic expression of auto-antigens
in the primary lymphoid organs.
Negative selection involves recognition of MHC-presented
peptides.
Positive selection requires recognition of self components.
Auto-antigens are
expressed in the thymus
under the influence of the
AIRE transcription factor.
Heavy or TCR chain allelic exclusion.
Light ( /) or TCR chain allelic exclusion.
90% of lymphocytes are lost prior to export to the periphery.
Development is completed in the periphery to allow for tolerance to antigens not expressed in the primary lymphoid organs
Many T cells display more
than one TCR chain.
Note: The comparisons in this table most accurately refer to the predominant T and follicular B-2 cell subsets, although many of the statements made are equally
valid for the minority lymphocyte subsets.
353
354
PA R T I V
|
Adaptive Immunity: Development
there to function as self-renewing populations until the death
of the host. In the adult animal, B-cell and T-cell development continue in the bone marrow, starting with hematopoietic stem cells. B and T cells share the early phases of their
developmental programs, as they pass through progressively
more differentiated stages as MPPs, LMPPs, and ELPs. At the
ELP and CLP stages, T-cell progenitors leave the bone marrow, and migrate to the thymus to complete their development, leaving B-cell progenitors behind. In mammals, B cells
do not have an organ analogous to the thymus in which to
develop into mature, functioning cells, although birds do
possess such an organ—the bursa of Fabricius. (Many students are unaware that it is from this organ, and not from the
bone marrow, that the “B” in B cells originates.)
With the initiation of V(D)J recombination, the B cell
irreversibly commits to its lineage, and begins the process of
receptor rearrangement, followed by B-cell selection and differentiation that will culminate in the formation of a complete repertoire of functioning peripheral B cells.
B cells and T cells must both pass through stages of positive selection, in which those cells capable of receiving survival signals are retained at the expense of those which
cannot. They must also survive the process of negative selection, in which lymphocytes with high affinity for self antigens are deleted. The process of positive selection in B
cells—its mechanism, and the BCR ligands involved—
remains one of the least well characterized processes in
B-cell development. Unlike T cells, however, B cells do not
undergo selection with respect to their ability to bind to self
MHC antigens.
With the expression of high levels of IgD on the cell surface and the necessary adhesion molecules to direct their
recirculation, development of the mature, follicular B-2 cell
is complete and, for a few weeks to months, it will recirculate, ready for antigen contact in the context of T-cell help
and subsequent differentiation to antibody production. For
the final, antigen-stimulated stages of B-cell differentiation,
the reader is directed to Chapter 12.
S U M M A R Y
■
■
■
■
■
■
■
Hematopoiesis in the embryo generates rapidly dividing
hematopoietic stem cells that populate the blood system of
the animal, providing red blood cells that supply the oxygen needs of the fetus, and generating the early precursors
of other blood-cell lineages. Early sites of blood-cell development include the yolk sac and the placenta, as well as
the aorta-gonad-metanephros (AGM) region and the fetal
liver.
The earliest B-cell progenitors in the fetal liver supply B-1
B cells that migrate into the pleural and peritoneal cavities and remain self-renewing throughout the life of the
animal.
In adult animals, hematopoiesis occurs in the bone marrow.
Progenitor stages of the conventional B-2 subset are
defined by the presence of particular cell-surface markers,
which include chemokine and lymphokine receptors and
proteins involved in adhesive interactions.
Progenitor stages are also defined by the status of immunoglobulin V region gene rearrangements. The heavychain V genes rearrange first, with D to JH recombination
occurring initially, followed by VH to DJH recombination.
The heavy chain is then expressed on the cell surface in
combination with the surrogate light chain, which is made
up of VpreB and 5. Together they form the pre-B-cell
receptor, which is expressed on the cell surface along with
the Ig,Ig signaling complex.
Signaling through the B-cell receptor stops VH gene rearrangement and calls for a few rounds of cell division. This
allows multiple B cells to use the same, successfully rear-
ranged heavy chain in combination with many different
light chains.
■
■
■
■
■
After light-chain rearrangement, and the expression of the
completed immunoglobulin receptor on the cell surface,
immature B cells specific for self antigens present in the
bone marrow are deleted by apoptosis.
Immature B cells emerge from the bone marrow as transitional 1 (T1) B cells and circulate to the spleen. Interaction
with self antigens in the spleen can give rise to apoptosis.
T1 B-2 B cells then enter the follicles, where the level of
IgD expression increases, they become T2 cells, and then
mature into either a follicular B-2 cell (a conventional B
cell) or a marginal-zone (MZ) B cell.
The three major subsets of B cells differ according to their
site of generation, their sites of maturation, their anatomical niches in the adult, the antigens to the which they
respond, their need for T-cell help in antibody production,
the diversity of their immunoglobulin repertoires, and
their abilities to undergo somatic hypermutation and
memory generation following antigenic stimulation.
Like T cells, developing B cells must undergo both positive
and negative selection. Unlike T cells, B cells need not be
selected for their ability to recognize antigens in the context of MHC antigens, nor is there a primary immune
organ aside from the bone marrow specialized for their
maturation. There is no equivalent of the AIRE protein
that has yet been discovered to provide for ectopic expression of antigens in the bone marrow in order to facilitate
clonal deletion of self antigen-specific B cells.
B-Cell Development
|
CHAPTER 10
355
R E F E R E N C E S
Allman, D., and Pillai, S. 2008. Peripheral B cell subsets. Current
Opinion in Immunology 20:149–157.
Cambier, J. C., S. B. Gauld, K. T. Merrell, and B. J. Vilen. 2007.
B-cell anergy: From transgenic models to naturally occurring
anergic B cells? Nature Reviews Immunology 7:633–643.
Carsetti, R., M. M. Rosado, and H. Wardmann, H. 2004. Peripheral development of B cells in mouse and man. Immunological
Reviews 197:179–191.
Casola, S. 2007. Control of peripheral B-cell development. Current Opinions in Immunology 19:143–149.
Chung, J. B., R. A. Sater, M. L. Fields, J. Erikson, and J. G. Monroe. 2002. CD23 defines two distinct subsets of immature B
cells which differ in their responses to T cell help signals.
International Immunology 14:157–166.
Dorshkind, K., and E. Montecino-Rodriguez. 2007. Fetal B-cell
lymphopoiesis and the emergence of B-1-cell potential. Nature
Reviews Immunology 7:213–219.
Dzierzak, E., and N. A. Speck. 2008. Of lineage and legacy: The
development of mammalian hematopoietic stem cells. Nature
Immunology 9:129–136.
Nemazee, D. 2006. Receptor editing in lymphocyte development and central tolerance. Nature Reviews Immunology
6:728–740.
Nutt, S. L., and Kee, B. L. 2007. The transcriptional regulation of
B cell lineage commitment. Immunity 26:715–725.
Palis, J., et al. 2001. Spatial and temporal emergence of high
proliferative potential hematopoietic precursors during
murine embryogenesis. Proceedings of the National Academy
of Sciences of the United States of America 98:4528–4533.
Perez-Vera, P., A. Reyes-Leon, and E. M. Fuentes-Panana. 2011.
Signaling proteins and transcription factors in normal and
malignant early B cell development. Bone Marrow Research
2011:502,751.
Pillai, S., and A. Cariappa. 2009. The follicular versus marginal
zone B lymphocyte cell fate decision. Nature Reviews Immunology 9:767–777.
Rajewsky, K., and H. von Boehmer. 2008. Lymphocyte development: Overview. Current Opinions in Immunology
20:127–130.
Fuxa, M., and J. A. Skok. 2007. Transcriptional regulation in
early B cell development. Current Opinions in Immunology
19:129–136.
Ramirez, J., K. Lukin, and J. Hagman. 2010. From hematopoietic progenitors to B cells: Mechanisms of lineage restriction
and commitment. Current Opinions in Immunology 22:177–
184.
Hoek, K. L., et al. 2006. Transitional B cell fate is associated with
developmental stage-specific regulation of diacylglycerol and
calcium signaling upon B cell receptor engagement. Journal of
Immunology 177:5405–5413.
Srivastava, B., W. J. Quinn, 3rd, K. Hazard, J. Erikson, and D.
Allman. 2005. Characterization of marginal zone B cell
precursors. Journal of Experimental Medicine 202:1225–
1234.
Kee, B. L. 2009. E and ID proteins branch out. Nature Reviews
Immunology 9:175–184.
Tokoyoda, K., T. Egawa, T. Sugiyama, B. I. Choi, and T. Nagasawa, T. 2004. Cellular niches controlling B lymphocyte behavior within bone marrow during development. Immunity
20:707–718.
Koralov, S. B., et al. 2008. Dicer ablation affects antibody diversity and cell survival in the B lymphocyte lineage. Cell
132:860–874.
Kurosaki, T., H. Shinohara, and Y. Baba. 2010. B cell signaling
and fate decision. Annual Review of Immunology 28:21–55.
Liao, D. 2009. Emerging roles of the EBF family of transcription
factors in tumor suppression. Molecular Cancer Research
7:1893–1901.
Mackay, F., and P. Schneider. 2009. Cracking the BAFF code.
Nature Reviews Immunology 9:491–502.
Malin, S., et al. 2010. Role of STAT5 in controlling cell survival
and immunoglobulin gene recombination during pro-B cell
development. Nature Immunology 11:171–179.
Mikkola, H. K., and S. H. Orkin. 2006. The journey of developing hematopoietic stem cells. Development 133:3733–3744.
Monroe, J. G., and K. Dorshkind. 2007. Fate decisions regulating bone marrow and peripheral B lymphocyte development.
Advances in Immunology 95:1–50.
Nagasawa, T. 2006. Microenvironmental niches in the bone
marrow required for B-cell development. Nature Reviews
Immunology 6:107–116.
Tung, J. W., and L. A. Herzenberg. 2007. Unraveling B-1 progenitors. Current Opinions in Immunology 19:150–155.
von Boehmer, H., and F. Melchers. 2010. Checkpoints in lymphocyte development and autoimmune disease. Nature Immunology 11:14–20.
Whitlock, C. A., and O. Witte. 1982. Long-term culture of B
lymphocytes and their precursors from murine bone marrow.
Proceedings of the National Academy of Sciences of the United
States of America 79:3608–3612.
Xiao, C., and K. Rajewsky. 2009. MicroRNA control in the
immune system: Basic principles. Cell 136:26–36.
Yin, T., and L. Li. 2006. The stem cell niches in bone. Journal of
Clinical Investigation 116:1195–1201.
Useful Web sites
www.bio.davidson.edu/courses/immunology/
Flash/Bcellmat.html An unusual animation of B-cell
development.
PA R T I V
356
S T U D Y
|
Adaptive Immunity: Development
Q U E S T I O N S
1. You wish to study the development of B-1 B cells in the
absence of the other two major B-cell subsets. You have a
recipient Rag1/ mouse that you have already repopulated with T cells. What would you choose to be your
source of B-1 progenitors and why? Which anatomical sites
would you expect to harvest the B-1 B cells from?
(b)
2. Describe the phenotypic and functional differences
between T1 and T2 immature B cells.
3. Following expression of the pre-B-cell receptor on the pro-
genitor B-cell surface, the B cell undergoes a few rounds of
cell division. What purpose does this round of division
serve in the development of the B-cell repertoire?
4. Immature B cells bearing potentially autoimmune recep-
tors can be managed in three ways to minimize the probability of disease. Describe these three strategies, noting
whether they are shared by T-cell progenitors.
5. You suspect that a new transcription factor is expressed at
the pre-pro-B-cell stage of development. How would you
test your hypothesis? What is the status of heavy-and lightchain rearrangement at this stage of development and how
would you test it?
6. How would you determine whether a particular stage of
B-cell development occurs in association with a stromal
cell that expresses CXCL12?
7. Describe the order in which B-cell receptor genes undergo
rearrangement, indicating at what steps you might expect
to see the B cell express one or both chains on the cell surface. In what sense(s) does this gene rearrangement process mimic the analogous progression in T cells, and in
what ways do the two processes differ?
ANALYZE THE DATA The two columns of data in the following
figure below are flow cytometric plots that describe the levels of
the antigens denoted on the x and y axes. The left column represents the antigens present on spleen (part A) and bone marrow (parts B) from wild-type (genetically normal) animals. The
plots represent all lymphocytes in the spleen (part A) or B-cell
progenitor and precursor cells in the bone marrow (part B).
The right column shows the same plots from animals in which
the Dicer gene has been knocked out. As you recall, the Dicer
gene is required for the maturation of controlling miRNAs.
a. For each pair of plots, describe the differences in the cell
populations, indicating whether the differences reflect
losses or gains in particular developing B-cell populations.
b. At what point(s) in B-cell development do you think
miRNAs are functioning?
Wild type
mouse
(a)
Dicer knockout
mouse
ANALYZE THE DATA The following figure is derived from the
same paper as those above. In this case the data are expressed
as histograms, in which the y axis represents the number of
cells binding the molecule shown on the x axis, Annexin V.
Annexin V binds to phosphatidyl serine on the outer leaflet
of cell membranes. Phosphatidyl serine is found on the outer
leaflet only in cells about to undergo apoptosis. The top two
panels represent cells from a wild-type animal, and the bottom
two panels represent cells from animals in which the Dicer
gene has been knocked out.
Pro-B cells
Pre-B cells
Wild
type
mouse
Dicer
knockout
mouse
a. Does the presence of Dicer have an effect on the frac-
tion of pro-B cells undergoing apoptosis? Explain your
reasoning.
b. Does the presence of Dicer have an effect on the fraction of pre-B cells undergoing apoptosis? Again explain
your reasoning.
c. Describe one function that you now think miRNAs
fulfill in B-cell development.
11
T-Cell Activation,
Differentiation,
and Memory
T
he interaction between a naïve T cell and an
antigen-presenting cell (APC) is the initiating
event of the adaptive immune response. Prior to
this, the innate immune system has been alerted
at the site of infection or tissue damage, and APCs,
typically dendritic cells, have been activated via their
pattern recognition receptors. These cells may have
engulfed extracellular (or opsonized intracellular)
pathogens, or they may have been infected by an
intracellular pathogen. In either case, they have
processed and presented peptides from these pathogens
in complex with surface MHC class I and class II
molecules, and have made their way to a local (draining)
lymph node and/or the spleen. The APCs have taken up
residence in the T-cell zones of the lymph node or
spleen to join networks of other cells that are continually
scanned by roving naïve CD8⫹ and CD4⫹ T cells, which
recognize MHC class I-peptide and MHC class IIpeptide complexes, respectively.
We have seen that each mature T cell expresses a
unique antigen receptor that has been assembled via
random gene rearrangement during T-cell development
in the thymus (Chapter 9). Because developing T cells
undergo selection events within the thymus, each mature,
naïve T cell is tolerant to self antigens, and restricted to
self-MHC (Chapter 9). Some naïve T cells have
committed to the CD8⫹ cytotoxic T-cell lineage, some to
the CD4⫹ helper T-cell lineage. When a naïve CD8⫹ or
CD4⫹ T cell binds tightly to an MHC-peptide complex
expressed by an activated dendritic cell, it becomes
activated by signals generated through the TCR (see
Chapter 3). These signals, in concert with signals from
other factors that we will describe below, stimulate the T
cell to proliferate and differentiate into an effector cell.
As you know, naïve CD8⫹ T cells become cytotoxic
cells in response to engagement of MHC class I-peptide
combinations. Although we refer to the activation of
CD8⫹ T cells in this chapter, we will discuss their effector
functions in detail in Chapter 13. Naïve CD4⫹ T cells
become helper cells in response to engagement of MHC
class II-peptide combinations (Overview Figure 11-1).
Dendritic cell (orange) interacting with T cells
(green). [M. Rohde, HZI, Braunschweig, Germany.]
■
T-Cell Activation and the Two Signal Hypothesis
■
T-Cell Differentiation
■
T-Cell Memory
This chapter focuses on this event, which is critical for the
development of both humoral and cell-mediated
immunity, as well as the development of B-cell and CD8⫹
T-cell memory. As discussed below, CD4⫹ T cells can
differentiate into a surprising number of distinct helper
subsets, each of which has a different function in
combating infection.
In this chapter, we briefly review the cellular and
molecular events that activate T cells and then deepen
your understanding of the costimulatory interactions that
play an important role in determining the outcome of
T cell-APC interactions. We then discuss the outcomes of
naïve T-cell activation—the development of effector and
memory T cells—focusing primarily on the different fates
and functions of the CD4⫹ helper T-cell subsets that drive
the adaptive response.
Which helper subset a naïve CD4⫹ T cell becomes
depends on the types of signals (e.g., cytokines,
costimulatory signals) they receive from the dendritic
cells they engage via their TCRs. And as described in
Chapter 5, the signals dendritic cells are able to deliver
depend in large part on the pathogen to which they have
been exposed (see Figure 5-18). Investigators are still
working to understand all the variables involved in
determining the lineage choices of T helper cells, but we
introduce you to the current thinking.
357
358
PA R T V
|
Adaptive Immunity: Effector Responses
11-1
OVERVIEW FIGURE
T-Cell Activation and Differentiation
Antigen
recognition
Clonal
expansion
Activation
Differentiation
Effector
functions
Effector
CD4+
T cell
Naïve
CD4+
T cell
Activation of
macrophages,
B cells, other
cells
Memory
CD4+
T cell
IL-2R
IL-2
Cytokines
APC
Other cellular
sources of
cyotkines
Naïve
CD8+
T cell
IL-2R
IL-2
Lymphoid organs
Activation of a naïve T cell in a secondary lymphoid organ results in
the generation of effector and memory T cells. Activation requires
several receptor-ligand interactions between the T cell and a dendritic cell, as well as signals through cytokines produced by the
We also describe the known functions of the
specialized helper cells, focusing on TH1, TH2, TH17, TFH,
and TREG cells. Finally, we close the chapter with a
discussion of T-cell memory, which is dependent on
CD4⫹ T cell help, and describe both what is known and
what is currently under investigation.
A Classic Experiment box and Clinical Focus box are
offered as a pair and describe the basic research behind
the discovery of the costimulatory molecule CD28, an
essential participant in naïve T-cell activation, and then
the development of a molecular therapy for autoimmune
diseases that takes advantage of what we know about the
biology of costimulation. These boxes, together, illustrate
the powerful connections between basic research and
clinical development, which underlie translational
research, an effort to bring bench scientific discovery to
the “bedside” that has captured the imagination of many
biomedical investigators.
Effector
CD8+ T cell
(CTL)
Memory
CD8+
T cell
Killing of
infected
“target cells”;
macrophage
activation
Peripheral tissues
activating APC, as well as other supportive cells in the lymphoid
organ. Effector CD4⫹ T cells become helper T cells (TH) and secrete
cytokines that enhance the activity of many other immune cells.
Effector CD8⫹ T cells are cytotoxic cells (TC) that kill infected cells.
The Advances box describes a more recent effort to
figure out precisely how many T-cell receptors must be
engaged to initiate T-cell activation. The answer was
initially surprising, yet in hindsight may not be surprising
at all. The final Clinical Focus box discusses how a
disease, an “experiment of nature,” has helped us to better
understand the basic biology and physiological function
of the effector cells introduced in this chapter.
T-Cell Activation and
the Two-Signal Hypothesis
CD4⫹ and CD8⫹ T cells leave the thymus and enter the circulation as resting cells in the G0 stage of the cell cycle. These
naïve T cells are mature, but they have not yet encountered
antigen. Their chromatin is condensed, they have very little
cytoplasm, and they exhibit little transcriptional activity.
T-Cell Activation, Differentiation, and Memory
However, they are mobile cells and recirculate continually
among the blood, lymph, and secondary lymphoid tissues,
including lymph nodes, browsing for antigen. It is estimated
that each naïve T cell recirculates from blood through lymph
nodes and back again every 12 to 24 hours. Because only
about 1 in 105 naïve T cells is likely to be specific for any
given antigen, this large-scale recirculation increases the
chances that a T cell will encounter appropriate antigen.
If a naïve T cell does not bind any of the MHC-peptide
complexes encountered as it browses the surfaces of stromal
cells of a lymph node, it exits through the efferent lymphatics,
ultimately draining into the thoracic duct and rejoining the
blood (see Chapter 2). However, if a naïve T cell does encounter an APC expressing an MHC-peptide to which it can bind,
it will initiate an activation program that produces a diverse
array of cells that orchestrate efforts to clear infection.
Recall from Chapter 3 that a successful T cell-APC interaction results in the stable organization of signaling molecules into an immune synapse (Figure 11-2). The TCR/
MHC-peptide complexes and coreceptors are aggregated in
the central part of this synapse (central supramolecular activating complex, or cSMAC). The intrinsic affinity between
the TCR and MHC-peptide surfaces is quite low (Kd ranges
from 10⫺4 M to 10⫺7 M) and is stabilized by the activity of
several molecules which together increase the avidity (the
combined affinity of all cell-cell interactions) of the cellular
interaction. The coreceptors CD4 and CD8, which are found
in the cSMAC, stabilize the interaction between TCR and
MHC by binding MHC class II and MHC class I molecules,
respectively. Interactions between adhesion molecules and
their ligands (e.g., LFA-1/ICAM-1 and CD2/LFA-3) help to
sustain the signals generated by allowing long-term cell
interactions. These molecules are organized around the central aggregate, forming the peripheral or “p” SMAC.
However, even the increased functional avidity offered by
coreceptors and adhesion molecules is still not sufficient to
fully activate a T cell. Interactions between costimulatory
receptors on T cells (e.g., CD28) and costimulatory ligands
on dendritic cells (e.g., CD80/86) provide a second, required
signal. In addition, as you will see below, a third set of signals, provided by local cytokines (Signal 3), directs T-cell
differentiation into distinct effector cell types.
Costimulatory Signals Are Required for
Optimal T-Cell Activation and Proliferation
What evidence pointed to a requirement for a second, costimulatory signal? In 1987, Helen Quill and Ron Schwartz recognized that, in the absence of functional APCs, isolated high
affinity TCR-MHC interactions actually led to T-cell nonresponsiveness rather than activation—a phenomenon they
called T-cell anergy. Their studies led to the simple but powerful notion that not one but two signals were required for full
T-cell activation: Signal 1 is provided by antigen-specific TCR
engagement (which can be enhanced by coreceptors and
adhesion molecules), and Signal 2 is provided by contact with
|
CHAPTER 11
359
a costimulatory ligand, which can only be expressed by a functional APC. When a T cell receives both Signal 1 and Signal 2,
it will be activated to produce cytokines that enhance entry
into cell cycle and proliferation (Figure 11-3).
It is now known that Signal 2 results from an interaction
between specific costimulatory receptors on T cells and
costimulatory ligands on dendritic cells (Table 11-1). Recall
from Chapter 5 that dendritic cells and other APCs become
activated by antigen binding to PRRs, to express costimulatory ligands (e.g., CD80 and CD86) and produce cytokines
that enhance their ability to activate T cells. CD28 is the
most commonly cited example of a costimulatory receptor,
but other related molecules that provide costimulatory signals during T-cell activation have since been identified and
are also described below. Because these molecules enhance
TCR signaling, they are collectively referred to as “positive”
costimulatory receptors and ligands.
Negative costimulatory receptors, which inhibit TCR
signaling, have also been identified. Although our understanding of their specific functions is incomplete, as a group
these play important roles in (1) maintaining peripheral T-cell
tolerance and (2) reducing inflammation both after the natural course of an infection and during responses to chronic
infection. As you can imagine, the expression and activity of
negative and positive costimulatory molecules must be carefully regulated temporally and spatially. Naïve T cells, for
example, do not express negative costimulatory receptors,
allowing them to be activated in secondary lymphoid tissue
during the initiation of an immune response. On the other
hand, effector T cells up-regulate negative costimulatory
receptors at the end of an immune response, when proliferation is no longer advantageous. However, these generalizations belie the complexity of regulation of this highly important
costimulatory network, and investigators are still working to
understand the details. Below we introduce aspects of the
structure, function, expression, and, when known, regulation
of several positive and negative costimulators.
Positive Costimulatory Receptors: CD28
CD28, a 44 kDa glycoprotein expressed as a homodimer, was
the first costimulatory molecule to be discovered (see Classic
Experiment Box 11-1). Expressed by all naïve and activated
human and murine CD4⫹ T cells, all murine CD8⫹ T cells,
and, interestingly, only 50% of human CD8⫹ T cells, it markedly enhances TCR-induced proliferation and survival by
cooperating with T-cell receptor signals to induce expression
of the pro-proliferative cytokine IL-2 and the prosurvival
bcl-2 family member, bcl-xL.
CD28 binds to two distinct ligands of the B7 family of
proteins: CD80 (B7-1) and CD86 (B7-2). These are members
of the immunoglobulin superfamily, which have similar
extracellular domains. Interestingly, their intracellular
regions differ, suggesting that they might not simply act as
passive ligands; rather, they may have the ability to generate
signals that influence the APC, a view that has some experimental support.
PA R T V
360
|
Adaptive Immunity: Effector Responses
(a)
cSMAC
TCR/CD3
Coreceptors CD4 or CD8
Costimulatory receptors
(CD28)
pSMAC
Adhesion
molecules
LFA-1/ICAM-1
LFA-3/CD2
(b)
Antigen-presenting cell
CD4+ T cell
CD8+ T cell
Antigen-presenting cell
Antigen
Antigen
pSMAC
Peptide
LFA-3
CD2
LFA-3
Peptide
TCR−
CD3
Class II
MHC
p56lck
CD4
Class I MHC
LFA-1
CD80 or
CD86
ICAM-1
FIGURE 11-2 Surface interactions responsible for T-cell
activation. (a) A successful T-cell/dendritic-cell interaction results in
the organization of signaling molecules into an immune synapse. A
scanning electron micrograph (left) shows the binding of a T cell (artificially colored yellow) and dendritic cell (artificially colored blue). A
fluorescent micrograph (right) shows a cross-section of the immune
synapse, where the TCR is stained with fluorescein (green) and adhesion molecules (specifically LFA-1) are stained with phycoerythrin
(red). Other molecules that can be found in the central part of the
synapse (central supramolecular activation complex [cSMAC]) and
the peripheral part of the synapse (pSMAC) are listed. (b) The interactions between a CD4⫹ (left) or CD8⫹ (right) T-cell and its activating
CD28
LFA-1
SS
CD8
SS
CD28
pSMAC
TCR−
CD3
SS
cSMAC
SS
CD2
CD80 or
CD86
ICAM-1
dendritic cell. A dendritic cell (to the right of each diagram) can
engulf an antigen and present peptide associated with MHC class
II to a CD4⫹ T cell or can load internal peptides into MHC class I and
present the combination to a CD8⫹ T cell. Binding of the TCR to MHCpeptide complexes is enhanced by the binding of coreceptors CD4
and CD8 to MHC class II and class I, respectively. CD28 interactions
with CD80/86 provide the required costimulatory signals. Adhesion
molecule interactions, two of which (LFA-1/ICAM-1, LFA-3/CD2) are
depicted, markedly strengthen the connection between the T cell
and APC or target cell so that signals can be sustained. [(a) right:
Michael L. Dustin, J Clin Invest. 2002; 109(2): 155, Fig.1. doi:10.1172/JCI14842. Left:
Dr. Olivier Schwartz, Institut Pasteur/Photo Researchers.]
T-Cell Activation, Differentiation, and Memory
Normal
APC
1 TCR signaling
Gene
expression
2 Costimulatory
interaction
Autocrine
(e.g., IL-2)
3 Cytokine signaling
Paracrine
(e.g., IL-12)
FIGURE 11-3 Three signals are required for activation of
a naïve T cell. The TCR/MHC-peptide interaction, along with CD4
and CD8 coreceptors and adhesion molecules, provide Signal 1. Costimulation by a separate set of molecules, including CD28 (or ICOS,
not shown) provide Signal 2. Together, Signal 1 and Signal 2 initiate a
signal transduction cascade that results in activation of transcription
factors and cytokines (Signal 3) that direct T-cell proliferation (IL-2)
and differentiation (polarizing cytokines). Cytokines can act in an autocrine manner, by stimulating the same cells that produce them, or
in a paracrine manner, by stimulating neighboring cells.
Although most T cells express CD28, most cells in the
body do not express its ligands. In fact, only professional
APCs have the capacity to express CD80/86. Mature dendritic cells, the best activator of naïve T cells, appear to con-
TABLE 11-1
|
CHAPTER 11
361
stitutively express CD80/86, and macrophages and B cells
have the capacity to up-regulate CD80/86 after they are activated by an encounter with pathogen (see Chapter 5).
Positive Costimulatory Receptors: ICOS
Since the discovery of CD28, several other structurally
related receptors have been identified. Like CD28, the closely
related inducible costimulator (ICOS) provides positive
costimulation for T-cell activation. However, rather than
binding CD80 and CD86, ICOS binds to another member of
the growing B7 family, ICOS-ligand (ICOS-L), which is also
expressed on a subset of activated APCs.
Differences in patterns of expression of CD28 and ICOS
indicate that these positive costimulatory molecules play
distinct roles in T-cell activation. Unlike CD28, ICOS is not
expressed on naïve T cells; rather, it is expressed on memory
and effector T cells. Investigations suggest that CD28 plays a
key costimulatory role during the initiation of activation and
ICOS plays a key role in maintaining the activity of already
differentiated effector and memory T cells.
Negative Costimulatory Receptors: CTLA-4
The discovery of CTLA-4 (CD152), the second member of
the CD28 family to be identified, caused a stir. Although
closely related in structure to CD28 and also capable of binding both CD80 and CD86, CTLA-4 did not act as a positive
costimulator. Instead, it antagonized T-cell activating signals
and is now referred to as a negative costimulatory receptor.
CTLA-4 is not expressed constitutively on resting T cells.
Rather, it is induced within 24 hours after activation of a
T-cell costimulatory molecules and their ligands
Costimulatory receptor on T cell
Costimulatory ligand
Activity
CD28
CD80 (B7-1) or CD86 (B7-2)
Expressed by professional APCs, (and medullary
thymic epithelium)
Activation of naïve T cells
ICOS
ICOS-L
Expressed by B cells, some APCs, and T cells
Maintenance of activity of differentiated T cells;
a feature of T-/B-cell interactions
CTLA-4
CD80 (B7-1) or CD86 (B7-2)
Expressed by professional APCs (and medullary
thymic epithelium)
Negative regulation of the immune response
(e.g., maintaining peripheral T-cell tolerance;
reducing inflammation; contracting T-cell pool
after infection is cleared)
PD-1
PD-L1 or PD-L2
Expressed by professional APCs, some
T and B cells, and tumor cells
Negative regulation of the immune response,
regulation of TREG differentiation
BTLA
HVEM
Expressed by some APCs, T and B cells
Negative regulation of the immune response,
regulation of TREG differentiation (?)
Positive costimulation
Negative costimulation
362
PA R T V
|
Adaptive Immunity: Effector Responses
CLASSIC EXPERIMENT
Discovery of the First Costimulatory Receptor: CD28
In 1989, Navy immunologist Carl June
took the first step toward filling (and
revealing) the considerable gaps in our
understanding of T-cell proliferation and
activation by introducing a new actor:
CD28. CD28 had been recently identified
as a dimeric glycoprotein expressed on all
human CD4⫹ T cells and half of human
CD8⫹ T cells, and preliminary data suggested that it enhanced T-cell activation.
June and his colleagues specifically wondered if CD28 might be related to the
Signal 2 that was known to be provided
by APCs (sometimes referred to as “accessory cells” in older literature).
June and his colleagues isolated T cells
from human blood by density gradient
centrifugation and by depleting a T-cellenriched population of cells that did not
express CD28 (negative selection). They
then measured the response of these
CD28⫹ T cells to TCR stimulation in the
presence or absence of CD28 engagement (Figure 1a).
To mimic the TCR-MHC interaction,
June and colleagues used either monoclonal antibodies to the CD3 complex or
the mitogen phorbol myristyl acetate
(PMA), a protein kinase C (PKC) activator.
To engage the CD28 molecule, they used
an anti-CD28 monoclonal antibody. They
included two negative controls: one population that was cultured in growth
medium with no additives and another
population that was exposed to phytohaemagglutinin (PHA), which was known
to activate T cells only in the presence of
APCs. This last control was clever and
would be used to demonstrate that the
researchers’ isolated populations were not
contaminated with APCs, which would
express their own sets of ligands and confound interpretation.
June and colleagues measured the
proliferation of each of these populations
by measuring incorporation of a radioactive (tritiated) nucleotide, [3H]Uridine,
which is incorporated by cells that are
synthesizing new RNA (and, hence, are
showing signs of activation). Their results
were striking, particularly when responses
to PMA were examined (Figure 1b). As
expected, T cells grown without stimulation or with incomplete stimulation (PHA)
naïve T cell and peaks in expression within 2 to 3 days poststimulation. Peak surface levels of CTLA-4 are lower than
peak CD28 levels, but because it binds CD80 and CD86 with
markedly higher affinity, CTLA-4 competes very favorably
with CD28. Interestingly, CTLA-4 expression levels increase
in proportion to the amount of CD28 costimulation, suggesting that CTLA-4 acts to “put the brakes on” the proproliferative influence of TCR-CD28 engagement. The
importance of this inhibitory function is underscored by the
phenotype of CTLA-4 knockout mice, whose T cells proliferate without control, leading to lymphadenopathy (greatly
enlarged lymph nodes), splenomegaly (enlarged spleen),
autoimmunity, and death within 3 to 4 weeks after birth.
Negative Costimulatory Receptors: PD-1 and BTLA
PD-1 (CD279) and B and T lymphocyte attenuator (BTLA
(CD272)) are relatively new additions to our list of negative
costimulatory receptors. Although more distantly related to
CD28 family members than CTLA-4, they also inhibit TCRmediated T-cell activation. program death-1 (PD-1) is
expressed by both B and T cells and binds to two ligands,
remained quiescent, exhibiting no nucleotide uptake. Cell groups treated with
stimuli that were known to cause T-cell
proliferation—anti-CD3 and PMA—
showed evidence of activation, with CD3
engagement producing relatively more of
a response than PMA at the time points
examined.
The cells treated with anti-CD28 only
were just as quiescent as the negative
control samples, indicating that engagement of CD28, alone, could not induce
activation. However, when CD28 was
engaged at the same time cells were
exposed to PMA, incorporation of [3H]Uridine increased markedly. T cells cotreated
with anti-CD28 and anti-CD3 also took up
more [3H]Uridine than those treated with
anti-CD3 alone.
What new RNA were these cells producing? Using Northern blot analysis and
functional assays (bioassays) to characterize the cytokines in culture supernatant of
the stimulated cells, June and colleagues
went on to show that CD28 stimulation
induced anti-CD3 stimulated T cells to
produce higher levels of cytokines
PD-L1 (B7-H1) and PD-L2 (B7-DC), which are also members of the CD80/86 family. PD-L2 is expressed predominantly on APCs; however, PD-L1 is expressed more broadly
and may help to mediate T-cell tolerance in nonlymphoid
tissues. Recent data suggest that PD-L1/PD-L1/2 interactions
regulate the differentiation of regulatory T cells.
BTLA is more broadly expressed: not only has it been
found on conventional TH cells, as well as ␥␦ T cells and
regulatory T cells, but it is also expressed on NK cells, some
macrophages and dendritic cells, and most highly on B cells.
Interestingly, BTLA’s primary ligand appears not to be a B7
family member, but a TNF receptor family member known
as herpesvirus-entry mediator (HVEM), which is also
expressed on many cell types. Studies on the role of this
interesting costimulatory receptor-ligand pair are ongoing,
but there are indications that BTLA-HVEM interactions also
play a role in down-regulating inflammatory and autoimmune responses.
