Abstract
Naturally occurring splice variants of human constitutive androstane receptor (hCAR) exist, including hCAR-SV23 (insertion of amino acids SPTV), hCAR-SV24 (APYLT), and hCAR-SV25 (SPTV and APYLT). An extract of Ginkgo biloba was reported to activate hCAR-SV24 and the wild type (hCAR-WT). However, it is not known whether it selectively affects hCAR splice variants, how it activates hCAR isoforms, and which chemical is responsible for the effects of the extract. Therefore, we evaluated the impact of G. biloba extract on the functionality of hCAR-SV23, hCAR-SV24, hCAR-SV25, and hCAR-WT and compared it with that of phenobarbital, di-(2-ethylhexyl)phthalate (DEHP), 6-(4-chlorophenyl)imidazo[2,1-b][1,3]thiazole-5-carbaldehyde O-(3,4-dichlorobenzyl)oxime (CITCO), and 1,4-bis-[2-(3,5-dichloropyridyloxy)]benzene (TCPOBOP) in cell-based reporter gene assays. Among the hCAR splice variants investigated, only hCAR-SV23 was activated by G. biloba extract, and this required cotransfection of a retinoid X receptor α (RXRα) expression plasmid. The extract activated hCAR-SV23 to a lesser extent than hCAR-WT, but ginkgolide A, ginkgolide B, ginkgolide C, ginkgolide J, and bilobalide were not responsible for the effects of the extract. CITCO activated hCAR-SV23, hCAR-SV24, and hCAR-WT. By comparison, phenobarbital activated hCAR-WT, whereas DEHP activated hCAR-SV23, hCAR-SV24 (with exogenous RXRα supplementation), and hCAR-WT. TCPOBOP did not affect the activity of any of the isoforms. G. biloba extract and phenobarbital did not bind or recruit coactivators to the ligand-binding domains of hCAR-WT and hCAR-SV23, whereas positive results were obtained with the controls (CITCO for hCAR-WT and DEHP for hCAR-SV23). In conclusion, G. biloba extract activates hCAR in an isoform-selective manner, and hCAR-SV23, hCAR-SV24, and hCAR-WT have overlapping, but distinct, sets of ligands.
Introduction
The constitutive androstane receptor (CAR), originally referred to as MB67 (Baes et al., 1994), belongs to the superfamily of nuclear hormone receptors. It is designated as NR1I3 (Germain et al., 2006). CAR regulates the expression of multiple genes involved not only in xenobiotic metabolism, transport, and toxicity, but also in various other functions, including gluconeogenesis, lipogenesis, and thyroid hormone homeostasis (Gao and Xie, 2010). It can be activated by a mechanism that involves direct binding of an agonist to the ligand-binding domain of the receptor and recruitment of coactivators to the ligand-receptor complex (Timsit and Negishi, 2007). Chemicals shown to activate CAR by this mode of action are 6-(4-chlorophenyl)imidazo[2,1-b][1,3]thiazole-5-carbaldehyde O-(3,4-dichlorobenzyl)oxime (CITCO), which is an agonist of human CAR (hCAR) (Maglich et al., 2003), and 1,4-bis-[2-(3,5-dichloropyridyloxy)]benzene (TCPOBOP), which is an agonist of mouse CAR (mCAR) (Tzameli et al., 2000). In contrast, phenobarbital activates CAR not by binding to the ligand-binding domain of the receptor (Moore et al., 2000), but by cellular and molecular signaling mechanisms in which the intricate details are still not well understood (Timsit and Negishi, 2007).
Naturally occurring splice variants of hCAR have been identified, including hCAR-SV23, hCAR-SV24, and hCAR-SV25 (Lamba et al., 2005, 2008). Table 1 shows the various nomenclatures of these splice variants and the reference isoform or the wild type (hCAR-WT). Tissue distribution studies have shown that these hCAR splice variants and hCAR-WT are expressed in the liver (Auerbach et al., 2003; Savkur et al., 2003; Jinno et al., 2004; Lamba et al., 2004; Ross et al., 2010). Hepatic hCAR-SV23, hCAR-SV24, and hCAR-SV25 mRNA have been reported to be expressed at levels of approximately 6 to 30, 20 to 42, and 2 to 10% of total hCAR transcripts, respectively (Savkur et al., 2003; Jinno et al., 2004; DeKeyser et al., 2009; Ross et al., 2010). As depicted in Fig. 1, the hCAR-SV23 splice variant has a 12-nucleotide insert between exons 6 and 7, resulting in the addition of four amino acids (SPTV), whereas the hCAR-SV24 splice variant has a 15-nucleotide insert between exons 7 and 8, leading to the addition of five amino acids (APYLT) (Auerbach et al., 2003). By comparison, the hCAR-SV25 splice variant has both the 12-nucleotide insert (identical to hCAR-SV23) and the 15-nucleotide insert (identical to hCAR-SV24) (Auerbach et al., 2003; Savkur et al., 2003). A limited number of functional studies have identified ligands of these splice variants. The only known hCAR-SV23 activators to date are di-(2-ethylhexyl)phthalate (DEHP) (DeKeyser et al., 2009), several other phthalates (DeKeyser et al., 2011), and CITCO (DeKeyser et al., 2009). hCAR-SV24 activators include CITCO (Auerbach et al., 2005; Faucette et al., 2007) and artemisinin (Faucette et al., 2007). To date, only two chemicals have been studied for their effects on hCAR-SV25 function. However, neither clotrimazole (Auerbach et al., 2003) nor CITCO (Jinno et al., 2004) activated this isoform in those studies.
