Potential conflict of interest: Nothing to report.
Organic anion transporting polypeptide 1B1 (OATP1B1) is a liver-enriched transporter involved in the hepatocellular uptake of many endogenous molecules and several structurally divergent drugs in clinical use. Although OATP1B1 coding region polymorphisms are known to make an impact on substrate drug disposition in humans, little is known regarding the mechanisms underlying the transcriptional regulation of this transporter. In this study, we note that messenger RNA (mRNA) expression of OATP1B1 in a large human liver bank exhibited marked interindividual variability that was not associated with coding region polymorphisms. Accordingly, we hypothesized that such variability in expression is reflective of nuclear receptor-mediated transcriptional regulation of this transporter. We tested prototypical ligands for the nuclear receptors pregnane X receptor (PXR), constitutive androstane receptor (CAR), liver X receptor (LXR) α, and farnesoid X receptor (FXR) in a human hepatoma-derived cell line and noted induction of OATP1B1 mRNA when the cells were treated with LXRα or FXR ligands. To confirm a direct role for LXRα and FXR to OATP1B1 expression, we performed detailed promoter analysis and cell-based reporter gene assays resulting in the identification of two functional FXR response elements and one LXRα response element. The direct interaction between nuclear receptors with the identified response elements was assessed using chromatin immunoprecipitation assays. Using isolated primary human hepatocytes, we show that LXRα or FXR agonists, but not PXR or CAR agonists, are capable of OATP1B1 induction. Conclusion: We note that OATP1B1 transcriptional regulation is under dual nuclear receptor control through the oxysterol sensing LXRα and the bile acid sensor FXR. Accordingly, the interplay between OATP1B1 and nuclear receptors may play an important and heretofore unrecognized role during cholestasis, drug-induced liver injury, and OATP1B1 induction–related drug interactions. (HEPATOLOGY 2010)
OATP1B1 is a member of the organic anion transporting polypeptide (OATP) family of membrane transporters, which is increasingly recognized as a critical determinant in mediating the hepatocellular uptake of many drugs in clinical use today.1 Studies from our laboratory and others have shown that OATP1B1 is highly expressed in human liver and likely functions as a rate-limiting step in the overall hepatobiliary clearance of several compounds.2-4 Indeed, several functionally relevant single-nucleotide polymorphisms (SNPs) in the coding region of this uptake transporter have been linked to significant alterations in the pharmacokinetic profiles of substrate drugs in humans.5, 6 However, in addition to coding region differences that affect its transport activity, it is likely that mechanisms that govern the transcriptional regulation of OATP1B1 are also important to the overall in vivo activity of OATP1B1-mediated hepatic drug uptake.
It is now widely understood that the superfamily of nuclear receptors that function as constitutive and ligand-activated intracellular receptors are capable of transcriptional activation of many genes, particularly those involved in drug metabolism and transport through binding to receptor-specific DNA motifs in the promoter region of target genes.7 Remarkably, despite the recognized importance of OATP1B1 in the transport of ligands of nuclear receptors such as bile acids8, 9 and drugs, few data exist regarding the role of ligand sensing activators such as pregnane X receptor (PXR), constitutive androstane receptor (CAR), farnesoid X receptor (FXR), and liver X receptor (LXR) to OATP1B1 expression.
In this study, we demonstrate that human OATP1B1 expression is regulated by both LXRα and FXR and not by PXR or CAR. Importantly, we show that treatment of liver-derived cell lines using known ligands of these receptors result in a significant induction of OATP1B1 messenger RNA (mRNA) expression and transport activity. Such an effect was demonstrated to be the result of a direct interaction between activated FXR and LXRα and specific DNA binding motifs in the 5′ untranslated region (UTR) of the SLCO1B1 gene. Furthermore, we show that LXRα and FXR agonists induce OATP1B1 expression in freshly isolated human hepatocytes, thus suggesting that nuclear receptors likely play an important role in vivo in the regulated expression of OATP1B1.
Rifampin, CITCO, thyroxine, 9-cis retinoic acid, and TO-901317 were purchased from Sigma-Aldrich (St. Louis, MO). GW4064, fexaramine, and GW3965 were obtained from Tocris Biosciences (Cedarlane Laboratories Ltd., Hornby, Ontario, Canada). Tritium-labeled taurocholic acid was obtained from PerkinElmer Life Sciences (Waltham, MA). Tritium-labeled rosuvastatin was obtained from American Radiolabeled Chemicals (St. Louis, MO).
