Phytochemistry and Biological Properties of Glabridin

2.1 Occurrence and biosynthesis

Natural isoflavonoids are mainly distributed in leguminous plants (Fabaceae) and play ecophysiological roles as defensive agents against other organisms ranging from animals to microbes (phytoalexins). Some isoflavonoids are considered symbiotic signals to rhizobial bacteria to form nitrogen-fixing root nodules [7, 8]. The isoflavans (syn. deoxyisoflavanones) make up a unique subclass considering that the heterocycle ring C does not contain a double bond between carbon 2 and 3, nor a carbonyl group (Figure 2). Therefore, no conjugated double bonds exist between the rings A and B. Isoflavonoids possessing pterocarpan and isoflavan skeletons are the most frequently found phytoalexins of leguminous plants [9, 10].

The biosynthesis of these two subclasses of isoflavonoids is considered to be closely related, because most, if not all, isoflavans bear a hydroxyl group at C-2′ in the B-ring, representing the functional group which originally participated in the ether linkage to C-4 to produce the pterocarpan skeleton. More precisely, as demonstrated for medicarpin and vestitol, isoflavan derivatives are the end products of pterocarpan reductase enzymes (PTR) [11] (Figure 2A). To date, the specific biosynthetic pathway that leads to Glab and its precursor remains unknown. However, PTR may also be present in the genus Glycyrrhiza and, therefore, the bioprecursor of Glab could be a pterocarpan. Based on its characteristic features, shinpterocarpin, which has also been described in licorice [12], could be considered as the immediate pterocarpan precursor of Glab. Additionally, a pterocarpan specific prenyltransferase (GmG4DT) has been identified in Soybean (Glycine max, also Fabaceae) [13]. Both shinpterocarpin and glabridin share a prenyl function linked to C-8 of the A ring, suggesting that shinpterocarpin could stem from desmethylmedicarpin and be formed through GmG4DT prenyltransferase activity (Figure 2B and 2C).

Interestingly, Glab has been demonstrated to be localized only in the cork layer and the decayed part of the roots of G. glabra [14], representing 0.08 to 0.35% of the roots’ dry weight [15]. In 1978, Glab was identified as a species-specific secondary metabolite of G. glabra, and this designation was later confirmed by Kondo et al. through the comparative analysis of genetic and chemical markers from the three principal licorice species: G. uralensis Fisch. ex DC., G. glabra L., and G. inflata Batalin [16].

Considering the established role of Glab as a phytoalexin, its quantity in the investigated material may reflect the environmental stress to which producing parts of G. glabra were exposed. Accordingly, Glab content may fluctuate between samples depending on the type of cultivation and the geographical area from which the roots were harvested.

2.2 Methods for G. glabra extraction and enrichment in glabridin

Glab is generally not consumed as a single chemical entity but rather as constituent of G. glabra crude extract. This reveals the importance of method development dedicated to the extraction of G. glabra roots and the enrichment in Glab. Most licorice extracts are produced for commercialization in the cosmetic, food or DSs markets, hence requiring the use of non-toxic solvents and environmentally safe methodologies. Extraction with ethanol (EtOH) [17], aqueous EtOH [18], supercritical fluid CO₂ [19], and more recently, imidazolium based ionic liquids (ILs) [20] are preferred for the production of G. glabra extracts enriched in Glab. Other enrichment methods resort to the use of diverse macroporous resins [21, 22].

A licorice flavonoid oil (LFO) has been prepared by extracting a G. glabra ethanolic extract (EtOH 95%, 5 volumes/1volume of roots) with medium-chain triglycerides (MCT, C8:C10 = 99:1) [17]. Accordingly, LFO represents the hydrophobic polyphenol-rich fraction of licorice ethanol extract. Standardized to a Glab concentration of 1% (w/w), the oily LFO product is commercialized under the name Glavonoid®, which has been recently accepted by the European Food Safety Authority as a Novel Food Ingredient [1].

A supercritical fluid extraction method of G. glabra roots using CO₂ and co-solvents such as MeOH or EtOH has been developed by Cho et al. [19]. The optimum conditions were found to be an extraction over 3 hours at 80 °C, 420 bar, and with 25% MeOH as co-solvent (which was demonstrated to be more effective than EtOH for the enrichment of Glab). The final Glab concentration in the supercritical CO₂ extract was not indicated in this study. Another supercritical fluid extract (LSC) was prepared by Ahn et al. [23] for pharmacological studies. Briefly, 154 g of roots were extracted during 1 hour, at 300 bar and 40 °C with a CO₂ flow rate of 150 g/min. The extraction yield was defined as 3.57% w/w, and one gram of LSC extract contained 45.12 ± 0.14 mg of Glab (4.5% w/w).

GutGuard^™ is a G. glabra extract developed by Natural Remedies, Pvt., Ltd., in Bangalore, is reported to be standardized in Glab (≥ 3.5% w/w), glabrol, and caffeic acid derivatives, and has a total flavonoid content of ≥ 10.5% w/w. However, the methods of extraction and further preparation of GutGuard have not been disclosed [24, 25].

Tian et al. [18] have analyzed the parameters enabling a concomitant extraction and enrichment of both glycyrrhizic acid and Glab. Different solvents (EtOH, MeOH, dichloromethane) at various temperatures and varying durations were tested and led to the conclusion that a mixture of EtOH:H₂O (30:70 v/v) and a dipping time of 60 min at 50 °C represented the optimum conditions. The same authors also compared the capacity of various solvents (Chloroform (CHCl₃), n-hexanes, H₂O, H₂O/MeOH and MeOH) to extract glycyrrhizic acid, Glab and liquiritin simultaneously. Pure MeOH turned out to be the best extraction solvent when used in a ratio of 1:70 (weight of plant material [g] per volume of solvent [mL]) during 120 min [26]. Nevertheless, considering general safety concerns of residual MeOH in extracts, extraction with EtOH should be preferred for commercial products.

