Modulating influence on HIV/AIDS by interactingRANTESgene variants

Modulating influence on HIV兾AIDS by interacting RANTES gene variants Ping An†, George W. Nelson†, Lihua Wang†, Sharyne Donfield‡, James J. Goedert§, John Phair¶, David Vlahov储, Susan Buchbinder**, William L. Farrar††, William Modi†, Stephen J. O’Brien‡‡, and Cheryl A. Winkler†§§ †Intramural Research Support Program, SAIC–Frederick, National Cancer Institute, Frederick, MD 21702; Laboratories of ‡‡Genomic Diversity and ††Molecular Immunoregulation, National Cancer Institute, Frederick, MD 21702; ‡Rho, Inc., 121 South Estes Drive, Suite 100, Chapel Hill, NC 27514; §Viral Epidemiology Branch, National Cancer Institute, 6120 Executive Boulevard, Bethesda, MD 20892; ¶Northwestern University Medical School, Comprehensive AIDS Center, 6980 North Lake Shore Drive, Suite 1106, Chicago, IL 60611; 储Department of Epidemiology, The Johns Hopkins Bloomberg School of Public Health, Baltimore, MD 21205; and **San Francisco Department of Public Health, San Francisco, CA 94102 Communicated by Robert C. Gallo, Institute of Human Virology, Baltimore, MD, May 24, 2002 (received for review January 2, 2002) RANTES (regulated on activation normal T cell expressed and secreted), a ligand for the CC chemokine receptor 5, potently inhibits HIV-1 replication in vitro. We tested the influence of four RANTES single nucleotide polymorphism (SNP) variants and their haplotypes on HIV-1 infection and AIDS progression in five AIDS cohorts. Three SNPs in the RANTES gene region on chromosome 17 (403A in the promoter, In1.1C in the first intron, and 3ⴕ222C in the 3ⴕ untranslated region) are associated with increased frequency of HIV-1 infection. The common In1.1C SNP allele is nested within an intronic regulatory sequence element that exhibits differential allele binding to nuclear proteins and a down-regulation of gene transcription. The In1.1C allele or haplotypes that include In1.1C display a strong dominant association with rapid progression to AIDS among HIV-1-infected individuals in African-American, European-American, and combined cohorts. The principal RANTES SNP genetic influence on AIDS progression derives from the downregulating RANTES In1.1C allele, although linkage disequilibrium with adjoining RANTES SNPs including a weaker up-regulating RANTES promoter allele (ⴚ28G), can modify the observed epidemiological patterns. The In1.1C-bearing genotypes account for 37% of the attributable risk for rapid progression among African Americans and may also be an important influence on AIDS progression in Africa. The diminished transcription of RANTES afforded by the In1.1C regulatory allele is consistent with increased HIV-1 spread in vivo, leading to accelerated progression to AIDS. T he entry of HIV-1 into CD4⫹ cells is mediated by interactions between the viral envelope glycoproteins, the CD4 receptor, and HIV-1 coreceptors. The primary HIV-1 coreceptors are the chemokine receptors CCR5 (CC chemokine receptor 5), used by R5 HIV-1 strains, and CXCR4, used by X4– HIV-1 strains that emerge during the later stages of infection. The CCR5 ligands, RANTES (regulated on activation normal T cell expressed and secreted), MIP-1␣, and MIP-1␤, and the CXCR4 ligand SDF-1␣ all are potent inhibitors of HIV-1 cell entry and replication (1). Variants in the HIV-1 coreceptors and their natural ligand genes have been shown to modify HIV-1 transmission and disease progression (2–9). RANTES inhibits CCR5-mediated entry of R5 strains by competitive binding and down-modulation of CCR5 (10, 11). HIV-1exposed, but uninfected, individuals produce high levels of RANTES from peripheral blood mononuclear cells or cultured CD4⫹ T cells (12–14), and in HIV-1-infected individuals, those with higher levels of RANTES postpone the onset of AIDS-defining pathologies (refs. 15–19; reviewed in ref. 20). Two single nucleotide polymorphism (SNP) sites, ⫺28C兾G and ⫺403G兾A, in the promoter region of RANTES have been identified (8). The ⫺28G variant, but not ⫺403A, was reported to up-regulate RANTES transcription in one study (8) whereas ⫺403A was reported to up-regulate RANTES transcription in a second study without consideration of ⫺28C兾G (21). The [⫺403A⫺28G] haplotype was shown to be associated with a slower rate of CD4⫹ T-cell depletion in HIV-1-infected Japanese (8). In European Americans 10002–10007 兩 PNAS 兩 July 23, 2002 兩 vol. 99 兩 no. 15 (EA) the compound genotype [⫺403G兾A⫺28C兾C] was reported to be susceptible to HIV-1 infection but resistant to AIDS progression when compared with genotype ⫺403G兾G⫺28C兾C in one study (9), but to be susceptible to both HIV-1 infection and AIDS progression in another (22). No effect on HIV-1 infection and AIDS progression by these variants has been reported in African Americans (AA) (22). Considering the potential interaction of these two RANTES gene polymorphisms and the complex nature of RANTES gene expression (8, 9, 21–24), we reasoned that it would be valuable to screen the entire RANTES gene region for nucleotide polymorphisms and to perform an infection and survival association analysis on the same five cohorts that have been used to discover eight AIDS restriction genes (25, 26). Here we describe a group of seven SNPs within the RANTES gene including one, In1.1T兾C, within a newly identified intronic RANTES regulatory element that modulates RANTES transcription, possibly influences HIV-1 infection, and affects the rate of progression to AIDS