As the genome continues to be explored, additional
costimulatory molecules—both negative and positive in
influence—are likely to be identified. Understanding their
T-Cell Activation, Differentiation, and Memory
|
CHAPTER 11
363
BOX 11-1
10
PMA or anti-CD3
1
2
Anti-CD28
FIGURE 1
(a) June’s experimental setup used anti-CD3 monoclonal antibodies or a
mitogen, PMA, to provide Signal 1, and anti-CD28 antibodies to provide
Signal 2. (b) June measured [3H]Uridine incorporation, an indicator of RNA
synthesis, in response to various treatments. Addition of stimulating antiCD28 antibody increased RNA synthesis in response to activation by PMA
or anti-CD3. [Part (b) adapted from C. Thompson et al., 1989, CD28 activation pathway
3H-Uridine
Gene
expression
incorporation (CPMx10-3)
9
8
7
6
5
4
3
2
1
0
MED
PHA
regulates the production of multiple T-cell-derived lymphokines/cytokines. Proceedings
of the National Academy of Sciences of the United States of America 86:1333.]
involved in antiviral, anti-tumor, and proliferative activity, generating an increase in
T-cell immune response.
When this paper was published, June
did not know the identity of the natural
ligand for CD28, or even if the homodimer could be activated in a natural
immune context. We now know that
CD28 binds to CD80/86 (B7), providing
the critical Signal 2 during naïve T-cell
activation, a signal required for optimal
up-regulation of IL-2 and the IL-2 receptor. Finding this second switch capable of
modulating T-cell activation was only the
beginning of a landslide of discoveries of
additional costimulatory signals—positive and negative—involved in T-cell activation, and the recognition that T cells are
regulation and function will continue to occupy the attention of the immunological community and has already provided the clinical community with new tools for manipulating
the immune response during transplantation and disease
(see Clinical Focus Box 11-2).
Clonal Anergy Results if a Costimulatory
Signal Is Absent
Experiments with cultured cells show that if a resting T cell’s
TCR is engaged (Signal 1) in the absence of a suitable
costimulatory signal (Signal 2), that T cell will become unresponsive to subsequent stimulation, a state referred to as
anergy. There is good evidence that both CD4⫹ and CD8⫹ T
cells can be anergized, but most studies of anergy have been
conducted with CD4⫹ TH cells.
Anergy can be demonstrated in vitro with systems
designed to engage the TCR in the absence of costimulatory
molecule engagement. For instance, T cells specific for a
MHC-peptide complex can be induced to proliferate in vitro
by incubation with activated APCs that express both the
αCD28
PMA
PMA
+
αCD28
αCD3
αCD3
+
αCD28
even more subtly perceptive than we
once appreciated.
Based on a contribution by Harper Hubbeling,
Haverford College, 2011.
Thompson, C., et al. 1989. CD28 activation
pathway regulates the production of multiple T-cell-derived lymphokines/cytokines.
Proceedings of the National Academy of Sciences of the United States of America
86(4):1333–1337.
appropriate MHC-peptide combination and CD80/86. However, glutaraldehyde-fixed APCs, which express MHC class
II-peptide complexes, but cannot be induced to express
CD80/86, render T cells unresponsive (Figure 11-4a). These
anergic T cells are no longer able to secrete cytokines or
proliferate in response to subsequent stimulation (Figure
11-4b). T-cell anergy can also be induced in vitro by incubating T cells with normal, CD80/86-expressing APCs in the
presence of the Fab portion of anti-CD28, which blocks the
interaction of CD28 with CD80/86 (Figure 11-4c).
In vivo, the requirement for costimulatory ligands to
fully activate a T cell decreases the probability that circulating autoreactive T cells will be activated and become dangerous. For instance, a naïve T cell expressing a T-cell
receptor specific for an MHC class I insulin-peptide complex would be rendered nonresponsive if it encountered a
-islet cell expressing this MHC class I-peptide complex.
Why? -islet cells cannot be induced to express costimulatory ligands, and the encounter would result in T-cell
anergy, preventing an immune attack on these insulinproducing cells.
364
PA R T V
|
Adaptive Immunity: Effector Responses
BOX 11-2
CLINICAL FOCUS
Costimulatory Blockade
Taking an idea from the research
setting to a therapeutic reality—from
“bench to bedside”—is a dream for many,
but a reality for few. Immunologists recognized the therapeutic promise of costimulatory receptors as soon as they were
discovered. Reagents that would block
these interactions could block the activation of destructive T cells, which were
known to be responsible for many autoimmune disorders and for transplantation
rejection.
Several investigators specifically recognized that soluble versions of CD28 and
CTLA-4 could be very valuable. By blocking the interaction between costimulatory receptors and their CD80/86 ligands,
soluble CD28 and CTLA-4 could inhibit
destructive T-cell responses (e.g., those
involved in autoimmune disease or transplantation rejection). Couple this idea
with technological advances in protein
design and you had a novel reagent—and
a potential drug.
In the early 1990s, at least two groups
converted human CTLA-4 into a soluble
protein by fusing the extracellular domain
of human CTLA-4 with the Fc portion of an
IgG1 antibody (Figure 1). The Fc portion
enhances (1) the ability to manipulate a
protein during production by taking
advantage of Fc binding as well as (2) the
distribution of the reagent in an organism
(the Fc portion confers some of the antibodies’ tissue distribution behaviors). The
Fc portion is modified so that it does not
CTLA4
IgG1
domain
FIGURE 1
Structure of CTLA-4 Ig The extracellular
domain of human CTLA-4 is fused to a modified
Fc region of human IgG1. [PDB IDs 1IGY and 3OSK.]
bind to Fc receptors and cause unintended
cytotoxicity. Using this new protein, Peter
Linsley et al. found that their soluble CTLA4-Ig bound CD80/86 with higher affinity
than CD28-Ig, and could therefore more
potently block costimulation.
The development of CTLA-4 Ig as a
drug did not proceed without difficulty.
Originally designed and tested as a treatment for T-cell-mediated transplantation
rejection, it did not originally perform as
well as expected. However, hope in its
utility was revived when it showed significant promise as a treatment for rheumatoid arthritis. The marketing of CTLA-4 Ig
was not without controversy, either. At
least two groups claimed patent rights,
and communication between companies
Interactions between negative costimulatory receptors
and ligands can also induce anergy. This phenomenon,
which would typically only apply to activated T cells that
have up-regulated negative costimulatory receptors, could
help curb T-cell proliferation when antigen is cleared.
Unfortunately, negative costimulation may also contribute to
the T cell “exhaustion” during chronic infection, such as that
caused by mycobacteria, HIV, hepatitis virus, and more. T
cells specific for these pathogens express high levels of PD-1
and BTLA, and are functionally anergic. Some recent therapies, in fact, are designed to block this interaction and allow
T cells to reactivate.
and basic researchers was not always
smooth.
In late 2005, the FDA approved the use
of CTLA-4 Ig (abatacept) for rheumatoid
arthritis (RA). As of 2012 it is marketed as
Orencia by Bristol Meyers Squibb, which
shares patent royalties with Repligen Corporation. Abatacept shows promise in
delaying the joint damage seen with RA,
and clinical trials are also underway to
evaluate its potential to ameliorate psoriasis, lupus, Type 1 diabetes, and more.
This true bench-to-bedside story is still
not finished and continues to be informed
by basic researchers’ growing knowledge
of the complexities of costimulation.
Investigators have already developed a
modified version of CTLA-4 Ig that differs
in two amino acids in the CD80/86 binding region and exhibits a higher affinity for
CD86, which may play a more important
role in transplantation rejection. This drug
is also showing promise in clinical trials
and may result in fewer side effects than
standard immunosuppressant therapies
(e.g., cyclosporine), which are not specific
for T cells. It is also important to recognize
that CTLA-4 Ig not only blocks positive
costimulatory reactions, but also has the
potential to inhibit negative ones and in
some circumstances could lead to
enhancement of T-cell activity. Researchers will continue to draw from basic and
clinical knowledge to determine how best
to use the drug and to improve its design
for enhanced safety and efficacy.
Although anergy is a well-established phenomenon, the
precise biochemical pathways that regulate this state of nonresponsiveness are still not fully understood. During the past few
years, microarray analyses (see Chapter 20) have identified
several key enzymes expressed by anergic T cells, including
ubiquitin ligases that appear to target key components of the
TCR signaling pathway for degradation by the proteasome.
Cytokines Provide Signal 3
We have now seen that naïve T cells will initiate activation
when they are costimulated by engagement with both
T-Cell Activation, Differentiation, and Memory
(a)
1
Fixed
APC
(no CD80/86)
Anergic
genes
|
CHAPTER 11
365
As we will see below, Signal 3 is also supplied by other
cytokines (produced by APCs, T cells, NK cells, and others),
known as polarizing cytokines, that play central roles not just
in enhancing proliferation, but also in determining what
types of effector cells naïve T cells will become.
Antigen-Presenting Cells Have Characteristic
Costimulatory Properties
(b)
Anergic
T cell
Normal
APC
No response
(c)
1
Anergic
genes
Normal
APC
Fab
anti-CD28
CD80/86
FIGURE 11-4 Experimental demonstration of clonal anergy versus clonal expansion. (a) Only Signal 1 is generated
when resting T cells are incubated with glutaraldehyde-fixed APCs,
which cannot be stimulated to up-regulate the costimulatory ligand
CD80/86. (b) Anergic cells cannot respond to subsequent challenge,
even when APCs can engage costimulatory receptors. (c) Anergy can
also be induced when naïve T cells are incubated with normal APCs
in the presence of the Fab portion of anti-CD28, which blocks interaction with CD80/86.
MHC-peptide complexes and costimulatory ligands on
dendritic cells. However, the consequences and extent of
T-cell activation are also critically shaped by the activity of
soluble cytokines produced by both APCs and T cells. These
assisting cytokines are referred to, by some, as Signal 3 (see
Figure 11-3).
Cytokines bind surface cytokine receptors, stimulating a
cascade of intracellular signals that can enhance both proliferation and/or survival (see Chapter 4). IL-2 is one of the
best-known cytokines involved in T-cell activation and plays
a key role in inducing optimal T-cell proliferation, particularly when antigen and/or costimulatory ligands are limiting.
Costimulatory signals induce transcription of genes for both
IL-2 and the ␣ chain (CD25) of the high-affinity IL-2 receptor. Signals also enhance the stability of the IL-2 mRNA. The
combined increase in IL-2 transcription and improved IL-2
mRNA stability results in a 100-fold increase in IL-2 production by the activated T cell. Secretion of IL-2 and its subsequent binding to the high-affinity IL-2 receptor induces
activated naïve T cells to proliferate vigorously.
Which cells are capable of providing both Signal 1 and Signal
2 to a naïve T cell? Although almost all cells in the body
express MHC class I, only professional APCs—dendritic
cells, activated macrophages, and activated B cells—express
the high levels of class II MHC molecules that are required
for T-cell activation. Importantly, these same professional
APCs are among the only two cell types capable of expressing costimulatory ligands. (The only other cell type known
to have this capacity is the thymic epithelial cell; see Chapter
9.) Professional APCs are more diverse in function and origin than originally imagined, and each subpopulation differs
both in the ability to display antigen and in the expression of
costimulatory ligands (Figure 11-5).
In the early stages of an immune response in secondary
lymphoid organs, T cells encounter two main types of
professional APCs: the dendritic cell and the activated B
cell. Mature dendritic cells that have been activated by
microbial components via their pattern recognition
receptors (PRRs) are present throughout the T-cell zones.
They express antigenic peptides in complex with high
levels of MHC class I and II molecules. They also express
costimulatory ligands and are the most potent activators
of naïve CD4⫹ and CD8⫹ T cells.
Resting B cells residing in the follicles gain the capacity to
activate T cells after they bind antigen through their B-cell
receptor (BCR). This engagement stimulates the up-regulation
of MHC class II and costimulatory CD80/86, allowing the
B cell to present antigen to CD4⫹ T cells they encounter at
the border between the follicle and T-cell zone. Because of
their unique ability to internalize antigen (e.g., pathogens) via
specific BCRs and present them in MHC class II, B cells are
most efficient at activating CD4⫹ T cells that are specific for
the same pathogen for which they are specific. This situation
serves the immune response very well, focusing the attention
of antigen-specific CD4⫹ T cells activated in the T-cell zone on
B cells activated by the same antigen in the neighboring follicle.
The pairing of B cells with their helper T cells occurs at the
junction between the B- and T-cell zones (see Chapter 14) and
allows T cells to deliver the help required for B-cell proliferation, differentiation, and memory generation (see Chapter 12).
Macrophages are also found in secondary lymphoid organs,
but can activate T cells in a wide range of other peripheral
tissues. They also must be activated to reveal their full antigenpresenting capacity. They up-regulate MHC molecules and
costimulatory ligands in response to interactions with pathogens, as well as in response to cytokines produced by other
cells, including IFN-␥.
PA R T V
366
|
Adaptive Immunity: Effector Responses
Dendritic cell
Resting
B lymphocyte
Macrophage
Activated
Activated
Resting
Activated
Resting
Class I
MHC
PAMPs and
cytokines
PAMPs
Class I
MHC
Class I
MHC
Ag
T cell
help
(IFN-γ)
CD80/86
Antigen-presenting
cell
Class II
MHC
Class I
Class I
Class II
MHC
MHC CD80/86 MHC
Class II
MHC
Class II
MHC
CD80/86
Dendritic cell
Macrophage
B cell
Antigen uptake
Endocytosis
Phagocytosis
Phagocytosis
Receptor-mediated endocytosis
Activation
Mediated by pattern recognition
receptors
Mediated by pattern recognition
receptors and enhanced by T-cell
help
Mediated by antigen recognition
MHC Class II
expression
Increases with activation (may
express low levels constitutively)
Increases with activation
Increases with activation (expresses
low levels of constitutively)
Costimulatory
activity
Up-regulation of CD80/86 with
activation
Up-regulation of CD80/86 with
activation
Up-regulation of CD80/86 with
activation
T-cell activation
Naïve T cells
Effector T cells
Memory T cells
Effector T cells
Memory T cells
Effector T cells
Memory T cells
Location
Resting:
Circulation
peripheral tissues
Activated:
SLOs (T-cell zones)
Tertiary tissues
Resting:
Circulation
peripheral tissues
Activated:
SLOs (subcapsular cortex of lymph
node, marginal zones of spleen)
Peripheral tissues
Resting:
Circulation
SLOs (follicles)
Activated:
SLOs (B cell/T-cell zone interface,
germinal centers, and marginal
zones)
FIGURE 11-5 Differences in the properties of professional
antigen-presenting cells that induce T-cell activation. This figure describes general features of three major classes of professional
APCs. Dendritic cells are the best activators of naïve T cells. This may
be due, in part, to relatively high levels of expression of MHC and costimulatory molecules when they are mature and activated. Activated B cells interact most efficiently with differentiated TH cells that are
specific for the same antigen that activated them. Macrophages play
several different roles, processing and distributing antigen in second-
It turns out that there are several different dendritic cell
and macrophage subtypes; however, their functions are
still being clarified. Some are likely to activate or induce
differentiation of specific effector T cells that travel to the
site of infection, and some may be involved in reactivating
memory cells that reside in tissues. Others may help to
quell, rather than to activate, responses. Full understanding awaits more research.
ary lymphoid tissues (SLOs) as well as interacting with effector cells in
the periphery. It is important to recognize that the distinctions shown
are rules of thumb only. Functions among the APC classes overlap,
and the field now recognizes different subsets within each major
group of APC, each of which may act independently on different
T-cell subsets. This diversity may be a consequence of activation by
different innate immune receptors or may reflect the existence of
independent cell lineages. Note that activation of effector and memory
T cells is not as dependent on costimulatory interactions.
Superantigens Are a Special Class
of T-Cell Activators
Superantigens are viral or bacterial proteins that bind
simultaneously to specific V regions of T-cell receptors
and to the ␣ chain of class II MHC molecules. V regions
are encoded by over 20 different V genes in mice and 65
different genes in humans. Each superantigen displays a
T-Cell Activation, Differentiation, and Memory
TH cell
Vβ
β
α
Peptide for which
TCR is not specific
Superantigen
Endogenous
superantigen is
membrane bound
TCR
α
β
Class II MHC
APC
FIGURE 11-6 Superantigen-mediated cross-linkage of Tcell receptor and class II MHC molecules. A superantigen binds
to all TCRs bearing a particular V sequence regardless of their
antigen specificity. Exogenous superantigens are soluble secreted
bacterial proteins, including various exotoxins. Endogenous superantigens are membrane-embedded proteins produced by certain
viruses; they include Mls antigens encoded by mouse mammary
tumor virus.
“specificity” for one of these V versions, which can be
expressed by up to 5% of T cells, regardless of their antigen
specificity. This clamp-like connection mimics a strong
TCR-MHC interaction and induces activation, bypassing
the need for TCR antigen specificity (Figure 11-6). Superantigen binding, however, does not bypass the need for
costimulation; professional APCs are still required for full
T-cell activation by these microbial proteins.
TABLE 11-2
|
CHAPTER 11
367
Both endogenous superantigens and exogenous superantigens have been identified. Exogenous superantigens are soluble proteins secreted by bacteria. Among them are a variety of
exotoxins secreted by Gram-positive bacteria, such as staphylococcal enterotoxins, toxic shock syndrome toxin, and exfoliative dermatitis toxin. Each of these exogenous superantigens
binds particular V sequences in T-cell receptors (Table 11-2)
and cross-links the TCR to a class II MHC molecule.
Endogenous superantigens are cell-membrane proteins
generated by specific viral genes that have integrated into
mammalian genomes. One group, encoded by mouse mammary tumor virus (MTV), a retrovirus that is integrated into
the DNA of certain inbred mouse strains, produces proteins
called minor lymphocyte-stimulating (Mls) determinants,
which bind particular V sequences in T-cell receptors and
cross-link the TCR to class II MHC molecules. Four Mls
superantigens, originating from distinct MTV strains, have
been identified.
Because superantigens bind outside the TCR antigenbinding cleft, any T cell expressing that particular V
sequence will be activated by a corresponding superantigen.
Hence, the activation is polyclonal and can result in massive
T-cell activation, resulting in overproduction of TH-cell
cytokines and systemic toxicity. Food poisoning induced by
staphylococcal enterotoxins and toxic shock induced by
toxic shock syndrome toxin are two examples of disorders
caused by superantigen-induced cytokine overproduction.
Given that superantigens result in T-cell activation and
proliferation, what adaptive value could they possibly have
for the pathogens that make them? The answer is not clear,
but there is evidence that such antigen-nonspecific T-cell
Exogenous superantigens and their V specificity
V SPECIFICITY
Superantigen
*
Disease
Mouse
Human
SEA
Food poisoning
1, 3, 10, 11, 12, 17
nd
SEB
Food poisoning
3, 8.1, 8.2, 8.3
3, 12, 14, 15, 17, 20
Staphylococcal enterotoxins
SEC1
Food poisoning
7, 8.2, 8.3, 11
12
SEC2
Food poisoning
8.2, 10
12, 13, 14, 15, 17, 20
SEC3
Food poisoning
7, 8.2
5, 12
SED
Food poisoning
3, 7, 8.3, 11, 17
5, 12
Food poisoning
11, 15, 17
5.1, 6.1–6.3, 8, 18
Toxic shock syndrome toxin (TSST1)
SEE
Toxic shock syndrome
15, 16
2
Exfoliative dermatitis toxin (ExFT)
Scalded skin syndrome
10, 11, 15
2
Mycoplasma arthritidis supernatant (MAS)
Arthritis, shock
6, 8.1–8.3
nd
Streptococcal pyrogenic exotoxins
(SPE-A, B, C, D)
Rheumatic fever, shock
nd
nd
*
Disease results from infection by bacteria that produce the indicated superantigens.
368
PA R T V
|
Adaptive Immunity: Effector Responses
ADVANCES
How Many TCR Complexes Must Be Engaged
to Trigger T-Cell Activation?
Until single-cell imaging techniques
were developed, indirect methods were
used to calculate how many ligands a T
cell must recognize in order to be activated. Mark Davis and colleagues at Stanford School of Medicine approached this
question using an acutely sensitive, twopart microscopic visualization technique
(Figure 1). APCs were cultured (pulsed)
briefly with peptide bound to a biotin
molecule. When APCs are exposed to
soluble peptides a small number will
exchange it with a peptide bound to an
MHC molecule on the cell surface. The
number of peptides that actually bound
to MHC could be determined in this system by adding a fluorescent (phycoerythrin) streptavidin conjugate, which
binds the biotin of the biotinylated peptide and can be detected and quantified
by fluorescent microscopy.
The response of T cells specific for the
complex could also be quantified using
another fluorescent tool: fura-2, a dye
that can enter cells and fluoresces when
intracellular calcium is released. By adding fura-2 “loaded” T cells to the APCs
bound to varying numbers of red fluorescent peptide (see Figure 1a), researchers could determine (1) if a T cell was
activated and (2) how many molecules of
peptide were present at the point of
contact between the T cell and the
APC—in other words, how many TCR/
MHC-biotinylated peptide complexes
were required to stimulate the release of
intracellular calcium.
One set of data from these experiments is illustrated in Figure 1b. An activated, fura-2-loaded T cell (blue) is shown
interacting with an APC in the upper left.
The fluorescent micrograph of the peptide at the junction of the T cell and APC is
shown in the top right image. The intensity of the red fluorescence varies with the
number of peptides bound. (The image is
“stretched” artificially because of the computer program used to quantify fluores-
activation and inflammation hampers the development of a
coordinated antigen-specific response that would most
effectively thwart infection. Some speculate that the largescale proliferation and cytokine production that occur
harm the cells and microenvironments that are required to
start a normal response; others argue that these events
induce T-cell tolerance to the pathogen.
T-Cell Differentiation
How does an interaction between a naïve T cell and a dendritic cell result in the generation of cells with effector functions? An activating, costimulatory interaction between a
naïve T cell and an APC typically lasts 6 to 8 hours, a period
that permits the development of a cascade of signaling
events (see Chapter 3) that alter gene programs and induce
differentiation into a variety of distinct effector and memory
cell subtypes. Just a few TCR-MHC interactions (as described
in Advances Box 11-3) stimulates a signaling cascade that,
cence.) The fluorescence intensity
calculated from this particular image indicated that only a single MHC-peptide
combination was at the T cell-APC synapse. The plot below these images shows
the intensity of fura-2 fluorescence (i.e.,
the increase in intracellular Ca2⫹) over
time after T cell-APC engagement. The
initial spike is an indicator that this single
MHC-peptide could inspire robust Ca2⫹
signals.
The investigators quantified many
interactions in this way and definitively
concluded that a single MHC-peptide
combination could stimulate significant
Ca2⫹ release. Maximal Ca2⫹ release was
achieved in CD4⫹ T cells when as few as
10 TCR complexes were engaged. Similar results were obtained with CD8⫹ T
cells.
Irvine, D. J., M. A. Purbhoo, M. Krogsgaard, and
M. M. Davis. 2002. Direct observation of
ligand recognition by T cells. Nature
419:845–849.
in combination with costimulation and signals received by
soluble cytokines, culminate in the activation of “effector”
molecules that regulate (1) cell survival, (2) cell cycle entry,
and, as we shall see below, (3) cell differentiation.
In 1 to 2 days after successful engagement with a dendritic cell in the T-cell zone of a secondary lymphoid organ,
a naïve T cell will enlarge into a blast cell and begin undergoing repeated rounds of cell division. Signals 1 plus 2
induce up-regulation of expression and activity of prosurvival genes (e.g., bcl-2), as well as the transcription of
genes for both IL-2 and the ␣ chain (CD25) of the highaffinity IL-2 receptor (Figure 11-7). The combined effect on
a naïve T cell is activation and robust proliferation. Activated
T cells divide 2 to 3 times per day for 4 to 5 days, generating
a clone of progeny cells that differentiate into memory and
effector T-cell populations.
Activated T cells and their progeny gain unique functional abilities, becoming effector helper or cytotoxic T cells
that indirectly and directly act to clear pathogen. CD8⫹ cytotoxic T cells leave the secondary lymphoid tissues and circulate
T-Cell Activation, Differentiation, and Memory
|
CHAPTER 11
369
BOX 11-3
(a)
T cell
(b)
T
Ca2+
APC
4
PE snapshot
3
Ca2+ signal
APC
2
1
0
-5
0
5
10
15
20
Time (minutes)
FIGURE 1
Engagement of a single T-cell receptor can induce Ca2ⴙ
signals in a T cell. (a) The experimental approach taken by the
investigators to determine the ability of a single TCR-MHC interaction to
generate Ca2⫹ signals. See text for details. (b) One example of original data on
Ca2⫹ signaling generated from a single T cell that has engaged a single MHCpeptide ligand on the surface of an APC. The interaction between the T cell
and APC is captured by bright field microscopy (top left) and by highresolution fluorescence microscopy (top right), where the arrow points to the
single TCR/MHC-peptide interaction. The Ca2⫹ signal generated within the
T cell over a 20-minute period is depicted in the graph and measured using
software that quantifies the intensity of fura-2 fluorescence, which increases
with a rise in cytosolic Ca2⫹ levels. [Irvine, D.J. et al. 2002 Nature 419:845–849.]
to sites of infection where they bind and kill infected cells.
CD4⫹ helper T cells secrete cytokines that orchestrate the
activity of several other cell types, including B cells, macrophages, and other T cells. Some CD4+ T cells, particularly
those that “help” B cells and those that generate lymphocyte
memory, stay within secondary lymphoid tissue to continue
to regulate the generation of the response. Others return to
the sites of infection and enhance the activity of macrophages and cytotoxic cells.
Effector cells tend to be short-lived and have life spans
that range from a few days to a few weeks. However, the
progeny of an activated T cell can also become long-lived
memory T cells that reside in secondary and tertiary tissues
for significant periods of time, providing protection against
a secondary infection.
Effector T cells come in more varieties than was originally
anticipated, and each subset plays a specific and important
role in the immune response. The first effector cell distinction to be recognized was between CD8⫹ T cells and CD4⫹
T cells. Activated CD8⫹ T cells acquire the ability to induce
the death of target cells, becoming “killer” or “cytotoxic” T
lymphocytes (CTL, or TC). Because cytotoxic CD8⫹ T cells
recognize peptide bound to MHC class I, which is expressed
by almost all cells in the body, they are perfectly poised to
clear the body of cells that have been internally infected by
the pathogen that resulted in their activation. On the other
hand, activated CD4⫹ T cells (helper T cells or TH) acquire
the ability to secrete factors that enhance the activation
and proliferation of other cells. Specifically, they regulate
activation and antibody production of B cells, enhance the
phagocytic, anti-microbial, cytolytic, and antigen-presenting
capacity of macrophages, and are required for the development of B-cell and CD8⫹ T-cell memory.
As immunologists developed and adopted more tools to
distinguish proteins expressed and secreted by helper T cells,
it became clear that CD4⫹ helper T cells were particularly
diverse, differentiating into several different subtypes, each
of which secretes a signature set of cytokines. The cytokines
secreted by CD4⫹ TH cells can either act directly on the same
cell that produced them (acting in an autocrine fashion) or
370
PA R T V
|
Adaptive Immunity: Effector Responses
Normal
APC
IL-2
IL-2 gene
IL-2R gene
1
2
CD28 CD80/86
Activation
IL-2
IL-2
receptor
G1
M
S
G2
Several
divisions
M
M
E
E
E
E
Population of memory and effector cells
FIGURE 11-7 Activation of a naïve T cell up-regulates
expression of IL-2 and the high-affinity IL-2 receptor. Signal 1
and Signal 2 cooperate to enhance transcription and stability of
mRNA for IL-2 and IL-2R. Secreted IL-2 will bind the IL-2R, which generates signals that enhance the entry of the T cell into the cell cycle and
initiates several rounds of proliferation. Most of the cells differentiate
into effector helper or cytotoxic cells, but some will differentiate into
effector or central memory cells.
can bind to receptors and act on cells in the vicinity (acting
in a paracrine fashion). Below we describe the major features
and functions of the best-characterized CD4⫹ helper T-cell
subsets. Although CD8⫹ cytotoxic T cells also secrete cytokines, and there are indications that CD8⫹ T cells may also
differentiate into more than one killer subtype, the diversity
of CD8⫹ T-cell effector functions are clearly more restricted
than those of CD4⫹ TH cells. The generation and activity of
CTL cells are described in more detail in Chapter 13.
Helper T Cells Can Be Divided
into Distinct Subsets
Tim Mosmann, Robert Coffman, and colleagues can be credited with one of the earliest experiments definitively demonstrating that helper CD4⫹ T cells were more heterogeneous in
phenotype and function than originally supposed. Earlier
investigations, showing that helper T cells produced a diverse
array of cytokines, hinted at this possibility. However, Mosmann and Coffman definitively identified two distinct functional subgroups, TH1 and TH2, each of which produced a
different set of cytokines. Furthermore, they showed that
these differences were properties of distinct T-cell clones:
each activated T cell expanded into a population of effector T
cells that secreted a distinct array of cytokines.
Specifically, these investigators developed over 50 individual T-cell clones from a mixture of T cells with different
receptor specificities (i.e., a polyclonal T-cell population) that
had been isolated from the spleen of an immunized mouse.
At a time when the community did not have the tools to
identify most cytokines directly, these researchers developed
clever (and elaborate) strategies to define each clone’s cytokine secretion pattern. They showed that the cytokines
secreted by each of the 50 clones fell into one of two broad
categories, which they named TH1 and TH2.
Because TH1 and TH2 subsets were originally identified
from in vitro studies of cloned T-cell lines, some researchers
doubted that they represented true in vivo subpopulations.
They suggested instead that these subsets might represent
different maturational stages of a single lineage. In addition,
the initial failure to locate either subset in humans led some
to believe that TH1, TH2, and other subsets of T helper cells
were not present in this species. Further research corrected
these views. In many in vivo systems, the full commitment
of populations of T cells to either a TH1 or TH2 phenotype
occurs late in an immune response. Hence, it was difficult to
find clear TH1 or TH2 subsets in studies employing healthy
human subjects, where these cells would not have developed.
Indeed, TH1 and TH2 populations in T cells were ultimately
isolated from humans during chronic infectious disease or
chronic episodes of allergy, and studies in humans and mice
definitively confirmed their lineage independence.
With the benefit of new tools and technology, we now
have a more detailed understanding of the panels of cytokines that each group produces. The TH1 subset secretes
IL-2, IFN-␥, and Lymphotoxin-␣ (TNF-), and is responsible for many classic cell-mediated functions, including activation of cytotoxic T lymphocytes and macrophages. The
TH2 subset secretes IL-4, IL-5, IL-6, IL-9, IL-10, and IL-13,
and regulates B-cell activity and differentiation.
These experiments set the stage for the discovery over the
last decade that CD4⫹ T cells can adopt not just two but at
least five distinct effector fates after activation. The TH1 and
TH2 subpopulations have been joined definitively by the
TH17 and TREG subpopulations, each of which produces a
distinct cytokine profile and regulates distinct activities
within the body. Yet another subpopulation, T follicular
helper cells (TFH), has recently been characterized, and has
achieved membership among the major helper subsets.
More are bound to reveal themselves, although it will be
important to determine whether each represents distinct
subgroups or variants within the major subgroups.
In retrospect, we probably should have anticipated the
heterogeneity of helper responses, which allows an organism
T-Cell Activation, Differentiation, and Memory
to “tailor” a response to a particular type of pathogen. The
type of effector TH cell that a naïve T cell (also called a TH0
cell) becomes depends largely on the kind of infection that
occurs. For example, extracellular bacterial infections result
in the differentiation of activated CD4⫹ T cells into TH2
cells, which help to activate B cells to secrete antibodies that
can opsonize bacteria and neutralize the toxins they produce. On the other hand, infection by an intracellular virus
or bacterium induces differentiation of CD4⫹ T cells into
TH1 helpers that enhance the cytolytic activity of macrophages and CD8⫹ T cells, which can then kill infected cells.
Responses to fungi stimulate the differentiation of different
helper responses than responses to worms, and so on. The
reality is, of course, more complex. Infections evoke the differentiation of more than one helper subtype, some of which
have overlapping roles. What regulates the differentiation of
each effector subset and what function each subset plays are
still being actively investigated. We describe the fundamentals of our current understanding below.
The Differentiation of T Helper Cell Subsets Is
Regulated by Polarizing Cytokines
As you know, T-cell activation requires TCR and costimulatory receptor engagement, both of which are supplied by an
activated APC. It is now clear that the functional fate of activated T cells is determined by signals they receive from
additional cytokines generated during the response. These
cytokines (Signal 3) are referred to as polarizing cytokines
because they are responsible for guiding a helper T cell
toward one of several different effector fates. For example, T
cells that are activated in the presence of IL-12 and IFN-␥
tend to differentiate, or polarize, to the TH1 lineage, whereas
T cells that are activated in the presence of IL-4 and IL-6
polarize to the TH2 lineage.
Polarizing cytokines can be generated by the stimulating
APC itself, or by neighboring immune cells that have also
been activated by antigen. Which cytokines are produced
during an immune response depends on (1) the cell of origin
(DC, macrophage, B cell, NK cell, etc.), (2) its maturation
and activation status, (3) which pathogens and other inflammatory mediators it encounters, and (4) in what tissue environment it encounters that pathogen. Innate interactions
therefore have a critical role in shaping adaptive responses
(see Figure 5-18). Specifically, by influencing APC secretions
and the surface and the microenvironmental landscape that
a T cell encounters, innate immune responses directly influence the functional fate of helper T cells.
Recall from Chapter 5 that APCs and other innate
immune cells are activated by interaction with pathogens
bearing pathogen-associated molecular patterns (PAMPs).
These PAMPs bind PRRs, including, but not limited to Tolllike receptors (TLRs). PRR interactions activate dendritic
cells by stimulating the up-regulation of MHC and costimulatory proteins. They also determine the type of cytokine(s)
that dendritic cells and other immune cells will secrete.
|
CHAPTER 11
PAMPs from pathogens
(or adjuvants)
371
PRR
Other cellular
sources of
cytokines
PRR
APC
Polarizing
cytokines
STATs
TCR
Effector
cytokines
Master gene
regulators
FIGURE 11-8 General events and factors that drive TH
subset polarization. Interaction of pathogen with pattern recognition receptors (PRRs) on dendritic cells and other neighboring
immune cells determines which polarizing cytokines are produced
and, hence, into which T helper subset a naïve cell will differentiate. In general, polarizing cytokines that arise from dendritic cells
or other neighboring cells interact with cytokine receptors and
generate signals that induce transcription of unique master gene
regulators. These master regulators, in turn, regulate expression of
various genes, including effector cytokines, which define the function of each subset.
Hence, PRR signals regulate the fate a T cell will adopt following activation (Figure 11-8).
For example, double-stranded RNA, a product of many
viruses, binds TLR3 receptors on dendritic cells, initiating a
signaling cascade that results in production of IL-12, which
directly promotes TH1 differentiation. On the other hand,
worms stimulate PRRs on innate immune cells, including
mast cells, which generate IL-4. IL-4 directly promotes
polarization of activated T cells to the TH2 subset, which
coordinates the IgE response to helminths (see Figure 11-8).
In this case, the main polarizing cytokine is not made by the
activating dendritic cell, but is generated by a neighboring
immune cell. The pathogen interactions that give rise to the
polarizing cytokines that drive T helper cell differentiation
are often complex and a very active area of research.