Ginkgo biloba is often used by consumers for the self-treatment of various medical conditions, including dementia (Weinmann et al., 2010). It contains bioactive terpene trilactones, such as ginkgolides and bilobalide (van Beek and Montoro, 2009). Their chemical structures are shown in Fig. 2. Commercial preparations of G. biloba extract contain approximately 6% (w/w) terpene trilactones and 24% (w/w) flavonol glycosides, although considerable variability exists (Kressmann et al., 2002). As determined in a cell-based reporter gene assay (Li et al., 2009), a single concentration (100 μg/ml) of a G. biloba extract of undefined abundance of terpene trilactones and flavonol glycosides was reported to activate hCAR1 (also known as hCAR-WT; Table 1) and hCAR3 (also referred to as hCAR-SV24; Table 1) by ≤2-fold over the control group. However, it remains to be investigated whether G. biloba extract selectively activates hCAR splice variants, how it activates hCAR isoforms, and which chemical constituent is responsible for the hCAR-activating effect of G. biloba extract.
In the current study, we compared the effects of G. biloba extract on the activity of hCAR-SV23, hCAR-SV24, hCAR-SV25, and hCAR-WT in cell-based reporter gene assays and evaluated the role of five individual terpene trilactones (ginkgolide A, ginkgolide B, ginkgolide C, ginkgolide J, and bilobalide) in the extract. As a way to determine whether the extract exhibits receptor agonism, additional experiments were performed to assess whether it transactivates the ligand-binding domain of the receptor and promotes recruitment of coactivators. For comparative purposes, we also characterized the effects of phenobarbital, DEHP, CITCO, and TCPOBOP in our cell-based assays. Overall, our novel results demonstrate selective activation of hCAR splice variants by G. biloba extract and isoform-specific ligand-activation profiles for hCAR-SV23, hCAR-SV24, and hCAR-WT.
Materials and Methods
G. biloba Extract, Chemicals, and Reagents.
Five individual lots of G. biloba extract, designated as lot A, lot B, lot C, lot D, and lot E, were supplied in dry powder form by Indena S.p.A. (Milan, Italy). The quantity of terpene trilactones and flavonol glycosides in each individual lot is shown in Table 2. Ginkgolide A (CAS 15291-75-5), ginkgolide B (CAS 15291-77-7), ginkgolide C (CAS 15291-76-6), and (−)-bilobalide (CAS 33570-04-6) were obtained from LKT Laboratories (St. Paul, MN), and ginkgolide J (CAS 107438-79-9) was from ChromaDex (Irvine, CA). Sodium phenobarbital (CAS 57-30-7), DEHP (CAS 117-81-7), TCPOBOP (CAS 76150-91-9), and 1-(2-chlorophenyl)-N-methyl-N-(1-methylpropyl)-3-isoquinoline-carboxamide (PK11195; CAS 85532-75-8) were purchased from Sigma-Aldrich (St. Louis, MO). CITCO (CAS 338404-52-7) was obtained from Enzo Life Sciences, Inc. (Plymouth Meeting, PA), and 5α-androstan-3α-ol (androstanol; CAS 7657-50-3) was from Steraloids (Newport, RI). Charcoal-stripped, heat-inactivated fetal bovine serum was bought from Thermo Fisher Scientific (Nepean, ON, Canada). Opti-MEM and all other cell culture reagents were obtained from Invitrogen (Carlsbad, CA). FuGENE 6 transfection reagent was purchased from Roche Diagnostics (Laval, QC, Canada), and Dual-Luciferase Reporter Assay System was from Promega (Madison, WI).