ILs are attractive “green” and functional solvents in the field of organic synthesis and in various separation and extraction processes of organic acids or phenolic compounds. Li et al. [20] have developed specific ILs in order to selectively extract Glab from its natural matrix and suggested that this type of “green” extraction could be used as a first efficient step for the purification of Glab. A series of 1-alkyl-3-methylimidazolium (ILs) with different anions and alkyl chain lengths of cations were selected for this purpose.

A solid phase extraction (SPE) method using Molecular Imprinted Polymers (MIPs) has been proposed as a first enrichment step for the analysis of Glab content in different types of matrices [21]. For this purpose, the crude extracts of G. glabra or other herbal mixtures were washed from the MIPs using a mixture of MeOH:H₂O (60:40 v/v). In the second step, the Glab retained on the polymer was eluted with acetonitrile (MeCN) containing 0.5% of acid trifluoroacetic (TFA). MIPs were claimed to be selective for conducting targeted analyte isolation. However, the reported Glab recovery was only 42 to 56%, and Glab was also found to be present in the washing fraction. Subsequently, Xu et al. [22] have developed an enrichment method by studying the adsorption and desorption properties of Glab on different macroporous resins including HPD100, HPD300, HPD800, NKA and H103. The content of Glab in the final enriched fraction, after using HPD100, was ~150-fold higher than in the starting crude extract. One major drawback of these enrichment techniques is the necessity to employ a solid phase interaction step, which is known to lead to potential irreversible adsorption. For example, the recovery of Glab after passage through HPD100 was 80%, indicating that 20% of Glab remains on the resin. On the other hand, the cited SPE techniques, using MIPs or macroporous resins, are well suited for the enrichment and subsequent isolation of Glab.

2.3 Isolation from its natural source/purification processes

Most of the techniques reported for the isolation of Glab require a minimum of three chromatographic steps, including at least one step of liquid/liquid (L/L) partition and two steps of silica gel chromatography with the frequent use of benzene/ether as final eluent [27, 32] (Figure 3). Considering the chemical complexity of G. Glabra roots, it is understandable that purifying this prenylated isoflavan in less than three to four chromatographic steps remains a challenging task. In addition to silica gel, the second type of stationary phase which has been frequently used was polyamide, although both materials are known for unwanted properties such as irreversible adsorption and chemical modification through oxidation. In summary, Saito et al. [29] were able to isolate Glab after two steps of silica gel chromatography and one final purification on polyamide using MeOH as eluent. After L/L partition of the crude extract, Vaya et al. [28] isolated Glab from the organic layer, following one step of aluminum oxide chromatography and two steps of silica gel flash chromatography. Like Mitscher et al. [31], Cui et al. [30] have reported a procedure involving only two steps of silica gel chromatography from the CHCl₃ fraction obtained by L/L partition, and one final step of crystallization. A method that combined SPE, to obtain an enriched fraction, and preparative HPLC has been developed by Jiao et al. [33]. Glab was also isolated in our laboratory, after three steps of Medium Pressure Liquid Chromatography (MPLC) on HW-40F, Sephadex LH-20 and a final purification step on a cyano-derivatized silica adsorbent (LiChroCN).

Numerous methods for the isolation of Glab have been patented. Most of them comprised four steps including two chromatographic steps in order to obtain a reported pure Glab. These patented methods follow the same processes as those described above. Most require the use of a polyamide column during the purification process [34, 38]. The literature survey also revealed that the purity of Glab was determined exclusively by HPLC-UV [28, 30, 39], and that with the exception of some patents, the purity of isolated Glab and/or Glab used for bioassays remained undisclosed.

Because all reports involve multiple chromatographic steps to obtain pure Glab from its natural source, it can be concluded that the purification processes remain complicated, time-consuming, and require large amounts of volatile organic solvents. Additionally, most procedures are hardly suited for large scale production of Glab, and experimental details provided in the reports are frequently insufficient to ensure reproducibility.

2.4 Structural characterization

Naturally occurring Glab was isolated and characterized for the first time in 1976 by Saito et al [29]. The authors determined the configuration of carbon C-3 to be R by optical rotary dispersion according to the positive Cotton Effect observed between 260 and 300 nm.

To date, the parameters required for an exhaustive chemical characterization of Glab are spread throughout the literature. Nuclear Magnetic Resonance (NMR) data of Glab were mostly acquired in deuterated chloroform (CDCl₃) [27] and dimethyl sulfoxide (DMSO-d₆), at 300 or 400 MHz [40]. Both solvents offer good signal dispersion, and spectra acquired in DMSO-d₆ are essentially free of resonances overlap when using 300–600 MHz instruments. The ¹H NMR spectra of Glab acquired in these solvents are not identical (Table 1, Appendices A and B), especially in the aromatic region, where the signals of protons H-6 and H-5′ are overlapping in CDCl₃. In some articles, the proton assignments have been inaccurate, particularly with regard to the aromatic protons which have been reported as signal ranges (e.g., “6.37–6.41 (2H, m)”) [28, 30].

An X-ray analysis of Glab, crystallized from MeOH, was published recently [41]. The crystalline sample obtained in MeOH was defined as orthorhombic P2₁2₁2₁ (Z = 4). The absolute configuration of Glab was assumed to be the same as deduced from previous studies. A circular dichroism (CD) spectrum was recorded by us from an enantio-pure crystalline sample and found to be in accordance with data acquired by Kim et al. [42] (Figure 4). These authors had also determined the absolute configuration of synthetic Glab enantiomers isolated by chiral chromatography using CD polarimetry. The same authors assigned the coupling constants of H-3 which was determined to be a dddd resonance with J_{H-2eq, 3} = 3.7 Hz, J_H-2ax,3 = 10.5 Hz, J_H3,4ax = 10.5 Hz, and J_H3,4eq = 5.5 Hz) after a simulation of the ¹H NMR spectrum acquired in CDCl₃ and use of the ACDLABS HNMR viewer program. The 1D ¹H and ¹³C NMR data of Glab in DMSO-d₆ and CDCl₃ at 600 MHz and 225 MHz have also been recorded in our lab and led to unequivocal assignments for all proton and carbon resonances (Table 1, Figure 4, Appendices A and B).