Huh-7 cells (clone JCRB0403) were purchased from the Japanese Collection of Research Bioresources (http://cellbank.nihs.go.jp) and HepG2 cells were obtained from American Tissue Culture Collection (Manassas, VA). Freshly isolated human hepatocytes from five different individuals were obtained from Lonza Verviers SPRL (Verviers Belgium).
Gene Expression Analysis from Human Liver Samples.
Human liver mRNA expression data were obtained from GEO GS39588, which was conducted using a custom Agilent 44,000 feature microarray composed of 39,280 oligonucleotide probes targeting transcripts representing 34,266 known and predicted genes, including high-confidence, noncoding RNA sequences samples.10 Four hundred twenty-three samples from the published dataset were included for analysis in this study and visualized using principle component analysis.
DNA from a subset of the human liver tissue samples (n = 60) provided by the Liver Tissue Procurement and Distribution System (NIH Contract #N01-DK-9-2310) and by the Cooperative Human Tissue Network and processed through the St. Jude Liver Resource at St. Jude Children's Research Hospital was genotyped for common SNPs of SLCO1B1. The SLCO1B1 haplotypes *1b (c.388A>G, rs2306283), *5 (c.521C>T, rs2306283), and *15 (c.388A>G & c.521C>T) were determined by way of direct sequencing.
RNA Isolation from Cultured Cells.
Phenol-chlorophorm extraction was performed to isolate RNA from in vitro experiments using Trizol (Invitrogen, Carlsbad, CA) following the manufacturer's instructions. The integrity and content of the RNA was determined using an Agilent Bioanalyzer (Agilent, Santa Clara, CA). RNA samples were stored at −80°C.
Real-Time Polymerase Chain Reaction.
Total RNA was reverse-transcribed in a 50-μL reaction volume containing 1,500 ng of RNA with a TaqMan Reverse Transcription Kit (Applied Biosystems, Foster City, CA). Primer pairs used were as follows: OATP1B1 forward, 5′-TGAACACCGTTGGAATTGC-3′; OATP1B1 reverse, 5′-TCTCTATGAGATGTCACTGGAT-3′; NTCP forward, 5′-ACTGGTCCTGGTTCTCATTCC-3′; NTCP reverse, 5′-GTGGCAATCAAGAGTGGTGTC-3′. 18S ribosomal RNA, ABCA1, BSEP (ABCB11), and OATP1B3 were quantified using a predeveloped TaqMan Assay (Hs03003631, Hs01059118, Hs00994811, and Hs00251986, Applied Biosystems). Transporter expression was normalized to that of 18S ribosomal RNA.
In Silico Scan for DNA Binding Motifs of Nuclear Receptors.
A 10-kb fragment upstream of the transcription start of the SLCO1B1 gene was screened for potential NR binding sites using the NUBIscan algorithm (http://www.nubiscan.unibas.ch).
The SHP1-Luc-pGL3-basic vector was prepared as described11 containing the −571 to +1 bp fragment of the 5′-UTR of SHP1 (NR0B2) gene inserted in the luciferase reporter containing vector pGL3-basic (Promega, Madison, WI). The BSEP promoter construct containing a 140-bp fragment of the 5′-UTR BSEP (ABCB11) has been described.12
Detailed outline of primers and subcloning strategies for the promoter fragments of the 5′-UTR of the SLCO1B1 from genomic DNA is summarized in Supporting Table 1.
OATP1B1 Reporter Gene Assays.
HepG2 cells were plated in 24-well plates. After 24 hours, the cells were transfected with 250 ng of the reporter vector (pGL3 basic variants), 25 ng of pRL-CMV (Promega) to normalize transfection efficiency, and 250 ng of the human nuclear receptors expression plasmid (LXRα- or FXR-pEF6) or vector control in 200 μL Opti-MEM (Invitrogen) using Lipofectin (Invitrogen). Cells were incubated for 16 hours with the transfection mixture, then treated for 24 hours with 1 μM of a tested compound. Luminescence was quantified using a plate reader (Fluoroskan Ascent FL, Thermo-Fisher, Waltham, MA). Luciferase activities in the presence of the nuclear receptor were expressed as the percentage of cells transfected with blank vector.