Plasmids.
pCMV6-XL4-hCAR-WT, pCMV6-neo-hCAR-SV23, pCMV6-XL4-hCAR-SV24, pCMV6-XL5-hCAR-SV25, pCMV6-XL4-hRXRα, pCMV6-XL4, pCMV6-neo, and pCMV6-XL5 were purchased from Origene (Rockville, MD). Renilla reniformis luciferase pGL4.74[hRluc/TK] was obtained from Promega. A pGL3-basic-CYP2B6-PBREM/XREM-luc reporter was constructed as described previously (Wang et al., 2003). The pVP16 and pM empty vectors were provided in the Matchmaker Mammalian Two-Hybrid Assay Kit (Clontech, Mountain View, CA). PathDetect pFR-luc trans-reporter plasmid was purchased from Agilent Technologies (Santa Clara, CA). To construct the pVP16-hCAR-WT-LBD (Gln-105 to Ser-348) and pM-hCAR-WT-LBD (Gln-105 to Ser-348) plasmids, the ligand-binding domain (Gln-105 to Ser-348) (Burk et al., 2005) of hCAR-WT was amplified from pCMV6-XL4-hCAR-WT and inserted into the pVP16 or pM vector. To construct the pVP16-hCAR-SV23-LBD (Gln-105 to Ser-352) and pM-hCAR-SV23-LBD (Gln-105 to Ser-352) plasmids, the ligand-binding domain (Gln-105 to Ser-352) (Arnold et al., 2004) of hCAR-SV23 was amplified from pCMV6-neo-hCAR-SV23 and inserted into the pVP16 or pM vector. To construct the pM-hSRC1-RID plasmid, the receptor-interacting domain of hSRC-1 containing three consensus LXXLL motifs (Asp-621 to Asn-765) (Chang et al., 1999) was amplified from pCMV6-XL5-NCOA1 (OriGene Technologies, Inc.) and cloned into the pM vector. To construct the pM-hSRC2-RID plasmid, the receptor-interacting domain of hSRC-2 containing three consensus LXXLL motifs (Lys-583 to Thr-779) (Arnold et al., 2004) was amplified from human liver QUICK-Clone cDNA (Clontech) and cloned into the pM vector. To construct the pM-hSRC3-RID plasmid, the receptor-interacting domain of hSRC-3 containing three consensus LXXLL motifs (Ser-582 to Asp-782) (Arnold et al., 2004) was amplified from human liver QUICK-Clone cDNA (Clontech) and cloned into the pM vector. The primers used to construct the plasmids are shown in Table 3. The plasmids were sequenced (Nucleic Acid Protein Service Unit, University of British Columbia, Vancouver, BC, Canada), and the identity of plasmids was confirmed by comparing their sequences with published sequences.
HepG2 Cell Culture.
HepG2 human hepatocellular carcinoma cells were purchased from the American Type Culture Collection (Manassas, VA) and cultured as described previously (Lau et al., 2010).
Transient Transfection and Reporter Gene Assays.
As reported previously, DEHP, which is a known activator of hCAR-SV23, is present in fetal bovine serum, but it is removable by charcoal treatment (DeKeyser et al., 2009). In the present study, all of the assays were conducted using HepG2 cells cultured in 10% (v/v) charcoal-stripped, heat-inactivated fetal bovine serum. HepG2 cells were seeded onto 24-well microplates at a density of 100,000 cells/well and in a volume of 0.5 ml of culture medium. hCAR-WT-, hCAR-SV23-, hCAR-SV24-, and hCAR-SV25-dependent reporter gene assays were conducted as follows. At 5 h after plating, HepG2 cells were transfected for 24 h with 20 μl of a transfection master mix containing FuGENE 6 transfection reagent (either 3 μl/μg of DNA or 0.4 μl/well, as specified in the figure legends), serum-free Opti-MEM (20 μl/well), pCMV6-XL4-hRXRα (10 ng/well, unless specified otherwise in the figure legends), pGL4.74[hRluc/TK] internal control vector (5 ng/well), pGL3-basic-CYP2B6-PBREM/XREM-luc (50 ng/well), and a hCAR expression plasmid (50 ng/well) or its respective empty vector (50 ng/well). The full-length hCAR expression plasmids were pCMV6-XL4-hCAR-WT, pCMV6-neo-hCAR-SV23, pCMV6-XL4-hCAR-SV24, and pCMV6-XL5-hCAR-SV25. In the hCAR-WT-LBD and hCAR-SV23-LBD assays, HepG2 cells were transfected with pM-hCAR-WT-LBD (Gln-105 to Ser-348; 40 ng/well), pM-hCAR-SV23-LBD (Gln-105 to Ser-352; 40 ng/well), or pM empty vector (40 ng/well) along with pCMV6-XL4-hRXRα (10 ng/well for hCAR-SV23-LBD assay and none for the hCAR-WT-LBD assay), pGL4.74[hRluc/TK] internal control plasmid (5 ng/well), and pFR-luc reporter plasmid (100 ng/well), using FuGENE 6 transfection reagent (3 μl/μg of DNA) diluted in 20 μl of serum-free Opti-MEM.