in HIV-1-infected individuals. Materials and Methods Study Population. The study group includes 964 seroconverters, 2,103 seroprevalents, and 1,101 seronegatives for a total of 4,168 (EA, 2,594; AA, 1,574) from five natural history longitudinal AIDS cohorts: AIDS Link to the Intravenous Drug Experience (27), Hemophilia Growth and Development Study (28), Multicenter AIDS Cohort Study (29), Multicenter Hemophiliac Cohort Study (30), and the San Francisco City Clinic Study (31). An additional 129 Han Chinese normal blood donors were genotyped for allele frequencies. Informed consent was obtained from all study participants. Seroconversion date was estimated as the midpoint between the last seronegative and the first seropositive HIV-1 antibody test date (mean interval 0.79 years, range 0.07 to 3.0 years). High-risk exposed uninfected subjects (n ⫽ 271) were those with high-risk exposure through sharing of injection equipment (32), anal receptive sex with multiple partners (33), or transfusions with factor VIII clotting factor before 1984 when heat treatment was initiated (34). Identification of DNA Polymorphisms. To identify nucleotide polymorphisms in RANTES we used a DNA panel consisting of 72 EA and 72 AA. A nonisotopic RNA cleavage assay (NIRCA) was performed to screen polymorphisms by using the Mismatch Detect II kit (Ambion, Austin, TX) according to the manufacturer’s instructions (35). Overlapping PCR primers were designed to cover the entire RANTES gene except a portion of intron 2 (positions 166995 to 170610 and 176293 to 177383 of the GenBank sequence Abbreviations: RANTES, regulated on activation normal T cell expressed and secreted; CCR5, CC chemokine receptor 5; SNP, single nucleotide polymorphism; EA, European Americans; AA, African Americans; OR, odds ratio; RH, relative hazard. Data deposition: The sequences reported in this paper have been deposited in the GenBank database (accession nos. AF336300 and AF336301). §§To whom reprint requests should be addressed. E-mail: [email protected]. www.pnas.org兾cgi兾doi兾10.1073兾pnas.142313799 Electrophoretic Mobility-Shift Assay (EMSA). Nuclear protein extraction and EMSA were performed as described (36). CD4⫹-enriched lymphocytes were obtained from human T cells by high affinity negative selection with a human T cell CD4 Subset Column Kit (R&D Systems). After activation with 5 ␮g兾ml phytohemagglutinin for 7 days in the presence of 20 units兾ml IL-2, CD4⫹-enriched lymphocytes were harvested for preparing nuclear extracts. The probe sequences were 5⬘-gatcagtttttctgtctttaaggtctacaccctcaa-3⬘ for In1.1T, and 5⬘-gatcagtttttctgtcttcaaggtctacaccctcaa-3⬘ for In1.1C. The probes were filling-labeled with 32P-dATP. Cloning of the Intron 1 Fragment. PCR products of intron1 fragments spanning nucleotides 168693 to 169104 were obtained by using primers RT-In-F-BamHI 5⬘-tcaaggatccgtaagtcctggtcttgaccacc-3⬘ and RT-In-R-SalI 5⬘-acgcgtcgacgaatatggctgtctcagggtct-3⬘ and Turbo Pfu polymerase (Stratagene) and were subsequently placed into the BamHI and SalI sites in pGL3-promoter vector (Promega). Haplotype Cloning of the Intron 1 and Promoter Fragments. RANTES promoter fragments spanning ⫺634 to ⫹45 bp [numbering according to Liu et al. (8)] containing ⫺403G兾A ⫺28C兾G were amplified from RANTES promoter haplotype vectors (8) (provided by Tatsuo Shioda, University of Tokyo) with Turbo Pfu polymerase, and subsequently placed into the MluI and XhoI sites upstream of the Luc⫹ gene in the pGL3-basic vector (Promega), whereas the 412-bp intron 1 fragments with In1.1T or In1.1C were placed into the downstream BamHI and SalI sites. All constructs were verified by sequencing. Transfection and Luciferase Assays. Transient transfections of Jurkat cells were performed as described (36). Cationic liposomes (Roche Molecular Biochemicals) were used to transfect the reporter plasmids into Jurkat cells. After stimulation by phytohemagglutinin兾 phorbol 12-myristate 13-acetate for 24 h at 37°C, cells were lysed and measured for luciferase activity in the Luciferase Assay System (Promega). Luciferase activity was normalized against protein concentration. Genotyping of RANTES Polymorphisms. Two methods, PCRrestriction fragment length polymorphism (PCR-RFLP) and 5⬘ nuclease PCR assays (TaqMan), were used for genotyping. SNPs ⫺28C兾G, In1.1T兾C, and 3⬘222T兾C were genotyped by PCR-RFLP with primers and endonuclease restriction enzymes as follows: ⫺28C兾G, forward, 5⬘-actcgaatttccggaggcta and reverse, 5⬘tctgcagctcaggctggccctttat using MnlI for digestion; In1.1T兾C, forward, 5⬘-cctggtcttgaccaccaca and reverse, 5⬘-gctgacaggcatgagtcaga using MboII for digestion; and 3⬘222T兾C, 5⬘-ctgtcccggtactgacaagg and 5⬘-cccgagtagctgggactaca using HphI for digestion. SNPs ⫺105 and ⫺109 were genotyped by forced PCR-RFLP with primers RT(⫺105)F 5⬘-ttggtgcttggtcaaagagg and RT(⫺105)R 5⬘-ccggtatcataagtgaaattcca using BslI for digestion and RT(⫺109)F 5⬘ggtgcttggtcaaagaggaa and RT(⫺109)R 5⬘-catggtacctgtggaagagg using EarI for digestion, respectively. The artificially introduced nucleotides are underlined. All restriction enzymes were purchased from New England Biolabs. TaqMan assays were performed by using PCR primers and TaqMan probes (Perkin–Elmer) in the ABI Prism 7700 Sequence Detector (Applied Biosystems) according to the manufacturer’s manual. SNP ⫺403G兾A was genotyped with primers RT (⫺403)TBF 5⬘-tccagaggaccctcctcaataa and RT(⫺403)TBR 5⬘ctgagtcactgagtcttcaaagttcc and MGB TaqMan probes RT(⫺403)MGB-G 