Cellular Uptake Experiments.
Huh-7 cells or primary hepatocytes were grown in 12-well plates and pretreated with agonists of LXRα or FXR or vehicle control, respectively, for 12 hours. Afterward, the cells were briefly washed with OptiMEM and then incubated with tritium-labeled taurocholate (0.4 μM) or rosuvastatin (1 μM). After 10 minutes of incubation at 37°C, the cells were washed twice with ice-cold PBS and lysed in the presence of 1% sodium dodecyl sulfate. Cellular uptake was determined using a liquid scintillation counter (Tri-Carb 2900TR, PerkinElmer Life Sciences).
Western Blot Analysis of OATP1B1.
After treatment of human hepatocytes, the cells were harvested and lysed by way of repeated thawing and freezing in 5 mM Tris HCl in the presence of Protease Inhibitors. Protein content was determined using bicinchoninic acid. Cell lysates were separated by way of sodium dodecyl sulfate–polyacrylamide gel electrophoresis and electrotransferred to a nitrocellulose membrane in a tank-blotting system (Bio-Rad, Munich, Germany). For detection of OATP1B1, a specific antiserum was used as described.4
Chromatin Immunoprecipitation Assay.
For DNA cross-linking and chromatin immunoprecipitation, the EZ ChIP Assay (Millipore, Billerica, MA) was used according tot the manufacturer's instructions. Briefly, Huh-7 cells were cultured in 10-cm dishes and treated for 24 hours with dimethyl sulfoxide, chenodeoxycholic acid (CDCA) (1 μM), fexaramine (1 μM), GW4065 (1 μM), TO-901317 (1 μM), or GW3965 (1 μM). DNA was cross-linked, sheared by sonication (Virsonic 100, Virtis, Gardiner, NY), then incubated with 5 μg antibody overnight at 4°C. Polyclonal anti-FXR, anti-LXRα (Santa Cruz Biotechnology, Santa Cruz, CA), or anti-RXRα (ab24363; Abcam, Cambridge, UK) sera were used. A control was performed using an anti–acetyl-histone (H3) antibody. The antibody/antigen/chromatin complex was gathered with protein G agarose and centrifugation. After several washing steps, the antibody/chromatin complex was eluted and bound DNA was released by incubation at 65° C overnight after adding 8 μL of 5 M NaCl, treated with RNaseA and ProteinaseK. Binding was confirmed by way of polymerase chain reaction (PCR) amplification (see Supporting Table 2 for primer sequences).
Statistical differences between group parameters were determined using a Student t test and Mann-Whitney U test using Prism software (GraphPad Software, Inc., San Diego, CA). P < 0.05 was considered the minimum level of statistical significance.
Interindividual Variability of OATP1B1 Expression in Human Liver.
OATP1B1 mRNA was assessed performing a genome-wide expression analysis using a custom Agilent 44,000 feature microarray of a human liver bank (n = 423) and revealed marked variability (Fig. 1A). However, there was no correlation between OATP1B1 mRNA expression and the presence of *1b, *5, or *15 SLCO1B1 SNPs (analysis of variance, P = 0.143) (Fig. 1B).
Treatment of Huh-7 Cells with Prototypical Nuclear Receptor Ligands.
When human hepatoma Huh-7 cells, which exhibit low but sufficiently detectible OATP1B1 expression (CT value of 34 compared to liver CT value of 21), were treated with rifampin (PXR), thyroxine (THR), CITCO (CAR), TO-901317 (LXRα), or CDCA (FXR), statistically significant induction of OATP1B1 mRNA was only seen in cells treated with the LXRα and FXR agonist (Fig. 2A). The synthetic FXR agonists GW4064 and fexaramine were also able to mediate a significant (four-fold) increase in OATP1B1 transcription (Fig. 2B). Although our data revealed a lack of PXR effect on OATP1B1 expression as assessed using rifampin as a prototypical ligand (Fig. 2A), because the LXRα agonist TO-901317 is thought to also possess PXR activation capacity,13 we confirmed the initial observation of OATP1B1 activation by LXRα by testing another synthetic LXRα agonist (GW3965) in a similar manner. Indeed, both LXRα agonists induced OATP1B1 expression by approximately three-fold in Huh-7 cells (Fig. 2C).