There is a demand for enantio-pure Glab, both as a chemical reference standard for quality control and identification of G. glabra materials and for biological/pharmacological investigations. As structurally related isoflavans are required for structure activity relationship (SAR) studies, methods for chemical synthesis of Glab and other structurally related isoflavans have been developed. This topic will be addressed in the following.

2.5 Chemical synthesis of glabridin and isoflavan derivatives

Frequently, chemical synthesis is considered an efficient way to obtain quickly a large amount of pure reference material. However, in the case of Glab containing a stereogenic center on carbon 3, achiral chemical synthesis will inevitably lead to the production of an enantiomeric mixture. In 2007, Yoo and Nahm have proposed a rapid and efficient synthesis of racemic Glab in four steps and with 85% yield [43, 44]. Two years later, Jirawattanapong et al. [40] published an article on the synthesis of Glab derivatives (3″, 4″-dihydro- and 2′, 4′-diesters derivatives) starting from isolated Glab. A patent has been issued for the synthesis of Glab and other isoflavans derivatives as PPAR-γ ligands [45]. Optical resolution of these synthetic isoflavans was achieved through chiral chromatography using optically active polysaccharides and amino acids [46]. Pure enantiomers from the synthetic Glab were also isolated by chiral chromatography (Table 2), and their absolute configuration was determined by CD [42]. In summary, while synthetic methods for enantiopure Glab have been described, they involve relatively elaborate protocols due to the need for chiral separation and, therefore, retain the preparative chromatographic steps otherwise required for ab initio purification of Glab from its natural source.

2.6 Glabridin stability: prodrugs and formulation

Ao et al. [47] have evaluated various factors influencing Glab stability, such as temperature, pH, light exposure, and humidity. While Glab was found to be unstable under basic conditions (pH > 7), it remained stable at neutral pH. Glab was stable between 4 and 60 °C, and started to degrade only at temperatures above 60 °C. Moreover, degradation was shown to occur at room temperature under natural light. It was finally concluded that, for long term storage, Glab has to be kept in a dry, dark, and low oxygen environment. The structural identity of degradation products remains yet unknown.

In order to enhance Glab solubility in aqueous media, but also to increase its stability in formulations, two different techniques have been developed. One approach consists of synthetizing 2′,4′-diester or -diether derivatives, the second utilizes a physical protection of Glab through encapsulation. Ether or ester derivatives of Glab can be considered as pro-drugs, especially when taking into account that 2′,4′-dimethoxyglabridin undergoes demethylation in vivo, as indicated by Rosenblat et al. [48]. Glabridin 2′,4′-O-di-β-D-glucopyranoside was synthetized as UV_B absorber for sunscreen formulation [49], as tyrosinase inhibitor for a skin whitening formulation [50], and as anti-irritant [51]. The 2′,4′-dihydroxyl functions can also be protected through an etherification with an alkyl residue [52], or by esterification with undecylenic acid. The resulting hydrophobic derivatives have been also claimed to increase the skin adsorption of Glab [53, 54].

A methodology for Glab encapsulation in nanoemulsions has been developed by Hsieh et al. [55]. Such nanoemulsions might have the potential to be readily absorbed by the skin, and easy to sterilize by filtration. The optimum conditions for preparing Glab nanoemulsions were found to consist in an emulsifier (Tween 80 + Span 80) concentration of 5.3%, a caprylic triglycerides concentration of 3.65%, with a homogenization pressure of 129 MPa. Other methodologies of encapsulation include the development of a chitosan nano-complex, [56] and the development of microsponges formed by emulsion solvent evaporation [57]. Both methods have been demonstrated to enhance Glab stability in topical formulations. In the chitosan nano-complex, the hydroxyl functions of encapsulated Glab were proposed to interact with the amino groups of chitosan, and the maximum encapsulation efficiency was reached when using 84% of N-propionic chitosan for 1% of Glab in the formulation [56]. In contrast, Glab was shown to be incorporated passively into the pores of the microsponges, and no specific physical interactions were observed [57].

2.7 Analytical studies and quantitation

Reproducible and precise analytical methods are needed for the quantitation of Glab in standardized G. glabra extracts, or in biological samples for the subsequent determination of pharmacokinetic (PK) parameters. Accurate methods are also required for the quality control of polyherbal mixtures containing licorice, and when performing Glycyrrhiza species identification (Table 2). In summary of the available literature, quantitation of Glab in different matrices has been carried out by HPLC following UV (230 or 280 nm) [54, 58] or MS detection [59], and recently by HPTLC [60] (Appendix D for HPTLC profiles). Except for the chiral separation of Glab enantiomers [42], most of the HPLC analyses have utilized reverse phase C₁₈ columns [54, [58, 59]. Some of the principal studies are detailed below.

Kamal et al. [58] have developed an HPLC method for the quantitation of Glab from polyherbal preparations and crude extracts. Both were analyzed after extraction with 30% aqueous ethanol (overnight soaking followed by extraction under reflux during 45 min) on a reverse phase C18 (Merck, LiChrosphere 100 RP18e, 250 × 4.6 mm, 5 μm) column with an MeCN/H₂O gradient from 50 to 80% MeCN at 1 mL/min. Under these conditions, Glab was detected at 230 nm with a retention time (R_t) of 14.90 ± 0.02 min. The Glab enantiomers could be separated on a chiral Sumichiral OA-7000 (Sumika Chemicals, Osaka, Japan) column, a stationary phase composed of derivatized beta-cyclodextrins [42]. Aoki et al. validated an analytical method for the determination of Glab in plasma after SPE using C8 cartridge, followed by an LC-MS/MS analysis and using mefenamic acid as internal standard [59]. Ester derivatives of Glab (glabridin diacetate [GDA] and dihexonate [GDH]), used in topical formulations, have been quantified by HPLC-UV (280 nm) on a reverse phase C18 column (250 × 4.6 mm, 5 μm; Thermo Hypersil-Keystone, Bellefonte, PA, USA), eluted with a mixture of MeCN/H₂O as follows: 76% during 9 min to 90% MeCN in 9 min and during 12 min at a flow rate of 1 mL/min [54]. Under these conditions, the R_t of Glab was 4.70 min, and the esters eluted at 7.76 and 25.04 min for GDA and GDH, respectively.