Transfected HepG2 cells were treated for 24 h with 0.5 ml of fresh supplemented culture medium containing G. biloba extract (50–800 μg/ml), ginkgolide A (8.8 μg/ml), ginkgolide B (2.4 μg/ml), ginkgolide C (11.2 μg/ml), ginkgolide J (4.8 μg/ml), bilobalide (22.4 μg/ml), sodium phenobarbital (1000 μM), CITCO (10 μM), DEHP (10 μM), TCPOBOP (0.25 μM), or vehicle, as detailed in each figure legend. The chemicals were dissolved in 100% DMSO and diluted with culture medium to give a final DMSO concentration of 0.1% (v/v). G. biloba extract was suspended directly in culture medium containing 0.1% (v/v) DMSO. In the hCAR-WT-dependent reporter gene assay, androstanol (10 μM), which is a hCAR-WT inverse agonist (Moore et al., 2000), was added to each treatment group to reduce the constitutive activity (Burk et al., 2005). In the hCAR-WT-LBD assay, PK11195 (10 μM) was used as the inverse agonist (Li et al., 2008) because androstanol was found to be ineffective. At the end of the treatment period, transfected HepG2 cells were lysed, and firefly luciferase and R. reniformis luciferase activities were determined by using a Dual-Luciferase Reporter Assay System. Luminescence was measured using a GloMax 96 microplate luminometer (Promega). Luciferase activity was expressed as a ratio of firefly luciferase to R. reniformis luciferase activity. Background luciferase activity was determined in HepG2 cells transfected with the corresponding empty vector (pCMV6-XL4, pCMV6-neo, or pCMV6-XL5).
Mammalian Two-Hybrid Assay.
A hCAR-dependent mammalian two-hybrid assay was performed as described in detail previously (Lau et al., 2011). At 5 h after plating, HepG2 cells were cotransfected with pVP16-hCAR-WT-LBD (40 ng/well), pVP16-hCAR-SV23-LBD (40 ng/well), or pVP16 empty vector (40 ng/well) along with a coactivator expression plasmid (10 ng/well), pCMV6-XL4-hRXRα (10 ng/well), pGL4.74[hRluc/TK] internal control plasmid (10 ng/well), and pFR-luc reporter plasmid (100 ng/well). The coactivator expression plasmids were pM-hSRC1-RID, pM-hSRC2-RID, and pM-hSRC3-RID. At 24 h after transfection, cells were treated for 24 h with 0.5 ml of fresh supplemented culture medium containing G. biloba extract, sodium phenobarbital, CITCO, DEHP, TCPOBOP, or the corresponding vehicle, as detailed in each figure legend. In the hCAR-WT-dependent mammalian two-hybrid assay, androstanol (10 μM; hCAR-WT inverse agonist) (Moore et al., 2000) was added to each treatment group to reduce the constitutive activity (Burk et al., 2005).
Statistical Analysis.
Data were analyzed by two-way analysis of variance, and when significant differences were detected, the Student Newman-Keuls multiple comparison test was performed (SigmaPlot 11.0; Systat Software, Inc., San Jose, CA). The level of statistical significance was set a priori at P < 0.05.
Results
G. biloba Extract Selectively Activates hCAR Isoforms: Comparison with Phenobarbital, DEHP, CITCO, and TCPOBOP.
A cell-based reporter gene assay (without cotransfecting a hRXRα expression plasmid) was conducted to evaluate the effect of G. biloba extract (800 μg/ml) on the activity of hCAR-WT, hCAR-SV23, hCAR-SV24, and hCAR-SV25. The extract increased hCAR-WT activity (Fig. 3A), but not that of hCAR-SV23 (Fig. 3B), hCAR-SV24 (Fig. 3C), or hCAR-SV25 (Fig. 3D), compared with both the corresponding empty vector-transfected, G. biloba extract-treated group and the hCAR-WT-transfected, vehicle-treated control group. By comparison, phenobarbital (1000 μM) activated hCAR-WT (Fig. 3A), but not the other hCAR isoforms (Fig. 3, B-D). As expected, among the hCAR isoforms investigated, DEHP (10 μM) increased only hCAR-WT (Fig. 3A) and hCAR-SV23 (Fig. 3B), whereas CITCO (10 μM) elevated the activity of hCAR-WT (Fig. 3A), hCAR-SV23 (Fig. 3B), and hCAR-SV24 (Fig. 3C). TCPOBOP at a maximal mCAR-activating concentration (0.25 μM; Tzameli et al., 2000) did not influence the activity of any of these hCAR isoforms.
Another finding in the experiment shown in Fig. 3 was that G. biloba extract, but not TCPOBOP, phenobarbital, DEHP, or CITCO, was capable of increasing luciferase activity in cells transfected with an empty vector (i.e., pCMV6-XL4, pCMV6-neo, or pCMV6-XL5). Therefore, to further explore the basis of this effect, an experiment was performed with various combinations of plasmids. As shown in Fig. 4, the reporter construct (pGL3-basic-CYP2B6-PBREM/XREM-luc) was responsible for the increase in luciferase activity in cells transfected with an empty vector and treated with G. biloba extract (800 μg/ml). Therefore, the design of all subsequent experiments included additional control groups in which the cells were transfected with the corresponding empty vector.