5⬘-FAM-aaaggaggtaagatctgtaat and RT(⫺403)MGB-A 5⬘-VIC-aaaggagataagatctgtaatg. SNP In1.2T兾C was genotyped by using primers RT(In1.2)-TMF 5⬘-gcctcagtttctgtAn et al. caaggaaga and RT(In1.2)-TMR 5⬘-agggagacccttttattcattgc and the TaqMan probes RT(In1.2)-TM-G 5⬘-FAM-tgtccagcacaatgtcaagtgtgca and RT(In1.2)-TM-A 5⬘-VIC-aagtgtccagcacaatatcaagtgtgcagta. SNP sites are in bold. Detailed protocols are available from the authors. Statistical Analysis. Haplotypes were inferred by the expectation maximization algorithm (37). Kaplan–Meier nonparametric survival statistics and the Cox proportional hazards model in the SAS package (SAS Institute, Cary, NC) were used for survival analysis. Three endpoints reflecting advancing HIV-1 disease progression were evaluated: time to less than 200 CD4⫹ cells per mm3 (CD4 ⬍200); AIDS-1987, as defined by the Centers for Disease Control and Prevention (38); and AIDS-related death. The censoring date was the earliest of the date of the last recorded visit, or July 31, 1997 (for the AIDS Link to the Intravenous Drug Experience cohort), or December 31, 1995 (for all other cohorts). Cox model analyses were performed both unadjusted (using the haplotype factor by itself) and adjusted (considering the RANTES haplotype factor as a covariate along with additional AIDS-modifying genetic factors: for EA, CCR5-⌬32, CCR2–64I, SDF1–3⬘A, HLA-B*27, HLA-B*57, HLA-B*35Px, and HLA class I zygosity; for AA, HLA-B*57, HLA-B*35Px, and HLA class I zygosity (2–5, 7, 25, 26, 39). Participants were stratified by ethnic group and sex (7.6% female) and age group at seroconversion (0–20, ⬎20–40, and over 40 years). Defined categorical analyses (DCA) compared seroconverters progressing more rapidly than the median progression time to the specified outcome to seroconverters and seroprevalents surviving outcome free for at least that long; for AA a shorter time of 7.5 years was used for AIDS-1987 and death, to allow for the relatively short follow-up time for this group. The two-tailed Fisher’s exact test was performed for the DCA and infection tests, except that the codominant model was tested by the Mantel–Haenszel ␹2, and Cochran–Mantel–Haenszel stratified analysis was used for combined EA and AA. The attributable fraction (AF) was computed by the formula AF ⫽ f(R ⫺1)兾[1 ⫹ f(R ⫺1)], where f is the frequency of the risk factor in the population and R is the measure of relative risk (40); relative risk was obtained from a DCA using only seroconverters. The unpaired t test was used to assess the difference in transcription activities in the gene reporter assay. Results Polymorphisms and Linkage Disequilibrium Within the RANTES Gene. The human RANTES gene spans 8.8 kb on chromosome 17q11.2q12 and has the characteristic three exon–two intron organization of the CC chemokine family. Nonisotopic RNA cleavage assay and DNA sequencing were used to detect SNPs in the entire RANTES gene except a part of intron 2. Seven SNPs were identified (Fig. 1a): four in the promoter, including the previously described ⫺403G兾A and ⫺28C兾G (8), two in the first intron, and one in the 3⬘ untranslated region. The genotype of each SNP was determined for 4,168 individuals enrolled in five HIV-AIDS prospective cohorts plus 129 Han Chinese Asians. The ⫺403A promoter allele was common in all populations tested with allele frequencies ( f ) of 0.18 in EA, 0.43 in AA, and 0.36 in Asians. The ⫺28G promoter allele was infrequent in all groups except Asians ( f ⫽ 0.13). Two rare promoter SNPs, ⫺105C兾T and ⫺109T兾C, were located within the five IFN-simulated response elements between positions ⫺124 and ⫺97. Two SNPs were found in intron 1, In1.1T兾C and In1.2G兾A. In1.1C was frequent in all groups ( f ⫽ 0.14 in EA, 0.20 in AA, and 0.34 in Asians), whereas In1.2A was infrequent in AA ( f ⫽ 0.06) and absent in other ethnic groups. SNP 3⬘222T兾C, located in an Alu-related repeat region of 3⬘ untranslated region, had a frequency of 0.11 in EA, 0.07 in AA, and 0.24 in Asians. Three rare SNPs (105C兾T, 109C兾T, and In1.2G兾A) were not considered further. Strong linkage disequilibrium was observed between all of the RANTES SNPs, and in particular between the alleles at the four PNAS 兩 July 23, 2002 兩 vol. 99 兩 no. 15 兩 10003 MEDICAL SCIENCES AF088219). Primer sequences and PCR conditions are available from the authors. PCR products that revealed aberrant bands by NIRCA analysis were purified and sequenced. most polymorphic sites, ⫺403, ⫺28, In1.1, and 3⬘222 (0.96 ⬍ D⬘ ⬍ 1, P ⬍ 10⫺6). Five of the 16 possible RANTES haplotypes (R1–R5, Fig. 1a) for these four SNP sites accounted for more than 98% of EA and AA chromosomes (Fig. 1a). Within the common haplotypes, In1.1C occurs always on a haplotype containing ⫺403A, whereas ⫺28G and 3⬘222C always occur on the haplotype containing ⫺403A and In1.1C variant alleles, but SNPs ⫺28G and 3⬘222C never occur together. In both races, the majority of individuals carry the common haplotype R1 (57.0% and 77.4% in AA and EA, respectively). RANTES In1.1C Occurs in an Up-Regulating Intron 1 Element. Two Fig. 1. (a) The genomic structure of the RANTES gene on chromosome 17q11.2 and the haplotype structure and frequency of the RANTES variants. Exons are shown as open boxes, and intron sizes are indicated. Locations of nucleotide variants are indicated by arrows. Variants in the promoter region are numbered according to the transcription start site, and the SNP in the 3⬘ untranslated region is numbered from