Adjuvants, which have been used for decades to enhance
the efficacy of vaccines, are now understood to exert their
influence on the innate immune system by regulating the
expression of costimulatory ligands and cytokines by APCs,
events that ultimately shape the consequences of T-cell activation. PAMPs and cytokines such as IL-12, produced by
APCs themselves, are considered natural adjuvants. Dead
mycobacteria, which clearly activate many PRRs, have long
PA R T V
372
TABLE 11-3
|
Adaptive Immunity: Effector Responses
Regulation and function of T helper subtypes
Polarizing cytokines
Master gene regulators
Effector cytokines
Functions
TH1
IL-12
IFN-␥
IL-18
T-Bet
IFN-␥
TNF
Enhances APC activity
Enhances TC activation
Protects against intracellular pathogens
Involved in delayed type hypersensitivity,
autoimmunity
TH2
IL-4
GATA-3
IL-4
IL-5
IL-13
Protects against extracellular pathogens
(particularly IgE responses)
Involved in allergy
TH17
TGF-
IL-6
(IL-23)
ROR␥
IL-17A
IL-17F
IL-22
Protects against some fungal and
bacterial infections
Contributes to inflammation,
autoimmunity
TREG
TGF-
IL-2
FoxP3
IL-10
TGF-
Inhibits inflammation
TFH
IL-6
IL-21
Bcl-6
IL-4
IL-21
B cell help in follicles and germinal
centers
been used as a very potent adjuvant for immune responses
in mice. Very few adjuvants are approved for human vaccination, but given our new and evolving understanding of the
molecules that stimulate PRRs and the consequences of that
stimulation, investigators expect that we will one day be able
to shape the effector response to vaccine antigens by varying
the adjuvants—natural and/or synthetic—included in vaccine preparations (see Chapter 17).
Effector T Helper Cell Subsets Are
Distinguished by Three Properties
Each helper T-cell subset is defined by an array of features,
the details of which can rapidly overwhelm a new (or old)
student of immunology. Understanding these specifics is an
important first step to clarifying the role each subset plays in
resolving infection and causing disease. Having a reference
to them is also helpful when deciphering primary literature
describing advances. However, some generalizations provide
a useful conceptual framework for organizing some of these
details. Consider the following:
• Each of the major T helper cell subsets is characterized
by (1) a distinct set of polarizing cytokines that induce
the expression of (2) a master gene regulator that
regulates expression of (3) a signature set of effector
cytokines the T-cell population produces once it is fully
differentiated (see Figure 11-8 and Table 11-3).
• Which effector subset an activated helper cell becomes
depends on the quality and quantity of signals its naïve
cell precursor receives from APCs in a secondary
lymphoid organ; that activity, in turn, depends on the
nature of the pathogen the APC encountered at the site
of infection.
• Broadly speaking, TH1 and TH17 cells regulate cellmediated immunity (CD8⫹ T cells and macrophages)
and TH2 and TFH cells regulate humoral immunity (B
cells). However, it is important to recognize that all
CD4⫹ effector T-cell subsets may have the potential to
provide help to B cells. TH1 and TH17 subsets generally
encourage B cells to produce antibodies that contribute
to cell-mediated immunity (e.g., isotypes like IgG2a that
can “arm” NK cells for cytotoxicity; see Chapter 13). TH2
cells encourage B cells to produce antibodies that
mediate the clearance of extracellular pathogens (e.g.,
isotypes like IgE that induce the release of molecules
that harm extracellular parasites).
• Helper T-cell subsets often “cross-regulate” each other.
The cytokines they secrete typically enhance their own
differentiation and expansion and inhibit commitment to
other helper T-cell lineages. This is particularly true of
the TH1 and TH2 pair, as well as the TH17 and TREG pair.
• Helper cell lineages may not be fixed; some subsets can
assume the cytokine secretion profile of other subsets if
exposed to a different set of cytokines, particularly early
in the differentiation process.
• The precise biological function and sites of differentiation
and activity of each subset continue to be actively
investigated. Much remains unknown.
We start our discussion of helper cell subset characteristics with the first two subsets to be identified: TH1 and TH2
cells. They provide an illustrative example of the features that
T-Cell Activation, Differentiation, and Memory
Pathogens inducing
cell-mediated immunity
TH1-polarizing
(most viruses, some
factors
bacteria and fungi)
IL-12
PRRs
Dendritic
cell
|
CHAPTER 11
Naïve TH cell
T H1
TCR
MHC Class IIpeptide
Signal 1
T-Bet
Signal 2
GATA-3
TH2-polarizing
Signal 3
TH2
TH2-polarizing
factors
FIGURE 11-9 Regulation of TH1 and TH2 subset differentiation. This figure depicts some of the cellular events that drive TH1
and TH2 lineage commitment in more detail. Intracellular pathogens
activate a cascade of signals that polarize cells to the TH1 lineage. For
example, viruses interact with PRRs (e.g., TLR-3) that induce dendritic
cells to generate IL-12. This binds to receptors on naïve T cells, activating a signal transduction pathway mediated by STAT4 that induces
expression of the transcription factor T-Bet. T-Bet, in turn, activates
expression of effector cytokines, including IFN-␥, which define the TH1
subset’s functional capacities (and can also enhance TH1 polarization).
distinguish T helper cells as well as the relationship between
subsets. We follow this section with summaries of what is
currently understood about the more recently characterized
helper subsets: TH17, TFH, and TREG.
The Differentiation and Function of TH1 and TH2 Cells
The key polarizing cytokines that induce differentiation of
naïve T cells into TH1 cells are IL-12, IL-18, and IFN-␥
(Figure 11-9). IL-12 is produced by dendritic cells after an
encounter with pathogens via PRRs (e.g., TLR4, TLR3). It is
also up-regulated in response to IFN-␥, which is generated
by activated T cells and activated NK cells. IL-18, which is
also produced by dendritic cells, promotes proliferation of
developing TH1 cells and enhances their own production of
IFN-␥. These polarizing cytokines trigger signaling pathways that up-regulate the expression of the master gene
regulator T-Bet. This master transcription factor induces
commitment to the TH1 lineage, inducing expression of the
signature TH1 effector cytokines, including IFN-␥ and TNF.
IFN-␥ is a particularly potent effector cytokine. It activates macrophages, stimulating these cells to increase microbicidal activity, up-regulate the level of class II MHC, and, as
mentioned above, secrete cytokines such as IL-12, which
further enhance TH1 differentiation. IFN-␥ secretion also
induces antibody class switching in B cells to IgG classes
(such as IgG2a in the mouse) that support phagocytosis and
fixation of complement. Finally, IFN-␥ secretion promotes
IFN-γ
TH1-polarizing
Signal 3
CD80/86 CD28
Pathogens inducing
humoral immunity,
particularly extracellular
parasites (e.g., worms)
373
IL-4
IL-4
IL-5
IL-13
Cellular
sources
of IL-4
On the other hand, extracellular pathogens activate signal cascades
that can polarize naïve T cells to the TH2 lineage. Parasitic worms interact with PRRs on neighboring immune cells (such as mast cells, basophils, or germinal center B cells), triggering the release of the signature
TH2 polarizing cytokine IL-4. This interacts with receptors on T cells that
activate STAT6, up-regulating expression of the transcriptional regulator GATA-3. GATA-3, in turn, induces expression of the TH2 effector cytokines, including IL-4, IL-5, and IL-13. [Adapted from M. L. Kapsenberg, 2003,
Dendritic-cell control of pathogen-driven T-cell polarization, Nature Reviews
Immunology 3:984.]
the differentiation of fully cytotoxic TC cells from CD8⫹ precursors by activating the dendritic cells that engage naïve TC
cells. These combined effects make the TH1 subset particularly suited to respond to viral infections and other intracellular pathogens. They also contribute to the pathological
effects of TH1 cells, which are also involved in the delayed
type hypersensitivity response to poison ivy (see Chapter 15).
Just as differentiation to the TH1 subset is promoted by
IL-12 and IFN-␥, differentiation to the TH2 subset is promoted by a defining polarizing cytokine, IL-4 (see Figure
11-9). Exposing naïve helper T cells to IL-4 at the beginning
of an immune response causes them to differentiate into TH2
cells. Interestingly, TH2 development is greatly favored over
TH1 development. Even in the presence of IFN-␥ and IL-12,
naïve T cells will differentiate into TH2 effectors if IL-4 is
present. IL-4 triggers a signaling pathway within the T cell
that up-regulates the master gene regulator GATA3, which,
in turn, regulates expression of TH2-specific cytokines,
including IL-4, IL-5, and IL-13.
Dendritic cells do not make IL-4, so from where does it
come? Mast cells, basophils, and NKT cells can be induced
to make IL-4 after exposure to pathogens and could influence helper T cell fate in the periphery. Germinal-center B
cells and TFH cells can also produce IL-4, which could influence helper T-cell polarization in the lymph nodes and
spleen. And TH2 cells themselves are an excellent source
of additional IL-4 that can enhance polarization events.
374
PA R T V
|
Adaptive Immunity: Effector Responses
Investigators, however, are still working to definitively identify the source of the IL-4 that initiates TH2 polarization in
secondary lymphoid tissues.
The effector cytokines produced by TH2 cells help to clear
extracellular parasitic infections, including those caused by
helminths. IL-4, the defining TH2 effector cytokine, acts on
both B cells and eosinophils. It induces eosinophil differentiation, activation, and migration and promotes B-cell activation and class switching to IgE. These effects act synergistically
because eosinophils express IgE receptors (FcR) which,
when cross-linked, release inflammatory mediators that are
particularly good at attacking roundworms. Thus, IgE antibody can form a bridge between the worm and the eosinophil, binding to worm antigens via its variable regions and
binding to FcR via its constant region. IgE antibodies also
mediate allergic reactions, and the role of TH2 activity in
these pathological responses is described in Chapter 15.
IL-5 can also induce B-cell class switching to IgG subclasses that do not activate the complement pathway (e.g.,
IgG1 in mice), and IL-13 functions largely overlap with IL-4.
Finally, IL-4 itself can suppress the expansion of TH1-cell
populations.
TH1 and TH2 Cross-regulation
The major effector cytokines produced by TH1 and TH2 subsets (IFN-␥ and IL-4, respectively) not only influence the
overall immune response, but also influence the helper T cell
subsets. First, they promote the growth (and in some cases
even the polarization) of the subset that produces them; second, they inhibit the development and activity of the opposite
subset, an effect known as cross-regulation. For instance,
IFN-␥ (secreted by the TH1 subset) inhibits proliferation of
the TH2 subset, and IL-4 (secreted by the TH2 subset) downregulates the secretion of IL-12 by APCs, thereby inhibiting
TH1 differentiation. Furthermore, IL-4 enhances TH2 cell
development by making TH cells less susceptible to the TH1
promoting cytokine signals (and vice versa).
Similarly, these cytokines have opposing effects on target
cells other than TH subsets. In mice, where the TH1 and TH2
subsets have been studied most extensively, the cytokines
have distinct effects on the type of antibody made by B cells.
Recall from Chapter 3 (and see Chapter 13) that the antibody
isotype IgG2a enhances cell-mediated immunity by arming
NKT cells, whereas the isotypes IgG1 and IgE enhance
humoral immunity by their activities on extracellular pathogens. IFN-␥ secreted by the TH1 subset promotes IgG2a
production by B cells but inhibits IgG1 and IgE production.
On the other hand, IL-4 secreted by the TH2 subset promotes
production of IgG1 and IgE and suppresses production of
IgG2a. Thus, these effects on antibody production are consistent with TH1 and TH2 subsets’ overall tendencies to promote
cell-mediated versus humoral immunity, respectively.
IL-10 secreted by TH2 cells also inhibits (cross-regulates)
TH1 cell development, but not directly. Instead, IL-10 acts on
monocytes and macrophages, interfering with their ability to
activate the TH1 subset by abrogating their activation,
APC
TCR
IL-12
Stat4
T-Bet −
IL-4
Stat6
− GATA-3
+ IFN-γ −
− IL-4 +
− IL-5 +
Promotes
TH1
Promotes
TH2
FIGURE 11-10 Cross-regulation of T helper cell subsets by
transcriptional regulators. GATA-3 and T-Bet reciprocally regulate
differentiation of TH1 and TH2 lineages. IL-12 promotes the expression
of the TH1-defining transcription factor, T-Bet, which induces expression of TH1 effector cytokines, including IFN-␥. At the same time, T-Bet
represses the expression of the TH2 defining master transcriptional
regulator, GATA-3, as well as expression of the effector cytokines IL-4
and IL-5. Reciprocally, IL-4 promotes expression of GATA-3, which
up-regulates the synthesis of IL-4 and IL-5, and at the same time
represses the expression of T-Bet and the TH1 effector cytokine IFN␥. [Adapted from J. Rengarajan, S. Szabo, and L. Glimcher, 2000, Transcriptional
regulation of Th1/Th2 polarization, Immunology Today 21:479.]
specifically by (1) inhibiting expression of class II MHC molecules, (2) suppressing production of bactericidal metabolites (e.g., nitric oxide) and various inflammatory mediators
(e.g., IL-1, IL-6, IL-8, GM-CSF, G-CSF, and TNF-), and
even by (3) inducing apoptosis.
The master regulators T-Bet and GATA-3 also play an
important role in cross-regulation. (Figure 11-10) Specifically, the expression of T-Bet drives cells to differentiate into
TH1 cells and suppresses their differentiation along the TH2
pathway. Expression of GATA-3 does the opposite, promoting the development of naïve T cells into TH2 cells while suppressing their differentiation into TH1 cells. Consequently,
cytokine signals that induce one of these transcription factors
sets in motion a chain of events that represses the other.
TH17 Cells
The discovery of the TH17 subset of T cells, which, like the
TH1 subset, is involved in cell-mediated immunity, arose in
part from a recognition that IL-12, one of the polarizing cytokines that induces TH1 development, was a member of a family of cytokines that shared a subunit (p40) with IL-23. The
p40 knockout mouse was a favorite model for studying the
importance of TH1 cells because, in the absence of IL-12, it
failed to generate TH1 cells. However, once it was understood
T-Cell Activation, Differentiation, and Memory
|
CHAPTER 11
375
11-11
OVERVIEW FIGURE
T Helper Subset Differentiation
Naïve
CD4+
T cell
Polarizing
cytokines
Master
transcriptional
regulator
Effector
cytokines
Effector
functions
IL-2,
TGF-β
IL-1,
IL-6,
IL-23,
TGF-β
IL-4
IL-6,
IL-21
Bcl-6
IL-12,
IFN-γ,
IL-18
T-Bet
FOXP3
RORγt
GATA3
Induced TReg cell
TH17 cell
TH2 cell
TFH cell
TH1 cell
IL-10, TGF-β
IL-17A, IL-17F, IL-22
IL-4, IL-5, IL-13
IL-4, IL-21
IFN-γ, TNF
Regulation,
suppression of
immune and
inflammatory
responses
Inflammation
Allergic and
anti-helminth
responses
B cell help
in germinal
centers
Cell-mediated
immunity,
macrophage
activation,
inflammation
This figure synthesizes current information about the distinguishing features of T helper subset differentiation and activity. Polarizing
cytokines, master transcriptional regulators, effector cytokines, and
broad functions in health and disease are depicted for each of the
major helper subsets. Neither cross-regulation nor the potential
that p40 was also required for IL-23 production, investigators
quickly realized that the results from these mice were no
longer unambiguous. In fact, it became clear that these mice
also failed to produce a T-cell subset that required IL-23.
Some of the functions originally attributed to IL-12-induced
TH1 cells were actually carried out by an IL-23-induced T
helper cell subpopulation now referred to as TH17 cells.
TH17 cells are generated when naïve T cells are activated
in the presence of IL-6 and TGF-, the key polarizing cytokine for iTREG differentiation (Overview Figure 11-11). IL-23
also plays a role in finalizing the commitment to the TH17
fate and is induced in APCs by interactions with PAMPs
including fungal wall components, with TLR2 and the CLR
Dectin-1. Like TH1 and TH2 differentiation, TH17 cell differentiation is also controlled by a master transcriptional regulator
whose expression is induced by polarizing cytokines. In this
case the master regulator is the orphan steroid receptor ROR␥t,
plasticity in differentiation among subsets is depicted, but both are
described in the text. [Adapted from S. L. Swain, K. K. McKinstry, and T. M.
Strutt, Expanding roles for CD41 T cells in immunity to viruses, Nature Reviews.
Immunology 12:136–148.]
which also plays a role in T-cell development. ROR␥t knockout mice have reduced severity of experimental autoimmune
encephalitis (EAE, a mouse model of multiple sclerosis)
apparently because of the reduction in TH17 cells.
TH17 cells are so named because they produce IL-17A, a
cytokine associated with chronic inflammatory and autoimmune responses, including those that result in inflammatory
bowel disease, arthritis, and multiple sclerosis. TH17 cells are
the dominant inflammatory cell type associated with chronic
autoimmune disorders. They also produce IL-17F and IL-22,
cytokines associated with tissue inflammation. We have only
begun to understand the true physiological function of TH17
cells, which in healthy humans have been found at epithelial
surfaces (e.g., lung and gut) and may play a role in warding
off fungal and extracellular bacterial infections (see Clinical
Focus Box 11-4). However, a full appreciation of their biological role awaits further investigations.
376
PA R T V
|
Adaptive Immunity: Effector Responses
CLINICAL FOCUS
What a Disease Reveals about the Physiological
Role of TH17 Cells
Experiments that reveal the inner
workings of normal healthy immune cells
have offered invaluable insights into what
can and does go wrong. However, at
times we need “experiments of nature”—
the diseases themselves—to clarify how
the immune system works in healthy individuals. Job syndrome, a rare disease in
which patients suffer from defects in
bone, teeth, and immune function, is
helping us solve the mystery of the physiological function of TH17 cells.
People with Job syndrome suffer from
recurrent infections, typically of the lung
and skin. The painful abscesses that are
often a feature of the disease and the trials
endured by its patients explain its name,
which comes from a biblical story of a man
(Job) who is subject to horrific hardship as
a test of his piety. Another cardinal feature
of the disease, elevated IgE levels in serum,
is the basis for its more formal name, Hyper
IgE Syndrome (HIES). HIES comes in two
major forms: Job syndrome, the most
common, is also referred to as Type 1 HIES.
Patients with Type 2 HIES, which we will
not discuss here, do not have trouble with
bone or dental development.
The abundance of IgE originally suggested to some investigators who were
savvy about the roles of helper T-cell subsets that Type 1 HIES symptoms may be
caused by an imbalance between TH1 and
TH2 responses. Initial work did not reveal a
difference in these activities, but as the
diversity of T helper subsets was revealed,
investigators pursued the possibility of a
helper imbalance more vigorously. A
number of groups from Japan and America independently discovered that lymphocytes from HIES patients were unable
to respond to select cytokines, including
IL-6, IL-10, and IL-23. An analysis of genes
that could explain this signaling failure
revealed a dominant negative mutation in
STAT3, a key cytokine signaling molecule
(see Chapter 4).
These investigators knew from the literature that the absence of signaling
through these cytokines suggested a specific problem with polarization to the TH17
helper subset. Indeed, circulating TH17 cells
were absent in HIES patients with STAT3
mutations. The investigators then directly
tested their hypothesis that this absence
was due to a failure of CD4⫹ T cells to polarize normally. Specifically, they removed
T cells from blood, and exposed them to
conditions that would ordinarily polarize
them to the TH17 lineage: T-cell receptor
stimulation (Signal 1) in the presence of
costimulatory signals (CD28, Signal 2) and
cytokines known to drive human TH17 differentiation (IL-1, IL-6, TGF-, and IL-23,
(Induced) TREG Cells
Another major CD4⫹ T-cell subset negatively regulates
T-cell responses and plays a critically important role in
peripheral tolerance by limiting autoimmune T-cell activity.
This subset of T cells, designated induced TREG (iTREG)
cells, is similar in function to the natural TREG cells
(nTREGs) that originate from the thymus (see Chapter 9).
Induced TREG cells, however, do not arise in the thymus, but
from naïve T cells that are activated in secondary lymphoid
tissue in the presence of TGF- (see Overview Figure
11-11). TGF- induces expression of FoxP3, the master
transcriptional regulator responsible for iTREG commitment.
The iTREG cells secrete the effector cytokines IL-10 and TGF, which down-regulate inflammation via their inhibitory
effects on APCs, and can also exert their suppressive func-
Signal 3) (Figure 1a). They stained the cells
for intracellular expression of TH17’s signature cytokine, IL-17, and performed flow
cytometry (Figure 1b; also see Chapter 20).
Although CD4⫹ T cells from HIES patients
were able to develop into other helper lineages (and, as you can see from the flow
cytometry contour plots, were also able to
make IFN-␥, indicating they could become
TH1 cells), they could not be induced to
secrete IL-17. Specifically, whereas 18.3% of
T cells from healthy patients that were
subject to these conditions stained with
antibodies against IL-17, 0.05% (essentially
none) of the T cells from HIES patients
stained with the same antibodies.
These striking observations suggest
that the recurrent infections HIES patients
experience are caused at least in part by
the absence of TH17 cells. Reciprocally,
they indicate that TH17 cells play an
important role in controlling the type of
infection that afflicts these patients,
including Staphylococcus aureus skin
infections and pneumonia. The critical
role of TH17 in controlling bacterial and
fungal infections at epithelial surfaces has
been supported by studies in mice, too.
J. D. Milner et al. 2008. Impaired TH17 cell
differentiation in subjects with autosomal
dominant Hyper-IgE syndrome. Nature
452:773–776.
tion by interacting directly with T cells. The depletion of
iTREG cells in otherwise healthy animals leads to multiple
autoimmune outbreaks, revealing that even healthy organisms are continually warding off autoimmune responses.
Recent data also indicate that iTREG cells are critically important for maintaining a mother’s tolerance to her fetus.
TH17 and TREG Cross-Regulation
Just as TH1 and TH2 cells reciprocally regulate each other,
TREG and TH17 cells also cross-regulate each other. TGF-
induces TREG differentiation; however, when accompanied
by IL-6, TGF- induces TH17 differentiation. Specifically
TGF- appears to up-regulate both FoxP3 and ROR␥
(which control TREG and TH17 differentiation, respectively). In combination with signals generated by IL-6,
T-Cell Activation, Differentiation, and Memory
|
CHAPTER 11
377
BOX 11-4
IL-17?
TCR and CD28 engagement
+/–
polarizing cytokines (IL-1/IL-23)
12 days
Naïve CD4+ T cells
TH17 cells?
(a)
(b)
FIGURE 1
CD4ⴙ T cells from Hyper-IgE patients do not differentiate into TH17 cells. (a) Condi-
tions used to polarize CD4⫹ T cells to the TH17 lineage in vitro. The ability of the cells to make IL-17 (a
feature of TH17 cells) or IFN-␥ (a property of TH1, not TH17 cells) after they were polarized was assessed
by staining for intracellular cytokines. (b) Flow cytometric analysis of intracellular staining. The boxes
outlined in red indicate the quadrant where IL-17⫹ (TH17) cells would appear. [Irvine, D.J. et al. 2002.
Nature 419: 845–849. Reprinted by permission from Macmillan Publishers]
signals generated by TGF- inhibit FoxP3 expression, letting ROR␥ dominate and induce TH17 development. The
TH17 versus iTREG relationship may be very adaptive. At
rest, a healthy organism may favor the development of an
anti-inflammatory iTREG population, which would be reinforced by the iTREG cell’s own production of TGF-.
Inflammation, however, would induce the generation of
acute response proteins, including IL-6. In the presence of
IL-6, TGF- activity would shift development of T cells
away from iTREGs toward the pro-inflammatory TH17, so a
proper defense could be mounted.
TFH Cells
Follicular helper T (TFH) cells are a very recent addition to
the helper T-cell subset family. Whether they represent an
independent lineage or a developmental option for other
helper lineages remains controversial. Like TH2 cells, TFH
cells play a central role in mediating B-cell help and are
found in B-cell follicles and germinal centers. However, the
effector cytokines secreted by follicular helper T cells differ
partially from those secreted by TH2 cells.
Cytokines that polarize activated T cells toward the TFH
lineage include IL-6 and IL-21. These polarizing cytokines
induce the expression of Bcl-6, a transcriptional repressor
that is thought to be TFH’s master transcriptional regulator
(see Overview Figure 11-11). Cross-regulation is also a
feature of TFH function: Bcl-6 expression inhibits T-bet,
GATA3, and ROR␥t expression, thus inhibiting TH1, TH2,
and TH17 differentiation, respectively, while inducing TFH
polarization.
378
PA R T V
|
Adaptive Immunity: Effector Responses
Although both TFH and TH2 cells secrete IL-4, TFH cells
are best characterized by their secretion of IL-21, which
induces B-cell differentiation. Interestingly, they can also
produce IFN-␥ (the defining TH1 cytokine). How TFH and
TH2 cells divide responsibilities for inducing B-cell antibody
production is still an open question.
Other Potential Helper T-Cell Subsets
Other T-cell subsets with distinct polarizing requirements and
unique cytokine secretion profiles have been identified (e.g.,
TH9 cells, which secrete IL-9 and IL-10). However, because
these subpopulations secrete cytokines that are also produced
by TH1, TH2, TH17, or iTREG cells, some speculate that these
cell types do not represent distinct subclasses but rather are
developmental or functional variants of one of the major subpopulations. This perspective has, indeed, been applied to the
follicular helper T-cell (TFH) subset, which also expresses cytokines shared by several other subtypes. However, this subset
has a distinct gene signature and a distinct master regulator
(bcl-6), so most now consider it a bona fide independent lineage. Clearly, our understanding of the developmental relationship among effector subtypes will continue to evolve.
Helper T Cells May Not Be Irrevocably
Committed to a Lineage
Investigations now suggest that the relationship among TH
cell subpopulations may be more plastic than previously
suspected. At early stages in differentiation, at least, helper
cells may be able to shift their commitment and produce
another subset’s signature cytokine(s). For example, when
exposed to IL-12, young TH2 cells can be induced to express
the signature TH1 cell cytokine, IFN-␥. Young TH1 cells can
also be induced to express the signature TH2 cytokine, IL-4,
under TH2 polarizing conditions.
Interestingly, TH1 and TH2 cells do not seem able to adopt
TH17 or iTREG characteristics. On the other hand, TH17 and
iTREG cells are more flexible and can adopt the cytokine
expression profiles of other subsets, including TH1 and TH2.
The TH17 subset may be the most unstable or “plastic” lineage
and can be induced to express IFN-␥ and IL-4, depending on
environmental input. Some iTREG cells can also be induced to
express IFN-␥, and some can be redirected toward a TH17
phenotype if exposed to IL-6 and TGF-. This fluidity among
subsets makes it difficult to definitively establish the independence of helper cell lineages. In fact, some of the emerging
subgroups may be variants of TH1, TH2, TH17, and iTREG subsets that have been exposed to other polarizing environments.
Helper T-Cell Subsets Play Critical Roles
in Immune Health and Disease
Studies in both mice and humans show that the balance of
activity among T-cell subsets can significantly influence the
outcome of the immune response. A now classic illustration
of the influence of T-cell subset balance on disease outcome
is provided by leprosy, which is caused by Mycobacterium
leprae, an intracellular pathogen that can survive within the
phagosomes of macrophages. Leprosy is not a single clinical
entity; rather, the disease presents as a spectrum of clinical
responses, with two major forms of disease, tuberculoid and
lepromatous, at each end of the spectrum. In tuberculoid
leprosy, cell-mediated immune responses destroy most of
the mycobacteria. Although skin and peripheral nerves are
damaged in tuberculoid leprosy, it progresses slowly and
patients usually survive. In contrast, lepromatous leprosy
is characterized by a humoral response; cell-mediated
immunity is depressed. The humoral response sometimes
results in markedly high levels of immunoglobulin (hypergammaglobulinemia). This response is not as effective in
inhibiting disease, and mycobacteria are widely disseminated in macrophages, often reaching numbers as high as
1010 per gram of tissue. Lepromatous leprosy progresses
into disseminated infection of the bone and cartilage with
extensive nerve and tissue damage.
The development of lepromatous or tuberculoid leprosy
depends in part on the balance between TH1 and TH2
responses (Figure 11-12). In tuberculoid leprosy, the immune
response is characterized by a TH1-type response and high
circulating levels of IL-2, IFN-␥, and Lymphotoxin-␣ (LT-␣).
In lepromatous leprosy, there is a TH2-type immune
response, with high circulating levels of IL-4 and IL-5 (and
IL-10, which can also be made by TH2 cells). This cytokine
profile explains the diminished cell-mediated immunity and
increased production of serum antibody in lepromatous
leprosy.
Presumably each of these patients was exposed to the
same pathogen. Why did some develop an effective TH1
response while others did not? Studies suggest that genetic
differences among human hosts play a role. For example, differences in susceptibility may correlate with individual differences in the expression of PRRs (TLR1 and TLR2) expressed
TH1 activity
Tuberculoid Lepromatous
TH2 activity
Tuberculoid Lepromatous
IL-2
IL-4
IFN-␥
IL-5
LT-␣
IL-10
FIGURE 11-12 Correlation between type of leprosy and
relative TH1 or TH2 activity. Messenger RNA isolated from lesions
from tuberculoid and lepromatous leprosy patients was analyzed by
Southern blotting using the cytokine probes indicated. Cytokines
characteristic of TH1 cells (IFN-␥ and TNF-, for instance) predominate
in the tuberculoid patients, whereas cytokines characteristic of TH2
cells (IL-4) predominate in the lepromatous patients. [From P. A. Sieling
and R. L. Modlin, 1994, Cytokine patterns at the site of mycobacterial infection,
Immunobiology 191:378.]
T-Cell Activation, Differentiation, and Memory
by innate cells. This makes sense given that interactions
between pathogen and innate immune cells determine the
cytokine environment that influences the outcome of T-cell
polarization. Differences in TLR expression or activity could
alter the quality or quantity of cytokines produced.
Progression of HIV infection to AIDS may also be influenced by T-cell subset balance. Early in the disease, TH1
activity is high, but as AIDS progresses, some have suggested
that a shift may occur from a TH1-like to a TH2-like response,
which is less effective at controlling viral infection. In addition, some pathogens may “purposely” influence the activity
of the TH subsets. The Epstein-Barr virus, for example, produces a homolog (mimic) of human IL-10 called viral IL-10
(vIL-10). Like cellular IL-10, vIL-10 tends to suppress TH1
activity by inhibiting the polarizing activation of macrophages. Some researchers have speculated that vIL-10 may
reduce the cell-mediated response to the Epstein-Barr virus,
thus conferring a survival advantage on the virus.
TH17 cells first received attention because of their association with chronic autoimmune disease. Mice that were
unable to make IL-23, a cytokine that contributes to TH17
polarization, were remarkably resistant to autoimmunity.
TH17 cells and their defining effector cytokine, IL-17, are
often found in inflamed tissue associated with rheumatoid
arthritis, inflammatory bowel disease, multiple sclerosis,
psoriasis, and asthma. However, the role TH17 cells played in
protecting organisms from infection was not immediately
obvious. Studies of individuals with an autosomal dominant
form of a disease known as Hyper-IgE syndrome or Job syndrome confirmed indications from mice that TH17 cells were
important in controlling infections by extracellular bacteria
and fungi (see Clinical Focus Box 11-4).
These disorders and those described in Chapters 15 and 16
are just some examples of the influence of helper T-cell subsets
on disease development. It is important to recognize that our
current perspectives of the roles of helper subsets in disease
and health are still simplistic. Our developing appreciation of
the complexity of the interplay among subsets will improve
and add more subtlety to our explanations in the future.
T-Cell Memory
T-cell activation results in a proliferative burst, effector cell
generation, and then a dramatic contraction of cell number.
At least 90% of effector cells die by apoptosis after pathogen
is cleared, leaving behind an all-important population of
antigen-specific memory T cells. Memory T cells are generally long-lived and quiescent, but respond with heightened
reactivity to a subsequent challenge with the same antigen.
This secondary immune response is both faster and more
robust, and hence more effective than a primary response.
Until recently, memory cells were difficult to distinguish
from effector T cells and naïve T cells by phenotype, and for
some time they were defined best by function. Like naïve T
cells, most memory T cells are resting in the G0 stage of the
cell cycle, but they appear to have less stringent requirements
|
CHAPTER 11
379
for activation than naïve T cells. For example, naïve T cells
are activated almost exclusively by dendritic cells, whereas
memory T cells can be activated by macrophages, dendritic
cells, or B cells. Memory cells express different patterns of
surface adhesion molecules and costimulatory receptors that
allow them to interact effectively with a broader spectrum of
APCs. They also appear to be more sensitive to stimulation
and respond more quickly. This may, in part, be due to their
ability to regulate gene expression more readily because of
differences in epigenetic organization that occurred during
their formation. Finally, memory cells display recirculation
patterns that differ from those of naïve or effector T cells.
Some stay for long periods of time in the lymph node and
other secondary lymphoid organs, some travel to and
remain in tertiary immune tissues where the original infection occurred, anticipating the possibility of another infection with the antigen to which they are specific.
As our ability to identify different cell surface proteins
improved, so has our ability to identify and distinguish memory cell subpopulations. We now broadly distinguish two
subsets of memory T cells, central memory T cells (TCM) and
effector memory T cells (TEM), on the basis of their location,
their patterns of expression of surface markers, and, to some
extent, their function. Recent work also reveals a great deal of
diversity within these subsets, whose relationship is still being
clarified. We describe some useful generalizations below, and
close with the many questions that remain.
Naïve, Effector, and Memory T Cells Display
Broad Differences in Surface Protein Expression
Three surfaces markers have been used to broadly distinguish naïve, effector, and memory T cells: CD44, which
increases in response to activation signals; CD62L, an adhesion protein; and CCR7, a chemokine receptor. Both CD62L
and CCR7 are involved in homing to secondary lymphoid
organs (Table 11-4). Naïve T cells express low levels of CD44,
reflecting their unactivated state, and high levels of the adhesion molecule CD62L, directing them to the lymph node or
spleen. In contrast, effector helper and cytotoxic T cells have
the reciprocal phenotype. They express high levels of CD44,
indicating that they have received TCR signals, and low levels of CD62L, which prevents them from recirculating to
secondary lymphoid tissue, allowing them to thoroughly
probe sites of infection in the periphery.
Both types of memory T cells also tend to express CD44,
indicating that they are antigen experienced (i.e., have
received signals through their TCR). Like naïve T cells, central memory cells (TCM) express CD62L and the chemokine
receptor, CCR7, consistent with their residence in secondary
lymphoid organs. Effector memory cells (TEM), which are
found in a variety of tissues, can express varying levels of
CD62L depending on their locale; however, they do not
express CCR7, reflecting their travels through and residence
in nonlymphoid tissues. Other markers have been used to
distinguish subtypes of memory cells, but these still provide
a useful starting point for gauging the status of a T cell.
380
PA R T V
TABLE 11-4
|
Adaptive Immunity: Effector Responses
Surface proteins that are used
to distinguish naïve, effector,
and memory T cells
Cell type
CD44
CD62L
CCR7
low
⫹
⫹
Effector T cell
⫹
low
⫺
Effector memory T cell
⫹
variable
⫺
Central memory T cell
⫹
⫹
⫹
Naïve T cell
TCM and TEM Are Distinguished by Their Locale
and Commitment to Effector Function
A small proportion (⬍10%) of the progeny of a naïve cell
that has proliferated robustly in response to antigen differentiates into TCM and TEM cells. In general, these two subsets
are distinguished by where they reside as well as their level
of commitment to a specific effector cell fate.
In general, TCM cells reside in and travel between secondary lymphoid tissues. They live longer and have the capacity
to undergo more divisions than their TEM counterparts.
When they reencounter their cognate pathogen in secondary lymphoid tissue, they are rapidly activated and have the
capacity to differentiate into a variety of effector T-cell subtypes, depending on the cytokine environment.