Transport Activity of OATP1B1 After Treatment with FXR and LXRα Ligands.
To confirm that the observed increase in OATP1B1 mRNA is reflected as functional transport activity, transport of the known OATP1B1 substrates taurocholate and rosuvastatin was assessed after treatment for 24 hours with FXR and LXRα agonists, respectively. Determining the [3H]taurocholate uptake in Huh-7 cells revealed significantly greater cellular accumulation after treatment with FXR or LXRα agonists, respectively (Fig. 2D). Similar results were seen for rosuvastatin (Fig. 2E).
In Silico Identification of Potential Nuclear Receptor Binding Sites Using NUBIscan.
Because our cell line data strongly suggested that both LXRα and FXR were involved in the transcriptional regulation of OATP1B1, we looked for potential nuclear receptor response elements in the SLCO1B1 gene. Because we now know that for many drug disposition genes, key nuclear receptor response elements are often far upstream of the transcription initiation site, we assessed a 10-kb fragment of the SLCO1B1 gene promoter using the NUBIscan algorithm14 and noted the presence of several potential nuclear receptor response elements. Included in the search were several DNA motifs of tandem hexameric repeats with various spacing and orientation. Species-related sequence homology is shown in Supporting Fig. 1. As shown in Fig. 3A, several potential binding sites were identified.
Identification of Functional FXR Response Elements in the 5′-UTR of the SLCO1B1 Gene.
Several inverted repeats with one spacing base pair (IR1) known to be potential binding sites for FXR in the 5′-UTR of SLCO1B1 were identified using the NUBIscan algorithm gene: IR1-1 (AGGTCAaAGAGCA) located at −1545 bp (P = 0.176); IR1-2 (AGGTTAtTTACCA) located at −1850 bp (P = 0.045); IR1-3 (AGGACAcTACCCT) located at −4041 bp (P = 0.051), and IR1-4 (GTGTTTgTGACCT) located at −4165 bp (P = 0.493).
Promoter constructs containing the −3040 bp to −4070 bp or the −1480 bp to −2500 bp fragment of the SLCO1B1 5′-UTR were significantly activated by FXR when treated with CDCA (Fig. 3B) or the synthetic FXR activators GW4064 (10 μM) or fexaramine (10 μM) (data not shown).
Interestingly, mutation of both IR1 DNA motifs resulted in the complete loss of CDCA-stimulated, FXR-dependent, luciferase reporter activity in HepG2 cells (Fig. 4A). We further confirmed the role of these IR1 elements in the inductive regulation of OATP1B1 using chromatin immunoprecipitation assay (Fig. 4B,C). These results demonstrate that activated FXR binds to the SLCO1B1 promoter and strongly suggest that the identified IR1 motifs are the key elements responsible for FXR-mediated transactivation of OATP1B1 expression.
Subsequently, we assessed for the effects of the heterodimerization partner retinoid X receptor (RXR) α on the CDCA-mediated transactivation of SLCO1B1 promoter constructs. As shown in Fig. 5A,B, we noted that the promoter constructs containing the −1480 bp to −2500 bp or the −3040 bp to −4070 bp upstream sequences showed a moderate increase in luciferase activity when transfected with RXRα and treated with CDCA (1 μM). Treatment with the RXR ligand 9-cis retinoic acid (RA) alone did not have any effect on promoter activation, even when RXRα was transfected. However, treatment with CDCA in the presence of 9-cis retinoic acid resulted in a statistically significant reduction of the FXR-mediated transactivation of the promoter constructs compared with CDCA alone. This phenomenon has been described before by Kassam et al.,15 who explained this phenomenon by a reduction in coactivator recruitment to result in decreased DNA binding of FXR. Our findings further support this observation. Indeed, we see decreased binding to the identified FXREs in cells concomitantly treated with CDCA and 9-cis retinoic acid performing chromatin immunoprecipitation analysis for RXRα (Fig. 5C,D).
Identification of Functional LXR Response Elements in the 5′-UTR of the SLCO1B1 Gene.
To evaluate whether the predicted binding sites were functional, promoter constructs containing these identified DR4 DNA motifs were screened for their activity in a cell-based reporter assay. Addition of the known LXR-activator TO-901317 (10 μM) resulted in LXR-dependent enhanced luciferase activation of all constructs containing the first part of +53 to at least −128 bp of the SLCO1B1 gene (Fig. 3C). However, we did not observe any enhancement of luciferase activity in the fragment containing the potential distal enhancer module (−5450 bp to −4620 bp). Similar results were obtained treating the HepG2 cells with the LXRα agonist GW3965 (Fig. 3C).