the first nucleotide of the 3⬘ untranslated region. Variants In1.1 and In1.2 are located at nucleotide positions 168923 and 170226, respectively, in GenBank sequence AF088219. (b) Competitive electrophoretic mobility-shift assay of DNA–nuclear protein binding at the In1.1T兾C site. Nuclear extracts from CD4⫹-enriched lymphocytes bound to the intron 1 fragment containing In1.1T (lanes A–D) or In1.1C (lanes E–H). Lanes A and E, no extract; lanes B and F, In1.1T and In1.1C probe, respectively, without competitors; lanes C and G, In1.1T and In1.1C probe, respectively, with a 100-fold excess of cold In1.1T probe as competitor; lanes D and H, In1.1T and In1.1C probe, respectively, with a 100-fold excess of cold In1.1C probe as a competitor. The arrowheads indicate the specific DNA–protein complexes associated with In1.1T兾C. (c) Luciferase activities of pBL3-promoter constructs containing RANTES intron 1 fragments with either In1.1T (pGL3-In1.1T) or In1.1C (pGL3-In1.1C). The pBL3-promoter vector served as 10004 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.142313799 promoter SNPs, ⫺28G and ⫺403A, have been reported to upregulate RANTES transcription (8, 21). We investigated the role of In1.1C by screening nuclear extracts for proteins that would bind alternative alleles and by quantifying their influence on RANTES gene transcription in luciferase gene expression constructs. Using an electrophoretic mobility-shift assay synthetic (36-bp) oligonucleotide probes containing In1.1C were incubated with nuclear extracts from human CD4⫹-enriched lymphocytes. Two distinctive complexes, I and II, were observed with probes containing In1.1T and In1.1C, respectively (Fig. 1b). Specificity of the binding was confirmed by cross-competition with unlabeled In1.1T or In1.1C probes: competitive binding of In1.1T, but not In1.1C, eliminated complex I, and correspondingly, competition with In1.1C, but not In1.1T, eliminated complex II. These results demonstrate that the In1.1T兾C alleles bind to different nuclear proteins or different forms of the same protein. To assess the role of the In1.I.T兾C alleles on gene transcription, a RANTES intron 1 fragment containing either In1.1T or In1.1C was inserted downstream of the luciferase (luc⫹) gene in a simian virus 40 promoter-containing construct and tested in Jurkat T cells. The construct containing In1.1C showed a 3-fold reduction in gene expression relative to the construct containing In1.1T (P ⬍ 0.001, Fig. 1c) demonstrating a down-regulating role for In1.1C on gene transcription. We next tested the transcriptional effects of combinations of the RANTES promoter (⫺403G兾C, ⫺28C兾G) and the putative intronic (In1.1T兾C) regulatory element alleles on gene transcription by placing the RANTES promoter and intron 1 fragments upstream and downstream, respectively, of the luc⫹ gene. The constructs containing the In1.1T intron 1 fragment up-regulated RANTES transcriptional activity for the three promoter-intron constructs (denoting the haplotype as the nucleotide at positions ⫺403, ⫺28, and In1.1) GCT (12-fold, P ⬍ 0.001), ACT (6-fold, P ⬍ 0.001), and AGT (3-fold, P ⬍ 0.001) relative to the promoter-alone constructs (positions ⫺403 and ⫺28) GC, AC, and AG, respectively (Fig. 1d). However, relative to the In1.1T allele, RANTES transcriptional activity was reduced by 2.5-to 3.4-fold in similar constructs containing the variant In1.1C allele: GCC (3-fold, P ⫽ 0.002), ACC (3.4-fold, P ⫽ 0.009), or AGC (2.5-fold, P ⬍ 0.001). These results demonstrate both that the intron 1 fragment is a strong regulatory a control. Bars indicate mean values with standard deviations; P values were determined by Student’s t test. Results are the mean of two experiments performed in triplicate. (d) Luciferase activities of the RANTES haplotype constructs containing the RANTES promoter (⫺403G兾A, ⫺28C兾G) and the intron 1 element (In1.1T兾C) allele combinations. The pGL3-basic vector constructs contain the promoter fragments only (hatched bar), promoter and intron 1 fragments carrying In1.1T (black bars), and promoter and intron 1 fragments carrying In1.1C (dotted bars). Haplotypes are denoted by the nucleotide at positions ⫺403, ⫺28, and In1.1, respectively. These constructs were also tested in the U937 monocyte cell and CCRF-SB B cell lines. The gene transcription activities were similar in the U937 cells but were much weaker in the CCRF-SB B cells, compared with those in Jurkat T cells (data not shown). The pBL3-basic vector (pB) served as a control. Note: haplotype AGT has not been observed in the population studied. An et al. Fig. 2. Comparison of frequency of genotypes carrying RANTES SNPs or the R4 haplotype between high-risk HIV-1-exposed but uninfected and HIV-1 seroconverters for EA (Left) and combined EA and AA (Right). Numbers above bars are the numbers of subjects in each group; * indicates a significant frequency difference (dominant model) by a two-sided Fisher’s exact test. (95% CI) P RH (95% CI) P RH (95% CI) P RH (95% CI) P RH (95% CI) P RH (95% CI) P CD4 ⬍ 200 AIDS-1987 Death CD4 ⬍ 200 AIDS-1987 Death CD4 ⬍ 200 AIDS-1987 Death 659 672 671 287 291 291 945 962 961 0.86 1.14 1.13 0.77 1.47 1.39 0.83 1.19 1.16 (0.69, 1.07) (0.9, 1.44) (0.88, 1.45) (0.51, 1.15) (0.82, 2.64) (0.63, 3.07) (0.69, 1.01) (0.96, 1.47) (0.92, 1.47) 0.17 0.27 0.33 0.20 0.19 0.41 0.07 0.11 0.21 0.93 1.24 1.20 N兾A N兾A N兾A 0.94 1.27 1.23 (0.58, 1.49) (0.78, 1.99) (0.72, 2.03) 0.76 0.36 0.48 (0.59, 