On the other hand, TEM cells travel to and between tertiary tissues (including skin, lung, liver, and intestine). They
are arguably better situated to contribute to the first line of
defense against reinfection because they have already committed to an effector lineage during the primary response
and exhibit their effector functions quite rapidly after reactivation by their cognate pathogen.
It is important to note that some of these generalizations
may not hold up to scientific scrutiny. For instance, some
investigators contest the definitive distinction between TCM
and TEM cells and emphasize the diversity and continuum of
variations among these subtypes. Hence, memory cell biology is one of the most active fields of investigation, one that
is critical to our ability to develop the best vaccines.
How and When Do Memory Cells Arise?
Current work suggests that memory cells arise very early in the
course of an immune response (e.g., within 3 days), but their
cell of origin remains controversial. Some investigations suggest that memory cells arise as soon as naïve T cells are activated. Others suggest that memory cells arise from more fully
differentiated naïve T cells. Still others raise the intriguing possibility that naïve T-cell activation generates a “memory stem
cell” that is self-renewing and gives rise to memory effector cell
populations. These models are not mutually exclusive, and it is
possible that memory cells can arise at several different stages
of T-cell activation throughout a primary response.
The relationship between TCM and TEM cells is also
debated. They may originate independently from naïve and
effector cells, respectively, or may give rise to each other.
Studies suggest, in fact, that TCM cells arise from TEM cells,
and one possible model of relationships is shown in Figure
11-13. Here, investigators speculate that TCM cells arise prior
to TEM cells, from cells at an earlier stage of differentiation
into effector (helper or cytotoxic) T cells. TEM cells arise late,
and also may develop from fully differentiated effector cells.
The model also suggests that effector cells can replenish central memory cells.
It should be stressed, however, that several other models
have also been advanced. For instance, recent work suggests
interactions experienced by effector cells determines their TCM
versus TEM fate. Effector cells that interact with B cells may
preferentially develop into central and not effector memory T
cells. New models may also need to incorporate intriguing
recent observations, including the possibilities that (1) memory
Proliferation and differentiation
Effector T cell
Naïve T cell
?
Central memory
T cell
FIGURE 11-13 One possible model for the development
of central and effector memory T cells. This model, only
one of several that have been advanced, suggests that central
memory cells arise early after naïve T-cell activation, perhaps from
the first divisions. Effector memory T cells may arise later, after
the progeny have divided more and have assumed at least some
?
Effector memory
T cell
effector cell features. The model also includes the possibilities
that (1) some effector memory T cells arise from fully differentiated effector T cells and (2) effector memory T cells can develop
into central memory T cells. [Adapted from D. Gray, 2002, A role for
antigen in the maintenance of immunological memory, Nature Reviews
Immunology 2:60.]
T-Cell Activation, Differentiation, and Memory
cells arise from the asymmetric cell division of activated T cells,
where one daughter cell becomes an effector cell, and another
contributes to the memory pool, and (2) that T-cell activation
generates a self-renewing memory stem cell population that
provides a long-term source of memory T cells.
What Signals Induce Memory Cell
Commitment?
Most investigators agree that T-cell help is critical to generating long-lasting memory. For instance, CD8⫹ T cells can be
activated in the absence of CD4⫹ T-cell help, but this “helpless” activation event does not yield long-lived memory
CD8⫹ T cells. The relative importance of other variables in
driving memory development is still under investigation.
Although intensity of T-cell receptor engagement was
thought to be a factor in memory cell commitment, recent
data suggest that even low-affinity interactions can generate
memory T cells. All studies, however, appear consistent with
the recognition that the more proliferation a response
inspires, the better the memory pool.
Do Memory Cells Reflect the Heterogeneity of
Effector Cells Generated during a Primary
Response?
We have seen that naïve T cells differentiate into a wide variety of effector T-cell subpopulations, largely determined by
the cytokine signals they receive during activation. Studies
indicate that the memory cell response is also very diverse,
in terms of both the T-cell receptor specificities and the array
|
CHAPTER 11
381
of cytokines produced. However, the cellular origin of this
diversity is still under investigation. Specifically, does this
diverse memory response strictly reflect the functional effector diversity generated during the primary response? Or
does it develop anew from central memory T cells responding to different environmental cues during rechallenge? The
answer is likely to be “Both,” but investigations continue.
Are There Differences between CD4⫹
and CD8⫹ Memory T Cells?
The simple answer is “Maybe.” Memory CD8⫹ T cells are
clearly more prevalent than memory CD4⫹ T cells. This is
partly because CD8⫹ T cells proliferate more robustly and
therefore generate proportionately more memory T cells. It
may also be due to differences in the life span of memory T
cells: CD4⫹ memory T cells may not be as long-lived as
CD8⫹ memory T cells.
How Are Memory Cells Maintained
over Many Years?
Whether memory cells can persist for years in the absence of
antigen remains controversial, although evidence seems to
favor the possibility that they do. Regardless, it does seem
that memory persistence depends on the input of cytokines
that induce occasional divisions, a process known as homeostatic proliferation, which maintains the pool size by balancing apoptotic events with cell division. Both IL-7 and IL-15
appear important in enhancing homeostatic proliferation,
but CD4⫹ and CD8⫹ memory T-cell requirements may differ.
S U M M A R Y
■
■
■
■
T-cell activation is the central event in the initiation of the
adaptive immune responses. It results from the interaction
in a secondary lymphoid tissue between a naïve T cell and
an APC, specifically a dendritic cell.
Activation of naïve T cells leads to the differentiation of effector cells, which regulate the response to pathogen, and of
long-lived memory T cells, which coordinate the stronger and
quicker response to future infections by the same pathogen.
Three distinct signals are required to induce naïve T-cell activation, proliferation, and differentiation. Signal 1 is generated
by the interaction of the TCR-CD3 complex with an MHCpeptide complex on a dendritic cell. Signal 2 is a costimulatory signal provided by the interaction between molecules of
the B7 family expressed by APC with the positive costimulatory molecules CD28 or ICOS expressed by T cells. Signal 3
is provided by soluble cytokines and plays a key role in determining the type of effector cell that a T cell becomes.
In the absence of a costimulatory signal (Signal 2), T-cell
receptor engagement results in T-cell inactivity or clonal
anergy.
■
■
■
■
CD8⫹ T cells, which recognize MHC class I-peptide complexes that are expressed by virtually all cells in the body,
become cytotoxic (TC) cells with the capacity to kill
infected cells.
CD4⫹ T cells, which recognize MHC class II-peptide
complexes that are expressed by professional APCs,
become helper (TH) cells, secreting cytokines that regulate
(positively and negatively) cells that clear infection,
including B cells, macrophages, and other T cells.
CD4⫹ T cells differentiate into at least five main subpopulations of effector cells: TH1, TH2, TH17, iTREG, and TFH. Each
subpopulation is characterized by (1) a unique set of polarizing cytokines that initiate differentiation, (2) a unique master transcriptional regulator that regulates the production of
helper-cell-specific genes, and (3) a distinct set of effector
cytokines that they secrete to regulate the immune response.
TH1 and TH17 cells generally enhance cell-mediated
immunity and inflammatory responses. TH2 and TFH cells
enhance humoral immunity and antibody production,
and induced TREG cells inhibit T-cell responses.
382
■
■
■
PA R T V
|
Adaptive Immunity: Effector Responses
Helper T-cell subsets have also been associated with disease and play a role in the development of autoimmunity
and allergy.
Memory T cells, which are more easily activated than
naïve cells, are responsible for secondary responses.
The generation of memory B cells as well as CD4⫹ and
CD8⫹ memory T cells requires T-cell help.
■
Two types of memory T cells have been described. Central
memory (TCM) cells are longer lived, reside in secondary
lymphoid tissues, and can differentiate into several different effector T cells. Effector memory (TEM) cells populate
the sites of infection (tertiary tissues) and immediately
reexpress their original effector function after reexposure
to antigen.
R E F E R E N C E S
Ahmed, R., M. Bevan, S. Reiner, and D. Fearon. 2009. The precursors of memory: Models and controversies. Nature Reviews.
Immunology 9:662–668.
Miossec, P., T. Korn, and V.K. Kuchroo. 2009. Interleukin 17
and Type 17 helper T cells. New England Journal of Medicine
361:888–898.
Gattinoni, L., et al. 2011. A human memory T cell subset with
stem cell-like properties. Nature Medicine 17:1290–1297.
Mosmann, T., H. Cherwinski, M. Bond, M. Giedlin, and R.
Coffman. 1986. Two types of murine helper T cell clone. I.
Definition according to profiles of lymphokine activities
and secreted proteins. Journal of Immunology 136:2348–
2357.
Hoyer, K. K., W. F. Kuswanto, E. Gallo, and A. K. Abbas. 2009.
Distinct roles of helper T-cell subsets in systemic autoimmune
disease. Blood 113:389–395.
Jameson, S., and D. Masopust. 2009. Diversity in T cell memory:
An embarrassment of riches. Immunity 31:859–871.
Jelley-Gibbs, D., T. Strutt, K. McKinstry, and S. Swain. Influencing the fates of CD4⫹ T cells on the path to memory: Lessons
from influenza. Immunology and Cell Biology 86:343–352.
Kaiko, G. E., J. C. Horvat, K. W. Beagley, and P.M. Hansbro.
2007. Immunological decision making: How does the immune
system decide to mount a helper T-cell response? Immunology
123: 326–338.
Kapsenberg, M. L. 2003. Dendritic cell control of pathogen-driven
T-cell polarization. Nature Reviews Immunology 3: 984–993.
Kassiotis, G., and A. O’Garra. 2009. Establishing the follicular
helper identity. Immunity 31:450–452.
Murphy, K., C. Nelson, and J. Sed. 2006. Balancing co-stimulation and inhibition with BTLA and HVEM. Nature Reviews
Immunology 6:671–681.
Palmer, M. T., and C. T. Weaver. 2010. Autoimmunity: Increasing suspects in the CD4⫹ T cell lineup. Nature Immunology
11:36–40.
Pepper, M., and M. K. Jenkins. 2011. Origins of CD4⫹ effector
and memory T cells. Nature Immunology 12:467–471.
Readinger, J., K. Mueller, A. Venegas, R. Horai, and P. Schwartzberg. 2009. Tec kinases regulate T-lymphocyte development
and function: New insights into the roles of Itk and Rlk/Txk.
Immunological Reviews 228:93–114.
Reiner, S. 2008. Inducing the T cell fates required for immunity.
Immunological Research 42:160–165.
Khoury, S., and M. Sayegh. 2004. The roles of the new negative
T cell costimulatory pathways in regulating autoimmunity.
Immunity 20:529–538.
Riley, J. 2009. PD-1 signaling in primary T cells. Immunology
Reviews 229:114–125.
King, C. 2009. New insights into the differentiation and function of
T follicular helper cells. Nature Reviews. Immunology 9:757–766.
Rudd, C., A. Taylor, and H. Schneider. 2009. CD28 and CTLA-4
coreceptor expression and signal transduction. Immunological
Reviews 229:12–26.
Korn, T., E. Bettelli, M. Oukka, and V. Kuchroo. 2009. IL-17 and
TH17 Cells. Annual Review of Immunology 27:485–517.
Lefrancois, L., and D. Masopust. 2009. The road not taken:
Memory T cell fate “decisions.” Nature Immunology 10:369–370.
Linsley, P., and S. Nadler. 2009. The clinical utility of inhibiting CD28-mediated costimulation. Immunological Reviews
229:307–321.
Malissen, B. 2009. Revisiting the follicular helper T cell paradigm. Nature Immunology 10:371–372.
Mazzoni, A., and D. Segal. 2004. Controlling the Toll road to dendritic cell polarization. Journal of Leukocyte Biology 75:721–730.
McGhee, J. 2005. The world of TH1/TH2 subsets: First proof.
Journal of Immunology 175:3–4.
Sallusto, F., and A. Lanzavecchia. 2009. Heterogeneity of CD4⫹
memory T cells: Functional modules for tailored immunity.
European Journal of Immunology 39:2076–2082.
Sharpe, A. 2009. Mechanisms of costimulation. Immunological
Reviews 229:5–11.
Smith-Garvin, J., G. Koretzky, and M. Jordan. 2009. T cell activation. Annual Review of Immunology 27:591–619.
Thomas, R. 2004. Signal 3 and its role in autoimmunity. Arthritis Research & Therapy 6:26–27.
Thompson, C., et al. 1989. CD28 activation pathway regulates
the production of multiple T-cell-derived lymphokines/
cytokines. Proceedings of the National Academy of Sciences of
the United States of America 86:1333–1337.
T-Cell Activation, Differentiation, and Memory
Wang, S., and L. Chen. 2004. T lymphocyte co-signaling pathways of the B7-CD28 family. Cellular & Molecular Immunology
1:37–42.
Weaver, C., and R. Hatton. 2009. Interplay between the TH17
and TREG cell lineages: A (co)evolutionary perspective. Nature
Reviews. Immunology 9:883–889.
Yu, D., et al. 2009. The transcriptional repressor Bcl-6 directs
T follicular helper cell lineage commitment. Immunity
31:457–468.
Zhou, L., M. Chong, and D. Littman. 2009. Plasticity of CD4⫹ T
cell lineage differentiation. Immunity 30:646–655.
Zhu, J., and W. Paul. 2010. Heterogeneity and plasticity of T
helper cells. Cell Research 20:4–12.
Useful Web Sites
The following are examples of well-organized Web sites
developed by undergraduate students, graduate students,
and teachers of immunology who have done an excellent job
S T U D Y
|
CHAPTER 11
383
of simplifying a complex topic: helper T-cell subset differentiation. The Web sites may not be continually updated, so, as
always with Internet sources, double-check the date and the
accuracy of the information.
http://wenliang.my web.uga.edu/mystudy/
immunology/ScienceOfImmunology/Biological
process(immuneresponses).html These Web notes from
a former graduate student at the University of Georgia provide a rudimentary, but accurate and accessible, description
of the T-cell subsets.
http://microbewiki.kenyon.edu/index.php/Host_
Dependency_of_Mycobacterium_leprae You will
find here a posting from MicrobeWiki, “a student-edited
microbiology resource” originating from Kenyon College.
http://users.rcn.com/jk imball.ma.ultranet/
B i o l o g y Pa g e s / T / T H 1 _ T H 2 . h t m l # Ty p e s _ o f _
Helper_T_Cells This selection is from immunologist and
teacher John Kimball’s online version of his textbook.
Q U E S T I O N S
CLINICAL AND EXPERIMENTAL FOCUS QUESTION Multiple
sclerosis is an autoimmune disease in which TH cells participate in the destruction of the protective myelin sheath around
neurons in the central nervous system. Each person with
this disease has different symptoms, depending on which
neurons are affected, but the disease can be very disabling.
Recent work in a mouse model of this disease suggests that
transplantation of cell precursors of neurons may be a good
therapy. Although these immature cells may work because
they can develop into neuronal cells that replace the lost myelin sheath, some investigators realized that they play another,
perhaps even more important role. These scientists showed
TH17
TH1
0.6
2.9
TREG
1.2
36.9
Control
22.1
that the neuronal cell precursors secrete a cytokine called
Leukemia Inhibiting Factor (LIF). In fact, the administration
of this factor, alone, ameliorated symptoms.
These investigators were curious to know if this cytokine had
an effect on T-cell activity. They added LIF to cultures of (normal) T cells that were being stimulated under different polarizing conditions (i.e., TCR and CD28 engagement in the presence
of cytokines that drive differentiation to distinct T helper subsets). They stained T cells for cytokine production and analyzed
the results by flow cytometry. The data below show their results
from normal mouse T cells polarized to the lineage indicated in
the absence (top, control) or presence (bottom) of LIF.
3.4
0.7
2.7
0.6
37.7
5.2
IFN-γ
29.1
FoxP3
IL-17
LIF
11.2
27.8
CD4
PA R T V
384
|
Adaptive Immunity: Effector Responses
Which T helper lineage(s) is(are) most affected by the addition of LIF? Explain your answer. Why might these results
explain the beneficial effect of LIF on the disease? Knowing
what you know now about the molecular events that influence
T-cell differentiation, speculate on the molecular basis for the
activity of LIF.
1. Which of the following conditions would lead to T-cell
anergy?
5. The following sentences are all false. Identify the error(s)
and correct.
a. Macrophages activate naïve T cells better than dendritic
cells.
b. ICOS enhances T-cell activation and is called a negative
coreceptor.
c. Virtually all cells in the body express costimulatory
ligands.
d. CD28 is the only costimulatory receptor that binds to
B7 family members.
a. A naïve T-cell interaction with a dendritic cell in the
presence of CTLA-4 Ig.
b. A naïve T cell stimulated with antibodies that bind both
the TCR and CD28.
c. A naïve T cell stimulated with antibodies that bind only
the TCR.
d. A naïve T cell stimulated with antibodies that bind only
CD28.
2. A virus enters a cut in the skin of a mouse and infects den-
dritic cells, stimulating a variety of PRRs both on and
within dendritic cells that induce it to produce IL-12. The
mouse subsequently mounts an immune response that successfully clears the infection. Which of the following statements is(are) likely to be true about the immune response
that occurred? Correct any that are false.
a. The infected dendritic cells up-regulated CD80/CD86
and MHC class II.
b. The dendritic cells encountered and activated naïve T
cells in the skin of the mouse.
c. Naïve T cells activated by these dendritic cells generated
signals that released internal Ca2⫹ stores.
d. Naïve T cells activated by these dendritic cells were
polarized to the TH2 lineage.
e. Only effector memory T cells were made in this mouse.
3. Your lab acquires mice that do not have the GATA-3 gene
(GATA-3 knockout mice). You discover that this mouse
has a difficult time clearing helminth (worm) infections.
Why might this be?
4. You isolate naïve T cells from your own blood and want to
polarize them to the TH1 lineage in vitro. You can use any of
the following reagents to do this. Which would you choose?
Anti-TCR antibody
IL-4
IFN␥
CTLA-4 Ig
anti-CD80 antibody
anti-CD28 antibody
IL-12
IL-17
e. Signal 3 is provided by negative costimulatory recep-
tors.
f. Toxic shock syndrome is an example of an autoimmune
disease.
g. Superantigens mimic TCR-MHC class I interactions.
⫹
⫹
h. CD4 T cells interact with MHC class I on CD8 T
cells.
Naïve T cells produce IFN-␥.
T-bet and GATA-3 are effector cytokines.
Polarizing cytokines are only produced by APCs.
Bcl-6 is involved in the delivery of costimulatory
signals.
m. TH17 and TFH cell subsets are the major sources of
B-cell help.
n. iTREG cells enhance inflammatory disease.
o. Effector cytokines act exclusively on T cells.
p. Central memory T cells tend to reside in the site of
infection.
q. Like naïve T cells, effector memory cells express CCR7.
i.
j.
k.
l.
6. Like TH1 and TH2 cells, TH17 and TREG cells cross-regulate
each other. Which of these two statements about this crossregulation is (are) true? Correct either, if false.
a. TGF- is a polarizing cytokine that stimulates up-
regulation of each of the master transcriptional regulators that polarize T cells to the TH17 and TREG lineages.
b. IL-6 inhibits polarization to the TREG lineage by inhibiting expression of ROR␥.
7. A new effector T-cell subset, TH9, has been recently identi-
fied. It secretes IL-9 and IL-10 and appears to play a role in
the protection against intestinal worm infection. What
other information about this subset would help you to
determine if it should be considered an independent helper
T-cell lineage?
12
B-Cell Activation,
Differentiation, and
Memory Generation
T
he function of a B cell is to secrete antibodies
capable of binding to any organism or molecule
that poses a threat to the host. The secreted
antibodies have antigen-binding sites identical
to those of the receptor molecules on the B-cell surface.
Antibodies belong to the class of proteins known as
immunoglobulins, and once secreted, they can protect
the host against the pathogenic effects of invading
viruses, bacteria, or parasites in a variety of ways, as
described in Chapters 1 and 13.
Our current understanding of B-cell clonal selection,
activation, proliferation, and deletion finds its beginnings
in a theoretical paper in the Australian Journal of Science,
written by Sir Frank MacFarlane Burnet (Figure 12-1) over
the course of a single weekend in 1957. In this paper, built
on prior work by Neils Jerne, David Talmage, Peter
Medawar and others, Burnet laid out the Clonal Selection
Hypothesis, which provided the conceptual underpinnings
for the entire field of immunology (see Figure 1-7). The
Clonal Selection Hypothesis suggested for the first time
that the receptor molecule on the lymphocyte surface and
the antibody products secreted by that cell had identical
antigen-binding specificities. Furthermore, it posited that
stimulation of a single B cell would result in the generation
of a clone of cells having the identical receptor specificity as
the original cell, and that these clones would migrate to,
and function within, the secondary lymphoid organs. The
daughter cells within each clone would not only be able to
secrete large amounts of antibody to neutralize the
pathogen, but some progeny cells would also remain viable
within the organism and available to neutralize a secondary
infection by the same pathogen.
In other words, in one brilliant paper Professor Burnet
gave generations of immunologists the basis for thinking
about B-cell receptor diversity, lymphocyte trafficking,
and immune memory—all this at a time when
practitioners in the field were still unsure of the
differences between T and B cells.
Burnet’s paper went on to predict the generation of the
vast array of antibody specificities known to exist today. It
is difficult for us now, in the second decade of the twentyfirst century, to even begin to appreciate the prescience of
The germinal center of the lymph node
contains B cells (green) in the dark zone and
follicular dendritic cells (red) in the light zone.
Naïve B cells (blue) define the follicular
mantle zone. [Reprinted from Victora, G.D. et al.,
2010, Germinal center dynamics revealed by
multiphoton microscopy with a photoactivatable
fluorescent reporter, Cell 143:592–605. Copyright
2010 with permission from Elsevier.]
■
T-Dependent B-Cell Responses
■
T-Independent B Cell Responses
■
Negative Regulation of B Cells
Burnet’s suggestion that “The theory requires at some
stage in early embryonic development a genetic process for
which there is no available precedent [italics added]. In
some way, we have to picture a ‘randomization’ of the
coding responsible for part of the specification of gamma
globulin molecules, so that after several cell generations,
in early mesenchymal cells, there are specifications in the
genomes for virtually every variant that can exist as a
gamma globulin molecule.” Just four years after the
eludication of the double helical structure of DNA, Burnet
was postulating that there are movable genetic segments!
The paper ends with a final flourish, in which Burnet
predicted the need for clonal deletion in the developing
lymphocyte repertoire, in order to eliminate B cells
bearing receptors with specificity for self antigens. His
formulation of the Clonal Selection Hypothesis, along with
the brilliant experimental work he performed with others
on the generation of immunological tolerance, resulted in
385
386
PA R T V
|
Adaptive Immunity: Effector Responses
FIGURE 12-1 Sir MacFarlane Burnet, 1899–1985. Professor
McFarlane Burnet, the author of the Clonal Selection Hypothesis,
shared the 1960 Nobel Prize for Physiology or Medicine with Sir Peter
Medawar of Britain for “the discovery of acquired immunological
tolerance.” [Sir Macfarlane Burnet 1960–61 by Sir William Dargie. Oil on composition board. Collection: National Portrait Gallery, Canberra. Purchased 1999]
his being awarded the Nobel Prize for Physiology or
Medicine in 1960, along with Sir Peter Medawar. The
essential tenets of the Clonal Selection Hypothesis are
summarized below, and a visual representation of the
events he predicted are shown in Figure 12-2.
• Immature B lymphocytes bear immunoglobulin (Ig)
receptors on their cell surfaces. All receptors on a single
B cell have identical specificity for antigen.
• Upon antigen stimulation, the B cell will mature and
migrate to the lymphoid organs. There it will replicate.
Its clonal descendants will bear the same receptor as the
parental B cell and secrete antibodies with an identical
specificity for antigen.
• At the close of the immune response, more B cells bearing receptors for the stimulating antigen will remain in
the host than were present before the antigenic challenge. These memory B cells will then be capable of
mounting an enhanced secondary response.
• B cells with receptors for self antigens are deleted during
embryonic development.
In summary, each B cell bears a single type of Ig
receptor generated by the processes described in Chapter 7,
and on stimulation will create a clone of cells bearing the
same antigen receptor as the original B cell. Many B cells
bearing receptors with specificity for self antigens are
eliminated from the primary B-cell repertoire as
described in Chapter 10. In this chapter, we describe what
happens when mature B cells, located in the peripheral
lymphoid organs, encounter antigen.
The two major types of B-cell responses are elicited
by structurally distinct types of antigens. The first type
of response that we will describe is generated upon
recognition of protein antigens and requires the
participation of CD4⫹ helper T cells. Because T cells are
involved, this class of B-cell response is therefore known
as a T-dependent (TD) response, and it is mediated by
B-2 B cells binding to TD antigens. Because B-2 B cells
represent the majority of B cells, we will routinely refer
to the B-2 B-cell subset simply as “B cells,” and
distinguish the other B-cell subclasses by their particular
names as B-1, marginal zone, or B-10 B cells.
The T-dependent response requires two distinct
signals. The first is generated when a multivalent antigen
binds and cross-links membrane immunoglobulin
receptors (mIg) (Figure 12-3a). The second signal is
provided by an activated T cell, which binds to the B cell
both through its antigen receptor and via a separate
interaction between CD40 on the B cell and CD40L
(CD154) on the activated TH cell. The bound T cell then
delivers cytokines and other signals to its partner B cell
to complete the activation process.
The second type of response, which we will describe
later in this chapter, is directed toward multivalent or
highly polymerized antigens, and does not require T-cell
help. This type of response is referred to as a
T-independent response, and the antigens that elicit such
responses are T-independent (TI) antigens. One class of
TI antigens (TI-1 antigens), exemplified by the
lipopolysaccharide moiety of Gram-negative bacteria,
interacts with the B cell via both mIg and innate immune
receptors. TI-1 antigens are mitogenic (induce
proliferation) for most B cells at high concentrations, as a
result of their ability to bind to pattern recognition
receptors (PRRs) on the surface of the B cell. However, at
lower concentrations they activate only those B cells that
bind antigen with their Ig receptors (Figure 12-3b). The
other class of TI antigens, TI-2 antigens, includes highly
repetitive antigens, such as bacterial capsular
polysaccharides. These antigens are not inherently
mitogenic, but their ability to cross-link a large fraction of
the Ig receptors on the surface of a B cell allows them to
deliver an activation signal in the absence of T-cell help.
Many TI-2 antigens are specifically and covalently bound
by the complement component C3d, and mice depleted of
C3d mount poor responses to TI-2 antigens (Figure 12-3c).
Bone marrow
Peripheral lymphoid tissue
Memory cell
2
IgM
IgD
1
Antibody
2
2
2
1
Plasma cells
Antigen 2
2
2
2
2
2
2
Gene
rearrangement
2
Stem
cell
2
3
3
2
2
2
4
Clonal
deletion
Immature
B cells
Mature
B cells
2
2
Antigen-dependent proliferation and differentiation
into plasma and memory cells
Maturation into immature
committed B cells
FIGURE 12-2 Maturation and clonal selection of B lymphocytes. B-cell maturation, which occurs in the absence of antigen,
first produces immature B cells bearing IgM receptors. Each B cell bears
receptors of one specificity only. Any B cell with receptors specific for
antigens expressed in the bone marrow are deleted at the immature
B cells stage (indicated by clone 4). Those B cells that do not express
self-reactive receptors mature to express both IgM and IgD receptors
(a) TD antigen
and are released into the periphery, where they recirculate among the
blood, lymph, and lymphoid organs. Clonal selection occurs when an
antigen binds to a B cell with a receptor specific for that antigen. Clonal
expansion of an antigen-activated B cell (number 2 in this example)
leads to a clone of effector B cells and memory B cells; all cells in the
expanded clone are specific for the original antigen. The effector,
plasma cells secrete antibody reactive with the activating antigen.
(b) TI-1 antigen
Cytokines
CD21
TH cell
1 3
1
2
2
B cell
(c) TI-2 antigen
C3d
CD40/CD40L
FIGURE 12-3 Different types of antigens signal through
different receptor units. (a) T-dependent (TD) antigens bind to the
Ig receptor of B cells. Some of the antigen is processed and presented
to helper T cells. T cells bind to the MHC-peptide antigen, and deliver
further activating signals to the B cell via interaction between CD40L
(on the T cells) and CD40 on the B cells. In addition, T cells secrete
activating cytokines, such as IL-2 and IL-4, which are recognized by
receptors on the B-cell surface. Cytokines deliver differentiation, prolif-
B cell
TLR
1
B cell
eration, and survival signals to the B cells. (b) Type 1 T-independent
(TI-1) antigens bind to B cells through both Ig and innate immune
receptors. For example, LPS from Gram negative organisms binds to
B cells via both mIg and TLR4, resulting in signaling from both receptors. (c) Type 2 T-independent (TI-2) antigens are frequently bound by
C3d complement components and cross-link both mIg and CD21
receptors on B cells. Cross-linking of between 12 and 16 Ig receptors
has been shown to be sufficient to deliver an activating signal.
388
PA R T V
|
Adaptive Immunity: Effector Responses
Most T-independent responses are mediated by B-1 and
marginal zone B-cell types.
T-Dependent B-Cell Responses
Following the completion of their maturation program (see
Chapter 10), B cells migrate to the lymphoid follicles (Figure 2-8). They are directed there by chemokine interactions
between CXCL13, which is secreted by follicular dendritic
cells (FDCs) and its receptor, CXCR5, which is expressed on
B cells. It is important to recognize, however, that once
mature B cells have reached the follicles, they do not just
remain there; rather, they recirculate through the blood and
lymphatic systems and back to the lymphoid follicles many
times over the course of their existence. B-cell survival in the
follicle is dependent on access to the TNF-family member
cytokine B-cell activation factor (BAFF), otherwise known
as B lymphocyte survival factor (BLyS), which is secreted by
the FDCs, as well as by many types of innate immune cells,
such as neutrophils, macrophages, monocytes, and dendritic
cells. Mature B cells unable to secure a sufficient supply of
BAFF die by apoptosis. Recirculating, mature B cells are
estimated to have a half-life of approximately 4.5 months.
At the initiation of a T-dependent B-cell response, the B cell
binds antigen via its Ig receptors. Some of that bound antigen
is internalized into specialized vesicles within the B cells, where
it is processed and re-expressed in the form of peptides presented in the antigen-binding groove of class II MHC molecules (see Figure 8-19). T cells that have been previously
activated by an encounter with antigen-bearing dendritic cells
now bind to the MHC-presented peptide on the surface of the
B cell. This induces a vectorial redistribution of the T cell’s
secretory apparatus, such that it releases its activating cytokines directly into the T-cell/B-cell interface (as described in
Chapter 4). Other interactions also occur between accessory
molecules on the T- and B-cell surfaces that enhance their
binding and provide further signals to the pair of cells. For
example, CD40L on the T cell binds to CD40 on the B cell (see
Figure 12-3a) and the CD28 coreceptor on the T cell engages
with CD80 and CD86 on the surface of the B cell.
The B cell then integrates the activating signals received
through its antigen receptor with those from its various cytokine receptors and coreceptors and moves into specialized
regions of the lymph node or spleen to begin the process of
differentiation into an antibody-secreting cell. Recall that lymphocytes first enter the lymph node in the region of the T-cell
zone, just inside the follicles (Figure 2-8). Some of the antigenactivated B cells then move into regions at the borders of the
T-cell and B-cell areas, where they differentiate into clusters of
activated B cells known as primary foci (see Overview Figure
12-4). There, they complete their differentiation into plasma
cells. At around 4 days post stimulation, when this differentiation is complete, they migrate into the medullary cord regions
of the lymph node, or to parts of the red pulp of the spleen
close to the T-cell zones (Figure 2-10), where they secrete large
quantities of antibodies. Some of these plasma cells die after
the initial primary response is completed, whereas others take
up long-term residence in the bone marrow, gut, or other
locations as long-lived plasma cells. These primary focus
plasma cells deliver relatively large quantities of IgM antibodies in the early phases of the B-cell response.
Some antigen-stimulated B cells do not enter the primary
foci but rather migrate into the follicles of the lymph nodes
and spleen where they undergo further differentiation. As the
follicles swell with antigen-specific lymphocytes (primarily
B cells), they change their appearance and become known as
germinal centers (GCs). In these germinal centers, B-cell
variable region genes undergo mutational processes that result
in the secretion of antibodies with altered sequences in their
antigen-combining sites. This process of somatic hypermutation (SHM) is then followed by antigen selection, culminating
in the production of B cells bearing receptors and secreting
antibodies whose affinity for antigen increases as the immune
response progresses. Antibodies with mutations in their variable regions begin to appear in the circulation 6 to 10 days
following the onset of the immune response.
In addition, ongoing signals from helper T cells direct
their cognate B cells to produce antibodies of isotypes
(classes) other than IgM, in a process known as class switch
recombination (CSR). Both SHM and CSR are dependent
on the activity of a germinal center enzyme: activationinduced cytidine deaminase (AID).
The end result of this extraordinary set of events is the
production of high-affinity antibody molecules that eliminate
the pathogenic threat by one or more of the means described
in Chapter 13. At the close of the primary immune response,
memory B cells remain that are the daughters (and more
distant progeny) of those cells that were stimulated during the
primary response. Many of these progeny B cells now carry
mutated and selected B-cell receptors (BCRs). On secondary
exposure to the same antigen, these memory B cells will be
stimulated more quickly and will deliver higher-affinity and
heavy-chain class-switched antibodies, resulting in faster
elimination of microbial pathogens. This improved memory
response underlies the scientific rationale for vaccination.
As the immune response to T-dependent antigens winds
down, the host organism is left with two sets of long-lived cells
that ensure the provision of long-term memory responses to the
antigen. Plasma cells in the bone marrow and elsewhere create
a reservoir of antibody-producing cells that can last for the lifetime of the host, and memory B cells circulating through the
lymphoid organs remain poised for subsequent stimulation by
the same antigen. Excess residual lymphocytes and plasma cells
from the primary response are eliminated by apoptosis.
T-Dependent Antigens Require T-Cell Help
to Generate an Antibody Response
The initial experiments proving that B cells required
“help” from T cells in order to complete their differentiation were performed by Miller, Mitchell, Mitchison, and
B-Cell Activation, Differentiation, and Memory Generation
|
CHAPTER 12
389
12-4
OVERVIEW FIGURE
B-Cell Activation
T cell zone
Germinal center
Differentiation
Selection
Low
affinity
Apoptosis
Proliferation,
SHM, and
CSR
Conventional
naïve B cell
TD antigen
activation
Plasma
cell
FDC
Class-switched
memory B cell
Improved
affinity
TD antigen
activation
Abortive
autoreactive
GC B cell
TH cell
Apoptosis
IgM+ IgD+
memory B cell
Primary focus
Dark zone
Light zone
Medullary cords
B cells engaged in T-dependent activation first encounter antigenspecific T cells outside of the B-cell follicles. Some activated B cells
differentiate into antibody-producing plasma cells in primary foci
that lie outside the follicles, and then migrate to the medullary
cords of the lymph node, or the bone marrow, where they continue
to secrete antigen-specific antibodies. Other antigen-activated
B cells enter the follicle where they divide and differentiate. As the
follicle fills with proliferating B cells, it develops into a germinal
center, characterized by a light zone, in which follicular dendritic
others in the 1960s, using the technique of adoptive transfer.
In an adoptive transfer experiment, a mouse is lethally irradiated in order to eliminate its immune system; the investigator then attempts to reconstitute the ability to generate an
immune response by adding back purified cells of different
types from genetically identical mouse donors.