3.1 Anti-inflammatory properties

Inflammation is the starting point for many chronic diseases including cardiovascular, degenerative diseases and cancer. Multiple factors (cytokines, prostaglandins PGs), enzymes (COX-2, iNOS), and nuclear factors (NF-κB, AP-1) are involved in the expression of inflammation. The inducible enzymes, cyclooxygenase-2 (COX-2), and nitric oxide synthase (iNOS) are involved in the production of PGs (e.g. PGE₂), thromboxanes (TXs) and NO, respectively. In macrophages, expression of iNOS-related genes is mainly regulated at the transcriptional level by regulators belonging to the (NF-κB)/Rel family of transcription factors. The latter also regulate the production of other inflammatory cytokines such as tumor necrosis factor-α (TNF-α) or interleukines (ILs; e.g., IL-1β, IL-6). Overproduction and cellular release of all these inflammatory mediators (PGE₂, NO, TNF-α, ILs) are responsible for the characteristic expression of inflammation and its subsequent side effects.

The anti-inflammatory effect of Glab was first demonstrated in 1998 by Yokota et al. [63]. At the concentration of 6.25 μg/mL, Glab inhibited 31.9% of COX activity stimulated by the addition of arachidonic acid (AA). Since this discovery, the anti-inflammatory properties of Glab have been evaluated in vitro on murine macrophages [64], microglia [65], promyelocytic [64], and dendritic cells [66], as well as in vivo on BALB/c mice [67] and BDF1 mice [68], which suffered from colonic inflammation or sepsis (Table 3).

Bacterial lipopolysaccharide (LPS), a component of the gram negative bacterial membrane, is known to induce the overexpression of various inflammatory mediators in vitro and in vivo. In murine macrophages and microglia cells, Glab has been demonstrated to inhibit LPS induced PGE-2 [64], IL-1, NO [69], and TNF-α cellular release [65]. All these effects were shown to be directly linked to a reduction of COX-2 and iNOS expression. The latter was associated with a down regulation of major transcription factors, NF-κB [69], in both macrophages and microglia cells, where down-regulation of AP-1 was also implied [65]. In HL-60 promyelocytic leukemia cells, Glab also inhibited the production of potent inflammatory mediators, leukotriene LTB4 and thromboxane TXB2, in a concentration-dependent manner [64]. Similarly, anti-inflammatory effects observed in dendritic cells (DCs) were related to an inhibition of IκBα/β degradation, nuclear translocation of NF-κB p65/p50, and phosphorylation of elements from MAPKs signaling [66]. DCs are potent antigen-presenting cells that play a pivotal role in the innate and adaptive immune responses. Anti-inflammatory effects of Glab were also linked to a decrease in functional maturation of DCs, which according to Kim et al. [66] could have some interest for the regulation of the immune system especially in auto-immune diseases. A patent related to this research was issued in 2009 [70].

Kang et al. [68] suggested that Glab inhibits NF-κB/Rel activation through non-cell specific reduction of intracellular reactive oxygen species (ROS). Inhibitory effects of Glab on NF-κB activation has been found in different cell types and, therefore, seems to be universal. Park et al. [65] further corroborated this observation by demonstrating that the effect of Glab on AP-1 activity was differential in microglia and macrophages cells, suggesting that the modulation of AP-1 by Glab is cell-type specific.

Sepsis is a systemic response to infection triggered by an exposure to LPS. In vivo, Glab (1 to 10 mg/kg) was demonstrated to protect BDF1 mice against LPS-induced sepsis by reducing the production of various inflammatory mediators, such as TNF-α and NO [68]. Furthermore, oral administration of Glab (10 or 50 mg/kg during 7 days) to female BALB/C mice resulted in an attenuation of colonic inflammation induced by dextran sulfate sodium (DSS) [67]. In this colitis model, the activity of the enzyme MPO was shown to increase and found to be responsible for the formation of ROS leading to tissue damages. Glab treatment was associated with a suppression of MPO activity, thus contributing to the attenuation of inflammatory progression in the DSS-induced colitic colon. MCP-1 is known to contribute to macrophage infiltration in tissue, thus, maintaining their degradation. Non-obese/diabetic rats treated with Glab displayed a lower MCP-1 level compared to non-treated animals, indicating that Glab could also mitigate inflammation and reduce its deleterious effects on tissue through a reduction of MCP-1 secretion [71].

In conclusion, down regulation of NF-κB, cell specific down-regulation of AP-1, and MAPKS signaling, combined with a reduction of COX-2 expression, all participate in the cellular anti-inflammatory effects of Glab. On the other hand, suppression of MPO activity and MCP-1 secretion contribute to its protective effect against inflammation induced tissue damages. Altogether, these properties contribute to the various biological effects of Glab.

3.2 Prevention of LDL oxidation and anti-atherogenic properties

By far, the most studied biological property of Glab is its prevention of Low Density Lipoprotein (LDL) oxidation and its subsequent anti-atherogenic properties, which represent 23% of the current scientific literature dedicated to the bioactivities of Glab (Figure 1B). Atherosclerosis is characterized by the accumulation of cholesterol and oxidized lipid in the arterial wall. LDL is the major cholesterol carrier in human serum. Oxidized-LDLs are cytotoxic to the arterial wall through the stimulation of inflammatory and thrombotic processes and, therefore, are considered as highly atherogenic [72]. LDL can be oxidized by the major cell types present in the vascular wall, endothelial cells, macrophages, and smooth muscle cells. The possible cellular sources of LDL oxidants are, among others, NADPH oxidase and P450 enzymes [73]. Macrophage-mediated oxidation of LDL is considered to be of major importance during early atherogenesis, and the formation of oxidized LDL can then induce cellular cholesterol accumulation and foam cell formation (accumulation of ox-LDL in macrophages and other phagocytes, Figure 5) [74].