1.5) (0.8, 2.03) (0.73, 2.06) 0.80 0.31 0.44 0.85 1.06 1.16 1.09 1.93 2.50 0.91 1.20 1.29 (0.67, 1.08) (0.83, 1.36) (0.89, 1.51) (0.74, 1.62) (1.18, 3.18) (1.28, 4.91) (0.74, 1.11) (0.96, 1.49) (1.01, 1.64) 0.19 0.65 0.27 0.66 0.009 0.008 0.34 0.11 0.04 0.85 0.96 1.10 1.14 1.52 1.89 0.88 1.02 1.16 (0.66, 1.1) (0.73, 1.27) (0.83, 1.47) (0.66, 1.96) (0.8, 2.88) (0.85, 4.21) (0.7, 1.11) (0.8, 1.32) (0.89, 1.52) 0.22 0.78 0.51 0.64 0.20 0.12 0.29 0.85 0.28 1.01 1.20 1.03 0.75 0.94 0.70 0.88 1.10 0.93 (0.71, 1.46) (0.82, 1.76) (0.67, 1.56) (0.51, 1.12) (0.58, 1.55) (0.35, 1.4) (0.67, 1.15) (0.81, 1.49) (0.64, 1.34) 0.94 0.35 0.91 0.16 0.82 0.31 0.35 0.55 0.70 4.64 2.10 1.99 1.01 1.70 2.30 1.18 1.74 2.18 (1.7, 12.63) (0.67, 6.6) (0.63, 6.29) (0.64, 1.58) (1.01, 2.87) (1.16, 4.54) (0.77, 1.8) (1.08, 2.81) (1.23, 3.88) 0.003 0.21 0.24 0.98 0.05 0.02 0.44 0.02 0.008 1.06 1.36 1.29 1.01 1.70 2.30 1.05 1.49 1.54 (0.69, 1.63) (0.88, 2.1) (0.8, 2.07) (0.64, 1.58) (1.01, 2.87) (1.16, 4.54) (0.77, 1.42) (1.08, 2.07) (1.06, 2.25) 0.79 0.16 0.29 0.98 0.05 0.02 0.78 0.02 0.02 CD4 ⬍ 200 AIDS-1987 Death CD4 ⬍ 200 AIDS-1987 Death CD4 ⬍ 200 AIDS-1987 Death 659 672 671 287 291 291 945 962 961 0.90 1.20 1.20 0.79 1.44 1.41 0.87 1.23 1.22 (0.72, 1.12) (0.96, 1.52) (0.94, 1.54) (0.53, 1.18) (0.81, 2.57) (0.64, 3.1) (0.72, 1.06) (1, 1.53) (0.96, 1.55) 0.34 0.12 0.15 0.26 0.22 0.39 0.17 0.05 0.10 1.19 1.73 1.51 N兾A N兾A N兾A 1.19 1.73 1.51 (0.75, 1.89) (1.09, 2.74) (0.9, 2.51) 0.46 0.02 0.12 (0.75, 1.89) (1.09, 2.74) (0.9, 2.51) 0.46 0.02 0.12 0.92 1.18 1.26 1.08 1.78 2.31 0.96 1.29 1.37 (0.73, 1.17) (0.92, 1.52) (0.97, 1.64) (0.73, 1.59) (1.09, 2.91) (1.19, 4.5) (0.79, 1.18) (1.03, 1.6) (1.07, 1.74) 0.51 0.18 0.09 0.71 0.02 0.01 0.70 0.02 0.01 0.86 0.99 1.12 1.06 1.46 1.85 0.89 1.05 1.18 (0.67, 1.11) (0.76, 1.31) (0.84, 1.49) (0.62, 1.8) (0.77, 2.76) (0.83, 4.11) (0.71, 1.13) (0.82, 1.35) (0.9, 1.54) 0.25 0.97 0.43 0.84 0.25 0.13 0.34 0.71 0.23 0.95 1.07 1.00 0.75 0.94 0.73 0.85 1.02 0.91 (0.66, 1.36) (0.73, 1.56) (0.65, 1.52) (0.51, 1.12) (0.57, 1.55) (0.37, 1.44) (0.65, 1.12) (0.75, 1.38) (0.63, 1.31) 0.78 0.74 0.98 0.16 0.81 0.37 0.24 0.91 0.61 4.77 2.23 2.20 1.02 1.57 2.07 1.18 1.66 2.10 (1.75, 12.98) (0.71, 6.98) (0.7, 6.91) (0.66, 1.59) (0.94, 2.63) (1.07, 4.03) (0.78, 1.8) (1.03, 2.67) (1.18, 3.75) 0.002 0.17 0.18 0.92 0.08 0.03 0.43 0.04 0.01 1.33 1.79 1.57 1.02 1.57 2.07 1.17 1.70 1.71 (0.87, 2.03) (1.17, 2.74) (0.98, 2.51) (0.66, 1.59) (0.94, 2.63) (1.07, 4.03) (0.86, 1.59) (1.22, 2.36) (1.18, 2.5) 0.18 0.008 0.06 0.92 0.08 0.03 0.32 0.002 0.005 CI, confidence interval; RH, relative hazard; N兾A, not available. *, Adjusted with covariates (see Materials and Methods). †, ⫺28G is equivalent to haplotype R5. ‡, 3⬘222C is virtually equivalent to haplotype R4 (99.7% of EA and all AA haplotypes carrying 3⬘222C are R4. §, Results for haplotypes R3 and R5 combined. PNAS 兩 July 23, 2002 兩 vol. 99 兩 no. 15 兩 10005 MEDICAL SCIENCES element and that In1.1C down-regulates RANTES transcriptional activity. By contrast, there was no difference in transcriptional activity between haplotypes GCT and ACT, or between GCC and ACC, demonstrating no obvious effect on transcription by ⫺403A. A moderate up-regulating role of ⫺28G was observed in the comparison of AGC to ACC (78% increase, P ⫽ 0.006) but not of AGT to ACT. These experiments demonstrate a strong down-regulation of RANTES transcription by In1.1C and a modest up-regulation transcription action of ⫺28G (Fig. 1d). RH Combined P AA (95% CI) Unadjusted EA RH Combined N AA Outcome Adjusted* EA R3 ⫹ R5§ R3 R2 3⬘222C (⬇R4)‡ ln1.1C ⫺28G (⫽R5)† ⫺403A Effect of RANTES SNPs and Haplotypes on HIV-1 Infection. We compared the frequencies of the RANTES variant alleles between high-risk HIV-1-exposed uninfected (HREU) individuals and HIV-1-infected seroconverters for EA and AA. The variant alleles, ⫺403A, In1.1C, and 3⬘222C each showed a diminished frequency in the HREU group [odds ratio (OR) ⫽ 1.45, 1.41, and 1.53; P ⫽ 0.06, 0.11, and 0.08, respectively; Fig. 2, Table 2, which is published as supporting information on the PNAS web site, www.pnas.org], suggesting an association with increased susceptibility to HIV-1 infection. For combined ethnic groups the ⫺403A, In1.1C, and 3⬘222C variant alleles showed significant association with increased risk of HIV-1 infection (OR ⫽ 1.44, 1.41, and 1.51; P ⫽ 0.02, 0.04, and 0.04, respectively, Fig. 2, Table 2). The haplotypes R2–R5 carrying the variant alleles were all nonsignificantly elevated in the HIV-1-infected study participants. The positive linkage disequilibrium between the four variants precludes implication of a single associated variant SNP with HIV-1 infection, but lower RANTES levels specified by In1.1C poses it is an attractive candidate. Effect of RANTES SNPs and Haplotypes on AIDS Progression. The influences of individual RANTES SNP alleles and multisite haplotypes on AIDS progression were evaluated for dominant, codominant, and recessive genetic models among EA (n ⫽ 673) and AA (n ⫽ 291) HIV-1 seroconverters by using a Cox proportional hazards model (41). Three AIDS endpoints reflecting advancing morbidity were evaluated: (i) CD4 ⬍ 200 cells per mm3, (ii) AIDS-1987, as defined by the Centers for Disease Control and Prevention (38); and (iii) AIDS-related death. Analyses with and without adjustments for the influence of described AIDSrestriction gene covariates are listed in Table 1 and illustrated in Fig. 3. A strong