Because all of the constructs include at least the +53 to −128 bp fragment of the SLCO1B1 gene, we focused on the two possible LXR response elements located in this region (Fig. 6A). Interestingly, mutation of the hexamere DR4 DNA motifs (localized at DR4-1 +22 to +37 and DR4-2 +32 to +47) resulted in the complete loss of agonist-stimulated, LXRα-dependent luciferase reporter activity in HepG2 cells (Fig. 6B) in all fragments. These findings suggest that this part of the SLCO1B1 promoter is transactivated by agonist-bound LXRα. To further confirm the role of the DR4 elements in the inductive regulation of OATP1B1 expression, we performed an LXR-specific chromatin immunoprecipitation assay (Fig. 6C-E). These results demonstrate that agonist-activated LXR binds to the SLCO1B1 promoter. In accordance with the findings concerning the RXRα–FXR interaction, experiments were conducted using LXRα. As shown in Supporting Fig. 2, there is no additional effect of RXRα on SLCO1B1 promoter response.
Treatment of Freshly Isolated Human Hepatocytes with FXR and LXRα Ligands.
In order to confirm that our cell line–based findings are reflective of an in vivo situation, we performed ex vivo experiments using freshly isolated human hepatocytes. The inductive capacity of the human hepatocyte preparations for LXRα and FXR agonists were confirmed by assessing the effects of such agonists to bona fide target genes such as BSEP16 (Fig. 7C) and OATP1B317, 18 (Fig. 7D) for FXR and ABCA1 for LXRα19 (Fig. 7B). As shown in Fig. 7A, treatment of freshly isolated human hepatocytes with CDCA and TO-901317 resulted in a significant induction of OATP1B1 expression. Addition of CDCA and TO-091317 to two additional human hepatocytes preparations revealed a moderate induction of OATP1B1 protein expression (Fig. 7E). Functional assessment of OATP1B1 in the same hepatocytes showed significantly greater cellular accumulation of [3H]taurocholate acid in the presence of CDCA or TO-091317, respectively (Fig. 7F). Uptake of CCK-8, a specific substrate of OATP1B3, was increased in hepatocytes treated with the FXR ligand CDCA (CCK-8 uptake percentage of dimethyl sulfoxide control 159.73 ± 19.07), but not TO-091317 (127.13 ± 24.12) (data not shown), suggesting the LXR effect is OATP1B1-specific.
OATP1B1 is now emerging as a key transporter for the hepatic uptake of many compounds.1 Although much is currently known regarding the functional relevance of coding region SNPs in this drug transporter, remarkably little is known regarding the mechanisms that mediate its transcriptional regulation. Indeed, mRNA expression analysis from a large human liver bank would suggest marked interindividual differences in OATP1B1 mRNA levels regardless of the presence of the known OATP1B1 SNPs (Fig. 1A). Recently, an additional polymorphism (−11187G>A, part of the SLCO1B1*17 haplotype that includes the 521T>C coding region SNP) in the SLCO1B1 gene has been identified. This SNP is located in the 5′-UTR of the SLCO1B1 gene,20 although as yet there is no evidence to suggest this SNP has a major impact on OATP1B1 expression.
It should be noted that although hepatocyte nuclear factor (HNF) 1 can interact with OATP1B1 and that HNF4α can affect HNF1 activity, thereby indirectly regulating OATP1B1 expression,21 there are no published data regarding direct regulation of OATP1B1 by ligand-activated nuclear receptors such as PXR, CAR, FXR, and LXRα. We report that OATP1B1 expression is positively regulated by the bile acid sensor FXR and the cholesterol sensor LXRα. The key FXR response element appears to be approximately −4 kb upstream of the transcription initiation site, whereas the key LXRα binding motif is in close proximity to the transcription initiation site.