Using this technique, investigators showed that, in order
for a mouse to produce antibodies against a protein antigen, it
must receive cells derived from both the bone marrow and
from the thymus of a healthy donor animal. Neither thymusderived nor bone marrow-derived cells were capable of reconstituting the responding animal’s immune system on their
own (Figure 12-5). We now know, of course, that the thymusderived cells active in this response were mature helper T cells,
cells reside, and a dark zone in which the cell density is particularly
high. Within the germinal center, the immunoglobulin genes
undergo class switching, in which constant regions are replaced
by constant regions of other isotypes. The variable region is subject
to somatic hypermutation, and the mutated variable regions are
subject to antigen-mediated selection within the germinal centers.
Low-affinity and autoimmune receptor-bearing B cells die, and
those B cells with enhanced receptors leave the germinal centers
for the periphery. (See the text for details.)
whereas the bone marrow-derived, antibody-producing cells
were mature B cells, recirculating through the bone marrow.
In this way it was demonstrated that the antibody response to
protein antigens required both B cells and T cells.
Having introduced the major players and briefly described
the geography of the landscape in which the cells are operating, we will now sequentially step through a T-dependent
B-cell response.
Antigen Recognition by Mature B Cells
Provides a Survival Signal
The first step in antibody production is antigen recognition
by the Ig receptors on the surface of a naïve B cell (a B cell
390
PA R T V
|
Adaptive Immunity: Effector Responses
1. Irradiated recipient mice
2. Reconstituted recipient
mice with:
Nothing
3. Measured ability of recipient
mice to mount an antibody
response to a T-dependent
antigen (sheep red blood cells)
Bone marrow
cells alone
––
––
Thymus
cells alone
––
Bone marrow
and thymus cells
++
FIGURE 12-5 Adoptive transfer experiments demonstrated the need for two cell populations during the generation of
antibodies to T-dependent antigens. Early adoptive transfer experiments reconstituted syngeneic irradiated mice, in which the immune
system had been ablated, with bone-marrow cells, thymus-derived cells, or a mixture of bone-marrow and thymus-derived cells. These mice
were then challenged with a T-dependent antigen. Only recipient mice reconstituted with both bone-marrow and thymus-derived cells were
able to mount an antibody response.
that has not yet encountered antigen). These receptors, along
with their coreceptors and signaling properties, were
described in Chapter 3. Naïve B cells circulate in the blood
and lymph and pass through the secondary lymphoid
organs, most notably to the spleen and lymph nodes (Chapter 2). If a B cell is activated by interaction with an antigen,
the cell will proliferate and differentiate as described below.
If a B cell does not bind antigen, it will die by apoptosis a few
months after its emergence from the bone marrow.
B Cells Encounter Antigen in the
Lymph Nodes and Spleen
When antigen is introduced into the body, it becomes concentrated in various peripheral lymphoid organs. Bloodborne antigen is filtered by the spleen, whereas antigen from
tissue spaces drained by the lymphatic system is filtered by
regional lymph nodes (or lymphoid nodules in the gut).
Here, we will focus on antigen presentation to B cells in the
lymph nodes.
Antigen enters the lymph nodes either alone or associated
with antigen-transporting cells. Recall that B cells are capable of recognizing antigenic determinants on native, unprocessed antigens, whereas T cells must recognize antigen as a
peptide presented in the context of MHC antigens. In addition, as described in Chapter 6, antigen is often covalently
modified with complement fragments, and complement
receptors on B cells play an important role in binding
complement-coupled antigen with sufficient affinity to trigger B-cell activation.
The mechanism of B-cell antigen acquisition varies
according to the size of the antigen. Small, soluble antigens
can be directly acquired from the lymphatic circulation by
follicular B cells, without the intervention of any other cells.
These antigens enter the lymph node via the afferent lymph
and pass into the subcapsular sinus region (Figure 12-6a,b).
Some small antigens may diffuse between the subcapsular
sinus (SCS) macrophages that line the sinus to reach the
B cells in the follicles. Others leave the sinus through a
reticular (netlike) network of conduits (cylindrical vessels).
These conduits are leaky, and follicular B cells can access
antigen through gaps in the layer of follicular reticular cells
that form the walls of the conduits (Figure 12-6b).
Larger, more complex antigens take a slightly different
route. The SCS macrophages, shown just beneath the subcapsular sinus in Figure 12-6a, are a distinct subpopulation
of lymph node macrophages with limited phagocytic ability,
and express high levels of cell-surface molecules able to bind
and retain unprocessed antigen. For example, bacteria,
viruses, particulates, and other complex antigens that have
been covalently linked to complement components are held
by complement receptors on the surfaces of these macrophages. Antigen-specific B cells within the follicles can
acquire the antigens directly from the macrophages and
become activated. In addition, since SCS macrophages, B cells,
and FDCs all bear high levels of complement receptors, it is
quite possible that antigen may be passed between the three
subtypes of cells until it is finally recognized by a B cell with
the matching Ig receptor. Other classes of cells within the
lymph node can also present unprocessed antigens to B cells.
For example, a population of dendritic cells that is located
close to the high endothelial venules in lymph nodes has also
been shown to be capable of presenting unprocessed antigen
to B cells.
Any discussion of antigen presentation to B cells would
be incomplete without consideration of the role of follicular
dendritic cells (FDCs). Early electron microscope investigations of lymph-node-derived cells revealed the dendrites of
FDCs to be studded with antigen-antibody complexes called
iccosomes (Figure 12-6c). Further analysis showed that
these complexes are retained on the surface of the FDC
through interaction either with complement or with Fc
receptors. Because of the high surface density of antigen on
FDCs, many scientists at first posited that iccosomes’ primary role may be in antigen presentation to naïve B cells.
However, current evidence suggests that their main function
|
B-Cell Activation, Differentiation, and Memory Generation
(a)
Subcapsular
sinus
CHAPTER 12
391
(b)
Subcapsular
sinus pore
Afferent
lymph
Chemokine
Small antigen
(<70 kDa)
Large antigen
(>70 kDa)
SSM
Follicular
conduit
Follicular
B cell
FRC
Antigen
specific
B cell
Follicular
dendritic cell
IgM
Lymph
IgD
Collagen
Antigen
non-specific
B cell
Follicle
FIGURE 12-6 Antigen presentation to follicular B cells in
(c)
the lymph node. (a) Lymphatic fluid containing antigens (red) and
cytokines and chemokines (blue) reaches the lymph node through
the afferent lymph vessel and enters the subcapsular sinus (SCS)
region. The SCS region is lined with a porous border of SCS macrophages (SSMs) that prevent the free diffusion of the lymph fluid into
the lymph node. Larger antigens and signaling molecules are bound
by surface receptors on the SCS macrophages and then presented
directly to B cells. Smaller antigens, and chemokines less than approximately 70 kDa in molecular weight, access B cells in the follicles
either by direct diffusion through pores in the SCS, or by passage
through conduits emanating from the sinus. (b) These conduits are
formed by follicular reticular cells (FRCs) wrapped around collagen
fibers, and B cells are able to access their contents through pores in
the sides of the conduits. (See the text for further details.) (c) A follicular dendritic cell with its dendrites studded with iccosomes. Follicular
dendritic cells bind antigen-antibody complexes via complement or
Fc receptors on their cell membranes. These complexes are visible in
the electron microscope, as iccosomes. [Parts a and b adapted from
Harwood, N.E., and Batista, F.D., 2009, The antigen expressway: follicular conduits carry antigen to B cells, Immunity 30:177–189. Part (c) Courtesy Andras
K. Szakal PhD.]
is to provide a reservoir of antigen for B cells to bind as they
undergo mutation, selection, and differentiation along the
path to a memory phenotype, rather than playing the primary role in the initial antigen presentation process. In addition, FDCs secrete survival factors that ensure the survival of
B cells within the lymph node.
B-Cell Recognition of Cell-Bound Antigen
Results in Membrane Spreading
Prior to contact with antigen, the majority of BCRs are
expressed in monomeric, bivalent form on the cell surface.
Interaction of these monomeric receptors on the surface of
the B cell with multivalent, cell-bound antigens then induces
a rather spectacular cellular response. First, a cluster of BCRs
and their cognate antigens forms at the initial site of contact.
Next, the B cell rapidly spreads over the target membrane,
before contracting back.
This spreading reaction can be clearly seen in Figure 12-7,
in which the cell-bound antigen was modeled using a protein
incorporated into a planar lipid bilayer. Membrane spreading
occurs 2 to 4 minutes after antigen contact, during which
time micro-clusters of antigen-receptor interactions can be
seen by time-lapse fluorescence microscopy at the membrane-lipid interface. After reaching a maximum surface area
of approximately 25 m2, the area of contact between the cell
and the artificial lipid membrane begins to contract, and the
antigen-receptor complex is gathered into a central, defined
PA R T V
392
|
Adaptive Immunity: Effector Responses
By the end of the contraction phase, the BCR, still in
contact with its cognate antigen on the presenting cell, is
clustered on the surface of the B cell.
What Causes the Clustering of the B-Cell
Receptors Upon Antigen Binding?
FIGURE 12-7 Antigen recognition by the BCR triggers
membrane spreading. Transgenic B cells expressing a BCR specific
for hen egg lysozyme (HEL) were settled onto planar lipid bilayers containing HEL and were incubated for the time period shown, followed
by fixation and visualization with scanning electron microscopy. At 2 to
4 minutes, the B-cell membrane can clearly be seen spreading over the
surface of the planar lipid bilayer (arrows). By 5 minutes, the membrane
begins to contract again, after which the BCR molecules are found
clustered on the cell surface. (See the text for details.) [Fleire, S.J., Goldman,
J.P., Carrasco Y.R., Weber, M., Bray, D. and Batista, F.D. B cell ligand discrimination
through a spreading and contraction response. Science 5 May 2006, Vol. 312,
no. 5774, pp. 738–741. © 2006 The American Association for the Advancement of
Science]
cluster with an area of approximately 16 m2. The contraction phase takes an additional 5 to 7 minutes. The effectiveness of the ultimate antibody response has been shown to
correlate with the extent of this spreading reaction.
Experiments using monovalent ligands incorporated into
target lipid membranes demonstrated that the antigen itself
need not be multivalent in order for clustering to begin.
Rather, it appears that antigen binding to the BCR at sufficiently high affinity causes a structural alteration in the BCR,
rendering it susceptible to clustering. Initially, antigen ligation of the Ig receptor is associated with a decrease in the rate
of diffusion of IgM receptors in the plane of the membrane.
In addition, antigen binding appears to induce a conformational change in the C4 constant region domains of occupied IgM membrane receptors. These two changes result in
an increase in the tendency of the bound IgM receptors to
bind to the corresponding domains of other antigen-bound
IgM receptor molecules (Figure 12-8). The deceleration in
the receptor’s movement in the membrane coupled with this
receptor oligomerization occurs whether or not the Ig␣,Ig
(CD79␣,) signaling components of the receptor are present,
indicating that the receptor clustering does not require signaling events. It is now thought that this BCR clustering may
be the initiating event in antigen signaling through the BCR.
Upon oligomerization, the BCR complex moves transiently into parts of the membrane characterized by highly
ordered, detergent-insoluble, sphingolipid- and cholesterolrich regions, colloquially designated as lipid rafts. Recall
that association of the BCR with the lipid rafts then brings
the ITAMs of the Ig␣ and Ig components into close apposition with the raft-tethered src-family member tyrosine
Planar lipid membrane
mIgM
Monovalent
antigen
Cµ4
domain
Igα/β
Lyn
Lyn
B cell membrane
Antigen-binding B cell
FIGURE 12-8 Antigen binding induces a conformational change in the C4 domain and subsequent oligomerization of
antigen-bound IgM molecules in the plane of the membrane. When monovalent antigens floating in an artificial bilayer encounter IgM
receptors, they bind and induce a conformational change in the C4 domain of the IgM heavy chain. This conformational change facilitates
oligomerization of the IgM receptors in the plane of the membrane. (See the text for details). [Adapted from P. Tolar et al., 2009, The molecular assembly and organization of signaling active B-cell receptor oligomers, Immunological Reviews 232:34–41, Figure 2.]
B-Cell Activation, Differentiation, and Memory Generation
kinase, Lyn. This sets off the B-cell signaling cascade shown
in Figure 3-28.
Antigen Receptor Clustering Induces
Internalization and Antigen Presentation
by the B Cell
Antigen binding by the B cell activates the signaling cascade
described in Figure 3-28, and this cascade induces several
changes within the cell. One of these changes is the formation
around the BCR of clathrin-coated pits that facilitate the
internalization of the receptor-antigen complexes. Indeed, we
now know that most of the antigen-occupied BCR molecules
are internalized, leaving just a few copies of the receptor on
the cell surface to serve as signaling scaffolds. Experiments
have shown that those antigen-bound BCR-Ig␣,Ig receptor
complexes that remain at the cell surface and provide the
B-cell signaling function are more highly phosphorylated
than are those receptors that are internalized.
Once internalized, the antigen is processed as described in
detail in Chapter 8. This results in an increased expression of
peptide-loaded class II MHC molecules on the B-cell surface,
as well as an up-regulation of the expression of the costimulatory molecules CD80 and CD86. Recall that CD80 and CD86
both bind to the T-cell costimulatory molecule CD28. These
changes in the expression of CD80 and CD86, which occur
within 1 to 2 hours after antigen recognition, therefore prepare the B cells for their subsequent interactions with T cells.
A B cell that has taken up its specific antigen via receptormediated endocytosis is up to 10,000-fold more efficient in
presenting antigen to cognate T cells than is a nonantigenspecific B cell that can only acquire the same antigen by
nonspecific pinocytotic mechanisms. Effectively, this means
that the only B cells that process and present antigen to T cells
in a physiologically relevant context are those B cells with the
capacity to make antibody to that antigen.
Now that the B cell is ready to present antigen to T cells,
we must ask how a B cell, with a receptor that is normally
expressed at extremely low frequency within the receptor
repertoire, can possibly find and bind to a T cell specific for
the same antigen, which is also present at low frequency
among all the available T cells?
Activated B Cells Migrate to Find
Antigen-Specific T Cells
We owe a great deal of our understanding of the movement of
cells and antigens through the lymph nodes to recent advances
in imaging cells within their biological context. Specifically, in
some recent experiments described in Chapter 14, lymph
nodes have been brought outside the body of living, anesthetized animals, and lymphocyte circulation through these
nodes has been studied using fluorescently tagged T, B, and
antigen-presenting cells. In these experiments, the blood and
lymphatic circulation to the lymph node is preserved, and the
lymph node is gently lifted onto a warmed, humidified
|
CHAPTER 12
393
microscope stage for observation. Sometimes the fluorescent
tags are built into the genome of the animal by placing fluorescent proteins, such as green fluorescent protein (GFP),
under the control of specific promoters that are active only
in specific lymphocyte subsets. In other experiments, the
animal is injected with fluorescent antibodies just prior to
being anesthetized. This technique is known as intravital
fluorescence microscopy (see Chapters 14 and 20) and has
contributed hugely to our understanding of the process of
lymphocyte activation in the lymph nodes.
Investigators have used immunohistochemical, immunofluorescence, and the aforementioned intravital imaging
techniques to visualize particular cell populations during the
course of an ongoing immune response. This has, in some
cases, required the generation of transgenic mice in which
all, or most, B and T cells have defined antigen specificity. All
of these techniques are described in Chapter 20. What follows is a distillation of the information gained from a large
number of such experiments using various antigen and
transgenic model systems.
As we have learned, B cells pick up their antigen in the
follicular regions of the lymph node or spleen. By approximately 2 hours post antigen contact, the B cell has internalized and processed its antigen and expressed antigenic
peptides on its surface in the context of class II MHC antigens. In response to antigen recognition by the BCR, the B cell
begins to express the chemokine receptor CCR7. CCR7
binds to the chemokines CCL19 and CCL21, which are
secreted by stromal cells in the T-cell zones of secondary
lymphoid organs. The B cell still also expresses CXCR5,
which binds CXCL13, expressed by FDCs in the B-cell follicles. Because they express receptors for both T- and B-cellzone-derived chemokines, by approximately 6 hours post
stimulation, antigen-engaged B cells move to the boundary of
the B-cell and T-cell zones. Figure 12-9a shows these various
lymph node zones in diagrammatic form. Activated B cells
also up-regulate the expression of receptors for cytokines
released by activated T cells, allowing B cells to receive the
signals to proliferate, to differentiate, and also to initiate an
anti-apoptotic program. The receptor for IL-4, an important
B-cell-specific cytokine, can be detected as early as 6 hours
post antigen contact and reaches maximal levels at 72 hours.
Antigen-stimulated B cells move about within the T-cell
zone of the lymph node until they make contact with an
antigen-specific T cell, an event that most probably occurs
over the 24 to 48 hours post antigen contact. By this time,
many of the important cell-surface changes characteristic of
B lymphocyte activation can be observed, such as increased
levels of class II MHC antigens and of the two costimulatory
molecules, CD80 and CD86. Once contact with an activated
T cell is made, these antigen-stimulated B cells engage with
their conjugate T-cell partners over extended periods of time,
ranging from a few minutes to several hours. During this
period, the activated helper T cell also expresses CD40L
(CD154), a cell-membrane-bound member of the TNF receptor family, which interacts with CD40 (a TNF family member)
394
PA R T V
|
Adaptive Immunity: Effector Responses
(a)
(b)
Naïve
T-cell zone
T-B border
Germinal center
Inter-follicular zone
Sub-SCS
Follicle
Day 1
FIGURE 12-9 Movement of antigen-specific T and B cells
within the lymph node after antigen encounter. (a) Section of
the lymph node in diagrammatic form, showing the various regions of
the lymph node. (b) The locations of antigen-specific B cells (green)
and antigen-specific T cells (red) within the lymph node were visualized at specified times after antigen stimulation. The antigen used in
the experiment was nitro-phenacetyl (NP), conjugated to ovalbumin
(OVA). Fluorescently labeled NP-specific B cells and OVA-specific T cells
were transferred into a recipient animal, which was immunized with
NP-OVA in the footpads 2 days after cell transfer. Draining popliteal
lymph nodes were excised at the stated times. Anti-B220 (blue) was
used to identify the B-cell follicles. The times refer to the days (D) post
immunization. D1 and D2: Antigen-specific B-T cell pairs could be
found both at the border of the T- and B-cell zones and in the interfollicular regions. D3: T cells have begun to enter the follicle, and many
B cells can be seen just underneath the subcapsular sinus of the
lymph node. D4: B cells have taken up residence in the follicle, and the
formation of the germinal center can be seen. [Fleire, S.J., Goldman, J.P.,
Day 2
Day 3
Carrasco Y.R., Weber, M., Bray, D. and Batista, F.D. B cell ligand discrimination through
a spreading and contraction response. Science 5 May 2006, Vol. 312, no. 5774,
pp. 738–741. ©2006 The American Association for the Advancement of Science]
(see Chapter 4) on the activated B cell. The CD40-CD40L
interactions, coupled with BCR-mediated antigen recognition
and interleukin signaling, elicit all of the downstream activation pathways shown in Figure 3-28, and together drive the B
cell into the proliferative phase of the cell cycle.
After this period of intense communication between the
activated T and B cells, activated B cells then down-regulate
CCR7 while maintaining high levels of expression of CXCR5.
This allows them to leave the T-cell areas and to migrate into
the inter- and outer-follicular regions. The activated B cells
appear to spend 1 or 2 days in the regions to the outside and
in between the follicles before finally entering the follicle
around 4 days post immunization (Figure 12-9b).
In the experiment depicted in Figure 12-9b, a transgenic
mouse has been immunized with the T-dependent antigen
nitrophenacetyl-ovalbumin (NP-OVA). Antigen-specific B cells
express GFP, and are therefore labeled green; antigenspecific T cells are labeled red; and the B-cell follicles are
stained blue. In the naïve animal, T cells can be seen clearly
localized in the T-cell zones, and occasional green B cells can
be seen scattered throughout the follicles. With time post
immunization, B and T cells can be seen migrating first into
the interfollicular regions between the follicles (days 1 and 2
post immunization). By day 3, B and T cells can be found
Day 4
B-Cell Activation, Differentiation, and Memory Generation
clustered in the outer- as well as the interfollicular regions,
and by 4 days post immunization most, but not all, of the
B cells have entered the follicles and the beginning of germinal center development can be seen.
Interestingly, B cells derived from mice deficient in CXCR5
or CXCL13 behave in a fashion that is not as different from
those derived from normal mice as might have been expected,
and this observation led to the discovery of a second ligandreceptor pair that induced the same migrating behavior in
B cells. EBI2-ligand is recognized by a G-protein-coupled
receptor, EBV-induced ligand 2 receptor (EBI2), that is present
on all mature B cells. The level of expression of EBI2 increases
markedly and quickly following B-cell activation, and interaction of EBI2 with its ligand results in the movement of B cells
into the outer- and interfollicular regions in a manner analogous to that instructed by CXCR5 interaction with CXCL13.
Biochemical experiments suggest that the ligand for EBI2 is
lipid in nature, and investigators are still working to determine
the nature of the cell types that express the ligand, the dynamics of its expression, and the way in which the two sets of chemoattractant signals interact to ensure the correct movement
of the antigen-stimulated B cell through the lymph node.
The outer- and interfollicular regions of the lymph node
contain high numbers of SCS macrophages, dendritic cells,
and a cell population referred to as marginal reticular cells
(MRCs), and the roles of each of these cell populations in the
migratory behavior of antigen-activated B cells is another
area of active investigation.
Activated B Cells Move Either into the
Extrafollicular Space or into the Follicles
to Form Germinal Centers
Over the first few days after their interaction with T cells,
daughter cells of the proliferating B cell elect one of two
fates. Some of the stimulated daughter cells in the extrafollicular spaces migrate into the borders of the T-cells zone,
form a primary focus, and differentiate into plasmablasts.
Plasmablasts are B cells that can still divide and present antigen to T cells, but they have already begun to secrete antibody. Plasmablasts and plasma cells in the primary focus
secrete measurable levels of IgM by approximately 4 days
after antigen contact and are responsible for the earliest
manifestations of the antibody response.
Other daughter cells generated by stimulation of the original B cells enter the B-cell follicles, as shown in Figure 12-9b
in the bottom panel. There, they divide rapidly and undergo
further differentiation. This movement of B cells into the follicles is facilitated by the expression of the transcription factor
Bcl-6, which represses the expression of EBI-2 and enables the
B cells to leave the outer- and interfollicular regions. As the
B cells differentiate under the influence of follicular
helper T cells (TFH) (see Chapter 11), the follicle becomes
larger and more dense, developing into a germinal center.
(The germinal center reaction will be discussed below.)
The precise nature of the mechanisms responsible for
determining which cells enter the primary focus and which
|
CHAPTER 12
395
elect to enter the follicle and establish a germinal center
reaction are unclear at this time and may be stochastic.
Below, we first describe the processes that lead to primary
focus formation and then go on to discuss the extraordinary
fate that awaits those B cells that enter the follicles, and eventually develop into germinal center B cells.
Plasma Cells Form within the Primary Focus
Plasma cells are essentially Ig-producing machines. Their surface Ig levels are close to zero, and they are no longer capable
of being further stimulated by antigen, or of presenting antigen
to responsive T cells. An activated B cell in the extrafollicular
regions of the lymph node that is initiating a program of differentiation toward the plasma cell endpoint begins to divide
rapidly and to secrete low levels of Ig. As the cell divides, it
decreases levels of membrane Ig, MHC proteins, and CD80/86,
and increases its rate of Ig secretion to become a plasmablast.
Eventually, the cell reaches a stage of terminal differentiation in
which it is no longer capable of cell division and has achieved
a maximal rate of Ig secretion: the plasma cell. Plasma cells are
found within the first 5 to 6 days of an immune response in the
medullary cord region of the lymph node.
Most primary focus plasma cells have short half-lives,
dying by apoptosis within 5 to 10 days of their generation,
and for many years it was thought that all plasma cells from
the primary foci endured this fate. However, recent experiments have suggested that some of these plasma cells may
migrate to the bone marrow or to other locations within the
body, where they provide long-lasting Ig memory.
There is currently a great deal of interest in the transcription
factors that control whether antigen-stimulated B cells differentiate along the plasma cell or, alternatively, the germinal center
route. Scientists now understand that these transcription factors
are linked in a mutually regulatory network (Figure 12-10).
Pax-5
IRF-4
Bcl-6
BLIMP-1
Germinal center fate
Plasma cell fate
SHM
CSR
FIGURE 12-10 A regulatory network of transcription factors controls the germinal center B cell/plasma cell decision
point. The transcription factors that control germinal center B cell
versus plasma cell states of differentiation are related to one another
through a mutually regulatory network. (See the text for details).
396
PA R T V
|
Adaptive Immunity: Effector Responses
Transcription factors that favor the generation of proliferating germinal center B cells, Pax-5 and Bcl-6, actively
repress the expression of the plasma cell transcription factor
BLIMP-1. Indeed, bcl-6 gene knockout animals are incapable
of forming germinal centers, a factor that demonstrates the
centrality of Bcl-6 in determination of the germinal center
phenotype. Immune responses from these knockout mice
also fail to undergo affinity maturation and class switch
recombination, processes that depends on the normal functioning of germinal centers.
In wild-type animals, as illustrated in Figure 12-10,
induction of BLIMP-1 and IRF-4 results in the inhibition of
Bcl-6 and Pax-5 expression. This antigen-induced expression
of two alternative sets of transcription factors that control the
decision between two potential lymphocyte cell fates is reminiscent of the induction of T-Bet and GATA-3 and their
respective effects on TH1 and TH2 induction (Figure 11-10).
In both cases, induction of the transcription factor(s) that
initiate one differentiation pathway results in the direct inhibition of the synthesis and activity of the transcription factors supporting the alternative pathway.
The transcription factor B-lymphocyte-induced maturation protein 1 (BLIMP-1) had long been considered to be the
master regulator of plasma cell differentiation. BLIMP-1
knockout animals lack all capacity to generate plasma cells,
while enforced expression of BLIMP-1 is sufficient to promote
plasma cell differentiation. In wild-type animals, BLIMP-1
expression peaks at the plasma cell stage of B-cell differentiation, and the highest levels of BLIMP-1 expression are seen in
long-lived plasma cells in the bone marrow and primary focus
plasma cells in the secondary lymphoid organs. Expression of
BLIMP-1 also decreases the levels of MHC expression on the
surface of the cell, consistent with the inability of plasma cells
to engage in presentation of antigen to T cells. BLIMP-1 also
promotes the alternative splicing of Ig mRNA, which enables
the generation of mature transcripts encoding the secreted
form of antibodies (see Figure 7-15).
However, when scientists asked whether BLIMP-1 was
the first transcription factor operative in the cellular process
that induces B cells to differentiate into plasma cells, the
answer was not as straightforward as had been expected.
Specifically, recent experiments have demonstrated that
some genes associated with plasma-cell formation are still
turned on in blimp-1 knockout mice, suggesting that some
factor other than BLIMP-1 may initiate the process of
plasma cell differentiation, even if BLIMP-1 is required for
its completion. What other factor(s) might be responsible for
initiating the plasma-cell fate pathway?
Current evidence points to the Interferon Regulatory
Factor 4 (IRF-4) as the most probable candidate (see Figure 12-10). If IRF-4 is controlling the expression of blimp-1,
we would expect it to be expressed before BLIMP-1 in the
developing B cell, and to control the expression of the
blimp-1 gene. Such is indeed the case. The irf-4 gene is
expressed before blimp-1, and the IRF-4 protein binds to
elements upstream of the blimp-1 gene, up-regulating its
transcription. Furthermore IRF-4 down-regulates the expression of the genes encoding both Pax-5 and Bcl-6. Finally, the
concentration of IRF-4 varies in B cells at different stages of
differentiation in a manner consistent with it being a determining factor in plasma cell differentiation. Thus, it appears
that IRF-4 may initiate differentiation to the plasma cell stage.
But is it required, or merely a useful adjunct to the process of
plasma-cell differentiation? For the answer to this question,
we must turn to knockout experiments.
Experiments with IRF-4-deficient mice have shown that
they completely lack Ig-secreting plasma cells; conversely,
and as for BLIMP-1, overexpression of IRF-4 promotes
plasma cell differentiation. IRF-4 has been shown to bind to
Ig enhancer regions and is important in the generation of
high levels of Ig secretion. Thus, it would appear that IRF-4
may be a determining factor, or possibly the determining
factor, in driving a B cell toward the plasma cell, rather than
the germinal center phenotype.
The generation of the plasma cells in the primary focus is
a critically important step in the antibody response, and the
rate of secretion of Ig by these cells is extraordinarily high. At
the peak of an antibody response, a single antibody-secreting
cell can release as much as 0.3 ng of Ig per hour, with fully
30% of its cellular protein synthesis devoted to the generation
of secreted antibody molecules. Therefore, these cells provide
high levels of antibody that can bind antigen, neutralizing it
and/or opsonizing it for phagocytosis by macrophages.
However, the antigen-binding affinities of the antibodies
secreted by the primary focus have not yet been optimized
by the related processes of somatic hypermutation and antigen selection, and their affinity for the antigen is often relatively low. In the case of a particularly virulent infection, the
antibodies secreted by the cells of the primary focus may
therefore serve primarily to contain microbial numbers until
the high-affinity antibodies generated by the germinal center
reaction are released. It is within the germinal center that the
next extraordinary events in B-cell differentiation will occur.
Other Activated B Cells Move into the Follicles
and Initiate a Germinal Center Response
As described above, 4 or 5 days after T-dependent B-cell activation occurs at the border between the T-cell zone and the
follicles, some antigen-specific B cells enter the follicles. There,
they undergo rapid proliferation, resulting in the formation of
large clusters of antigen-specific B cells. The follicle, which
contains both TFH and follicular dendritic cells (FDCs),
becomes larger as the entering antigen-specific cells proliferate, and the teeming follicle in which an immune response is
actively ongoing is referred to as a germinal center (GC) (Figure 12-11). Interaction between CD40 on the B cell and its
ligand, CD40L, on the T cell is necessary for germinal center
formation, and no germinal centers are generated in mice in
which the genes for either CD40 or CD40L have been deleted.
In the germinal centers, B cells undergo a period of intense
proliferation, and then their Ig genes are subjected to some of
B-Cell Activation, Differentiation, and Memory Generation
(a)
|
CHAPTER 12
397
(b)
Follicular mantle zone
Light zone
Dark zone
Dark zone
Light zone
FIGURE 12-11 The germinal center. (a) A mouse lymph
node germinal center 7 days after secondary immunization, in
which 10% of the germinal center B cells express green fluorescent
proteins (and are therefore labeled green). These proteins are seen
most intensely in the dark zone of the germinal center, but are also
present less frequently in the light zone. Follicular dendritic cells,
labeled red with an antibody to the FDC marker CD35, mark the
light zone of the germinal center. The blue staining marks IgDthe most extraordinary processes in biology. First, the Ig variable (but not the constant) region genes undergo extremely
high rates of mutagenesis, on the order of one mutation per
thousand base pairs per generation. (Contrast this with the
background cosmic ray-induced mutation rate of one mutation per hundred million base pairs per generation.) The
mutated genes are then expressed, and the encoded Ig receptors are tested to see if their affinities for antigen have been
altered for the better. Those B cells bearing higher-affinity
receptors than those on the parental cells are selected for further rounds of mutation and selection, while those with no, or
decreased, affinity for antigen are allowed to die by apoptosis.
Second, also within the germinal center, interleukin signals from follicular T cells drive the process of Ig CSR: the
replacement of the constant region gene segments with
segments encoding other classes of constant regions. We are
just beginning to understand the molecular biology of these
processes, which take place in a rapidly modulating anatomical environment. Given the active mutational and selection events that occur within the germinal center, it should
come as no surprise that it has been referred to by some as a
“Darwinian Microcosm.”1 Below, we will describe the biology of the germinal center and then look in turn at the two
genetic processes that are unique to the immune system:
somatic hypermutation and the Ig class switch.
1
Garnett Kelsoe. (1998). V(D)J hypermutation and DNA mismatch repair:
Vexed by fixation. Proceedings of the National Academy of Sciences of the
United States of America 95:6576.
bearing B cells not specific for the immunizing antigen. (b) Histochemical staining of a germinal center, illustrating the remarkably
high centroblast cell density within the dark zone. [(a) Reprinted
from Victora, G.D. et al., 2010, Germinal center dynamics revealed by multiphoton microscopy with a photoactivatable fluorescent reporter, Cell
143:592–605. Copyright 2010 with permission from Elsevier. (b) Courtesy
Dr. Roger C. Wagner, Professor Emeritus of Biological Sciences, University of
Delaware]
Germinal Center Formation
When the activated B cells enter the B-cell follicles, they
proliferate in an environment provided by a network of
FDCs and TFH cells. TFH cells were described in detail in
Chapter 11, and both FDCs and TFHs provide survival signals to the rapidly dividing germinal center B cells.
B-cell proliferation results initially in a decrease in IgD
expression on the cell surface, although IgM expression is
maintained. As the GC expands, nonresponding IgD⫹ B cells
are displaced into the border region of the follicle, forming a
corona of naïve B cells referred to as the follicular mantle
zone. This is illustrated rather dramatically in Figure 12-11a
as a blue border to the germinal center and is also illustrated
in Figure 12-11b.
Germinal center formation requires a number of cytokines, including Lymphotoxin-␣ (TNF-), that are produced
by the FDCs, as well as by the TFH and potentially by other
cells within the lymph node. These soluble signaling molecules help to repress genes such as blimp-1, and promote the
expression of genes such as AID, which have important roles
in the developing germinal centers.
Dark- and Light-Zone Development
As the germinal center matures, two zones become visible: the
light zone and the dark zone (see Figure 12-11). The “darkness” of the dark zone results from a dense packing of rapidly
proliferating B cells (centroblasts), whereas the “lighter”
nature of the light zone results from the less dense distribution
398
PA R T V
|
Adaptive Immunity: Effector Responses
of B cells (centrocytes) within a network of FDCs; indeed,
FDCs are largely absent from the dark zone. (Note that in
Figure 12-11a, not all the B cells have been labeled, in order to
allow for clarity in the image. The green dark zone is, in fact,
densely packed with dividing centroblasts, a characteristic
that is better visualized in Figure 12-11b.)
The distribution of B cells between the two zones is
dependent on their relative levels of expression of the chemokine receptors CXCR4 and CXCR5, which interact with
chemokines preferentially expressed in each of the two
zones. CXCR4 is up-regulated on centroblasts, whereas
CXCR5 has higher expression on centrocytes.
Somatic Hypermutation and Affinity Selection
Occur within the Germinal Center
The term somatic hypermutation describes one of the most
extraordinary processes in biology. The word somatic tells us
that the mutational processes are occurring outside of the
germ line (egg and sperm) cells. Hypermutation alludes to
the fact that the mutational processes occur extremely rapidly. Somatic hypermutation (SHM) in mice and humans
occurs only following antigen contact, affects only the variable regions of the antibody heavy and light chains, and
requires the engagement of T cells, which must be able to
interact with B cells via CD40-CD40L binding.
The possibility that SHM of Ig genes might play a role in
antibody diversification was first implied by amino acid
sequencing studies conducted in the 1970s by Cesari, Weigert, Cohn, and others. Their studies focused on antibodies
produced by mouse myeloma tumors expressing light
chains. The mouse locus has been severely truncated, and
as a result, mice have very few different chain variable
regions. These investigators were therefore able to compare
each of their myeloma light-chain sequences to a known
germ-line sequence. They showed that the myeloma tumors
expressed point mutations that were restricted to the variable
regions of the chains and clustered in the complementarity
determining regions. With the advent of nucleic acid sequencing technology, these data were confirmed and extended by
others who showed that SHM affects the variable regions, but
not the constant regions, of both the heavy and the light
chains of Igs; that the frequency of somatic hypermutations
increases with time post immunization and further; and that
somatic hypermutation, followed by antigen selection, results
in the secretion of antibodies whose affinity for the immunizing antigen increases with time post immunization.