SNP genotypic association in the Cox model analysis involved AA seroconverters carrying one or two copies of the In1.1C. This allele conferred more rapid progression to AIDS-1987 [relative hazard (RH) ⫽ 1.9, P ⫽ 0.009] or death [RH ⫽ 2.5, P ⫽ 0.008 and RH ⫽ 4.6, P ⫽ 0.002 for dominant and recessive (not shown) models, respectively] (Fig. 3b, Table 1). The EA cohorts did not show the In1.1C influence when tested separately (Fig. 3a, Table 1; but see below). No significant associations for rapid AIDS progression were observed for ⫺403A, ⫺28G, or 3⬘222C-bearing An et al. Table 1. Survival analysis of RANTES allele-bearing genotypes and haplotype-bearing genotypes for association with progression to AIDS endpoints by using the Cox proportional hazard model (dominant model) genotypes, in analyses adjusted for known AIDS restriction genes (Table 1). To clarify the association of the individual SNPs, we next analyzed the influence of composite SNP haplotypes R1–R5 (Fig. 1a) on the rate of progression to AIDS. Evaluation of haplotype association with the AIDS progression also implicated the In1.1T兾C site as responsible for the observed allele association (Table 1, Fig. 3 c–f ). The R3 haplotype (which contains In1.1C) showed significant AIDS accelerating influence in AA (RH ⫽ 1.7, P ⫽ 0.05, for AIDS-1987 and RH ⫽ 2.3, P ⫽ 0.02 for AIDS-related death), in EA (RH ⫽ 4.6, P ⫽ 0.003 for CD4 ⬍ 200), and in combined cohorts (RH ⫽ 1.7, P ⫽ 0.02 for AIDS-1987; RH ⫽ 2.2; P ⫽ 0.008 for AIDS-related death). A weaker association for the In1.1C-containing R4 haplotype carriers with rapid AIDS progression was evident in AA (Fig. 3 d and f ) but not in EA (c and e). The strong AIDS accelerating influence of In1.1C in AA but not in EA apparently derives from the difference in haplotype frequencies between the two groups: the strongly accelerating haplotype R3 represents 64% of In1.1C-carrying haplotypes for AA, but only 3% of In1.1C haplotypes for EA. The EA R5 haplotype, (which is infrequent in AA), carries both In1.1C and ⫺28G, regulating alleles that influence transcription in opposite directions (Fig. 1d). The counteracting influence of In1.1C and ⫺28G is illustrated epide- Fig. 3. Kaplan–Meier survival curves for progression to AIDS-1987 (a–d), CD4 ⬍ 200, AIDS related death comparing the influence of the RANTES In1.1C allele, and In1.1C-carrying haplotypes on progression to AIDS by HIV-1-infected EA (Left) and AA (Right). Cox model RH and Kaplan–Meier Wilcoxon P values comparing each variant genotype or haplotype to the no-In1.1C group are shown for each factor. (a and b) Rate of progression to AIDS-1987 based on In1.1C genotype: black, no In1.1C; blue, one copy; and purple, two copies of In1.1C for EA (a) and AA (b). (c–f ) Rate of progression to AIDS partitioned according to In1.1Ccontaining haplotypes: black, no In1.1C; red, R3; green, R4; and orange, R5. (c and e) Progression of EA to AIDS-1987 (c) and CD4 ⬍ 200 (e). (d and f ) Progression of AA to AIDS-1987 (d) and AIDS-related death ( f). Subjects carrying two different In1.1C haplotypes (two EA and six AA) are omitted. 10006 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.142313799 Fig. 4. Defined disease category analysis of RANTES In1.1C and the In1.1Ccarrying R3 haplotype comparing slow versus rapid progressors to AIDS-1987 and AIDS-related death. Bars show frequencies of homozygotes (black) and heterozygotes (gray) for In1.1C or the R3 haplotype in the slow and rapid groups for AA (Left), EA (Center), and combined ethnic groups (Right). Number of subjects considered and significance for a codominant model by Fisher’s exact test are indicated (*, P ⬍ 0.05; **; P ⬍ 0.01; ***, P ⬍ 0.001; ****, P ⬍ 0.0001). miologically in an AIDS survival analysis of EA that compares R5 to R3 haplotypes (Fig. 3 c and e). R5 haplotypes that retain the offsetting In1.1C and ⫺28G alleles only slightly accelerate AIDS, whereas R3 haplotype with In1.1C and the wild-type ⫺28C allele show a much stronger influence toward rapid AIDS progression. The detrimental effects of In1.1C on HIV-1-infected individuals were also apparent in categorical analyses of slow versus rapid progressors to AIDS. This approach allows the inclusion of seroprevalent individuals (those whose seroconversion date is unknown because they were HIV-1 antibody-positive at the time of study enrollment) in the slow兾nonprogressor category (26). An elevation in In1.1C allele frequency was observed among the rapid progressor groups for AIDS-1987 and AIDS-death in both dominant and codominant (Fig. 4) models. The association was highly significant in AA for AIDS-1987 and death (OR ⫽ 2.54, P ⫽ 0.002 and OR ⫽ 3.53, P ⫽ 0.0009, respectively, Fisher’s exact test for the dominant model). An association for the R3 haplotype was also apparent (Fig. 4) and was highly significant both for AA (OR ⫽ 2.47, 3.42; P ⫽ 0.005, 0.002) and for combined EA and AA (OR ⫽ 2.45, 3.20; P ⫽ 0.002, 0.0004), respectively, for AIDS-1987 and AIDS-related death, dominant model. The Mantel–Haenszel trend test showed stronger associations for the codominant than for the dominant model for almost all cases, suggesting a gene dose effect. These survival and categorical results strongly implicate a role for the intronic regulatory allele In1.1C polymorphism in promoting the rate of AIDS progression. A quantitative estimate for the In1.1C-bearing haplotypes on AIDS progression in the study population can be determined by computing