In a study by Vaya et al. [28], Glab (30 μM) has been demonstrated to possess antioxidant activity, compared to β-carotene consumption, against 2, 2′-azobis (2-amidino-propane) dihydrochloride (AAPH) induced LDL oxidation. Furhman et al. [75] demonstrated that Glab, when administered to atherosclerotic apolipoprotein E (E⁰) deficient mice for a period of 6–12 weeks (20 μg/day/mouse), was not only absorbed but also bound to the LDL particles. The compound reduced the formation of macrophage foam cells and prevented the development of atherosclerotic lesion areas. Moreover, the same authors demonstrated that Glab reduced the atherogenic modification of human LDLs (oxidation and aggregation) [75]. Furthermore, Belinky et al. [76] also analyzed the potential of Glab (50 μg/day/mouse) to inhibit the oxidation of LDLs obtained from E⁰ deficient mice. The authors confirmed that Glab effectively bound to LDLs and further concluded its capacity to inhibit the formation of lipids peroxides and oxysterol derivatives in LDL, regardless whether the oxidation was induced by copper ions or AAPH. According to preliminary structure activity relationship, the same authors suggested that antioxidant properties of Glab were mainly associated with its 2′,4′-dihydroxyl functions, whereas its hydrophobic prenyl moiety was assumed to favor its lipophilic interactions and inclusion in the LDL particles [39].

The protective effect of Glab against LDL oxidation was further investigated by Rosenblat et al. [48], who suggested that the 50% reduction of the aortas atherosclerotic lesions observed in treated E⁰ mice was, in fact, related to the effect of Glab on macrophages mediated LDL oxidation. Therefore, the authors also analyzed the macrophage ability to take up Glab and studied the mechanisms by which cellular accumulation of Glab can affect macrophage-mediated oxidation of LDLs [48]. Interestingly, the results indicated that Glab accumulated in macrophages both in vitro and in vivo, and reduced the cell-mediated oxidation of LDLs at almost ~ 90%. This effect was linked to an intracellular inhibition of protein kinase C (PKC) activity (70% inhibition), which had a subsequent inhibitory effect on the translocation of the NADPH oxidase P-47 cytosolic component to the plasma membrane and, hence, on macrophage superoxide production. Additionally, Kent et al. [77] reported the inhibition of cytochrome P450 3A4 by Glab and concluded that this effect may contribute to the protective effect of Glab against LDL oxidation and its subsequent antiatherosclerotic properties (see section 4.1). The in vivo inhibitory effect of Glab on human LDL oxidation was investigated by Carmelli et al. [78]. In this study, administration of Glab (60 mg) through a standardized crude extract from G. glabra (Licolife^™) to 22 healthy volunteers and during 6 months resulted in a 20% reduction of plasma LDL oxidation.

Paraoxonase 2 (PON2) is expressed in many different types of tissues and cells, including cells of the arterial wall, e.g., macrophages, and demonstrates antioxidant properties [79]. Glab (0.1 μM) was shown to have the potential to up-regulate antioxidant enzymes including manganese sodium oxide dismutase (SOD), catalase (CAT) and PON2 under glucose stress in human monocytes (THP-1), which are precursors of macrophages [80]. Together with glutathione peroxidase (GPX), SOD and CAT are the major vascular antioxidant defense enzymes. The authors suggested that Glab could, therefore, strengthen the antioxidant defense mechanism and may serve as an anti-atherogenic agent. Paroxonase 1 (PON1) is another anti-atherogenic enzyme forming part of the circulating HDL (High Density Lipoprotein). Its activity is inhibited through an oxidation process initiated by the lipid fraction of the human carotide plaque or by linoleic acid hydroperoxide [79]. Hatrahimovich et al. [81] demonstrated that Glab specifically interacted with the enzyme and consequently reduced or prevented its inhibition by linoleic acid hydroperoxide.

In relation with its estrogenic-like properties (see section 3.4 and figure 6), Glab has been demonstrated to stimulate the proliferation of vascular smooth muscle and endothelial cells, two fundamental mechanisms for the healing of damaged vascular endothelium [82].

Intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1) are protein adhesion markers which play pivotal roles in the development of atherosclerosis and in the plaque destabilization. These proteins are known to recruit the circulatory inflammation cells and facilitate their migration into the arterial walls [5]. When cultured human umbilical vein endothelial cells (HUVEC) were exposed to TNF-α, Glab (3 mM) produced a 50% inhibition of the endothelial cell expression of ICAM-1 and VCAM-1 [83].

In summary, the antiatherosclerotic properties of Glab result from its beneficial effects on various targets involved in the maintenance of vascular integrity. This literature survey presents ample evidences that Glab is capable of down-regulating intracellular pro-oxidant proteins (PKC, NADPH oxidase), up-regulating antioxidant defense systems (SOD, CAT, PON2), and interacting with anti-atherogenic PON1, thus inhibiting cellular mediated LDL peroxidation. Moreover, Glab has been shown to inhibit the expression of protein markers ICAM-1 and VCAM-1 at elevated concentrations, while generally increasing the proliferation of vascular endothelial cells and, therefore, reducing vascular inflammation and favoring healing process of atherosclerotic lesion (Figure 5).

3.3 Neuroprotective effects

Oxidative stress, inflammation and subsequent neuronal apoptosis (programmed cell death) can occur after a brain ischemic insult. Excessive production of inflammatory mediators in brain causes neurotoxicity in various neurodegenerative diseases such as Alzheimer’s, and Parkinson’s diseases or senile dementia. Neuroprotective effects of Glab were demonstrated to be associated mainly to its anti-inflammatory functions [65] and its capacity to down-regulate intercellular oxidative stress by increasing the expression of antioxidant makers [84]. Hence, Glab was shown to protect cellular (neuronal) functions against damages (DNA fragmentations, protein oxidations, mitochondrial peroxidation) that lead to apoptosis [85].