These experiments defined the process of SHM, but left
open the question of where it occurred. A series of papers
published in the early 1990s, and representing a technical
tour de force, proved that somatic hypermutation occurs
within the germinal centers of an active lymph node (see
Classic Experiment Box 12-1).
Once investigators knew where mutation was occurring,
their next question, inevitably, was how? This question presented huge experimental challenges, and for almost 30 years
investigators made only incremental progress in defining the
mutational mechanism. Certain nucleotide sequences were
defined that appeared to be particularly susceptible to mutation, and immunogeneticists were able to demonstrate that
the mutational process appeared to be remarkably focused on
the Ig variable regions. Other studies showed that SHM
required the involvement of T cells, that it occurred only after
antigen contact, and that mutation appeared to affect only
genes that were being rapidly transcribed. However, the
nature of the enzymes that mediate the response, and the
molecular mechanisms involved, remained obscure.
The breakthrough came, as so often happens, from an
unexpected direction. In experiments designed to detect the
genes and enzymes that mediate CSR, Takasu Honjo and
colleagues isolated a cell line in which the antibody genes did
not undergo CSR. This type of finding is often the crucial
first step in identifying those proteins that are implicated in
a particular process, as scientists could now compare those
cells that could and could not undertake CSR and look for
differences. However, importantly, they then noticed that the
antibody genes in these cell lines were also devoid of somatic
mutations in the variable regions. Could it be that the same
enzyme was implicated in both SHM and CSR?
Members of Honjo’s laboratory quickly isolated the gene that
had been mutated in this cell line and proved that it encoded the
enzyme Activation-Induced Cytidine Deaminase (AID),
which was subsequently shown to be necessary for both
somatic hypermutation and CSR. But now the scientists were
faced with the new questions: How does AID induce hypermutation, and how is it restricted only to antibody variable regions?
AID-Mediated Somatic Hypermutation
Although all the details of the SHM mechanism are not yet
completely understood, several parts of the process are now
well established. AID-induced loss of the amino group from
cytidine residues located in mutation hot spots results in the
formation of uridine, as shown in Figure 12-12. This creates
a U-G mismatch. Several alternative mechanisms then come
into play that participate in the resolution of the original
mismatch, leading to the creation of shorter or longer
stretches of mutated DNA (Figure 12-13).
NH2
O
N
O
HN
N
O
HO
OH
H
Deoxycytidine
O
Activation induced
cytidine deaminase
N
O
HO
OH
H
Deoxyuridine
FIGURE 12-12 Activation-induced Cytidine deaminase
(AID) mediates the deamination of cytidine and the formation of uridine.
B-Cell Activation, Differentiation, and Memory Generation
|
CHAPTER 12
399
BOX 12-1
CLASSIC EXPERIMENT
Experimental Proof That Somatic Hypermutation and AntigenInduced Selection Occurred within the Germinal Centers
In a series of papers written in the
early 1990s, the laboratory of Garnett Kelsoe conclusively demonstrated that the
phenomena of somatic hypermutation and
antigen-induced selection occurred within
the germinal center regions of the secondary lymphoid organs. The researchers
made use of the fact that the immune
response to the T-dependent haptenprotein antigen—4-hydroxy-3-nitrophenylacetyl (NP)-hemocyanin—generates an
immune response that uses a predictable
combination of heavy- and light-chain
variable regions. This enabled them to
study the changes in antibody genes as
the immune response progressed, whereas
a similar study of a response that used, for
example, a hundred or more different
B-cell clones would have been technically
impossible.
The researchers generated serial tissue
sections from lymph nodes at different
time points during an immune response
to the T-dependent antigen. The tissue
samples were derived from the primary
foci as well as from the germinal centers
of antigen-responding lymph nodes. By
scraping the primary foci or germinal
center cells off the slides using microdissection tools, isolating the mRNA, and
then subjecting the mRNA to RT polymerase chain reaction (PCR) and nucleic
acid sequencing, mRNA from 100–300
B cells per sample were obtained. The
results can be summarized as follows:
• Each focus and germinal center is initiated by seeding with one to six antigenspecific B cells. Statistical analysis of the
immunoglobulin variable region
sequences obtained from each sam-
ple showed that each focus or germinal center contained few B cells at the
onset of the response. With time, the
number of distinct clones of B cells per
focus or germinal center decreased.
• Somatic hypermutation is restricted to
the germinal center B cells. Nucleic acid
sequencing of immunoglobulin variable regions obtained from B cells in
both primary foci and germinal centers demonstrated that foci-derived
sequences did not show any evidence
of mutation. In contrast, mutations
could be detected in germinal centerderived, V-region genes and accumulated as the response progressed.
• With time post immunization, mutations
in the complementarity-determining
regions are selected for; mutations in
framework regions are selected against.
Somatic hypermutation is an antigendriven process, unlike the other mechanisms for the generation of antibody
diversity. As discussed in Chapter 7,
the variable region of an immunoglobulin gene can be divided into
framework and complementaritydetermining regions. Jacobs and colleagues analyzed what fraction of the
mutations they observed were present
in framework versus complementaritydetermining regions. Those mutations
affecting framework residues might be
expected to be either neutral with
respect to antigen binding or, if they
affected the folding of the antibody
variable region, to be deleterious. In
contrast, those in the complementaritydetermining regions of the gene,
which encode the antigen-binding
Briefly, the simplest mechanism is the interpretation of the
deoxyuridine as a deoxythymidine by the DNA replication
apparatus. In this case, one of the daughter cells would have an
A-T pair instead of the original G-C pair found in the parent
cell (Figure 12-13, left). Alternatively, the mismatched uridine
could be excised by a DNA uridine glycosylase enzyme. Errorprone polymerases would then fill the gap as part of the cell’s
region of the antibody molecule, have
the potential to be advantageous.
Jacob and Kelsoe asked whether
selective mechanisms were at play
that increased the frequency of B cells
with enhanced antigen-binding capacity. Their results showed that, whereas
at 8 days post immunization, mutations
were randomly distributed throughout
the Ig variable region gene, by the end
of the primary immune response, a
greater fraction of the mutations was
located in the complementaritydetermining regions, despite the fact
that these regions comprised only 21%
of the V region sequence.
• Affinity measurements of antibodies
generated in vitro from PCR-amplified
samples derived from foci and germinal
center cells showed that mutated and
selected antibodies displayed increased
affinity for antigen.
Thus, not only did Kelsoe’s lab demonstrate that the germinal center was the
site of somatic hypermutation; their
experiments also showed that mutations
in the complementarity-determining
regions that led to increased affinity for
antigen were positively selected.
Jacob, J., C. Miller, and G. Kelsoe. 1992. In situ
studies of the antigen-driven somatic
hypermutation of immunoglobulin genes.
Immunology and Cell Biology 70:145–152.
Jacob, J., and G. Kelsoe. 1992. In situ studies of
the primary immune response to
(4-hydroxy-3-nitrophenyl)acetyl. II. A common clonal origin for periarteriolar lymphoid sheath-associated foci and germinal
centers. Journal of Experimental Medicine
176:679–687.
short-patch base excision repair mechanism (Figure 12-13,
middle). Third, mismatch repair (MMR) mechanisms could
be invoked that result in the excision of a longer stretch of
DNA surrounding the mismatch. The excised strand could
then be repaired by error-prone DNA polymerases, such as
DNA polymerase , leading to a lengthier series of mutations
in the region of the original mismatch (Figure 12-13, right).
PA R T V
400
|
Adaptive Immunity: Effector Responses
C
G
AID
U
G
Replication
Uridine
glycosylase
U
G
MMR
G
Short-patch
base excision
repair
Error prone
polymerases
A
C
T
T
A
G
G
FIGURE 12-13 The generation of somatic cell mutations
in Ig genes by AID. AID deaminates a cytidine residue, creating a
uridine-guanosine (U-G) mismatch. Resolution of this mismatch may
be facilitated by any one of several pathways that may compete with
one another. On the left, the deoxyuridine is interpreted by the DNA
replication machinery as if it were a deoxythymidine, resulting in the
creation of an A-T pair in place of the original G-C pair in one of the
daughter cells. In the center example, the mismatched uridine has
been excised, most probably by one of the uridine DNA glycosylase
enzymes, leaving an abasic site. The uridine can then be replaced by
any of the four bases, in a reaction known as short-patch base excision repair, which can be catalyzed by one of a number of error-prone
polymerases. Finally, as shown on the right, mismatch repair enzymes
can detect the mismatch and excise a longer stretch of the DNA surrounding the U-G couple. Error-prone polymerases are then recruited
to the hypermutable site by proliferating cell nuclear antigen (PCNA),
and these polymerases can introduce a number of mutations around
the original mismatch. Thus, depending on the repair mechanism, a
mutation may occur only in the originally altered base or in one or
more bases surrounding it. [Adapted from J. U. Peled et al., 2008, The Biochemistry of Somatic Hypermutation. Annual Reviews Immunology 26:481–511.]
Mutational Apparatus Targeting
Since the rate of hypermutation is orders of magnitude
higher in Ig variable region DNA than in other genes in germinal center B cells, some mechanism must exist to direct
the mutational machinery to the correct chromosomal location. In addition, a variation on this mechanism should also
direct AID to those parts of the constant region genes that
are recognized during CSR.
When DNA is being actively transcribed, localized
regions of DNA become transiently detached from their
partner strands. The presence of single-stranded DNA
appears to be necessary for SHM to occur, and the number
of mutations that accumulate in Ig DNA is roughly proportional to the rate of Ig transcription. However, genes other
than Igs are transcribed in germinal center B cells, as are the
constant regions of the immunoglobulin genes, and they do
not undergo mutation. Clearly this distinction alone does
not account for the fact that mutations concentrate almost
exclusively in the variable regions of Ig heavy and light
chains, and something other than active transcription must
be targeting AID to the variable regions.
Careful analysis of Ig variable region sequences revealed
that some sequence motifs were more likely than others to
be targeted by the mutational apparatus, and these are
referred to as mutational hot spots. In particular, it was
noted that the sequence motif DGYW/WRCH was frequently targeted for mutation at the underlined G-C pair. (In
this description of the sequence, the four nucleotides after
the backslash represent the inverted complement of the
DGYW sequence, such that the underlined G and C are
paired with one another.) The code used to describe the targeted hot spot sequence is as follows:
D ⫽ A/G/T
Y ⫽ C/T
R ⫽ A/G
W ⫽ A/T
H ⫽ T/C/A
The DGYW motif is also found frequently in the class
switch regions, and so it appears to be an important sequence
that directs AID binding to certain parts of the DNA.
Antigen-Induced Selection of B Cells
with Higher Affinity
In this section, we will describe how those B cells with
higher-affinity antigen receptors successfully compete with
their lower-affinity counterparts for survival and proliferation signals delivered to them by TFH cells in the germinal
center.
Seminal experiments using intravital fluorescence microscopy in transgenic mouse models have enabled investigators
to address the question of how higher-affinity B cells are
allowed to survive, while their lower-affinity precursors and
cousins succumb to apoptosis. In these mice, antigen-specific
T and B cells as well as FDCs were labeled with different fluorescent dyes, and the movement of B and T cells during an
ongoing germinal center reaction was observed with timelapse fluorescence microscopy. Investigators observed that
B cells spend a relatively short amount of time in the light zones
of the germinal centers, where they must take the opportunity
to come into contact with TFH cells, in order to receive T-cellderived growth and survival signals. They further noticed that
there are many more B cells than TFH cells in the light zones,
which indicated that B cells may have to compete with one
another for the privilege of interacting with T cells.
A well-accepted current model suggests that B cells bearing higher-affinity receptors capture and process antigen
B-Cell Activation, Differentiation, and Memory Generation
more effectively than do competing B cells with lower-affinity
receptors. The more antigen that a B cell processes, the more
antigen it will present on its class II MHC molecules for
recognition by cognate TFH cells. Therefore, a B cell that
has undergone an advantageous mutation in the antigenbinding region of its Ig genes will be better able to interact
with TFH cells than will its competitors, and therefore it will
receive more proliferative and survival signals from those
TFH cells.
Additional experiments have indicated that B-cell competition for antigen within the germinal center may be more
direct than previously thought, as some B cells were observed
actually stripping antigen from other B cells. This suggests
that higher-affinity B cells actually steal antigen from their
lower-affinity counterparts. In this way B cells with higheraffinity receptors present more antigen to T cells, and enjoy
better interactions with them that lead directly to enhanced
survival. In the absence of positive survival signals, loweraffinity B cells in the germinal center undergo apoptosis.
Genetic dissection of this process has shown that signals
delivered from the T cell through CD40 on the B-cell surface
provide an indispensable component of the signal that T cells
deliver to successfully competing B cells. In addition, BCRinduced PI3 kinase activation (see Chapter 3) in germinal
center B cells results in the activation of the serine/threonine
kinase Akt. Akt is a pleiotropic (has many effects) kinase,
which not only promotes cell survival and inhibits apoptotic
proteins, but also promotes the degradation of p53, thus
allowing the GC B cells to cycle.
Thus, those B cells that bind, process, and present more
antigen to T cells will win out over those B cells that cannot
express antigen as effectively. Low-affinity B cells will lose
out in the competition for antigen, and will die because of
the absence of T-cell-mediated survival signals.
Since mutation is a random process, some B cells may
acquire self-reactive receptors. One proposed mechanism
for the destruction of such mutation-generated, self-specific
B cells relies on the fact that self molecules, such as serum
proteins, will be expressed at extremely high concentrations in the lymph nodes, whereas foreign antigens will be
expressed at much lower levels. One might therefore expect
that all the BCRs on a self-reactive B cell will be occupied.
Full receptor occupancy leads to rapid internalization of
the vast majority of the BCRs on the cell surface. In follicular B cells, such loss of cell surface BCR expression would
result, in turn, in the loss of signaling through the BCR,
exit from the cell cycle, and the induction of apoptosis. In
contrast, on B cells specific for foreign antigens, only a
relatively small proportion of their receptors will be occupied by antigen and therefore, on these cells, enough BCRs
will remain on the cell surface to provide the scaffold for
the signaling cascade.
Inevitably, some self-specific B cells will escape into the
periphery. Recall, however, that these B cells will lack cognate T-cell help and, in the absence of stimulation, will simply be lost by neglect.
|
CHAPTER 12
401
As described above, both SHM and CSR depend on the
activity of the AID enzyme. However, analysis of mutated
forms of AID has demonstrated that different parts of the
same molecule catalyze the two different processes. Furthermore, mutations in different parts of the molecule lead to
different immunodeficiency states.
Specifically, investigators analyzed the structure of AID
genes isolated from a series of patients with Hyper IgM syndrome. In this immunodeficiency disease, patients generate
only IgM antibodies that fail to undergo either SHM or CSR.
Such patients suffer from recurrent, severe infections, thus
emphasizing the physiological importance of both the mutational and the class switching processes to the fully functioning immune system. However, some unusual patients were
found to have AID genes with premature stop codons near
the 3⬘ end. These patients generated antibodies that could
undergo SHM, but were severely compromised in their ability to perform CSR. Such individuals displayed only mild
symptoms, and their immunodeficiency was often not diagnosed until adulthood. This implies that different sections of
the AID protein are necessary for SHM and for CSR and
further suggests that the ability to generate high-affinity
antibodies may be more functionally relevant than the
capacity to synthesize antibodies of different heavy-chain
classes.
Class Switch Recombination Occurs Within the
Germinal Center after Antigen Contact
In Chapter 7, we noted that naïve B cells could simultaneously express both mIgM and mIgD: both proteins are
encoded on the same long transcript, and the decision to
translate (IgM) versus ␦ (IgD) heavy chains is made at the
level of RNA splicing. In contrast, the decision to switch
from the expression of IgM to expression of any of the other
classes of antibodies is made at the level of DNA recombination, and the process by which it occurs is referred to as class
switch recombination (CSR). The switch to the expression of
any heavy-chain class other than ␦ results in the irreversible
loss of the intervening DNA.
The Ig heavy-chain locus is approximately 200 kb in
length. The formation of ␥, , and ␣ heavy-chain genes
requires cutting and rejoining of the heavy-chain DNA (Figure 12-14) in such a way that the desired constant region lies
directly downstream from the rearranged VDJ region. Class
switching occurs by the induction of recombination between
donor and acceptor switch (S) regions located 2 kb to 3 kb
upstream from each CH region (except for C␦). The donor S
region is the S region upstream from the antibody heavychain constant region gene expressed prior to the class
switch (which is normally , except for those instances in
which a B cell undergoes more than one class switch). The
acceptor S region is the S region upstream of the antibody
heavy-chain constant region that the B cell will express next.
Switch regions consist of tandem repeats of short, G-rich
sequences, 20 bp to 80 bp in length, that differ slightly for
PA R T V
402
|
Adaptive Immunity: Effector Responses
Heavy chain genes in IgM-expressing cells
δ mRNA
Germline
transcript
µ mRNA
AID
5'
V DJ
AID
Cµ
Cδ
C γ3
3'
C γ1
S
AID
Makes double-strand
breaks in S regions
C γ 2b
S
S
C γ 2a
Cα
Cε
S
S
AID
C γ1
C γ3
Class switch
recombination
(requires DSBs)
C γ 2b
Cδ
C γ 2a
Cµ
+
Heavy chain genes
in IgE-expressing cell
5'
V DJ
Cα
Cε
3'
S
FIGURE 12-14 Class switch recombination from a C to a C heavy-chain constant region gene. The activation induced cytidine deaminase (AID) enzyme initiates CSR by deaminating cytidine residues within the switch (S) regions upstream of C and C on both
strands. This leads to the formation of double-strand breaks within both S regions that are then resolved by DNA repair mechanisms, with the
loss of the intervening DNA sequence, as described in the text. [Adapted from J. Stavnezer et al., 2008, Mechanism and Regulation of Class Switch Recombination. Annual Review of Immunology 26:261–292.].
each isotype and contain targeting sites for AID. In CSR,
genetic analysis has indicated that the critical sequence motif
required for AID binding is a pair of WGCW overlapping
motifs on the top and bottom strand, where W represents
either adenine or thymine. (Note that WGCW merely represents a subset of the group of sequences described above as
DGYW.) The overall length of the switch regions vary from
1 kb to 10 kb, and CSR can occur anywhere within or near
the S regions.
Signals for Class Switch Recombination
B cells must receive costimulatory signals from CD40 or,
occasionally, B-cell Toll-like receptors in order to engage in
CSR. The importance of CD40-CD40L interactions in the
mediation of CSR is illustrated in patients suffering from
X-linked Hyper-IgM syndrome, an immunodeficiency disorder in which TH cells fail to express CD40L. Patients suffering from this disorder express IgM, but no other isotypes.
Such patients also fail to form germinal centers, they fail to
generate memory cell populations, and their antibodies do
not show evidence of SHM.
The cytokine signal received by the B cell determines
which class of Ig it will make (see Table 12-1). These cytokines signal the B cells to induce transcription from germline promoters located upstream (5⬘) of the respective donor
and acceptor switch regions. The resulting germ-line transcripts do not encode proteins and are therefore referred to
as sterile RNAs. Importantly, no CSR can occur in the
absence of this transcriptional activity, which is probably
important in the creation of localized regions of singlestranded DNA recognized by AID. The germ-line promoters
express the appropriate cytokine-responsive elements. For
example, germ-line ␥1 and promoters, which are induced
by IL-4, have binding sites for the IL-4-induced transcriptional activator Stat6. But how does a B cell that receives a
signal from IL-4 know whether to switch to IgG1 or IgE? The
answer to this, and similar questions, is not yet clear, and
there is still much to learn regarding the details of the signals
that differentially regulate switching to particular classes of
antibodies.
The Molecular Mechanism of Class Switch
Recombination
CSR occurs by an end-joining mechanism and, like SHM,
the process is initiated by AID. AID deaminates several cytosines within both the donor and acceptor S sites that have
been previously activated as a result of cytokine signaling.
DNA uridine glycosylase enzymes remove the U, created by
the deamination of cytidine, and then apurinic/apyrimidinic
endonucleases nick the DNA backbone at the abasic sites,
B-Cell Activation, Differentiation, and Memory Generation
TABLE 12-1
Specific cytokines signal B cells
to undergo CSR to different
heavy-chain classes
Cytokine signal
Isotype synthesized by target B cell
IL-4
IgG1, IgE
TGF-
IgA, IgG2b
IL-5
IgA
IFN-␥
IgG3, IgG2a
|
CHAPTER 12
403
• What roles do residual antigen, follicular dendritic cells,
and T-cell signaling play in the survival of B cells after
the primary response is over?
• Since some cells will survive this primary response as
either memory B cells or long-lived plasma cells, how
are B cells within a single clone selected for survival versus cell death? Is the decision made earlier or later during the course of the response, and is it random or
directed in some as yet unknown way?
Some Germinal Center Cells Complete Their
Maturation as Plasma Cells
creating single-strand breaks at multiple points in the donor
and acceptor S sites. Mismatch DNA repair enzymes then
convert the single-strand DNA breaks into double-strand
breaks. In the final step of the process, the cell’s doublestrand break repair machinery steps in and ligates the two
switch regions, resulting in the excision of the intervening
sequence. The process of CSR can occur more than once
during the lifetime of the cell. For example, an initial CSR
event can switch the cell from making IgM to synthesizing
IgG1, and a second CSR can switch it to making IgE or IgA.
Most Newly Generated B Cells Are Lost at the
End of the Primary Immune Response
Between 14 and 18 days after its initiation, the primary
immune response winds down, and the immune system is
faced with a problem of excess. Rapid proliferation of antigenspecific B cells over the course of the immune response leads
to the generation of expanded clones of cells, and if all the
cells from each clone were allowed to survive at the close of
every immune response, there would soon be no room for
new B cells emerging from the bone marrow and seeking to
circulate through the lymphoid follicles and receive their
survival signals.
Although we know that most B cells are lost by apoptosis
at the end of the immune response, the exact mechanism by
which this occurs has not yet been fully characterized. As
antigen levels wane, the balance between survival signals
and death signals experienced by the B cell in the lymph
node may be tipped in favor of apoptosis. Certainly, the B cell
no longer receives survival signals via antigen binding to
the BCR. In addition, animals deficient in the genes encoding Fas have excess numbers of B cells, implicating the
Fas-FasL interaction in the control of B-cell numbers. But
we do not yet know the whole story. This is an area of active
research, and a number of important questions remain to be
answered:
• Does the cessation of B-cell signaling simply lead to a
decrease in anti-apoptotic proteins, such that the cell is
no longer protected from these death-inducing signals?
Or is a more active switch engaged that leads to the cell’s
demise?
At some point in the ongoing immune response, approximately 5 to 15 days after antigen encounter, a fraction of
germinal center B cells will begin to up-regulate IRF-4
expression, heralding the beginning of their differentiation
into antibody-secreting plasma cells. As described earlier,
IRF-4 expression then induces the generation of the transcriptional repressor, BLIMP-1, which down-regulates those
genes important to B-cell proliferation, CSR, and SHM, and
up-regulates the rate of the synthesis and secretion of Ig
genes. As the germinal center B cell differentiates into a fully
mature plasma cell, it reduces the level of expression of the
chemokine receptor CXCR5, which has been responsible for
retaining it within the germinal center. Instead the nascent
plasma cell begins to express CXCR4, which enables it to
leave the lymph node and circulate within the peripheral tissues. These germinal center-derived plasma cells differ from
those generated from the primary focus in that their Ig genes
have undergone both SHM and CSR, and hence the antibodies they secrete will be of high affinity and may be class
switched.
For many years, it was thought that plasma cells localized
primarily in the medullary cords of the lymph nodes (the
inner parts of the kidney-shaped lymph nodes), or the red
pulp in the spleen, and that they were relatively short-lived.
However in the past decade or so, we have come to understand that plasma cells can home to several other locations
and that the 10% to 20% of plasma cells that home to the
bone marrow can be very long-lived. Indeed, smallpox-specific
serum antibodies have been identified 75 or more years after
immunization with smallpox vaccine, suggesting that the
plasma cells secreting these antibodies may persist for the
lifetime of the host. We now know that these long-lived
plasma cells derive from both the plasma cells of the primary
focus as well as from the B cells that have passed through the
germinal center and have undergone CSR and SHM.
Within the bone marrow, the niches occupied by fully
differentiated plasma cells differ from those inhabited by
developing B cells (Figure 12-15). Experiments using in vitro
culture techniques to determine the survival requirements
for long-lived bone marrow plasma cells have highlighted
the need for CXCL12 (recognized by CXCR4 on the plasma
cell) and the TNF family cytokine member APRIL (recognized
404
PA R T V
|
Adaptive Immunity: Effector Responses
Periphery
Germinal
center
Plasmablast
High-affinity
antibodies
B cell
CXCR4
APRIL
Plasma cell
(CXCR4+)
Eosinophil
(CXCR4+)
BCMA
Megakaryocyte
Stromal cells
(CXCL12+)
FIGURE 12-15 The bone marrow niche occupied by
plasma cells is supported by eosinophils and megakaryocytes, as well as by mesenchymal stromal cells. Plasmablasts
and plasma cells that have passed through the germinal center
reaction and entered the circulation take up residence in the bone
marrow, where they seek niches adjacent to eosinophils and
by the plasma cell APRIL receptor, BCMA). CXCL12 is produced in the bone marrow by mesenchymal stromal cells,
and APRIL by both eosinophils and megakaryocytes. Characterization of the bone marrow plasma cell niche continues
to be a rapidly advancing area of research.
Other plasma cells, generated in the lymphoid tissues of
the gut, remain associated with the mucosal tissues and
secrete large amounts of IgA. IgA-secreting plasma cells are
generated both in the Peyer’s patches, areas of lymphoid concentration within the gut tissues, or in isolated lymphoid
follicles in the lamina propria of the gut (see Figures 2-11 and
2-12). These gut-associated, IgA-secreting plasma cells have
significant functional differences from the IgG-secreting
plasma cells in other lymphoid tissues, and share some
important characteristics with cells of the granulocytemonocyte lineages. In particular, they produce the antimicrobial mediators TNF-␣ and inducible nitric oxide synthase
(iNOS), molecules normally associated with monocyte and
granulocyte activation. In order to continue producing these
Bone marrow
megakaryocytes, as well as to the traditional mesenchymally derived
stromal cells. The eosinophils and megakaryocytes provide the longlived plasma cell with the survival factor APRIL, a TNF family member
recognized by the receptor BCMA, whereas the stromal cells release
CXCL12, recognized by the receptor CXCR4 on plasma cells.
mediators, gut IgA-producing plasma cells must remain in
contact with gut stromal cells and be subject to microbial
costimulation. Deletion of TNF-␣ and iNOS in B-lineage
cells resulted in poor clearance of gut pathogens and has been
associated with a concomitant reduction in IgA synthesis.
B-Cell Memory Provides a Rapid and Strong
Response to Secondary Infection
The first recorded concept of immunological memory
appears in the writings of Thucydides, around 430 bce.
Describing the plague in Athens he wrote, “Only those who
had recovered from the plague could nurse the sick, because
they would not contract the disease.” Implicit in this statement was the fact that those who had suffered and recovered
had immunity to the plague, and were therefore able to
mount a stronger and faster response to future infection.
Figure 12-16 illustrates the classic conception of a memory immune response. The index of immune responsiveness
B-Cell Activation, Differentiation, and Memory Generation
Antibody concentration in serum,
units per ml
100
Total
10
Primary
response
Secondary
response
1.0
Total
IgG
IgM
0.1
IgG
IgM
lag
0.01
1° Ag
2° Ag
Time after immunization
FIGURE 12-16 Concentration and isotype of serum antibody following primary and secondary immunization with
antigen. The antibody concentrations are plotted on a logarithmic
scale. The time units are not specified because the kinetics differ
somewhat with type of antigen, administration route, presence or
absence of adjuvant, and the species or strain of animal.
in this figure is the production of serum antibody. The primary response is characterized by a lag period, which
reflects the time required for the division and differentiation
of B cells within the primary foci and their movement into
the germinal center. The B cells of the primary foci then
release IgM, and after a short delay B cells that have migrated
into the germinal center join the response with the concomitant release of both IgM and IgG. As the primary
response draws to a close, somatically hypermutated receptors make their first appearance, and then a selected subpopulation of B cells leaves the germinal centers and enters
the memory B-cell compartment.
The secondary response to the antigen is both faster and
stronger than the first. Antigen-specific B-cell division has
already occurred, and an expanded set of memory B cells bearing high-affinity receptors is available for immediate differentiation to high-affinity IgG secretion. Interestingly, somatically
TABLE 12-2
|
CHAPTER 12
405
hypermutated B cells can undergo further hypermutation with
additional antigen exposure, such that the average affinity of
antigen-specific antibodies increases through the third, and
even a fourth, immunization with the same antigen.
The actual kinetics of both primary and secondary B-cell
responses varies with the nature, dose, and route of administration of the antigen. For example, immunization of mice
with an antigen such as sheep red blood cells typically results
in a lag phase of 4 to 5 days before antibody is reliably
detected in serum, and peak serum antibody levels are
attained by 7 to 10 days. In contrast, the lag phase for soluble
protein antigens is a little longer, often lasting about a week,
and peak serum titers do not occur until around 14 days.
What properties distinguish memory from naïve B cells
(Table 12-2)? Memory and naïve B cells share the characteristics of being nonsecreting cells that need further antigen
stimulation prior to expansion and differentiation to plasma
cells. However, memory B cells differ from their naïve counterparts in terms of their sensitivity to antigen activation;
memory B cells are able to respond to lower concentrations
of antigen than primary naïve B cells can, as a consequence
of having generated higher-affinity Ig receptors during the
process of SHM. They also respond more quickly to antigen
activation, and therefore there is a shorter lag between antigen contact and antibody synthesis. Memory B cells do not
express IgD and usually carry Ig receptors that are class
switched (although some IgM-bearing memory cells may
persist). Memory B cells have longer half-lives than primary
B cells and recirculate among the lymphoid tissues.
Despite the fact that scientists have known about the existence of these cells for decades, surprisingly little is known
about the factors that control their generation and maintenance. Questions regarding memory B cells that still await
resolution include the following:
• At what stage in B-cell differentiation is the B-cell fate
determined, and what are the signals that control that
fate? It seems clear that at some point in the maturation
process of a B cell, individual daughter cells of the same
stimulated clone of primary B cells will be selected for
memory generation (life) or death. What mechanism
Functional differences between primary and secondary B cells
Naïve B cell
Memory B cell
Lag period after antigen
administration
4–7 days (depends on antigen)
1–3 days (depends on antigen)
Time of peak response
7–10 days
3–5 days
Magnitude of peak antibody
response
Varies, depending on antigen
Generally 10 to 1000 times higher than primary
response
Antibody isotype produced
IgM predominates in early primary response
IgG predominates (IgA in the mucosal tissues)
Antigens
Thymus independent and thymus dependent
Primarily thymus dependent
Antibody affinity
Low
High
Life span of cells
Short-lived (days to weeks)
Long-lived, up to life span of animal host
Recirculation
Yes
Yes
406
PA R T V
|
Adaptive Immunity: Effector Responses
allows for the differential survival of members of the
same clone of B cells? Exciting, ongoing experiments
that study asymmetric cell division in antigen-stimulated
B cells are beginning to unravel the threads of this
particular immunological knot.
• Where, in the body, do memory cells bearing differentially class-switched receptors reside? In mice, some
memory B cells localize to the marginal zones of the
spleen, and others have been located in the follicles; in
humans, memory cells appear to localize to the marginal
zones and to sub- and intraepithelial surfaces. Interestingly, during an ongoing infection or in an individual
suffering from an autoimmune disease, ectopic reservoirs
of organized lymphoid tissues can develop in multiple
organs not usually associated with immune responsiveness, including the kidneys, pancreas, lungs, thyroid, and
so on. The contributions of such ectopic tissues, as well
as of organized lymphoid organs within the bone marrow, gut, and other mucosal tissues, to memory responses
have yet to be investigated.
• How does the balance of pro- and anti-apoptotic molecules differ between primary and memory B cells, and
are particular cytokine signals required for persistent
memory B-cell survival?
• Are there differences in the efficiency of signal transduction in the two kinds of cells that may help to explain
the variations in their lag times to antibody synthesis?
T-Independent B-Cell Responses
Not all antibody-producing responses require the participation of T cells, and certain subsets of B cells have evolved
mechanisms so as to respond with antibody production to
particular classes of antigens, without T-cell help (see Figure 12-3). Antigens capable of eliciting T-independent
antibody responses tend to have polyvalent, repeating determinants that are shared among many microbial species.
Many of these antigens are recognized by B cells of the B-1
as well as the marginal zone subsets (discussed below). B-1
B cells secrete mainly IgM antibodies that are not subject to
SHM. Because of the shared nature of their antigens and
their oligoclonal antibody response, B-1 B cells may be considered to belong to an “innate-like” category of lymphocytes. Our understanding of the physiology of MZ is still
developing, but it is clear that B cells of this unusual subset
of cells may receive help from cell types other than T cells, as
is discussed below (see Advances Box 12-2).
T-Independent Antigens Stimulate Antibody
Production without the Need for T-Cell Help
The nude (nu/nu) mouse is one of the most bizarre accidents
of nature. Devoid of body hair, its ears seem oversized, and it
appears absurdly vulnerable (Figure 12-17). These mice have
FIGURE 12-17 The nude mouse. Nude mice have a mutation
in the Foxn1 gene that results in hair loss (alopecia) and interferes
with thymic development, such that they possess very few T cells and
only a thymic rudiment. Nude mice provided useful early models for
the exploration of a T-cell-depleted immune response. [David A.
Nortcott/Corbis]
a mutation in the gene for the transcription factor Foxn1 that,
in addition to affecting hair growth, also results in athymia:
mice and humans with this mutation have only thymic rudiments and possess few mature recirculating T cells.
Using nude mice as well as mice whose thymuses were
surgically removed early in life (referred to as neonatally
thymectomized), immunologists showed that most protein
antigens fail to elicit an antibody response in such animals
even while the response to many carbohydrate antigens was
unaffected. Those antigens capable of generating antibody
responses in athymic mice were referred to as T-independent
antigens because they did not require T-cell help to generate
an antibody response. Since T-cell interactions are required
for the induction of AID, TI antigens elicit predominantly
low-affinity antibody responses from B cells expressing only
the IgM isotype. We now know that T-independent antigens
fall into two subclasses (Table 12-3).
TI-1 Antigens
The first class of T-independent antigens (TI-1 antigens) is
exemplified by the bacterial lipopolysaccharide (LPS). TI-1
antigens bind to innate immune receptors on the surface of
all B cells (including the majority B-2 B-cell population), and
are capable, at high-antigen doses, of being mitogenic for all
B cells bearing the responding innate receptors. Since B-cell
stimulation in this instance is occurring through the innate
receptor (TLR4), only a small minority of the antibodies
produced will be able to bind directly to the TI-1 antigen.
The huge in vivo polyclonal TI-1 responses generated in
response to high levels of Gram-negative organisms can be
catastrophic for an individual and are associated with the
phenomenon we know as septic shock (see Chapter 15).