the attributable fraction, a parameter that combines the strength of the epidemiological influence (relative risk) and the frequency of the protective genotype (40). To estimate the attributable fraction of In1.1C for progression to AIDS-related death, relative risk was obtained from a categorical analysis with AA seroconverters. The calculated relative risk (2.61, confidence interval 1.38–4.95) combined with the In1.1C allele frequency (36%) in AA indicate that 37% (confidence interval: 12–59%) of rapid (within 7.5 years) progression to clinical AIDS can be attributed to the detrimental effect of the RANTES In1.1C allele. Discussion We describe here the validation of seven SNPs within the R ANTES gene and analyze the four common variants An et al. (⫺403G兾A, ⫺28C兾G, In1.1T兾C, and 3⬘222T兾C) for genetic association with HIV-1 transmission, AIDS progression, and gene transcription. In1.1T兾C is shown to differentially bind nuclear proteins (Fig. 1b) and have powerful regulatory activity on gene expression (Fig. 1 c and d). The In1.1C allele results in reduced RANTES transcription and is associated with more rapid progression to AIDS (Figs. 3 and 4, Table 1). The rapid disease course of an estimated 37% of AA AIDS patients whose disease progresses within 7.5 years after HIV-1 infection can be attributed to their RANTES–In1.1C-bearing genotype. Because 36% of AA carry the In1.1C allele, it is likely that In1.1C may have a significant impact on the AIDS epidemic in sub-Saharan Africa. The results suggested that In1.1C also increases susceptibility to HIV-1 infection and accelerates progression to AIDS in EA, but these effects are confounded and partially quenched by the nearly total positive linkage disequilibrium between In1.1C and the other RANTES variants studied. Three common RANTES variants and their inclusive haplotype R4 were associated with increased risk of infection (Fig. 2). That association plus the demonstrated downregulation of RANTES transcription by In1.1C (Fig. 1 b–d) implicate In1.1C as regulating in the HIV-1 infection process. In1.1C shows little influence on AIDS progression in EA (Fig. 3a); however, the R3 haplotype, carrying In1.1C, accelerates AIDS progression in both EA and AA (Table 1, Fig. 3 c–f ). In EA this haplotype is rare, and its effect is diluted by the effects of the R5 haplotype, which carries the ⫺28G variant that we show to counter the down-regulating effect of In1.1C, and the R4 haplotype, in which the effect of In1.1C appears to be countered by unknown factors. The haplotype analysis, along with the demonstrated downregulating role of the In1.1C allele on RANTES transcription, strongly implicates In1.1C as a highly associated genetic risk factor for progression to AIDS and AIDS-related death. The R2 haplotype (carrying ⫺403A but lacking In1.1C) has no accelerating effect (RH ⬍ 1.15, P ⬎ 0.4, Table 1), arguing against ⫺403A as a causative factor (9, 22). Because CCR5 ligands have been shown to competitively bind to and reduce surface expression of CCR5, down-regulation of RANTES may increase the number of binding targets for HIV-1, thus promoting HIV-1 replication (1, 10–20). The susceptible influence of In1.1C is likely caused by an in vivo reduction in RANTES–CCR5 binding, which may increase HIV-1 replication and rate of progression to AIDS. The association of the downregulating variant of RANTES with accelerated progression to AIDS is clinically relevant because modified forms of RANTES are under active investigation as HIV-1-antiviral agents (41). The in vitro and in vivo evidence in this study argues for a beneficial role of high RANTES levels in limiting progression to AIDS in HIV1-infected individuals. Our results would predict that RANTES levels would be diminished in individuals with an In1.1C-bearing genotype. The consistent genetic association shown here between susceptibility to HIV-1 transmission and disease progression by a physiologically plausible down-regulating In1.1C variant clarifies the underlying complex mechanisms of RANTES genetic influence on HIV-1 pathogenesis and provides valuable information in evaluating RANTES for therapeutic purposes. 1. Berger, E. A., Murphy, P. M. & Farber, J. M. (1999) Annu. Rev. Immunol. 17, 657–700. 2. Dean, M., Carrington, M., Winkler, C., Huttley, G. A., Smith, M. W., Allikmets, R., Goedert, J. J., Buchbinder, S. P., Vittinghoff, E., Gomperts, E., et al. (1996) Science 273, 1856–1862. 3. Smith, M. W., Dean, M., Carrington, M., Winkler, C., Huttley, G. A., Lomb, D. A., Goedert, J. J., O’Brien, T. R., Jacobson, L. P., Kaslow, R., et al. (1997) Science 277, 959–965. 4. Winkler, C., Modi, W., Smith, M. W., Nelson, G. W., Wu, X., Carrington, M., Dean, M., Honjo, T., Tashiro, K., Yabe, D., et al. (1998) Science 279, 389–393. 5. Martin, M. P., Dean, M., Smith, M. W., Winkler, C., Gerrard, B., Michael, N. L., Lee, B., Doms, R. W., Margolick, J., Buchbinder, S., et al. (1998) Science 282, 1907–1911. 6. McDermott, D. H., Zimmerman, P. A., Guignard, F., Kleeberger, C. A., Leitman, S. F. & Murphy, P. M. (1998) Lancet 352, 866–870. 7. An, P., Martin, M. P., Nelson, G. W., Carrington, M., Smith, M. W., Gong, K., Vlahov, D., O’Brien, S. J. & Winkler, C. A. (2000) AIDS 14, 2117–2122. 8. Liu, H., Chao, D., Nakayama, E. E., Taguchi, H., Goto, M., Xin, X., Takamatsu, J. K., Saito, H., Ishikawa, Y., Akaza, T., et al. (1999) Proc. Natl. Acad. Sci. USA 96, 4581–4585. 