On cortical neurons, Glab was shown to inhibit staurosporine induce DNA fragmentation and apoptosis [84]. Its protective effects were further confirmed in vivo on a reversible model of cerebral ischemia (i.e, Middle Cerebral Artery Occlusion [MCAO] rat models), where Glab treated rats displayed a reduced total infarct volume associated with a reduction of malondialdehyde (MDA, maker of lipid peroxidation) levels and an increase in antioxidant markers such as SOD and glutathione [84]. Altogether, these effects were associated with inhibition of oxidative stress induced neuronal damages.

Investigations carried out by Cui et al. [30] demonstrated that mice treated with Glab (2 and 4 mg/kg) during 3 days presented a general improvement of their memory. Interestingly, these positive effects were associated with an increase in brain acetylcholine levels and a decrease in choline esterase activity. Similarly, Glab was also demonstrated to reverse learning deficits of streptozotocine-induced diabetic rats, without altering their hyperglycemic status [86]. It was, therefore, assumed that Glab might exert a beneficial effect on various neurodegenerative diseases, including cognitive impairment linked to mellitus diabetes and Alzheimer’s disease, not only through the protection of neuronal cells from oxidative stress and inflammatory damages, but also via potential anticholinesterase properties. A composition for the treatment of mental health disorders containing Glab has been patented in 2006 [87].

Additionally, Glab was found to be an effective inhibitor of serotonin re-uptake in HEK 293 cells [88]. An approximate 60% inhibition of serotonin re-uptake was observed at 50 μM, compared to a 74% inhibition for imipramine at 5 μM. Selective blockage of the serotonin receptors in the human central nervous system (CNS) is considered to be the initial step in the pharmacological improvement of a wide variety of disorders, including major depression [89]. The authors suggested that Glab could have a beneficial effect on the mild to moderate depression, especially among pre-post-menopausal women.

Nevertheless, despite interesting neuroprotective activities, the bioavailability of Glab in the brain could be compromised due to its interaction with P-glycoprotein (P-gp), which is expressed notably in capillary endothelial cells. As demonstrated by Yu et al., this protein limits Glab penetration through the blood-brain barrier and consequently reduces its bioavailability in the CNS [90]. The interaction of Glab with P-gp would translate into a higher dosage required to reach efficacious levels of Glab for the treatment of CNS diseases, and might also be responsible for Glab-drug interactions (see section 4.2).

3.4 Estrogenic properties: is glabridin a potential PhytoSERM?

Diverse estrogen-like activities have been reported for Glab, which has been compared in structure and lipophilicity with 17β-estradiol (Figure 6). Glab binding to the estrogen receptors (ERs) has been evaluated in vitro on human breast cancer cells [91] and in yeast-based assays [92]. Its estrogenic-like properties have also been evaluated, in vitro and in vivo, on skeletal and cardiovascular tissues of female rats [82, 93] by measuring the induction of the specific activity of creatinine kinase (CK), an estrogen-induced enzyme. The present literature survey clearly indicates that Glab shows assay-dependent estrogenicity (Table 3).

Tamir et al. [91] have investigated the capacity of Glab to bind to ERs and measured its effects on estrogen responsive human breast cancer cells (MCF-7, MDA-MB-231 and T-47D). The authors found that Glab was a weak ligand to the ER (IC₅₀ ~5 μM without differentiation of ER sub-type) and has a biphasic effect on the proliferation of breast cancer cells (MCF-7 and T47D cells). In vivo observations indicate that Glab induces the estrogen response marker CK in rat skeletal, cardiovascular tissues, and promotes uterus wet weight gain. The authors concluded that Glab represents a phytoestrogen which acts as estrogen agonist in estrogen-responsive tissues such as bone and vascular tissues (Figure 6).

Completing these picture of observations on the estrogenic-like properties of Glab, Somjen et al. further analyzed the capacity of Glab to induce CK in the same vascular, and skeletal tissues [82, 93]. In both studies, Glab showed effects similar to 17β-estradiol (E2), although at higher concentration, by inducing the estrogen response marker CK in vitro and in vivo. Glab stimulated DNA synthesis in endothelial cells ECV-304 (30 to 3000 nM), and displayed a bi-phasic effect on the proliferation of VSMC smooth muscle cells (3 to 300 nM). Following all these results on the estrogen-like activities of Glab and its derivatives, Vaya filed a patent on the phytoestrogenic pharmaceutical preparations comprising flavonoids from licorice, particularly Glab and its derivatives [94].

Using a yeast estrogen bioassay, in which observed responses are specific for effects directly mediated by the ERs, Simons et al. tested the estrogenic-like activities of Glab [92]. As a result, Glab did not initiate any significant estrogenic response on both ER subtypes. However, Glab was demonstrated to be a selective ERα antagonist with 80% inhibition of the estrogen response at 6 μM. Considering the agonistic activity of Glab in the MCF-7 cell proliferation assay [91], in two in vivo animal models [82, 93], and its ERα-selective antagonistic activity in the yeast estrogen bio-assays, the authors concluded that Glab might act as a SERM regarding its dual mode of action [92, 95]. These tentative conclusions regarding the potential phytoSERM properties of Glab were shared by Somjen et al. [93], who corroborated previous finding by studies demonstrating that Glab has potent osteogenic activities.

3.5 Anti-osteoporotic properties

Estrogen deficiency is known to be involved in osteoporosis. According to Somjen [93], Glab can act as a phytoestrogen and yield the same beneficial effects on bone tissues as E₂ or genistein, an isoflavone. Compared to genistein, Glab is effective at lower concentration differing by a factor 10, which emphasizes the potential of isoflavan derivatives to modulate bone disorders in post-menopausal women.