Cotransfection of a Retinoid X Receptor-α Expression Plasmid Leads to Activation of hCAR-SV23 but Not hCAR-SV24 or hCAR-SV25 by G. biloba Extract.
The lack of an effect by G. biloba extract on the hCAR splice variants SV23 (Fig. 3B), SV24 (Fig. 3C), and SV25 (Fig. 3D) may reflect a need for exogenous supplementation of HepG2 cells with RXRα. Therefore, a concentration-response experiment was conducted in which the cells were cotransfected with a human RXRα (hRXRα) expression plasmid (pCMV6-XL4-hRXRα; 1–50 ng/well). Cotransfection of pCMV6-XL4-hRXRα led to activation of hCAR-SV23 (Fig. 5C), but not hCAR-SV24 (Fig. 5E) or hCAR-SV25 (Fig. 5G), and it further elevated the extent of hCAR-WT activation by G. biloba extract (800 μg/ml; Fig. 5A). By comparison, cotransfection of a hRXRα expression plasmid resulted in enhanced activation of hCAR-SV23 (Fig. 5D) by DEHP (10 μM) and CITCO (10 μM) and hCAR-SV24 by CITCO (10 μM) and DEHP (10 μM) (Fig. 5F). In contrast, it did not have any effect in hCAR-SV23-transfected cells treated with TCPOBOP (0.25 μM) or phenobarbital (1000 μM) (Fig. 5D) or hCAR-SV24-transfected cells treated with TCPOBOP or phenobarbital at the above concentrations (Fig. 5F). Likewise, exogenous hRXRα supplementation did not influence hCAR-WT (Fig. 5B) or hCAR-SV25 (Fig. 5H) activity in cells treated with each of these individual chemicals, as analyzed by two-way analysis of variance. Given that hRXRα supplementation was necessary in hCAR-WT and hCAR-SV23 activation by G. biloba extract, subsequent reporter gene assays were conducted in cells cotransfected with the hRXRα expression plasmid (10 ng/well).
Multiple Lots of G. biloba Extract Activate hCAR-WT and hCAR-SV23, but Not hCAR-SV24 or hCAR-SV25.
We also compared the effect of five individual lots of G. biloba extract on hCAR-WT, hCAR-SV23, hCAR-SV24, and hCAR-SV25 activities. As shown in Fig. 6, A and B, each of the lots increased hCAR-WT and hCAR-SV23 activities to a comparable extent. In contrast, none of the lots affected hCAR-SV24 (Fig. 6C) or hCAR-SV25 (Fig. 6D) activity. Based on the similarity of results among multiple lots of extract, subsequent experiments were conducted using lot A. Given that the extract had no effect on hCAR-SV24 or hCAR-SV25 activity, the remaining experiments focused only on hCAR-WT and hCAR-SV23.
Concentration-Response Relationship in hCAR-WT and hCAR-SV23 Activation by G. biloba Extract.
A detailed concentration-response experiment was conducted to compare the effect of G. biloba extract (50–800 μg/ml) on hCAR-WT and hCAR-SV23 activities. Concentrations greater than 800 μg/ml of extract were not investigated because of solubility problems. Our previous data indicated that G. biloba extract at concentrations up to 800 μg/ml did not increase the release of lactate dehydrogenase (a marker of cytotoxicity) in HepG2 cells (Lau et al., 2010). As shown in Fig. 7A, the extract at concentrations of 50, 100, and 200 μg/ml had no effect on hCAR-WT activity, whereas at 400, 600, and 800 μg/ml, it increased hCAR-WT activity. By comparison, the extract at concentrations of 50, 100, 200, and 400 μg/ml had no effect on hCAR-SV23 activity, whereas at 600 and 800 μg/ml, it increased hCAR-SV23 activity (Fig. 7B). Overall, hCAR-WT was activated to a greater extent than hCAR-SV23 by G. biloba extract.
Ginkgolide A, Ginkgolide B, Ginkgolide C, Ginkgolide J, and Bilobalide Do Not Contribute to the Activation of hCAR-WT or hCAR-SV23 by G. biloba Extract.