At lower doses, the innate immune receptors are unable to
bind sufficient antigen to be stimulatory for the B cell. However,
B-Cell Activation, Differentiation, and Memory Generation
TABLE 12-3
|
CHAPTER 12
407
Properties of thymus-dependent and thymus-independent antigens
TI antigens
Property
TD antigens
Type 1
Type 2
Chemical nature
Soluble protein
Bacterial cell-wall
components (e.g., LPS)
Polymeric protein antigens;
capsular polysaccharides
Isotype switching
Yes
No
Limited
Affinity maturation
Yes
No
No
Immunologic memory
Yes
No
No
Polyclonal activation
No
Yes (high doses)
No
Humoral response
in those B cells that bind to the TI-1 antigen through their Ig
receptors, the TI-1 antigen may still be able to cross-link the
Ig and innate receptors, thereby eliciting an oligoclonal (few
clones) B-cell response that remains independent of T-cell
involvement. Under these circumstances, all the secreted
antibodies will be specific for the TI-1 antigen.
Given the number of commensal bacteria in the gut,
many of which bear these TI-1 antigens, why aren’t our gut
B cells in a constant state of activation? Part of the answer to
this question lies in the anatomy of the gut. The gut is lined
by a mucous layer, which prevents access of most commensal bacteria to gut-resident lymphocytes. Second, regulatory
T cells in the gut help to tone down the immune response,
thus preventing the inflammation that would ensue were
organisms to mount a constant immune response to commensal organisms.
TI-2 Antigens
Unlike TI-1 antigens, type 2 T-independent antigens (TI-2
antigens), such as capsular bacterial polysaccharides or polymeric flagellin, are not mitogenic at high concentrations.
Their capacity to activate B cells in the absence of T-cell help
results from their ability to present antigenic determinants
in a flexible and remarkably multivalent array, which causes
extensive cross-linking of the BCR. In addition, most naturally occurring TI-2 antigens are characterized by the ability
to bind to the complement fragments, C3d and C3dg (see
Chapter 6). As a result, TI-2 antigens are capable of activating B cells by cross-linking the BCR and CD21 receptors on
the B-cell surface.
TI-2 antigens can only partially stimulate B cells in the
complete absence of help from other cells. Monocytes, macrophages, and dendritic cells have been shown to facilitate B-cell
responses to TI-2 antigens by expressing a molecule known as
BAFF—a membrane-bound homolog of tumor necrosis factor—to which mature B cells bind though the TACI (transmembrane activator and CAML interactor) receptor. This
interaction activates important transcription factors that
promote B-cell survival, maturation, and antibody secretion.
T cells can also enhance B-cell activation by producing cytokines that push TI-2 antigen-activated B cells from activation
to antibody production. Recognition by B-cell-bound receptors of these T-cell cytokines may stimulate the B cell to
secrete antibody classes other than IgM in response to TI-2
antigen stimulation. Unlike TI-1 antigens, TI-2 antigens
cannot stimulate immature B cells and do not act as polyclonal activators.
Two Novel Subclasses of B Cells Mediate
the Response to T-Independent Antigens
All B cells bear Ig receptors and secrete antibodies, but
recent research has demonstrated that there are multiple
subpopulations of B cells varying in locations, phenotypes,
and functions (Table 12-4). Some of these subpopulations
(the transitional B-cell populations, T1 and T2) represent
temporal stages of B-cell development that occur after the
B cell leaves the bone marrow. An additional transitional
subset, T3, appears to represent an anergic subpopulation of
B cells (see Chapter 10). Others (B-1a, B-1b, B-2, and marginal zone B cells) represent different subpopulations of
mature B cells, each characterized by a preferential location
and range of functional capacities. The T-dependent B-cell
response described above is conducted by the B-2 cell population, also known as follicular B cells.
The presence of B cells with properties distinct from
those of the majority B-2 subset was first suggested when
scientists noticed that B cells responding to T-independent
antigens differed in several important ways from B cells recognizing T-dependent antigens. Those B cells with specificity for T-independent antigens can be divided into two
major subtypes that differ in aspects of their development,
their anatomical location, and their cell surface markers. We
will describe each of them in turn.
B-1 B Cells
Like ␥␦ intraepithelial T cells, B-1 B cells occupy a functional
niche midway between those of the innate and adaptive
408
PA R T V
|
Adaptive Immunity: Effector Responses
ADVANCES
New Ideas on B-Cell Help: Not All Cells That Help B Cells Make
Antibodies Are T Cells
numbers of neutrophils (neutropenic
individuals) or with other neutrophil disorders. Such patients show reduced levels
of IgM, IgG, and IgA antibodies to microbial TI antigens normally recognized by
MZ B cells, whereas the amounts of these
Igs with specificity for T-dependent (TD)
antigens were normal. Furthermore, in
cases where somatic hypermutation
could be assessed, patients with neutrophil disorders showed less evidence of
V-region mutations in MZ B cells than did
healthy control subjects. These findings
all suggest that neutrophils may play an
important role in the survival and activation of MZ B cells in vivo, a suggestion
borne out by the careful microscopic and
in vitro culture experiments described by
Puga et al.
Neutrophils were observed colonizing the regions of the spleen around the
MZ, even in the absence of infection or
inflammation. Although microscopic
observation and flow cytometry failed to
detect major morphological distinctions
between the MZ splenic neutrophils and
immune systems; many representatives of both cell types are
located outside the classical secondary lymphoid tissues, and
the repertoires of their antigen receptors are less diverse than
those of other, more conventional adaptive immune lymphocyte subsets. Both ␥␦ intraepithelial T cells and B-1 B cells are
self-renewing in the periphery—their populations are maintained without the need to be continuously reseeded from
bone-marrow-derived precursors. Finally, both of these
subsets respond rapidly to antigen challenge with relatively
low-affinity responses.
The existence of the B-1 B-cell subpopulation was first
described in 1983, when the lab of Leonard and Leonore
Herzenberg discovered a set of B cells bearing the CD5 antigen, whose expression had previously been thought to be
restricted to T cells. These CD5⫹ B cells were termed B-1 B cells
to distinguish them from the conventional B-2 B-cell population. (The numbering reflects the order of appearance of
the two B-cell subsets in ontogeny.)
In humans and mice, B-1 B cells make up only about 5%
of B cells, although in some species such as rabbits and cattle,
those in the circulation, subsequent
functional assays showed that the splenic
neutrophils were able to induce IgM
secretion by splenic MZ cells, even in the
absence of prior neutrophil activation by
antigen. Surprisingly, the splenic neutrophils were more effective in the induction of IgM secretion by MZ B cells than
CD4⫹ T cells (Figure 1). Culture of the
1,200
IgM (ng/ml)
Since the classical experiments of
Miller, Mitchell, Mitchison, and others in
the 1960s, immunologists investigating
the sources of help for B-cell activation
and antibody production have been conditioned to think in terms of T lymphocytes. However, we are gradually coming
to appreciate that cells other than T lymphocytes may promote B-cell activation
and survival. We now know that mesenchymal stromal cells and eosinophils
cooperate in ensuring plasma cell survival
in the bone marrow. Recently, in what can
be considered a major shift in our thinking about aiding B-cell function, Puga and
colleagues have demonstrated that marginal zone (MZ) B cells may receive help in
activation, somatic hypermutation (SHM),
class-switch recombination (CSR), and
antibody secretion from an unexpected
source: neutrophils.
Since many cells are capable of exerting effects in vitro that are not replicated
when tested in vivo, we first note that the
numbers and/or activity of MZ B cells are
diminished in patients with reduced
600
0
NBH
DCS
MS
TS
FIGURE 1
NBHs enhance IgM secretion by MZ B
cells even more effectively than do
helper T cells. Splenic MZ B cells were
cultured for 6 days with NBH cells, splenic
dendritic cells (DCs), macrophages (Ms), or
CD4⫹ T cells (Ts). The concentration of IgM
released into the culture supernatant was
assessed by enzyme-linked immunosorbent
assay (ELISA, see Chapter 20).
B-1-like cells represent the major subset. However, even in
humans and mice, B-1 cells predominate in the pleural and
peritoneal cavities, and it is probable that their major function is to protect these body cavities from bacterial infection.
Because of their priority in the B-cell developmental
sequence, B-1 B cells are found in relatively high numbers in
fetal and neonatal life. Cells having the functional characteristics of B-1 cells but lacking expression of the CD5 molecule
were identified at a later stage and termed B-1b B cells.
Because the lack of T-cell involvement in their stimulation
means that AID is never activated, B-1 B cells secrete antibodies of relatively low affinity that are primarily of the IgM class.
Since B-1 B cells derive from a limited number of B-cell clones
generated early in ontogeny, the antibodies they secrete are
also significantly less diverse than the antibodies secreted by
B-2 B cells. Antibodies secreted by B-1 B cells have evolved
to recognize antigenic determinants expressed by gut and
respiratory system bacteria and are primarily directed toward
such common repeated antigens as phosphatidyl choline (a
component of pneumococcal cell walls), lipopolysaccharide,
B-Cell Activation, Differentiation, and Memory Generation
|
CHAPTER 12
409
BOX 12-2
of soluble factors, or does cell-cell contact also play a role? The answer to this
question appears to be both. Comparing
the mRNA and protein expression, as
well as the cell surface presence of a
variety of helper B-cell factors between
the circulating, inactive neutrophils (NC)
and the NBHs, Puga et al. showed that the
helper B-cell neutrophils produced
higher quantities of soluble BAFF and
APRIL, as well as of IL-21 than did the NC
*
50
Relative amount
of AID mRNA
splenic neutrophils, termed B helper
neutrophils, or NBH by the authors, with
CD4⫹ T cells resulted in suppression of
T-cell proliferation following stimulation
with anti-CD3 and IL-2. This shows that
the neutrophils are not stimulating B cells
indirectly by first activating T cells, and
also allows for the possibility that the
neutrophils may be selectively biasing
splenic B cells toward T-independent
responses.
Using qPCR, Puga and colleagues
demonstrated the presence of aicda
mRNA (encoding the AID protein) in MZ
B cells. They further showed that culturing B cells in medium conditioned by the
growth of NBH cells in it (NBH-conditioned
medium) resulted in the up-regulation
of aicda mRNA expression in those B cells.
The amount of aicda mRNA in the MZ
B cells was greater than in naïve B cells,
but not as much as in B-2 B cells engaged
in an active germinal center reaction.
Culturing MZ B cells with NBH cells, or
with NBH-conditioned medium, resulted
in both SHM and CSR and confirmed
that NBH-derived help was sufficient to
drive both processes in the MZ B cells
(Figure 2).
What form does this neutrophilderived help take? Is it purely in the form
25
0
Ctrl
NBH
FIGURE 2
Quantitative RT PCR analysis of the
expression of the aicda gene in MZ B
cells cultured for 2 days in control
medium (left) or in medium
conditioned by NBH cells (right). The
cells. They also expressed more BAFF on
their membranes. Investigators reduced
the amount of each of these cytokines
that was available to B cells by adding
soluble forms of the cytokine receptors
to in vitro B cell cultures. The soluble
receptors act by blocking B cell recognition. Under these conditions, IgM secretion, and CSR to IgG and IgA were both
significantly reduced, thus demonstrating the relevance of the three cytokine
signals to MZ B cell activation. The
authors also note that NBH cells express
surface CD40L, although at lower levels
than those found on CD4+ T cells.
Thus, at least for MZ B cells, the role
played by helper neutrophils in B-cell activation may be even more relevant than
that played by helper T cells. Paradigm
shifts such as this one enliven a field in
delightful and stimulating ways.
Source: I. Puga, et al. 2011. B cell-helper neutrophils stimulate the diversification and production of immunoglobulin in the marginal zone
of the spleen. Nature Immunology 13:170–180.
relative amounts of mRNA encoding the AID
protein were assessed by qPCR. Gene
expression was normalized to the expression of
pax5 mRNA. Culture of MZ B cells in NBH⫺
conditioned medium was shown to enhance
the expression of the aicda gene by
approximately 35-fold.
and influenza viruses. B-1 B cells represent the majority component of the B-cell responses to TI antigens.
Very early in the history of immunology, investigators
noted that the serum of unimmunized mice and humans
contains so-called natural IgM antibodies that bind a broad
spectrum of antigens with relatively low affinity. These antibodies derive mainly from B-1 B cells, and their presence in
unimmunized animals suggests that B-1 B cells may exist in
a partially activated state and constitutively secrete low levels
of natural antibodies. In addition, a relatively high frequency
of these IgM antibodies display autoimmune reactivities,
although their affinities for self antigens are sufficiently low
that they rarely induce disease. Low-level interactions with
self antigens during development may therefore be important in the development of B-1 B-cell function and in the
maintenance of their partially activated phenotype.
Although antibody secretion by B-1 B cells is not dependent on T-cell help, it can be enhanced in the presence of
T-cell cytokines. Indeed, recent data suggests that, in the
presence of T-cell help, B-1b cells may express certain attri-
butes of B-2 cells, such as Ig class switching (with the resultant production of IgA antibodies), SHM, and the generation
of a long-lasting antibody response.
Marginal Zone B Cells
The second class of B cells capable of responding to TI
antigens is the marginal zone (MZ) subset, which resides
in the marginal zone of the spleen (Figure 12-18). As for
B-1 B cells, maintenance of physiological levels of MZ B
cells appears to depend on their capacity to receive lowlevel signals through the BCR. Again, like B-1 B cells, MZ
B cells have the capacity for self-renewal in the periphery
and do not need to be constantly replenished from bone
marrow precursors. MZ B cells derive originally from the
transitional T2 B-cell population, and it has recently been
shown that the Notch2 signaling system plays a role in
sending B cells down the MZ pathway. This is notable
because Notch signaling was previously thought to be
restricted in its activity among lymphocytes to T cells. MZ
B cells bear unusually high levels of CD21 (CR2), enabling
410
TABLE 12-4
PA R T V
|
Adaptive Immunity: Effector Responses
Functional differences among mature B-cell subsets
Attribute
Conventional B2 B cells
B-1 B cells
Marginal zone B cells
Major sites
Secondary lymphoid organs
Pleural and peritoneal cavity;
also spleen
Marginal zones of spleen in
mice; primates also have MZ
cells in other locations
V region diversity
Highly diverse
More restricted diversity
Moderate diversity
Rapidity of antibody
response
Slow; plasma cells appear 7–10
days post stimulation
Rapid; plasma cells appear as
early as 3 days after stimulation
Rapid; plasma cells appear as
early as 3 days after stimulation
Surface IgD?
High levels of IgD
Low levels of IgD
Low levels of IgD
Somatic hypermutation
Yes
No
Yes in primates; possibly in
rodents
Requirements for help
from other cell types
Provided by T cells
No, although T and other cells
can enhance response
Dendritic cells and neutrophils
can enhance response (see
Box 12-2)
Participate in germinal
center reaction?
Yes
No
Possibly, although with
slower kinetics than follicular
B cells
Isotypes produced
All isotypes
Predominantly IgM
Predominantly IgM
Immunological memory
Yes
Very little
Yes; source of IgM-producing
memory cells
them to bind very efficiently to antigens covalently conjugated to C3d or C3dg. Since one characteristic of TI-2
antigens is an enhanced tendency to bind C3d, MZ B cells
are particularly important in the host protection against
pathogens bearing TI-2 antigens.
MZ B cells are specialized to respond to blood-borne antigens that enter the immune system via the splenic MZ. Antigen stimulation of MZ B cells results in their movement from
the MZ to the bridging channels and red pulp of the spleen,
where they undergo a burst of proliferation, forming foci of
plasmablasts not unlike the primary foci formed in the lymph
Follicle
Marginal
sinus
Follicular
arteriole
nodes upon antigen challenge. These cells produce high levels
of antigen-specific IgM within 3 to 4 days after antigenic
stimulation. Very recent work has suggested that MZ B cells
may be helped in antibody secretion, SHM, and CSR not by
T cells (as we have come to expect), but instead by neutrophils, which we normally think of as participants in the
innate arm, rather than the adaptive arm, of immunity. These
experiments are described in Advances Box 12-2.
Some of the characteristics of MZ B cells, as well as their
local environments, differ between rodents and primates. Of
particular interest is the possibility that the variable regions
Marginal zone
Periarteriolar
lymphoid sheath
Central
arteriole
Venous
drainage
Arterial supply
FIGURE 12-18 Longitudinal section through the mouse spleen, showing its blood supply and the location of lymphoid
populations. The periarteriolar lymphoid sheath (PALS), containing the T-cell zone, can be seen surrounding the central arteriole, outside of
which are the lymphoid follicles. Outside the follicles lies the marginal sinus, and the marginal zone separates the sinus from the red pulp of the
spleen. The anatomy of the marginal regions varies somewhat between primate and rodent spleens. [Adapted from S. Pillai et al., 2005, Marginal Zone
B Cells. Annual Review of Immunology 23:161–196, Figure 1.]
B-Cell Activation, Differentiation, and Memory Generation
of antibodies derived from human MZ B cells may undergo
SHM in the absence of an obvious germinal center reaction,
and conceivably even in the absence of antigenic stimulation. Furthermore, in rodents, B cells with the characteristics
of MZ cells are restricted to the MZs of the spleen, whereas
in primates they can be found in other peripheral lymphoid
tissues such as the tonsils.
Negative Regulation of B Cells
Up until this point, we have been discussing how B cells are
activated, and the functional correlates of that activation.
However, antigen stimulation of B cells results in a proliferative response that is as rapid as any observed in vertebrate
organisms; an activated lymphocyte may divide once every 6
hours. Control mechanisms have therefore evolved to ensure
that B-cell proliferation is slowed down once sufficient specific B cells have been generated and that most of the B cells
enter into an apoptotic program once the pathogen has been
eliminated. In this section, we will address negative regulation of B-cell activation that is mediated through two different molecules on the B-cell surface.
Negative Signaling through CD22 Shuts Down
Unnecessary BCR Signaling
In addition to CD19/CD21 and CD81, the BCR of resting
B cells is also associated with an additional transmembrane
molecule, CD22. CD22 bears an Immunoreceptor Tyrosinebased Inhibitory Motif (ITIM), similar in structure to the
ITAM motifs introduced in Chapter 3, but mediating inhibitory, rather than activating functions. Activation of B cells
results in phosphorylation of the ITIM, thus allowing association of the SHP-1 tyrosine phosphatase with the cytoplasmic tail of CD22. SHP-1 can then strip activating
phosphates from the tyrosine of neighboring signaling
complexes.
For as long as the BCR signaling pathway is being activated by antigen engagement, phosphate groups are reattached to the tyrosine residues of adapter molecules and
other signaling intermediates as fast as the phosphatases can
strip them off. However, once antigen levels begin to decrease,
receptor-associated tyrosine kinase activity slows down, and
signaling through CD22 can then induce the removal of any
residual activating phosphates. CD22 thus functions as a
negative regulator of B-cell activation, and its presence and
activity ensure that signaling from the BCR is shut down
when antigen is no longer bound to the BCR. Consistent with
this negative feedback role, levels of B-cell activation are elevated in CD22 knockout mice, and aging CD22 knockout
animals have increased levels of autoimmunity.
CD22 is a cell-surface receptor molecule that recognizes
N-glycolyl neuraminic acid residues on serum glycoproteins
and other cell surfaces and can thus double as an adhesion
molecule. It is expressed in mature B cells that bear both
mIgM and mIgD Ig receptors.
|
CHAPTER 12
411
Negative Signaling through the Fc␥RIIb
Receptor Inhibits B-Cell Activation
It has long been known that the presence of circulating,
specific, antigen-IgG complexes is inhibitory for further
B-cell activation, and this phenomenon has now been
explained at the molecular level by the characterization of
the Fc␥RIIb receptor (also known as CD32). Fc␥RIIb recognizes immune complexes containing IgG and, like CD22,
bears a cytoplasmic ITIM domain. Co-ligation of the B cell’s
Fc␥RIIb receptor molecules with the BCR by a specific
antigen-antibody immune complex results in activation of
the Fc␥RIIb signaling cascade, and phosphorylation of
Fc␥RIIb’s ITIM. The phosphorylated ITIM serves as a docking site for the inositol phosphatase SHIP, which binds to
the ITIM via its SH2 domain. SHIP then hydrolyzes PIP3 to
PIP2, thus interfering with the membrane localization of the
important signaling molecules Btk and PLC␥2 (see Chapter 3)
and causing the effective abrogation of B-cell signaling. Signaling through Fc␥RIIb also results in decreased phosphorylation of CD19 and reduced recruitment of PI3 kinase
to the membrane.
Negative signaling through the Fc␥RIIb receptor makes
intuitive sense, as the presence of immune complexes containing the antigen for which a B cell is specific signals the
presence of high levels of antigen-specific antibody, and
hence a reduced need for further B-cell differentiation.
B-10 B Cells Act as Negative Regulators
by Secreting IL-10
Recently, an unusual population of B cells has been discovered that appears to be capable of negatively regulating
potentially inflammatory immune responses by secreting
the cytokine IL-10 upon stimulation.
Working with two mouse autoimmune models, investigators demonstrated that B cells capable of secreting the
immunoregulatory cytokine IL-10 could alleviate the symptoms of a mouse suffering from a form of the antibodymediated autoimmune disease multiple sclerosis. Recall
from Chapter 11 that IL-10 is a cytokine normally associated with regulatory T cells. It has pleiotropic effects on
other immune system cells, which include the suppression
of T-cell production of the cytokines IL-2, IL-5, and TNF-␣.
Furthermore, IL-10 interacts with antigen-presenting cells
in such a manner as to reduce the cell surface expression of
MHC antigens. The finding that B cells could be capable of
secreting this immunoregulatory cytokine represents the
first indication that they, as distinct from T cells, might
have the capacity to down-regulate the function of other
immune system cells. A small population of splenic B cells
appears to account for almost all of the B-cell-derived
IL-10. However, at this point, we do not know whether this
IL-10 secreting B-cell population truly represents a single
developmental B-cell lineage. For example, B cells producing IL-10 have been identified among both B-1 and B-2
412
PA R T V
|
Adaptive Immunity: Effector Responses
B-cell populations. In addition, some, but not all B cells
producing IL-10 bear markers typical of the transitional T2
B-cell subset.
Importantly, all B-10 B cells appear to demonstrate the
capacity to secrete a diverse repertoire of antibodies, with
specificities for both foreign and auto-antigens. It is thought
that the function of these cells may be to limit and control
inflammation during the course of an ongoing immune
response. A great deal of work is still required to tease out
the lineage relationships among these various IL-10-secreting
and other B-cell subpopulations.
S U M M A R Y
■
■
■
■
■
■
■
The Clonal Selection Hypothesis states that each B cell
bears a single antigen receptor. When antigen interacts
with a B-cell receptor (BCR) specific for that antigen, it
induces proliferation of that B cell to form a clone of cells
bearing the identical specificity. Cells from this clone
secrete antibodies having the same specificity as the antigen receptor. At the close of the immune response, there
will be more cells bearing that specificity than existed
prior to the response, and these so-called “memory B cells”
will generate a faster and higher response upon secondary
antigen exposure. Those cells bearing receptors with
specificity for self antigens are eliminated from the B-cell
repertoire during development.
The B-cell response to some antigens requires help from
T cells. These T-dependent responses are mounted by B-2
or follicular B cells. The response to some other antigens
can occur in the absence of T-cell help. These T-independent
responses are primarily mounted by B-1 or marginal zone
(MZ) B cells.
Mature B-2 B cells migrate to the lymphoid follicles under
the influence of the chemokine CXCL13. Survival of
mature B-2 B cells depends upon access to the survival
factor BAFF.
B cells may acquire antigens directly. Some antigens enter
the lymph node by squeezing between the subcapsular
sinus (SCS) macrophages that line the lymph node,
whereas others enter via a leaky network of conduits that
are sampled by the follicular B cells. Yet others are taken
up first by the SCS macrophages or follicular dendritic
cells and passed on to the B cells.
When the BCR recognizes its cognate antigen, the receptors on the B-cell membrane briefly spread over the antigen surface, and then contract, resulting in B-cell receptor
clustering. This represents the earliest phase of B-cell
activation.
Receptor clustering results in internalization of the receptorantigen complex followed by antigen presentation by the
B cell to the T cell. The T-cell/B-cell interaction occurs
primarily at the border between the T- and B-cell zones of
the lymph node.
T cells interact with their cognate B cells by binding to the
processed antigen with their T-cell receptor (TCR), as
well as by interactions between T cell CD28 with B cell
CD80 and CD86, and between T cell CD40L and B cell
CD40. These interactions facilitate the directional secretion of T-cell cytokines that are necessary for full B-cell
activation.
■
■
■
■
■
■
■
■
Following stimulation of primary B cells at the T-cell/
B-cell border within the lymph node, some B cells differentiate quickly into plasma cells that form primary foci
and secrete an initial wave of IgM antibodies. This requires
the up-regulation of the plasma cell transcription factors
IRF4 and BLIMP-1.
Other B cells from the antigen-stimulate clones migrate to
the primary follicles and form germinal centers.
Within germinal centers, B-cell differentiation continues,
with the generation of somatically hypermutated receptors
that are then subject to antigen selection, with the eventual generation of high-affinity antibodies.
Somatic hypermutation (SHM) affects certain sequences
called mutational hot spots, in the variable region genes of
antibody molecules.
Also within the germinal center, the constant region of the
heavy chain of the antibody genes undergoes class switch
recombination (CSR). This results in the formation of
antibodies bearing mutated variable regions and constant
regions other than . CSR occurs between switch regions
upstream of each heavy chain constant region gene
(except for the ␦ constant region).
Both somatic hypermutation and class switch recombination are mediated by the enzyme activation-induced cytidine deaminase (AID), followed by DNA repair.
At the close of the B-cell response, long-term memory
exists in two forms. Recirculating B memory cells must be
reactivated by antigen in order to yield a higher, faster, and
stronger response than the primary response. In addition,
long-lived plasma cells residing in the bone marrow and
other locations continually secrete antibodies and ensure
that antibodies to commonly encountered antigens constantly circulate within the blood.
T-independent responses generated by B-1 and marginal
zone (MZ) B cells give rise to relatively low-affinity, primarily IgM antibodies.
B-Cell Activation, Differentiation, and Memory Generation
■
TI-1 antigens interact with B cells via both the BCR and
innate immune receptors, whereas TI-2 receptors are
highly polymerized antigens that do not have an intrinsic,
mitogenic activity. Both types of T-independent responses
are enhanced by interactions with other cell types, including T cells, macrophages, and monocytes, and possibly,
neutrophils.
■
■
|
CHAPTER 12
413
B-10 B cells release the interleukin IL-10 upon antigenic
stimulation, and may serve to reduce inflammation during
an ongoing immune response.
B cells can be negatively signaled through CD22 and
Fc␥RIIb—a receptor that recognizes the presence of IgGcontaining immune complexes in the blood.
R E F E R E N C E S
Ada, G. L., and P. Byrt. 1969. Specific inactivation of antigenreactive cells with 125I-labelled antigen. Nature 222:1291–1292.
Allen, C. D., T. Okada, and J. G. Cyster. 2007. Germinal-center
organization and cellular dynamics. Immunity 27:190–202.
Allen, C. D., T. Okada, H. L. Tang, and J. G. Cyster. 2007. Imaging of germinal center selection events during affinity maturation. Science 315:528–531.
Berek, C., G. M. Griffiths, and C. Milstein. 1985. Molecular
events during maturation of the immune response to oxazolone. Nature 316:412–418.
Burnet, F. M. 1957. A modification of Jerne’s theory of antibody
production using the concept of clonal selection. The Australian Journal of Science 20:67–69.
Catron, D. M., et al. 2004. Visualizing the first 50 hr of the primary
immune response to a soluble antigen. Immunity 21:341–347.
Dal Porto, J. M., et al. 2004. B cell antigen receptor signaling
101. Molecular Immunology 41:599–613.
Dykstra, M., et al.2003. Location is everything: Lipid rafts and
immune cell signaling. Annual Review of Immunology
21:457–481.
Eisen, H. N., and G. W. Siskind. 1964. Variations in affinities
of antibodies during the immune response. Biochemistry
3:996–1008.
Flajnik, M. F., and M. Kasahara. 2010. Origin and evolution of
the adaptive immune system: Genetic events and selective
pressures. Nature Reviews Genetics 11:47–59.
Harwood, N. E., and F. D. Batista. 2009. The antigen expressway: Follicular conduits carry antigen to B cells. Immunity
30:177–179.
Hayakawa, K., R. R. Hardy, D. R. Parks, and L. A. Herzenberg.
1983. The “Ly-1 B” cell subpopulation in normal immunodefective, and autoimmune mice. Journal of Experimental Medicine 157:202–218.
Jacob, J., and G. Kelsoe. 1992. In situ studies of the primary
immune response to (4-hydroxy-3-nitrophenyl)acetyl. II. A
common clonal origin for periarteriolar lymphoid sheathassociated foci and germinal centers. Journal of Experimental
Medicine 176:679–687.
Jacob, J., J. Przylepa, C. Miller, and G. Kelsoe. 1993. In situ
studies of the primary immune response to (4-hydroxy-3nitrophenyl)acetyl. III. The kinetics of V region mutation
and selection in germinal center B cells. Journal of Experimental Medicine 178:1293–1307.
Kohler, G., and C. Milstein. 1975. Continuous cultures of fused
cells secreting antibody of predefined specificity. Nature
256:495–497.
Koshland, M. E. 1975. Structure and function of the J chain.
Advances in Immunology 20:41–69.
LeBien, T. W., and T. F. Tedder. 2008. B lymphocytes: How they
develop and function. Blood 112:1570–1580.
Martin, F., and J. F. Kearney. 2002. Marginal-zone B cells.
Nature Reviews Immunology 2:323–335.
Fleire, S. J., et al. 2006. B cell ligand discrimination through a
spreading and contraction response. Science 312:738–741.
Mitchell, G. F., and J. F. Miller. 1968. Cell to cell interaction in the
immune response. II. The source of hemolysin-forming cells in
irradiated mice given bone marrow and thymus or thoracic duct
lymphocytes. Journal of Experimental Medicine 128:821–837.
Germain, R. N., et al. 2008. Making friends in out-of-the-way
places: How cells of the immune system get together and how
they conduct their business as revealed by intravital imaging.
Immunological Reviews 221:163–181.
Muramatsu, M., et al. 2007. Discovery of activation-induced
cytidine deaminase, the engraver of antibody memory.
Advances in Immunology 94:1–36.
Haas, K. M., J. C. Poe, D. A. Steeber, and T. F. Tedder, T.F. 2005.
B-1a and B-1b cells exhibit distinct developmental requirements and have unique functional roles in innate and adaptive
immunity to S. pneumoniae. Immunity 23:7–18.
Hannum, L. G., A. M. Haberman, S. M. Anderson, and M. J.
Shlomchik. 2000. Germinal center initiation, variable gene
region hypermutation, and mutant B cell selection without
detectable immune complexes on follicular dendritic cells.
Journal of Experimental Medicine 192:931–942.
Pape, K. A., D. M. Catron, A. A. Itano, and M. K. Jenkins. 2007.
The humoral immune response is initiated in lymph nodes by
B cells that acquire soluble antigen directly in the follicles.
Immunity 26:491–502.
Raff, M. C., M. Feldmann, and S. De Petris. 1973. Monospecificity of bone marrow-derived lymphocytes. Journal of Experimental Medicine 137:1024–1030.
Schneider, P. 2005. The role of APRIL and BAFF in lymphocyte
activation. Current Opinion in Immunology 17:282–289.
414
PA R T V
|
Adaptive Immunity: Effector Responses
Shapiro-Shelef, M., and K. Calame. 2005. Regulation of
plasma-cell development. Nature Reviews Immunology
5:230–242.
Sixt, M., et al. 2005. The conduit system transports soluble antigens from the afferent lymph to resident dendritic cells in the
T cell area of the lymph node. Immunity 22:19–29.
Weigert, M. G., I. M. Cesari, S. J. Yonkovich, and M. Cohn.
1970. Variability in the lambda light chain sequences of mouse
antibody. Nature 228:1045–1047.
Weill, J. C., S. Weller, and C. A. Reynaud. 2009. Human marginal zone B cells. Annual Review of Immunology 27:267–285.
Useful Websites
http://bio-alive.com/seminars/immunology.
htm This Bio Alive Web site is an excellent source of lec-
loaded for personal use. The series contains lectures pertinent
to this chapter, including seminars entitled Immunological
Memory; Regulating B Cell Immunity; Moviemaking and
Modeling (this one by Ronald Germain, one of the pioneers
of the application of sophisticated imaging technology to the
study of the immune system); Molecular Mechanisms of Leukocyte Migration; and Somatic Hypermutation.
www.sciencedirect.com/science/ar ticle/pii/
S1074761304002389 Two movies from a paper by D.
M. Catron et al. (2004, Visualizing the first 50 hr of the primary immune response to a soluble antigen, Immunity
21:341–347) that is discussed in Chapter 14 can be found at
this Web site. The first portrays the movements of T cells and
dendritic cells prior to antigen entry into the observed
lymph node. The second shows the movements of T cells,
B cells, and dendritic cells through a nearby lymph node,
following injection of antigen into the footpad of a mouse.
tures by accomplished immunologists that can be down-
S T U D Y
Q U E S T I O N S
1. Name one distinguishing feature of TI-1 and TI-2 antigens.
2. In the following flow cytometric dot plots, draw a circle
where you expect to see the designated cell populations.
(a) B2 (follicular) B cells
(b) B-1 B cells
4. You are looking at lymph node sections of a mouse that
has had its Lymphotoxin-␣ gene disrupted, and you are
comparing it with sections from a mouse with an intact
Lymphotoxin-␣ gene. Both mice were immunized approximately 10 days previously with a T-dependent antigen.
What is the most striking feature that is missing in your
knockout animal?
CD5
CD5
5. You have immunized two mice with a T-dependent anti-
IgM
(c) B2 (follicular) B cells
IgM
(d) MZ B cells
gen. One of them is a wild-type mouse, and the other
belongs to the same strain as the first but is a knockout
mouse that does not have the gene encoding activationinduced cytidine deaminase (AID). Upon primary immunization, both mice express similar titer (concentrations)
of IgM antibodies. You allow the mice to rest for 6 weeks
and then reimmunize. Do you expect the secondary antibodies to be similar or different between the two animals?
Explain your answer.
6. Draw the biochemical reaction catalyzed by AID, and
CD21
CD21
describe how subsequent DNA repair mechanisms can
lead to somatic hypermutation (SHM) and to class switch
recombination (CSR).
7. Describe one mechanism by which higher-affinity B cells
may gain a selective survival advantage within the germinal center.
8. The presence of circulating antigen-antibody complexes
IgM
IgM
3. You have generated a knockout mouse that does not
express the CD40 antigen. On stimulation with a T-dependent
antigen, do you expect this mouse to be able to:
a. secrete IgG antibodies specific for the antigen?
b. secrete antigen-specific antibodies that have undergone
somatic hypermutation?
has been shown to result in the down-regulation of B-cell
activation. State why this would be advantageous to the
organism, and offer one mechanism by which the downregulation may occur.
9. Recent evidence suggests that not all regulatory cytokines
secreted by lymphocytes derive from T cells. Explain.
10. Describe three mechanisms by which antigen can enter the
lymph node and make contact with the B-cell receptor.
13
Effector Responses: Celland Antibody-Mediated
Immunity
I
n previous chapters, we have focused on how the
immune response is initiated. Here, we will finally
describe how the targets of the immune system—
pathogen, infected cells, and even tumor cells—are
actually cleared from the body.
We have already seen how the innate immune system
initiates the response to pathogen and alerts the adaptive
immune system to the presence and nature of that
pathogen (Chapter 5). In Chapters 9 through 12, we
described the development and activation of the antigenspecific cells of the adaptive immune system, B and T
lymphocytes. You were also introduced to the
differentiation and activity of helper T lymphocytes, a
type of effector cell that regulates the activity and
function of cytotoxic T cells, B cells, and other antigenpresenting cells. Here we focus on the effector cells and
molecules of both the cell-mediated and the humoral
(antibody-mediated) immune responses that directly rid
the organism of pathogens and abnormal cells. These
effector responses are arguably the most important
manifestations of the immune response: they protect the
host from infection and rid the host of pathogens that
have breached defenses.