The osteogenic activities of Glab has been analyzed in vitro on murine osteoblast cells MC3T3-E1 [96–98], but also on human female derived bone cells (hObs) [99]. In these studies, Glab was demonstrated to increase the function of osteoblasts, which are key cells in bone matrix formation and calcification. Beneficial effects were demonstrated to be ER mediated [96, 99] and associated with cellular protection against oxidative burst, through an up-regulation of antioxidant enzymes (SOD I and GPX 4) [97], thus, preventing the alteration of mitochondrial functions [100, 98].

Optimum bone growth and prevention of osteoporosis in postmenopausal women require adequate concentrations not only of estrogens but also of vitamin D3 (1, 25(OH)₂D₃). When testing on human female osteoblasts at 300 nM, Glab and its derivative, glabrene, have been demonstrated to favor the production of 1,25(OH)₂D₃ which in turn up-regulated cellular sensitivity to E2 [99]. Moreover, Glab has been shown to restore the expression of osteoblastic differentiation genes, and increase the expression of phosphatidylinositol-3′kinase (PI3K) and protein kinase B 2 (AKT2) genes [98, 97], which play a critical role in cellular plasticity, growth, and survival [101]. Anti-inflammatory properties of Glab, tested on murine osteoblasts, also contribute to its beneficial effects through the inhibition of PGE₂ and NO production and the prevention of osteoblasts apoptosis induced by TNF-α [96].

Acting in a complementary manner to its effects on osteoblasts, Glab has been demonstrated to inhibit RANKL induced osteoclastogenesis in murine macrophages, RAW 264.7 [102]. In these differentiated bone resorbing cells, Glab was shown to down-regulate the expression of osteoclast related genes and reduce the expression of PI3K, AKT2. These negative effects on osteoclast were proposed to be related to an inhibition of the transcription factors AP-1/c-Fos via blockage of MAPKs signaling.

In summary, Glab exerts antioxidant activities in osteoblasts and improves their differentiation function, whereas in osteoclast it reduces the expression of genes related to their bone resorbing properties. Proven ER-dependent osteogenic effects and its potential to reduce inflammation suggest that Glab might play a positive role in the prevention of osteoporosis induced by estrogen deficiency, and in the regulation of localized bone destruction associated with inflammatory bone diseases.

3.6 Regulation of Energy Expenditure and Metabolism

Disturbances in energy homoeostasis underlie a number of disease states in humans such as type-2 diabetes and obesity, which can lead to hypertension and cancer. Glab has been proven to participate into the regulation of energy expenditure through mainly two types of interactions, the stimulation of AMPK activity and the up-regulation of PPAR-γ. Adenosine monophosphate activated protein kinase (AMPK) is a master energy sensor, which regulates diverse metabolic pathways, increases mitochondrial activity and biogenesis in muscle, and is responsible for the inhibition of adipogenesis [103, 104]. PPAR-γ is a ligand-dependent transcriptional regulatory factor, known to regulate the expression of a group of genes that maintain glucose and lipid metabolism. PPAR-γ also participates into the programmed differentiation of adipocytes [105]. Altogether, the stimulation of AMPK and up-regulation of PPAR-γ lead to the regulation of genes implied in lipid metabolism and adipocytes differentiation, eventually decreasing adiposity and reducing insulin resistance (Table 3).

The earliest reports on the regulation of energy metabolism by Glab were focused on the biological evaluation of LFO extract standardized to ~1% w/w of Glab [1, 17, 106]. This extract has been tested on two mice models: fat-fed induced mice (C57BL/6J mice) and obese diabetic KK-A^y mice. LFO extract has been reported to suppress the expression of genes involved in lipogenesis [106], leading to a decrease in both abdominal-fat and blood sugars [17]. However, no unambiguous evidence could be obtained that Glab might be associated or mainly responsible for these beneficial effects.

Lee et al. have studied the direct effect of Glab on AMPK activity in obese rodents models (C57BL/6J mice)[107]. They observed that Glab decreases adiposity and ameliorates lipid dysregulation as well as insulin resistance, thus, lowering body weight. The molecular mechanism underlying these effects were demonstrated to be associated with an attenuation of mitochondrial ATP production, leading to an increase in the cellular AMP contents, hence, promoting AMPK activation. In muscles, Glab was shown to regulate genes associated with fatty acid oxidation, and to reduce the expression of lipogenic genes in liver and adipose tissues, where it also decreased the expression of pro-inflammatory genes (inos, mcp-1).

A total of 39 licorice constituents have been tested for their PPAR-γ ligand binding activity. Among them, 14 phenolics including Glab and shinpterocarpin displayed significant PPAR-γ binding, suggesting that these compounds, and the hydrophobic crude extract containing them, can potentially have anti-diabetes and anti-obesity properties mediated by the activation of PPAR-γ receptors [12]. Recently, Ahn et al. confirmed this hypothesis by demonstrating that Glab (25 μM) inhibited the adipogenic differentiation of mice fibroblast 3T3-L1. However, there is no evidence that Glab alone can effectively up-regulate PPAR-γ in this cellular model. A supercritical CO₂ extract of G. glabra (LSC) enriched in Glab (4.5% w/w Glab) was shown to inhibit adipogenic differentiation of fibroblasts through an up-regulation of PPAR-γ and C/EBPα. In vivo, LSC (0.1 and 0.25% w/w during 8 weeks) was responsible for a dose-dependent reduction of weight gain of treated C57BL/6J mice, and was demonstrated to regulate genes involved in hepatic metabolism [23].

In complement of these effects on cellular regulation of energy metabolism, Glab has also been shown to be effective in the reduction of mice physical fatigue associated with exercise by delaying lactic acid production and increasing glycogen storage in muscle and liver [108]. Additionally, direct hypoglycemic effects of Glab were investigated in vivo using an animal model of diabetes mellitus [109]. Data revealed that mice treated with Glab over a period of 28 days displayed a better glucose tolerance associated with a surprising increase in their body weight, compared to other studies. Moreover, the sodium oxide dismutase activity was increased, while the level of fast blood glucose (FBG) and MDA were decreased in the liver, kidney, and pancreas. Taken altogether, these results indicate clearly that Glab itself possesses hypoglycemic effects. However, the mechanism of action underlying these properties remains to be investigated. Finally, licorice flavonoids have been tested for their capacity to inhibit the pancreatic lipase and reduce intestinal lipid absorption. Glab did not show any significant inhibition (IC₅₀: 485.6 μM) of this enzymatic activity [110].