To investigate whether any of the five individual terpene trilactones contributes to the activation of hCAR-WT and hCAR-SV23 by the extract, a reporter gene assay was conducted on transfected HepG2 cells treated with each of the chemicals at a level equivalent to those present in a hCAR-WT- or hCAR-SV23-activating concentration of the extract. Based on the concentration-response curves, an 800 μg/ml concentration of the extract was chosen because it was a concentration that activated both hCAR-WT (Fig. 7A) and hCAR-SV23 (Fig. 7B). Our results indicated that at a level present in an 800 μg/ml concentration of the extract, ginkgolide A (8.8 μg/ml), ginkgolide B (2.4 μg/ml), ginkgolide C (11.2 μg/ml), ginkgolide J (4.8 μg/ml), and bilobalide (22.4 μg/ml) did not activate hCAR-WT (Fig. 7C) or hCAR-SV23 (Fig. 7D). The terpene trilactones are present together in the G. biloba extract (Table 2). Therefore, to investigate the possibility of pharmacological synergism, we determined the effect of a combination of the five individual terpene trilactones at the above concentrations. However, it also had no effect, whereas in the same experiment, G. biloba extract (800 μg/ml) did activate hCAR-WT (Fig. 7C) and hCAR-SV23 (Fig. 7D).
G. biloba Extract Does Not Transactivate the Ligand-Binding Domains of hCAR-WT or hCAR-SV23.
To determine whether G. biloba extract transactivates the ligand-binding domains of hCAR-WT and hCAR-SV23, a reporter gene assay was conducted on HepG2 cells transfected with pM-hCAR-WT-LBD (Gln-105 to Ser-348) or pM-hCAR-SV23-LBD (Gln-105 to Ser-352). Whereas CITCO (10 μM) transactivated hCAR-WT-LBD (Fig. 8A) and DEHP (10 μM) transactivated hCAR-SV23-LBD (Fig. 8B), G. biloba extract (800 μg/ml), phenobarbital (1000 μM), and TCPOBOP (0.25 μM) had no effect (Fig. 8). Control analysis indicated that none of the treatment groups affected the activity in cells transfected with the pM empty vector (Fig. 8).
G. biloba Extract Does Not Promote Recruitment of Coactivators to hCAR-WT or hCAR-SV23.
A mammalian two-hybrid assay was performed to determine whether G. biloba extract is capable of recruiting coactivators (hSRC-1, hSRC-2, and hSRC-3) to the ligand-binding domains of hCAR-WT and hCAR-SV23. In HepG2 cells cotransfected with pVP16-hCAR-WT-LBD and pM-hSRC1-RID (Fig. 9A), pM-hSRC2-RID (Fig. 9B), or pM-hSRC1-RID (Fig. 9C), G. biloba extract (800 μg/ml) did not affect the luciferase activity, whereas an increase was obtained with CITCO (10 μM; a positive control). In contrast to DEHP (10 μM; a positive control), a lack of effect was obtained with the extract in cells cotransfected with pVP16-hCAR-SV23-LBD and pM-hSRC1-RID (Fig. 9D), pM-hSRC2-RID (Fig. 9E), or pM-hSRC1-RID (Fig. 9F). Phenobarbital (1000 μM) and TCPOBOP (0.25 μM) also had no effect on hSRC-1, hSRC-2, or hSRC-3 recruitment to hCAR-WT (Fig. 9, A–C) or hCAR-SV23 (Fig. 9, D–F). Control analysis indicated a lack of an effect of the extract and chemicals in cells cotransfected with pVP16 (empty vector) and pM-hSRC1-RID, pM-hSRC2-RID, or pM-hSRC3-RID (Fig. 9).
Discussion
The four-amino acid (SPTV) insertion in hCAR-SV23 has been postulated to extend the loop between helices 6 and 7, modify the properties of the ligand-binding pocket, and alter the ligand-binding specificity (Auerbach et al., 2003; DeKeyser et al., 2011). A novel finding in the present study is that among the hCAR splice variants investigated hCAR-SV23 was the only isoform activated by G. biloba extract, and this was verified with multiple lots of the extract. To date, G. biloba extract is one of the few activators of hCAR-SV23. The others are CITCO (DeKeyser et al., 2009; present study) and some of the phthalates, such as DEHP (DeKeyser et al., 2009; present study) and di-isononyl phthalate (DeKeyser et al., 2011). As shown for the first time in the present study, phenobarbital and TCPOBOP at concentrations known to activate hCAR-WT (Fig. 3A) and mCAR (Tzameli et al., 2000), respectively, were not capable of activating hCAR-SV23, as evaluated in our cell-based reporter gene assays. In contrast to DEHP, G. biloba extract did not transactivate the ligand-binding domain of hCAR-SV23 or promote recruitment of coactivators (hSRC-1, hSRC-2, and hSRC-3), indicating that the extract did not act as an agonist of this hCAR isoform. These results provide the first demonstration of an indirect mechanism of ligand-mediated activation of hCAR-SV23.