The effector functions of cell-mediated and humoral
branches of the immune system assume different,
although overlapping, roles in clearing infection from a
host. The effector molecules of the humoral branch are
antibodies, the secreted version of the highly specific
receptor on the surface of B cells. Antibodies secreted into
extracellular spaces are exquisitely antigen specific and
have several methods at their disposal to rid a body of
pathogen. How an antibody contributes to clearing
infection depends on its isotype, which determines
whether it can recruit complement (recall Chapter 6). The
isotype also determines which receptors an antibody can
bind. Antibody-binding receptors, which bind to the
constant regions of antibodies and are therefore called Fc
receptors or FcRs, determine which cells an antibody can
recruit to aid in its destructive mission, as well as the
tissues to which it can gain entry.
Two cytotoxic T lymphocytes bind to a
tumor-specific antigen on the surface of a
cancer cell, deliver the “kiss of death,” and
induce apoptosis. [Steve Gschmeissner/Photo
Researchers]
■
Antibody-Mediated Effector Functions
■
Cell-Mediated Effector Responses
■
Experimental Assessment of Cell-Mediated
Cytotoxicity
If antibodies were the only agents of immunity,
intracellular pathogens, which occupy spaces that
antibody cannot access, would likely escape the immune
system. Fortunately there is another effector branch of our
immune system, cell-mediated immunity, which detects
and kills cells that harbor intracellular pathogens. Cellmediated immunity consists of both helper T cells
(CD4⫹ TH) and several types of cytotoxic cells. As you
have seen in Chapter 11, TH cells exert their effector
functions indirectly, by contributing to the activation of
antigen-presenting cells, B cells, and cytotoxic T cells via
receptor-ligand interactions and soluble cytokines and
chemokines. On the other hand, cytotoxic cells exert their
effector functions directly, by attacking infected cells and,
in some cases, the pathogens themselves.
Effector cytotoxic cells arise from both the adaptive
and innate immune systems and, therefore, include both
antigen-specific and -nonspecific cells. Antigennonspecific (innate immune) cells that contribute to the
415
416
PA R T V
|
Adaptive Immunity: Effector Responses
clearance of infected cells include NK cells and
nonlymphoid cell types such as macrophages,
neutrophils, and eosinophils. Antigen-specific cytotoxic
cells include CD8⫹ T lymphocytes (CTLs or TC cells), as
well as the CD4⫹ NKT cell subpopulation, which,
although derived from the T-cell lineage, displays some
useful features of innate immune cell types, too.
Populations of cytotoxic CD4⫹ TH cells have also been
described and may contribute to delayed-type
hypersensitivity. The discussion of DTH reactions and
the role of CD4⫹ T cells in their orchestration appears in
Chapter 15.
The humoral and cell-mediated immune systems also
cooperate effectively. Cells such as macrophages, NK cells,
neutrophils, and eosinophils all express Fc receptors,
which induce phagocytosis of antibody-antigen
complexes as well as direct killing of target cells via a
process known as antibody-dependent cell-mediated
cytotoxicity.
It is also important to note that the cell-mediated
immune system plays a role in recognizing and
eliminating not only infected cells, but tumor cells,
which often have undergone genetic modifications that
lead to surface expression of antigens not typical of
normal cells.
In this chapter we will first focus on antibodymediated activities, describing not only their ability to fix
complement but also the multiple functions they acquire
by interacting with cells expressing Fc receptors. We will
then describe cytotoxic effector mechanisms mediated by
cells of the innate and adaptive immune system. We will
close with a discussion of experimental assays of
cytotoxicity.
A Clinical Focus feature and a Classic Experiment
feature are also included in this chapter. The first
describes how knowledge of antibody structure and
function has led to new and successful disease
therapies. The second introduces recent data showing
that lymphocytes are not the only cell type that can
develop a memory; natural killer cells, members of the
innate immune system, also appear to have this
capacity.
Antibody-Mediated Effector
Functions
Antibodies generated by activated B cells (see Chapter 12)
protect the body against the pathogen in several ways: they
can neutralize pathogen by binding to and blocking receptors that the pathogen uses to gain entry into a cell; they can
opsonize pathogen by binding and recruiting phagocytic
cells; and they can fix complement by binding to pathogen
and initiating the complement cascade, which punctures
cell membranes (see Chapter 6). In addition, they can cooperate with the cellular branch of the immune system by
binding and recruiting the activities of cytotoxic cells, specifically natural killer (NK) cells, in a process called
antibody-dependent cell-mediated cytotoxicity (ADCC)
(Figure 13-1 and Table 13-1). The ability of antibodies to
mediate these responses is dependent not only on their
antigen specificity, but also on their isotype (see Chapter 3
and Table 13-2), which determines whether an antibody
will “fix” complement, to which Fc receptors they will bind,
and therefore, which cellular effects they will have.
Collectively, these mechanisms result in the destruction of
pathogen, either directly or indirectly, by stimulating phagocytosis and digestion of pathogen-antibody complexes, directly
destroying pathogenic cell membranes, or inducing suicide of
the infected cell.
Antibodies Mediate the Clearance and
Destruction of Pathogen in a Variety of Ways
Antibodies are versatile effector molecules that play a direct
role in resolving infection by (1) blocking pathogen entry
into cells (neutralization) and (2) recruiting cytotoxic molecules and cytotoxic cells to kill pathogen (via complement
fixation, opsonization, and ADCC).
Neutralization
Most viruses and some bacteria gain entry into a cell by
binding specifically to one or more cell-surface proteins,
stimulating endocytosis. For example, the human immunodeficiency virus (HIV) expresses a coat protein, gp120, that
specifically binds the CD4 coreceptor. The influenza virus
expresses a protein, hemagglutinin, that binds to sugarmodified proteins on epithelial cells. Antibodies that bind
such proteins are particularly potent effector molecules
because they can prevent a pathogen from ever initiating an
infection. They are referred to as neutralizing antibodies.
Pathogens disarmed by neutralizing antibodies are typically
phagocytosed by macrophages.
Neutralizing antibodies can also block entry of toxins
into cells. For example, tetanus toxoid, a product of the Clostridium tetani bacteria, is a neurotoxin that can result in
uncontrolled muscle contraction and death. The tetanus vaccine contains an inactive version of this toxin that stimulates
B-cell production of anti-tetanus toxoid antibodies. These
antibodies bind and potently inhibit the entry of the toxin
into nerve cells. Neutralizing antibodies have also been
raised against snake venom toxins and are effective therapies
for some snakebites.
Physiologically, neutralization is the most effective
mode of protection against infection. However, because
microbes proliferate rapidly, they can generate genetic
Effector Responses: Cell- and Antibody-Mediated Immunity
|
CHAPTER 13
417
Variable regions (Fab)
Constant regions (Fc)
C1q
C1s
C1r
NK cell
FcR
(1) Virus and toxin
neutralization
Tumor
cell
Macrophage
MAC
(2) Opsonization
Prevents
pathogen-host
binding
Phagocytosis
Infected
or tumor
cell
(3) Complement fixation
and formation of the
membrane attack complex
(4) Antibody-Dependent
Cell-mediated
Cytotoxicity (ADCC)
Phagocytosis
or lysis
NK-induced
apoptosis
functions. Antibody binding can enhance the clearance of pathogens by (1) binding to proteins that pathogens use to infect cells
(neutralization), (2) interacting with Fc receptors on the surface of
phagocytic cells and inducing internalization and degradation (opso-
nization), (3) recruiting complement proteins that can directly kill
pathogen or enhance its phagocytosis, and (4) binding to FcRs on
cytotoxic cells (e.g., NK cells) and directing their activity to infected
cells, tumor cells, and/or pathogens (antibody-dependent cell-mediated cytotoxicity [ADCC]).
variants able to evade neutralizing antibodies. For example,
influenza viral variants expressing modified hemagglutinins that do not bind circulating antibodies have a distinct
reproductive advantage and can rapidly take over an infection. HIV is one of the most adept viruses at evading anti-
body responses. In theory, gp120 is an ideal target for
neutralizing antibodies; however, few individuals generate
such antibodies. This is in part because the region of gp120
that binds CD4 is concealed from the body until just before
viral entry.
FIGURE 13-1 Four broad categories of antibody effector
TABLE 13-1
Ways antibodies can protect the host against invasion by pathogens
Mode of antibody-mediated protection
How it works
Neutralization
Antibody binds to sites on the pathogen or a toxin that interact with host proteins, masking
them, and inhibiting entry of that pathogen or toxin into the host. Antibody-pathogen complexes are then eliminated, often after phagocytosis.
Opsonization
Antibody binds to pathogen and is then bound by Fc receptors on phagocytic cells. The antibodyantigen binding to FcRs induces internalization and destruction by the phagocytic cell.
Complement fixation
The antibody-antigen complex becomes bound by complement components in serum and is
either phagocytosed via cells expressing C3 receptors or lysed as a result of pore formation
by the complement components C7, C8, and C9.
Antibody-dependent cell-mediated
cytotoxicity
Antibody-antigen complexes are bound by Fc receptors on NK cells and granulocytes, thus
directing the cytotoxicity of these cells toward the antigen targeted by the antibody (e.g.,
viral proteins on the surface of an infected cell), and inducing apoptosis of the target cell.
418
PA R T V
TABLE 13-2
|
Adaptive Immunity: Effector Responses
Properties and biological activities* of classes and subclasses of human
serum immunoglobulins
IgG1
IgG2
IgG3
IgG4†
IgA1
IgA2
IgM§
IgE
IgD
Molecular weight‡
150,000
150,000
150,000
150,000
150,000–
600,000
150,000–
600,000
900,000
190,000
150,000
Heavy-chain
component
␥1
␥2
␥3
␥4
␣1
␣2
␦
Normal serum
level (mg/ml)
9
3
1
0.5
3.0
0.5
1.5
0.0003
0.03
In vivo serum
half-life (days)
23
23
8
23
6
6
5
2.5
3
Activates classical
complement
pathway
⫹
⫹/⫺
⫹⫹
⫺
⫺
⫺
⫹⫹
⫺
⫺
Crosses placenta
⫹
⫹Ⲑ⫺
⫹
⫹
⫺
⫺
⫺
⫺
⫺
Present on
membrane of
mature B cells
⫺
⫺
⫺
⫺
⫺
⫺
⫹
⫺
⫹
Binds to Fc
receptors of
phagocytes
⫹⫹
⫹Ⲑ⫺
⫹⫹
⫹
⫺
⫺
?
⫺
⫺
Mucosal transport
⫺
⫺
⫺
⫺
⫹⫹
⫹⫹
⫹
⫺
⫺
Induces mast cell
degranulation
⫺
⫺
⫺
⫺
⫺
⫺
⫺
⫹
⫺
*
Activity levels are indicated as follows: ⫹⫹ ⫽ high; ⫹⫽ moderate; ⫹/⫺ ⫽ minimal; ⫺⫽ none; ? ⫽ questionable.
Note that mice do not make IgG4 but generate two different versions of IgG2 (IgG2a and IgG2b).
‡
IgG, IgE, and IgD always exist as monomers; IgA can exist as a monomer, dimer, trimer, or tetramer. Membrane-bound IgM is a monomer, but secreted IgM in serum
is a pentamer.
§
IgM is the first isotype produced by the neonate and during a primary immune response.
†
Opsonization
From the Greek word for “to make tasty,” opsonization
refers to the ability of antibodies to promote and/or enhance
the engulfment of antigens by phagocytes. Opsonizing antibodies coat antigen and interact with Fc receptors expressed
by phagocytic cells.
Thus, phagocytic cells of the innate immune system (e.g.,
macrophages and neutrophils) have two sets of receptors
capable of binding pathogens: they can bind them directly
using innate immune receptors (pattern recognition
receptors such as Toll-like receptors), or they can bind
antigens bound in immune complexes with antibodies, via
Fc receptors.
Innate immune receptors are important in activating
antigen-presenting cells and shaping the signals that they
will convey to lymphocytes (see Chapter 5 and Chapter 11).
Fc receptors also generate signals and regulate the effector
responses of innate immune cells (see below). In the case of
opsonization, binding of pathogen (antigen)-antibody complexes to an Fc receptor on phagocytes will induce internal-
ization of the complex and internal digestion of the pathogen
in lysosomes.
Complement Fixation
As we saw in Chapter 6, antigen-antibody complexes can
also induce a complement cascade. Specifically, when antibodies that associate with complement bind to the surface of
bacteria and some (enveloped) viruses, they can initiate a
cascade of reactions that results in the generation of the
membrane attack complex (MAC), perforating pathogen
membranes and killing the microbe. This capacity to stimulate complement-mediated membrane damage is referred to
as fixing complement and is a property of specific antibody
isotypes, including some IgGs and IgM.
Activation of the early part of the complement cascade can
also protect the host by opsonizing pathogens in an antibodyindependent manner. This complement function does not
require the later components (C4–C9) of the complement cascade. Instead, the complement protein fragment C3b, which is
produced early on in the complement reaction, binds to the
Effector Responses: Cell- and Antibody-Mediated Immunity
|
CHAPTER 13
419
surface of the pathogen and is recognized by the many different blood cells that have C3b receptors. Binding of C3b on the
surface of a pathogen by a macrophage results in phagocytosis
and destruction of that pathogen by mechanisms similar to
those described above for antibody-mediated phagocytosis.
Interestingly, binding of C3b-bound pathogen by red blood
cells results in the delivery of the pathogens to the liver or
spleen, where the pathogen is removed from the red blood cell,
without affecting the viability of the red blood cell, followed by
phagocytosis of the pathogen by a macrophage.
are often pentavalent (bound together in groups of five) they
can be very effective in binding pathogen. Interestingly, most
circulating IgM antibodies come not from conventional B cells
but from B-1 cells, which constitutively produce IgM antibodies that appear to protect us from common pathogens, particularly those that can come from the gut and other mucosal areas.
IgM antibodies are very good at fixing complement and
efficiently induce lysis of pathogens to which they bind. They
are also good at forming dense antibody-pathogen complexes that are efficiently engulfed by macrophages.
Antibody-Dependent Cell-Mediated Cytotoxicity
Antibodies can recruit the activities of multiple cytotoxic cells,
including NK cells (lymphoid members of the innate immune
system) and granulocytes (myeloid members of the innate
immune system). Unlike cytotoxic T lymphocytes, these cells
do not express antigen-specific T-cell receptors. However,
they do express Fc receptors (specifically, Fc␥RIII) and can
use these to “arm” themselves with antibodies that allow them
to “adopt” an antigen specificity. Antibodies bound to NK
cells then direct the cytotoxic activity to a specific cellular
target—for example, an infected cell that expresses viral
proteins on its surface or a tumor cell that expresses tumorspecific proteins on its surface (see Figure 13-1). This phenomenon, referred to as ADCC, is under intense investigation
because of its therapeutic potential. Monoclonal antibodies
(mAbs) developed as cancer treatments are thought to act in
part by taking advantage of ADCC mechanisms and directing
NK cytotoxicity toward tumor cell targets (see Clinical Focus
Box 13-1). As with complement fixation, the ability to mediate
ADCC is antibody-isotype dependent, and murine IgG2a and
human IgG1 are most potent in this capacity (see below).
Effector Functions of IgG Antibodies
IgG antibodies are the most common antibody isotype in the
serum. They are also the most diverse, and include several
subclasses (IgG1, IgG2, IgG3, and IgG4 in humans; IgG1,
IgG2a, IgG2b, and IgG3 in mice), each of which has distinct
effector capabilities. All of the IgG variants bind to Fc receptors and can enhance phagocytosis by macrophages (opsonization). Human IgG1 and IgG3 are particularly good at
fixing complement. Mouse IgG2a and human IgG1 are particularly good at mediating ADCC by NK cells.
Specific IgG isotypes are also associated with disease
pathologies. For example, in a mouse lupus model, IgG2a
and IgG2b isotypes dominate. In human lupus, IgG1 and
IgG3 are involved in anti-DNA reactions, and IgG2 and IgG3
are associated with kidney dysfunction. Investigators are
making excellent use of our growing knowledge of IgG effector functions and are tailoring the isotypes of mAbs used in
therapies to enhance their effect. For example, IgG1 antibodies are most commonly used in tumor therapy because they
can both fix complement and mediate ADCC by NK cells.
However, recent data suggest that IgG2 is a potent mediator
of ADCC by myeloid cells (including neutrophils) and may
offer some therapeutic advantages.
Antibody Isotypes Mediate Different
Effector Functions
As you learned in Chapters 3 and 12, activated B cells develop
into plasma cells, which are remarkably productive antibodyproducing cellular factories. Antibody specificity is defined
by their Fab regions, and their isotype or class is defined by
their Fc region (see Figure 3-21). Depending on the type of T
cell help they received during activation, as well as the cytokines to which they were exposed, B cells will secrete one of
four major classes of antibody: IgM, IgG, IgA, or IgE.
Although immature B cells express IgD, little if any is secreted
and its function remains a bit of a mystery. The structure of
each of these antibodies is reviewed in detail in Chapter 3.
Each plays a distinct role in thwarting and clearing infection,
as described below and depicted in Figure 13-2.
Effector Functions of IgM Antibodies
IgM antibodies are the first class of antibodies to be produced
during a primary immune response. They tend to be lowaffinity antibodies because the B cells that produce them have
not gone through affinity maturation. However, because they
Effector Functions of IgA Antibodies
Although IgA antibodies are also found in circulation, they
are the major isotype found in secretions, including mucus
in the gut, milk from mammary glands, tears, and saliva. In
these secretions, IgA can neutralize both toxins and pathogens, continually interacting with the resident (commensal)
bacteria that colonize our mucosal surfaces and preventing
them from entering the bloodstream. Because IgA cannot fix
complement, these interactions do not induce inflammation,
which is advantageous given that IgA acts continually on
antigens and pathogens that typically pose no threat.
IgA’s half-life in secretions is relatively long because the
amino acid sequence of the Fc region is resistant to many of
the proteases that are present. Interestingly, several microbes,
including those that cause gonorrhea and strep infections,
produce proteases that do degrade IgA, allowing them to
evade this form of immune protection.
IgA does exist in monomeric form (particularly in circulation) but also forms dimers and polymers in mucosal
tissues. The dimeric and polymeric forms bind to receptors
420
PA R T V
|
Adaptive Immunity: Effector Responses
CLINICAL FOCUS
Monoclonal Antibodies in the Treatment of Cancer
Cancer is a disease in which a single
cell multiplies rapidly and uncontrollably
to form a tumor. As the tumor cells replicate, the DNA repair functions that normally ensure faithful replication of DNA
sequence during cell division may begin
to fail, and new mutations begin to accumulate, leading to a worsening prognosis.
Cancer remains one of the worst killers in
the developed world, and despite decades
of intensive and creative work by research
scientists and clinicians, a diagnosis of
cancer is still followed all too frequently by
a shortened life span, preceded by a
course of weakening and toxic treatment.
Attempts to develop effective anticancer drug strategies have had to address
two main problems. First, because many
cancer cells are derived from self and do
not “look” too different from healthy cells,
any drug that binds to the cancer cell is
likely also to bind and enter the patient’s
other, healthy cells. The trick is to find a
molecule on the surface of a cancer cell
that is either not expressed in normal
cells, or is expressed at much lower concentrations in normal cells compared to
cancer cells, and then to design a drug
that will preferentially bind to the malignant cell. Second, because cancer cells
and normal dividing cells use the same
metabolic strategies to generate ATP, any
untargeted drug that affects the metabolism or growth of a cancer cell will also
affect normal dividing cells.
Most of the first generation of anticancer drugs, however, are focused on inhibiting cancer cell division. Drugs such as
daunorubicin, cis-platin, and others enter
all cells of the body and interfere with the
process of cell division. As mentioned
above, such drugs will also target cells
that divide normally as part of their dayto-day function, and the patient’s wellbeing will be compromised not only by
the tumor they carry, but also by the treatment they endure. Cells of the alimentary
tract, skin, hair follicles, and immune system all divide more frequently than others
such as the liver, kidney, and brain and
tend to be more sensitive to these drugs.
This is why conventional chemotherapy
leads to hair loss, digestive distress, and
impaired immune function.
The development of monoclonal antibodies, with their extraordinary capacity to
target a single 6- to 8-amino-acid length of
a protein was seized upon by scientists as a
potential opportunity to specifically target
tumor cells, while leaving the neighboring
healthy cells undamaged. Immunotherapeutics, the production and refinement of
mAbs that will specifically target cancer
cells for death and healthy neighbors, is a
burgeoning field. Now part of the therapeutic arsenal, mAbs have met with some
success. Some examples of current
reagents and their mechanisms of action
are described below.
How Do Antibodies Kill Cancer Cells?
• Competitive binding to cell surface receptors and preventing their activation. Antibody binding to molecules on the
surface of a cell can block interactions
that may be necessary to allow growth
of that cell. For example, many tumor
cells become dependent for their
growth on binding of specific growth
factors such as Epidermal growth factor
(EGF). Several of the mAbs currently in
clinical use, such as erbitux, bind to the
EGF receptor (EGFR) and interfere with
its function. Herceptin is an example of
an extremely successful drug that targets the Her-2 EGFR on advanced
breast cancer cells. All EGFRs must
dimerize on binding to EGF so that
they can send their growth signal to
the nucleus. Herceptin interferes with
this dimerization reaction and, so, interrupts the EGFR signal.
• Interference with the generation of new
blood vessels. Another group of drugs,
including bevacizumab (Avastin), targets activity of vascular endothelial
growth factor (VEGF). VEGF interacts
with VEGF receptors (VEGFR1 and
VEGFR2) located on blood vessels.
Binding stimulates the formation of
new blood vessels that can be used to
supply the nascent tumor cells with
nutrients. Avastin binds to VEGF and
prevents it from interacting with its
receptors, thereby inhibiting the
development of new blood vessels
and compromising tumor growth.
• Direct binding to receptors with induction of apoptosis. Some mAbs, such as
rituximab (Rituxan), may kill their target cells by binding to the surface of
the cell and mimicking the binding of
natural ligands. Binding of rituximab to
follicular B lymphoma cells results in
interference with cell cycle regulation
and the induction of apoptosis.
• Induction of ADCC. Binding of certain
isotypes of antibodies to the surface of
a cell results in the induction of antibody-dependent cell-mediated cytotoxicity (ADCC). If a tumor cell becomes
coated with antibodies of human IgG1
or IgG3 subclasses, Fc␥RIII molecules
on the surface of natural killer (NK)
cells and other immune cells will bind
to the Fc regions of these antibodies.
Once bound to the NK cell, the tumor
cell becomes the target of secreted
perforin and granzymes, which induce
apoptosis in the affected cell. In addition, macrophages will recognize the
Fc portion of antibody clusters of the
IgG1 isotype that are bound to a cell
surface, and be induced to phagocytose the antibody-cell complex and/or
release toxic metabolites.
• Induction of complement fixation. IgG3
and IgG1 antibody binding to the surface of a cell can also result in the attachment (fixation) of a set of serum proteins
known as the complement cascade to
the cell surface. By a complex set of
reactions detailed in Chapter 6, this can
result in either the dissolution of the cell
following disruption of the integrity of
its membrane, or enhanced phagocytosis of that cell by macrophages.
• Delivery of toxins. Very early on in the
development of mAbs as clinical tools,
Effector Responses: Cell- and Antibody-Mediated Immunity
|
CHAPTER 13
421
BOX 13-1
scientists began modifying them with
toxins such as radioactive isotopes and
cytotoxic reagents. The isotopes 90Y and
131
I are both capable of delivering shortrange cytotoxic doses of radiation to
anything bound by the conjugated
antibody and the immuno-conjugates
Zevalin and Bexxar were approved for
use in humans for the treatment of
relapsed or refractory non-Hodgkin’s
lymphoma in 2002 and 2003, respectively. Other toxins that have been studied in the lab and/or used in the clinic
include modified bacterial toxins such
as Pseudomonas exotoxin A, which has
been conjugated to an mAb directed
toward the CD22 molecule and has
proven promising in the treatment of
chemoresistant hairy cell leukemia.
Ricin, another bacterial toxin, has also
TABLE 1
been successfully conjugated with antibodies directed toward CD22 and
CD19, and these molecules are currently being developed for the treatment of pediatric and acute
lymphoblastic leukemia.
Currently, scientists are working to
improve the biological efficacy of the antibodies employed in cancer treatment by
subjecting proven antibodies to mutation
in vitro followed by selection using phage
display technology to improve the affinity
and selectivity of binding to tumor cells.
Others are combining multiple antibodies
with different specificity to the same cell in
a single treatment, or combining mAbs
with more conventional chemotherapeutic approaches. The Human Genome Project has provided a wealth of information
regarding the number and structure of
cell-surface signaling molecules, and it is
also being exploited in the design of novel
drugs and toxins, including antibodies and
antibody fragments.
For further reading on this topic, see
the references below. Also keep an eye on
the news. Immunotherapeutics is one of
the fastest-moving and most exciting
areas of research at the confluence of the
laboratory and the clinic. New ideas arise
every day, and new reagents are likely to
follow.
Rivera, F., et al. 2008. Current situation of panitumumab, matauzumab, nimotuzumab and
zalutumumab. Acta Oncologia 47:9–19.
Strome, S. E., et al. 2007. A mechanistic perspective of monoclonal antibodies in cancer
beyond target-related effects. The Oncologist 12:1084–1095.
Some monoclonal antibodies in clinical use
mAb product
(trade name)
Nature of
antibody
Target (antibody
specificity)
Modification of
antibody
Rituximab,
(Rituxan)
Chimeric
CD20 (mouse B-cell antigen)
None
Relapsed or refractory
non-Hodgkin’s lymphoma
Trastuzumab
(Herceptin)
Humanized
Human epidermal growth
factor receptor 2 (HER-2)
None
HER-2 receptor positive
advanced breast cancers
Alemtuzumab
(Campath)
Humanized
CD52 (an antigen on many
types of leukocytes)
None
Chronic lymphocytic
leukemia
Bevacizumab
(Avastin)
Humanized
Vascular endothelial growth
factor (VEGF)
None
Colorectal cancer
Cetuximab (Erbitux)
Chimeric
EGFR
None
Colorectal cancer
Panitumumab
(Vectibix)
Human
EGFR
None
Colorectal cancer
Ibritumomab
Mouse
CD20
None
Relapsed or refractory
non-Hodgkin’s lymphoma
Ibritumomab
tiuxetan (Zevalin)
Mouse
CD20
90
Relapsed or refractory
non-Hodgkin’s lymphoma
Tositumomab
(Bexxar)
Mouse
CD20
131
Gemtuzumab
Humanized
ozogamicin (Mylotarg)
CD33 (glycoprotein antigens
on myeloid progenitor cells
and monocytes)
Attached to an anti-tumor agent
that cleaves double-stranded
DNA at specific sequences
Acute myelogenous
leukemia
Epratuzumab
CD22 (glycoprotein antigens
on mature and neoplastic B
cells)
None
Relapsed or refractory
non-Hodgkin’s lymphoma
Humanized
Y
I
Treatment
Relapsed or refractory
non-Hodgkin’s lymphoma
PA R T V
422
|
Adaptive Immunity: Effector Responses
FcR
Eosinophil
Helminth
(a) Degranulation
(FcεR and IgE)
(b) Opsonization of
bacteria and phagocytosis
(FcγR and FcαR
and IgG or IgA)
Digestion
in lysosome
Phagocyte
Blood
(c) Maintaining serum
levels of antibodies
(FcRn and IgG)
Plasma cell
Dissociation at
physiological pH
Acidic
endosome
Endothelial cell
(d) ADCC
(FcγR and IgG)
NK cell
Tumor cell
(e) Transcytosis into secretions
(PolyIgR and IgA)
Lumen (GI tract)
Tumor
Ag
IgA
Poly-Ig receptor
Epithelial cell
Blood
Secreted IgA
Secretory component
FIGURE 13-2 Functions of Fc receptors. Fc receptors (FcRs)
come in a variety of types and are expressed by many different cell
types. Some of their major functions are illustrated here. (a) When
bound by IgE-pathogen (e.g., worm) complexes, FcRs expressed
by granulocytes can induce the release of histamine and proteases
(degranulation). (b) When bound by antibody-pathogen complexes, Fc␣R and Fc␥R can induce macrophage activation and
phagocytosis. (c) The neonatal FcR (FcRn) binds antibody that has
been nonspecifically engulfed by endothelial cells and returns the
antibody, intact, to the blood. (d) When bound by IgG antibodies
coating infected or tumor cells, Fc␥Rs activate the cytolytic activity
of natural (NK) cells. (e) PolyIg receptors (PolyIgR) expressed by the
inner (basolateral) surface of epithelial cells (facing the blood) will
bind dimers and multimers of IgA and IgM antibodies and transfer
them through the cell to their apical (outer) surface and into the
lumen of an organ (e.g., the GI tract). This is a process referred to as
transcytosis and is responsible for the accumulation of antibodies
in bodily secretions.
Effector Responses: Cell- and Antibody-Mediated Immunity
|
CHAPTER 13
423
lysis, many of the effector functions of the antibodies described
above depend on their ability to interact with receptors on cells
in the innate and adaptive immune systems as well as cells in
various epithelial and endothelial tissues. These antibody receptors were originally discovered over 40 years ago by early antibody biochemists, who referred to them as Fc receptors because
they bound specifically to the Fc portion of the antibody.
FcRs allow nonspecific immune cells to take advantage of
the exquisite specificity of antibodies to focus their cellular
functions on specific antigens and pathogens. They also
provide a physical bridge between the humoral and cellmediated immune systems. When bound by antibody alone,
FcRs do not trigger signals; however, when cross-linked by
multiple antibody-antigen complexes, FcRs initiate effector
responses (see Figure 13-2).
The diversity of FcRs, most of which are members of the
immunoglobulin superfamily (Figure 13-3), is now fully
appreciated. Each differs by (1) the antibody class (isotype)
to which they bind (e.g., Fc␥Rs bind to several IgG isotypes,
FcRs bind to the IgE isotype, and Fc␣Rs bind to IgA isotypes), (2) the cells that express them (macrophages, granulocytes, NK cells, T and B cells), and (3) their signaling
properties. Together these properties define a distinct effector function that an antibody can inspire (Table 13-3).
on epithelial cells, an event that triggers endocytosis and
transport of the molecules from the basolateral (inner) to the
apical (outer) sides of the epithelial cell and into the tissue
lumen (see Figure 13-2).
Two subclasses of IgA—IgA1 and IgA2—are found in
humans. IgA1 is more prevalent in the serum (and typically
monomeric), and IgA2 is more prevalent in secretions. Both
isotypes can mediate ADCC by binding FcRs on NK T cells,
and both can trigger degranulation of granulocytes.
Effector Functions of IgE Antibodies
IgE antibodies are best known for their role in allergy and
asthma. However, research suggests that they play an important role in protection against parasitic helminths (worms)
and protozoa. They are made in very small quantities but
have a very potent effect, inducing degranulation of eosinophils and basophils, and release of molecules such as histamine that do permanent damage to a large pathogen.
Fc Receptors Mediate Many Effector
Functions of Antibodies
Although several antibody isotypes have a direct effect on the
viability of a pathogen by initiating complement-mediated
Poly IgR
S
S
S
S
S
S
Fcγ RI
FcRN
S
S
S
S
S
S
S
S
S
S
S
S
β2m
S
S
Fcγ RIIB
Fcγ RIII
FcαR
FcεRI
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
ITIM
ITAM
γ γ
ITAM
CD32
β
γ γ
CD16
CD64
ITAM
γ γ
ITAM
β
γ γ
CD89
Inhibitory
FIGURE 13-3 The structure of a number of human Fc
receptors. The antibody-binding polypeptides are shown in blue
and, where present, accessory signal-transducing polypeptides are
shown in green. Most FcRs are activating receptors and are associated with signaling proteins that contain the Immuno-receptor
Tyrosine Activation Motif (ITAM) or include ITAM(s) in their own intracellular regions. Fc␥RIIB is an inhibiting FcR and includes an Immunoreceptor Tyrosine Inhibition Motif (ITIM) in its intracellular region. All
FcRs shown in this figure are members of the immunoglobulin supergene family; their extracellular regions include two or more immunoglobulin folds. These molecules are expressed on the surface of
various cell types as cell-surface antigens and, as indicated in the
figure, some have been assigned cluster of differentiation (CD) designations (for clusters of differentiation, see Appendix 1). [Adapted from
M. Daeron, 1999, Fc Receptors, in The Antibodies, M. Zanetti and J. D. Capra,
eds., vol. 5, p. 53, Amsterdam: Harwood Academic Publishers.]
424
PA R T V
TABLE 13-3
|
Adaptive Immunity: Effector Responses
Expression and function of FcRs
FcR
Isotypes they bind
Cells that express them
Function
Fc␥RI
(CD64)
IgG2a in mice, IgG1 and IgG3 in
humans
High-affinity receptor
Dendritic cells, monocytes,
macrophages, granulocytes, B
lymphocytes
Phagocytosis
Cell activation
Fc␥RII
(CD32)
IgG
Dendritic cells, monocytes,
macrophages, granulocytes, B
lymphocytes, some immature
lymphocytes
Inhibitory receptor
Traps antigen-antibody
complexes in germinal center
Abrogates B-cell activation
Fc␥RIII
(CD16)
Humans generate two versions:
Fc␥RIIIA (CD16a) and
Fc␥RIIIB (CD16b)
IgG1, IgG2a, and IgG2b in mouse;
IgG1 in human
Low -affinity receptor
Only FcR that binds mouse IgG1
Dendritic cells, monocytes,
macrophages, granulocytes, B
lymphocytes
Only FcR expressed by NK cells
ADCC
Cell activation
Fc␥RIV (in mouse, with some
similarity to human Fc␥RIIIA
and/or human FcRI)
IgG2a and IgG2b in mice; IgG1 in Monocytes, macrophages,
humans
granulocytes
Intermediate affinity receptor,
Not on lymphocytes
although exhibits higher affinity
for human IgG1 than FcgRIIA.
ADCC
Cell activation
FcRI
IgE
Eosinophils, basophils,
mast cells
Degranulation of granulocytes,
including eosinophils, basophils,
mast cells
FcRII
(CD23)
IgE (low affinity)
B lymphocytes
Regulation of B-cell production
of IgE
Transport of IgE-antigen
complexes to B-cell follicles
Fc␣RI
(CD89)
IgA
Dendritic cells, monocytes,
macrophages, granulocytes,
some liver cells
Phagocytosis
Cell activation
ADCC
pIgR
IgA and IgM
Multiple epithelial cells
Transport of antibody from
blood to the lumens of GI,
respiratory, and reproductive
tracts (transcytosis)
FcRn
(neonatal FcR)
IgG
Epithelial cells (including
intestinal epithelium)
Endothelial cells of mature
animals
Transport of antibodies from
milk to blood (transcytosis)
Transport of antibody-pathogen
complexes from gut to mucosal
immune tissue
Phagocytosis
Maintenance of levels of serum
IgG and albumin
For example, antibody-antigen complexes binding to FcRs
on eosinophils and other granulocytes trigger the release of
histamine and proteases from granules (degranulation). Fc␥Rs
and Fc␣Rs on macrophages trigger a signaling cascade that
induces the internalization of the pathogen into a phagocyte,
which can destroy it via a variety of mechanisms, including
oxidative damage, enzyme digestion, and membrane-disrupting
effects of antibacterial peptides. “Neonatal” FcR (FcRn)
molecules on cells that line the bloodstream help maintain
antibody levels in serum by protecting them from degradation. Other Fc␥Rs expressed by cytotoxic cells, including NK
cells