In summary, there is strong in vitro and in vivo evidence that Glab and licorice extracts enriched in this compound can stimulate AMPK and participate to the up-regulation of PPAR-γ. Both effects contribute to the reduction of adipogenesis and favor insulin sensitivity. In vivo, Glab has been demonstrated to enhance glucose tolerance of mice as well as glycogen storage in liver and muscle. Consequently, both pure Glab and Glab-enriched extracts were found to have therapeutic potency for obesity-related metabolic diseases, which is further supported by the observation that Glab itself exhibits anti-atherogenic and anti-inflammatory activities. The latter are known to be linked to numerous secondary clinical effects of obesity. Considering the importance of metabolic related diseases in developed countries and their impact on the DS market, several extracts of licorice enriched in Glab have been patented for the treatment of metabolic disorders [111][112][113].

3.7 Less frequently reported biological activities

4.1 Metabolism and Cytochromes P450

Kent et al. have investigated the effect of G. glabra extract and Glab, on the enzymatic activities of several human P450 enzymes such as P450 3A4, 2B6 and 2C9, which play a key role in the metabolism of clinically used drugs [77]. The isoform P450 3A4, which represents the major hepatic and intestinal P450 enzyme in humans, is known to metabolize more than 50% of clinically used drugs. In the cited study, cytochromes were expressed and purified from E. coli, then reconstituted for the enzymatic assays which were carried out with different concentrations of Glab. Interestingly, the reconstituted cytochromes 3A4 and 2B6 were inhibited by Glab in a time-, concentration- and NADPH-dependent manner, and the concentration required to obtain 50% inhibition was determined to be 7 and 12 μM, respectively. Both inhibitions were found to be irreversible through dialysis, and for the isoform 3A4 the inhibition was associated with the destruction of the heme moiety. Additionally, the authors suggest that metabolism of Glab by the cytochrome 2B6 could generate reactive intermediates that might bind to the apoprotein. Nevertheless, the identity of the reactive intermediates produced by the cytochromes has not been defined. As 2′,4′-dimethylgabridin does not interact with the isoform 3A4, Kent et al. suggested that the two hydroxyl groups on the 2′ and 4′ position of the B ring, which are believed to be essential for most of Glab biological activities [39, 130], are also required for P450 3A4 inactivation. Further investigations will be necessary to define the interactions of Glab with other P450 isoforms and identify its metabolites.

4.2 Metabolism and Intestinal Absorption

A study by Cao et al. [131] reported that Glab is a substrate of the intestinal p-glycoprotein, P-gp/MDR1. Glab metabolism in rat intestinal and hepatic microsomes was also investigated and revealed that Glab was mainly metabolized by glucuronidation, and that the metabolic capacity of intestinal microsomes was 1/15 to 1/20 of that of liver microsomes. Systemic bioavailability of Glab after oral administration at 5 and 20 mg/kg was found to be very low, only about 7.5% of the Glab was absorbed. The same authors concluded that low water solubility, intestinal transporter-mediated efflux via P-gp, and first-pass metabolism through glucuronidation could be considered as factors responsible for the overall low oral bioavailability of Glab.

Different results were presented by Ito et al. [71], who evaluated the absorption of Glab by a human colon adenocarcinoma cell line (Caco-2). The authors demonstrated that intact Glab was released to the basolateral side mostly without metabolic conversion and with a 90% recovery rate when administered alone, or at a rate of 82% when administered in the form of LFO enriched extract. In rats, dietary Glab (oral dose of 10 mg (30 μmol)/ kg) was partly incorporated into the blood and liver in an unchanged form. Glab showed a maximum concentration 1 h after the dose administration at 87 nmol/L for standard non-conjugated Glab, and decreased gradually over 24 h after dosing. In vivo, the intestinal absorption of Glab was higher when administered in the form of LFO compared to the pure compound. Moreover, Glab seemed to escape conjugation during intestinal absorption and appeared in the blood stream as the unaltered aglycone. Based on these results, the authors concluded that Glab possesses high oral bioavailability.

These two apparently conflicting studies indicate that further investigations on Glab intestinal absorption and metabolism are warranted for a better understanding of its bioavailability. Interestingly, absorption or PK studies of Glab following administration of powdered roots or other commercial extracts from G. glabra have not been reported to date. This type of study would be very interesting to consider, given that commercial DSs mostly contain either powdered roots or their crude extracts rather than a single chemical entity.

4.3 Pharmacokinetics in healthy humans

Aoki et al. have also studied some pharmacokinetic parameters in healthy humans when administering Glab in the form of LFO (1% w/w of Glab) [132]. On the basis of the hematological (e.g., hematocrit, platelet, red blood cells) and relevant biochemical (e.g., total protein, albumin, bilirubin, creatinine) results obtained from healthy volunteers, the authors concluded that a dose of LFO up to 1200 mg/day appeared to be safe and well tolerated. Additionally, their results indicated that Glab exhibited linear pharmacokinetics over the dose range of 300 to 1200 mg/individual of LFO. The steady state plasma level of Glab (1.42–4.4 ng/mL) was reached after two weeks of daily administration. Regarding these plasma concentrations, Aoki suggested that inactivation of the major cytochrome P450s by Glab should be minimal when daily doses of at least 300 to 1200 mg of LFO would be administered for 4 weeks. However, the same team also recognized that other flavonoids of the extract may additively or synergistically inactivate cytochromes P450, leading to an overall complex situation.

Despite several documented studies, the PK parameters of Glab and determination of its metabolism are only in their beginnings, and further studies will be necessary to better define its bioavailability, existence of potential bioactive metabolites, and precise profile of its P450 interactions.