Moreover, previous studies have suggested bile acids suppress OATP1B1 expression and that this mechanism is a part of the FXR-mediated protection against hepatocellular injury induced by cytotoxic bile acids.22 This hypothesis was based on findings that showed reduced basal SLCO1B1 promoter (−950 bp to +21 bp) gene activity in Huh-7 cells and decreased OATP1B1 expression in human liver slices, associated with bile acid treatment.23, 24 Furthermore, it has been shown that OATP1B1 expression is negatively correlated with bile acid levels in human patients where lower OATP1B1 expression is observed in livers of patients with inflammation-induced cholestasis.25 Similar observations have been reported for patients with chronic cholestatic liver diseases such as progressive familial intrahepatic cholestasis type II (hereditary insufficiency of the apical bile acid pump BSEP) and type III (hereditary insufficiency in the apical-located phosphatidylcholin flipase MDR3).26 However, as noted by Ostrow and Tiribelli,27 in such studies, the control groups were composed of adult livers with different hepatic diseases, and in such cases, there was a significant modulation of OATP1B1 expression, whereas liver tissue samples of healthy adults did not show any difference in expression.
Our result showing a direct regulation of OATP1B1 expression by FXR is consistent with the findings from our laboratory in which we observed an association between a commonly occurring FXR polymorphism and OATP1B1 expression in human liver samples.28 Indeed, our current results provide the mechanistic basis for our previous observation. It has been hypothesized before that down-regulation of OATP1B1 is a part of a mechanism to protect hepatocytes from the accumulation of bile acids due to its ability to transport bile acids in a sodium-independent manner.29 However, based on our current observation, it is plausible that OATP1B1 functions as a key pathway in the network for modulation of hepatic bile acid concentration through its ability to mediate the sodium-independent hepatic uptake of bile acids and thereby enhance bile acid sensing through FXR and modulation of target gene expression. It seems noteworthy that another hepatic OATP capable of bile acid uptake, OATP1B3,29 has been shown to also be positively regulated by FXR.17, 18 Indeed, in human hepatocytes, we were able to confirm that treatment with CDCA results in OATP1B3 induction (Fig. 7). However, unlike OATP1B1, OATP1B3 does not appear to be regulated by LXRα (Fig. 7).
We hypothesize that regulation of the bile acid transporters is multifactorial and includes several components (Fig. 8). Protein kinase A exhibits cyclic adenosine monophosphate (cAMP)-dependent catalytic activity and is involved in the regulation of several intracellular processes, including the activity of transcription factors such as HNF4α as well as OATP1B1.30 HNF4α not only regulates OATP1B1, but also the expression of NTCP (SLC10A1), the sodium-dependent transporter for bile acids.31, 32 Introducing an additional factor to this network of OATP1B1 expression is the G protein–coupled receptor TGR5, which induces intracellular cAMP levels upon binding of bile acids.33 Thus, increased bile acid levels would reduce the expression of both bile acid transporters through suppression of HNF4α activity and expression of the transporters that facilitate the uptake of bile acid FXR ligands. This notion is supported by findings showing that cAMP protects against hepatocellular apoptosis induced by hydrophilic bile acids such as GCDCA,34 although expression and function of TGR5 in human hepatocytes is controversial.35, 36 There are reports suggesting moderate but functional expression of TGR5 in hepatocytes.29
In terms of LXRα, there has been significant progress using LXRα as a therapeutic target to treat metabolic disorders and atherosclerosis.37 Indeed, our observed effects of an LXRα agonist in human hepatocytes suggest that such a strategy might result in the induction of hepatic drug transporters such as OATP1B1 (Fig. 6), which for drugs such as the statin class of HMG-Co-A reductase inhibitors would result in a higher liver concentration of the drug while lowering systemic exposure. This may be viewed as a therapeutically beneficial effect of LXRα. Given the importance of regulated conversion of cholesterol to bile acids by LXRα target genes, regulation of OATP1B1 by LXRα is consistent with an important physiological role of OATP1B1 to hepatic cholesterol and bile acid homeostasis
In conclusion, we show for the first time that OATP1B1 is dual nuclear receptor–regulated through the actions of the bile acid sensor FXR and the cholesterol sensor LXRα, but not by the typical xenobiotic receptors such as PXR and CAR. Because the substrate specificity for OATP1B1 is quite broad and includes several endobiotics and xenobiotics, its regulation by FXR and LXRα sheds important new insights regarding the role of this transporter in drug disposition and suggests that intersubject variability in FXR and LXRα activity may be a heretofore unrecognized determinant of OATP1B1 expression and function.