9. McDermott, D., Beecroft, M., Kleeberger, C., Al-Sharif, F., Ollier, W., Zimmerman, P., Boatin, B., Leitman, S., Detels, R., Hajeer, A. & Murphy, P. (2000) AIDS 14, 2671–2678. 10. Cocchi, F., DeVico, A. L., Garzino-Demo, A., Arya, S. K., Gallo, R. C. & Lusso, P. (1995) Science 270, 1811–1815. 11. Trkola, A., Paxton, W. A., Monard, S. P., Hoxie, J. A., Siani, M. A., Thompson, D. A., Wu, L., Mackay, C. R., Horuk, R. & Moore, J. P. (1998) J. Virol. 72, 396–404. 12. Paxton, W. A., Martin, S. R., Tse, D., O’Brien, T. R., Skurnick, J., VanDevanter, N. L., Padian, N., Braun, J. F., Kotler, D. P., Wolinsky, S. M. & Koup, R. A. (1996) Nat. Med. 2, 412–417. 13. Schwartz, D. H., Castillo, R. C., Arango-Jaramillo, S., Sharma, U. K., Song, H. F. & Sridharan, G. (1997) J. Infect. Dis. 176, 1168–1174. 14. Zagury, D., Lachgar, A., Chams, V., Fall, L. S., Bernard, J., Zagury, J. F., Bizzini, B., Gringeri, A., Santagostino, E., Rappaport, J., et al. (1998) Proc. Natl. Acad. Sci. USA 95, 3857–3861. 15. Aukrust, P., Muller, F. & Froland, S. S. (1998) J. Infect. Dis. 177, 1091–1096. 16. Saha, K., Bentsman, G., Chess, L. & Volsky, D. J. (1998) J. Virol. 72, 876–881. 17. Than, S., Tumaliuan, J. A., Tamma, S., Morales, M., Bakshi, S. & Pahwa, S. (1999) AIDS Res. Hum. Retroviruses 15, 1279–1285. 18. Paxton, W. A., Neumann, A. U., Kang, S., Deutch, L., Brown, R. C., Koup, R. A. & Wolinsky, S. M. (2001) J. Infect. Dis. 183, 1678–1681. 19. Zanussi, S., D’Andrea, M., Simonelli, C., Tirelli, U. & De Paoli, P. (1996) AIDS 10, 1431–1432. 20. Gallo, R. C., Garzino-Demo, A. & DeVico, A. L. (1999) J. Clin. Immunol. 19, 293–299. 21. Nickel, R. G., Casolaro, V., Wahn, U., Beyer, K., Barnes, K. C., Plunkett, B. S., Freidhoff, L. R., Sengler, C., Plitt, J. R., Schleimer, R. P., et al. (2000) J. Immunol. 164, 1612–1616. 22. Gonzalez, E., Dhanda, R., Bamshad, M., Mummidi, S., Geevarghese, R., Catano, G., Anderson, S. A., Walter, E. A., Stephan, K. T., Hammer, M. F., et al. (2001) Proc. Natl. Acad. Sci. USA 98, 5199–5204. 23. Nelson, P. J., Kim, H. T., Manning, W. C., Goralski, T. J. & Krensky, A. M. (1993) J. Immunol. 151, 2601–2612. 24. Schall, T. J., Jongstra, J., Dyer, B. J., Jorgensen, J., Clayberger, C., Davis, M. M. & Krensky, A. M. (1988) J. Immunol. 141, 1018–1025. 25. O’Brien, S. J. & Moore, J. P. (2000) Immunol. Rev. 177, 99–111. 26. O’Brien, S. J., Nelson, G. W., Winkler, C. A. & Smith, M. W. (2000) Annu. Rev. Genet. 34, 563–591. 27. Vlahov, D., Anthony, J. C., Munoz, A., Margolick, J., Nelson, K. E., Celentano, D. D., Solomon, L. & Polk, B. F. (1991) in Longitudinal Studies of HIV Infection Among Intravenous Drug Users, eds. Hartsock, P. I. & Genser, S. G. (National Institute on Drug Abuse, Rockville, MD), pp. 75–100. 28. Hilgartner, M. W., Donfield, S. M., Willoughby, A., Contant, C. F., Jr., Evatt, B. L., Gomperts, E. D., Hoots, W. K., Jason, J., Loveland, K. A., McKinlay, S. M., et al. (1993) Am. J. Pediatr. Hematol. Oncol. 15, 208–218. 29. Phair, J., Jacobson, L., Detels, R., Rinaldo, C., Saah, A., Schrager, L. & Munoz, A. (1992) J. Acquired Immune Defic. Syndr. 5, 490–496. 30. Goedert, J. J., Kessler, C. M., Aledort, L. M., Biggar, R. J., Andes, W. A., White, G. C., II, Drummond, J. E., Vaidya, K., Mann, D. L., Eyster, M. E., et al. (1989) N. Engl. J. Med. 321, 1141–1148. 31. Buchbinder, S. P., Katz, M. H., Hessol, N. A., O’Malley, P. M. & Holmberg, S. D. (1994) AIDS 8, 1123–1128. 32. Vlahov, D. & Polk, B. F. (1988) Psychopharmacol. Bull. 24, 325–329. 33. Detels, R., Liu, Z., Hennessey, K., Kan, J., Visscher, B. R., Taylor, J. M., Hoover, D. R., Rinaldo, C. R., Jr., Phair, J. P., Saah, A. J., et al. (1994) J. Acquired Immune Defic. Syndr. 7, 1263–1269. 34. Salkowitz, J. R., Purvis, S. F., Meyerson, H., Zimmerman, P., O’Brien, T. R., Aledort, L., Eyster, M. E., Hilgartner, M., Kessler, C., Konkle, B. A., et al. (2001) Clin. Immunol. 98, 200–211. 35. Goldrick, M. M., Kimball, G. R., Liu, Q., Martin, L. A., Sommer, S. S. & Tseng, J. Y. (1996) BioTechniques 21, 106–112. 36. Wang, L. H., Kirken, R. A., Erwin, R. A., Yu, C. R. & Farrar, W. L. (1999) J. Immunol. 162, 3897–3904. 37. Long, J. C., Williams, R. C. & Urbanek, M. (1995) Am. J. Hum. Genet. 56, 799–810. 38. Centers for Disease Control and Prevention (1987) Morbid. Mortal. Wkly. Rep. 36, Suppl. 1, 1S–15S. 39. Gao, X., Nelson, G. W., Karacki, P., Martin, M. P., Phair, J., Kaslow, R., Goedert, J. J., Buchbinder, S., Hoots, K., Vlahov, D., et al. (2001) N. Eng. J. Med. 344, 1668–1675. 40. Adams, M. J., Jr., Khoury, M. J. & James, L. M. (1989) J. Clin. Epidemiol. 42, 659–662. 41. Simmons, G., Clapham, P. R., Picard, L., Offord, R. E., Rosenkilde, M. M., Schwartz, T. W., Buser, R., Wells, T. N. C. & Proudfoot, A. E. (1997) Science 276, 276–279. An et al. PNAS 兩 July 23, 2002 兩 vol. 99 兩 no. 15 兩 10007 MEDICAL SCIENCES We gratefully acknowledge the participants, their families, and their clinicians who participated in the AIDS Link to the Intravenous Drug Experience, Multicenter AIDS Cohort Study, Hemophilia Growth and Development Study, Multicenter Hemophiliac Cohort Study, and San Francisco City Clinic Study cohort studies. We thank Dr. Tatsuo Shioda for kind gifts of RANTES promoter constructs, Kui Gong and Beth Binns-Roemer for excellent technical assistance, Dr. Xiaoyi Yang for helpful discussions, and Dr. Jeffery Kopp for critically reading the manuscript. This study is supported by federal funds from the National Cancer Institute, National Institutes of Health, under Contract NO1CO-12400.