Our result indicating activation of hCAR-WT by G. biloba extract is in agreement with the only previous finding in which a 2-fold increase was obtained with an 100 μg/ml concentration of an extract (Li et al., 2009). However, based on two-way analysis of variance of our dose-response data, statistically significant activation of hCAR-WT was not obtained until the extract concentrations were 400, 600, and 800 μg/ml. In contrast to CITCO but the same as phenobarbital, G. biloba extract activates hCAR-WT by an indirect mechanism rather than by receptor agonism. This conclusion is based on the ability of CITCO and inability of phenobarbital and the extract to transactivate the ligand-binding domain of hCAR-WT and promote the recruitment of coactivators (hSRC-1, hSRC-2, and hSRC-3). Overall, it seems that G. biloba extract is a phenobarbital type rather than a CITCO type of hCAR-WT activator. The mechanism of CAR activation by phenobarbital is still not well understood. Based on the cumulative experimental evidence obtained for this drug, a key step is facilitating the translocation of the cytoplasmic receptor to the nucleus (Timsit and Negishi, 2007), and this has been shown to be associated with dephosphorylation of an amino acid residue in CAR (threonine 38 in hCAR and threonine 48 in mCAR) (Mutoh et al., 2009).
The G. biloba extract used in our study contained known quantities of ginkgolide A, ginkgolide B, ginkgolide C, ginkgolide J, and bilobalide. As assessed by reporter gene assays, these five individual terpene trilactones, when tested individually or in combination, were shown not to be responsible for the activation of hCAR-WT or hCAR-SV23 by the extract. In a previous study, the same experimental approach led to the identification of ginkgolide A as a partial contributor to human pregnane X receptor agonism by G. biloba extract (Lau et al., 2010). Currently, it is not known whether other chemical constituents in the extract contribute to hCAR-WT and hCAR-SV23 activation. Flavonol glycosides, such as those of quercetin, kaempferol, and isorhamnetin, are another class of chemicals present in G. biloba extract (van Beek and Montoro, 2009). However, it remains to be investigated whether they contribute to the hCAR-WT- and hCAR-SV23-activating effects of the extract, but it has been shown that quercetin aglycone (10 μM) does not activate hCAR-WT (Yao et al., 2010).
The five-amino acid (APYLT) insertion in hCAR-SV24 has been hypothesized to expand the loop between helices 8 and 9, cause steric hindrance during heterodimerization of hCAR-SV24 and hRXRα, and render little or no influence on the structure of the ligand-binding domain (Auerbach et al., 2003). In the present study, G. biloba extract did not activate hCAR-SV24 even in cells supplemented with exogenous hRXRα. However, in a previous study, the extract was shown to minimally (<2-fold) activate hCAR-SV24 in a cell-based reporter gene assay in which the empty vector-transfected, G. biloba extract-treated group was not included (Li et al., 2009). A possible reason for the apparent discrepancy may be related to our finding that G. biloba extract was capable of increasing the background luciferase activity in empty vector-transfected cells. In our study performed in cultured HepG2 cells, the increase could be attributed to the interaction between G. biloba extract and an endogenous receptor that is functionally compatible with our reporter plasmid (pGL3-basic-CYP2B6-PBREM/XREM-luc). However, the identity of that endogenous receptor is not pregnane X receptor even though a small amount of it may be present in HepG2 cells. This proposal is based on our data showing that ginkgolide A and phenobarbital, which are agonists of human pregnane X receptor (Moore et al., 2000; Lau et al., 2010), did not increase the reporter activity in cells transfected with the empty vector (i.e., in the absence of a receptor expression plasmid; Fig. 3). Overall, our results highlight the importance of conducting reporter gene assays with all of the appropriate control groups.
hCAR-WT and hCAR-SV24 have common ligands, including CITCO (Maglich et al., 2003; Faucette et al., 2007), artemisinin (Burk et al., 2005; Faucette et al., 2007), and phenytoin (Wang et al., 2004; Faucette et al., 2007). However, G. biloba extract (present study), DEHP (DeKeyser et al., 2011; present study), and resveratrol (Dring et al., 2010) activate hCAR-WT, but not hCAR-SV24. Conversely, other chemicals, such as 7-(acetoxy)-6-(p-methoxyphenyl)pyrrolo-[2,1-d][1,5]benzothiazepine (also referred to as NF49), nonylbenzene, doxylamine, and pheniramine, activate hCAR-SV24, but not hCAR-WT (Dring et al., 2010; Anderson et al., 2011). Collectively, these results indicate that hCAR-WT and hCAR-SV24 have overlapping, but distinct, sets of ligands, suggesting that the use of hCAR-SV24 in cell-based reporter gene assays as a screening tool (Faucette et al., 2007) may lead to false-positive and false-negative identification of hCAR-WT ligands.
hCAR-SV25 contains both the four-amino acid (SPTV) and five-amino acid (APYLT) insertions (Auerbach et al., 2003). Previously, only two chemicals were investigated for their potential interaction with hCAR-SV25, but both clotrimazole (Auerbach et al., 2003) and CITCO (Jinno et al., 2004) were shown not to activate hCAR-SV25. Our data confirmed the lack of effect of CITCO and showed for the first time that G. biloba extract, DEHP, phenobarbital, and TCPOBOP are not capable of activating hCAR-SV25 even in the presence of exogenous hRXRα. Overall, the experimental data indicate that the insertion of both the SPTV and APYLT amino acids leads to a loss of ligand recognition by hCAR-SV25, in agreement with predictions based on computer modeling studies (Auerbach et al., 2003; Savkur et al., 2003).
Heterodimerization of RXRα with CAR is a key step in the mammalian CAR activation pathway (Timsit and Negishi, 2007). A conclusion from the current study on the functionality of hCAR-SV23, hCAR-SV24, and hCAR-WT is that the necessity of exogenous RXRα supplementation in HepG2 cells depends on the specific ligand and hCAR isoform under investigation. This conclusion is based on the following patterns obtained in our experiments: 1) exogenous RXRα must be added to elicit receptor activation (e.g., G. biloba extract/hCAR-SV23 and DEHP/hCAR-SV24); 2) exogenous RXRα is not obligatory, but the additional input further enhances activity (e.g., CITCO/hCAR-SV23, DEHP/hCAR-SV23, CITCO/hCAR-SV24, and G. biloba extract/hCAR-WT); and 3) exogenous RXRα is not required, and supplementation does not further enhance activity (e.g., phenobarbital/hCAR-WT, DEHP/hCAR-WT, and CITCO/hCAR-WT). Therefore, before a study, preliminary experiments should be conducted to assess whether it is necessary to cotransfect a hRXRα expression plasmid in the cell type of interest when investigating ligand activation of hCAR isoforms in reporter gene assays.
In conclusion, G. biloba extract interacts with hCAR in an isoform-selective manner. It activated hCAR-WT and hCAR-SV23, but not hCAR-SV24 or hCAR-SV25. These effects of the extract were not caused by ginkgolide A, ginkgolide B, ginkgolide C, ginkgolide J, or bilobalide. In contrast to CITCO and DEHP (DeKeyser et al., 2009), G. biloba extract did not activate hCAR-WT or hCAR-SV23 by receptor agonism, but by an indirect mechanism that did not involve binding to the ligand-binding domain of the receptor or stimulating the recruitment of coactivators. The demonstration of hCAR activation by G. biloba extract may provide a mechanistic insight for the neuroprotective effect of this herbal medicine against β-amyloid toxicity, which has been reported in various experimental models (Luo et al., 2002). As shown in in vitro and ex vivo experiments, chemical activation of CAR up-regulates P-glycoprotein expression in rodent brain capillaries (Wang et al., 2010), and P-glycoprotein is known to reduce the accumulation of brain β-amyloid (Hartz et al., 2010), which is believed to play a role in the etiology of Alzheimer's disease (Zlokovic, 2005). Finally, our data along with those reported recently (Dring et al., 2010; Anderson et al., 2011) support the conclusion that hCAR-SV23, hCAR-SV24, and hCAR-WT have overlapping, but distinct, sets of ligands.
Authorship Contributions
Participated in research design: Lau and Chang.
Conducted experiments: Lau and Yang.
Performed data analysis: Lau.
Wrote or contributed to the writing of the manuscript: Lau and Chang.
Acknowledgments
We thank Indena S.p.A. (Milan, Italy) for the generous provision of G. biloba extracts.
Footnotes
This research was supported by the Canadian Institutes of Health Research [Grant MOP-84581]. T.K.H.C. received a Senior Scholar Award from the Michael Smith Foundation for Health Research.
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
doi:10.1124/jpet.111.186130.
-
ABBREVIATIONS:
- CAR
- constitutive androstane receptor
- hCAR
- human CAR
- mCAR
- mouse CAR
- CITCO
- 6-(4-chlorophenyl)imidazo[2,1-b][1,3]thiazole-5-carbaldehyde O-(3,4-dichlorobenzyl)oxime
- DEHP
- di-(2-ethylhexyl)phthalate
- DMSO
- dimethyl sulfoxide
- RXRα
- retinoid X receptor α
- hRXRα
- human RXRα
- hSRC
- human steroid receptor coactivator
- PK11195
- 1-(2-chlorophenyl)-N-methyl-N-(1-methylpropyl)-3-isoquinoline-carboxamide
- TCPOBOP
- 1,4-bis-[2-(3,5-dichloropyridyloxy)]benzene
- WT
- wild type
- CAS
- Chemical Abstracts Service
- bp
- base pair(s)
- aa
- amino acids.
- Received July 18, 2011.
- Accepted August 22, 2011.
- Copyright © 2011 by The American Society for Pharmacology and Experimental Therapeutics