Abstract
The pregnenolone X receptor (PXR), a new member of the nuclear hormone receptor superfamily, was recently demonstrated to mediate glucocorticoid agonist and antagonist activation of a hormone response element spaced by three nucleotides (DR-3) within the ratCYP3A23 promoter. Because many other steroids and xenobiotics can up-regulate CYP3A23 expression, we determined whether some of these other regulators used PXR to activate theCYP3A23 DR-3. Transient co-transfection of LLC-PK1 cells with (CYP3A23)2-tk-CAT and mouse PXR demonstrated that the organochlorine pesticides transnonachlor and chlordane and the nonplanar polychlorinated biphenyls (PCBs) each induced the CYP3A23 DR-3 element, and this activation required PXR. Additionally, this study found that PXR is activated to induce (CYP3A23)2-tk-CAT by antihormones of several steroid classes including the antimineralocorticoid spironolactone and the antiandrogen cyproterone acetate. These studies reveal that PXR is involved in the induction of CYP3A23 by pharmacologically and structurally distinct steroids and xenobiotics. Moreover, PXR-mediated PCB activation of the (CYP3A23)2-tk-CAT may serve as a rapid assay for effects of nonplanar PCBs.
The human and rat CYP3A forms are induced by numerous steroidal hormones; a hallmark feature is their nonclassical induction by glucocorticoids (Schuetz and Guzelian, 1984; Schuetz et al., 1984). Characteristics of this nonclassical induction include requirements for pharmacological amounts of steroid to induce CYP3A as well as the paradoxical induction of CYP3A231 by both the glucocorticoid agonist dexamethasone and the antiglucocorticoids PCN (Schuetz et al., 1984) and RU486 (Kocarek et al., 1995). We previously hypothesized that other classes of antihormones might share regulatory characteristics and induce CYP3A as well (Kocarek et al., 1995). Further studies revealed that antihormone representatives of many of the major steroid classes, including the antimineralocorticoid spironolactone and the antiandrogen cyproterone acetate, were each able to induce hepatic CYPA3A in rat and human hepatocytes (Kocarek et al., 1995) and led us to speculate about a common regulatory factor (Kocareket al., 1995). Additionally, CYP3A23 is also induced by important environmental xenobiotics including organochlorine pesticides (e.g., transnonachlor and chlordane), and some polychlorinated biphenyls, particularly those with ortho-chlorines (Schuetz et al., 1986).
A new member of the nuclear hormone receptor superfamily, the PXR that mediates glucocorticoid agonist and glucocorticoid antagonist induction of the CYP3A23 gene has recently been identified (Kliewer et al., 1998). The glucocorticoid agonist or antagonist activated PXR:RXR heterodimer bound to a hormone response element, an AGTTCAtgaAGTTCA direct repeat with a three-nucleotide spacer (DR-3), in the 5′-flanking region of the CYP3A23gene to induce its transcription. A number of other hormones (e.g., pregnenolone and progesterone) were also identified as activators of PXR. We have tested the hypothesis that other classes of steroidal antihormones and environmental chemicals might represent additional activators of PXR. We found that transcriptional activation of theCYP3A23 DR-3 hormone response element by many of these structurally diverse compounds is mediated by a single PXR induction process.
Experimental Procedures
Materials.
(CYP3A23) 2-tk-CAT and mouse PXR.1 (hereafter referred to as PXR) were kindly provided by Dr. Steven Kliewer (Glaxo Wellcome Research and Development, Research Triangle Park, NC). Dexamethasone-t-butylacetate was from Research Plus (Bayonne, NJ), organochlorine pesticides from Velsicol (Chicago, IL), and PCBs were given by Stephen Safe (Texas A & M, College Station, TX).
Cell culture.
LLC-PK1 pig kidney epithelial cells were obtained from American Type Culture Collection (Rockville, MD) and cultured in medium 199 (Life Technologies, Gaithersburg, MD) containing 10% fetal calf serum.
Transfection assays.
LLC-PK1 cells were plated in 24-well dishes at 2.0 × 105 cells per well. Twenty-four hours later, medium was changed and cells transfected with 200 ng of (CYP3A23) 2-tk-CAT, 66 ng of mPXR and 300 ng of TK-Luciferase by calcium phosphate overnight. The next day, cells were washed once with medium and refed with medium containing 10% charcoal-stripped delipidated calf serum (Sigma) and xenobiotics or steroids. All steroids and xenobiotics were dissolved in dimethyl sulfoxide, with the dimethyl concentration in medium not exceeding 0.1%. Twenty-four hours later, cells were washed once with phosphate-buffered saline and lysed in 100 μl of 1× luciferase lysis buffer according to manufacturer’s instructions (Promega); 35 μl was assayed for luciferase activity and 35 μl was assayed for CAT activity as previously described (Burger et al., 1992). Transfections containing (CYP3A23) 2-tk-CAT were normalized to TK-luciferase activities. In some cases, BioRad (Richmond, CA) protein assays were performed on 10 μl of remaining lysate according to the manufacturer’s instructions.
Results
LLC-PK1 cells were chosen for transfection studies because multiple derivative LLC-PK1 cell lines stably expressing a variety of drug and steroid efflux transporters have been made (Schinkel et al., 1995; Evers et al., 1996). Therefore, if LLC-PK1 cells proved suitable for these transfection studies of PXR, the parent and derivative cell lines could be used in future experiments examining the influence of these transporters on PXR activation. Immunoblot analysis of LLC-PK1 cell lysates with anti-RXR IgG (Affinity Bioreagents) revealed immunodetectable RXR (data not shown). Hybridizable PXR mRNA transcripts were only detectable in lysates of LLC-PK1 cells transfected with the PXR expression vector, but not lysates of nontransfected LLC-PK1 cells (data not shown). We first determined whether PXR could induce the CYP3A23 DR-3 in LLC-PK1 cells treated with established PXR ligands (Kliewer et al., 1998). Similar to findings in CV-1 cells (Kliewer et al., 1998), the CYP3A23 DR-3 was induced in LLC-PK1 cells co-transfected with PXR and treated with either dexamethasone-t-butylacetate, RU486, progesterone, 1,16-dimethylpregnenolone, pregnenolone, and 5β-pregnane-3,20-dione (Fig. 1), although the fold-increase was not quite as robust as in CV-1 cells.(CYP3A23) 2-tk-CAT was not transcriptionally activated by these steroids in the absence of co-transfected PXR (data not shown), thus demonstrating PXR’s essential role in transactivation by these agents and confirming the original results in CV-1 cells (Kliewer et al., 1998). Moreover, because neither PXR nor RXR alone can bind to the CYP3A23 DR-3 alone (Klieweret al., 1998), but only binds as a heterodimer, the transcriptional activation of the CYP3A23 DR-3 in LLC-PK1 cells further confirms the presence of RXR.
Steroids activate the CYP3A23 DR-3 in LLC-PK1 cells co-transfected with PXR. LLC-PK1 cells co-transfected with mouse PXR, (CYP3A23) 2-tk-CAT, and TK-LUC were treated with 50 μm of the indicated steroids (control, no steroid treatment) for 24 hr and cell extracts were assayed for CAT and Luciferase activities. Data was graphed as -fold increase over untreated control and represented CAT activity in treated cells normalized to co-transfected TK-LUC in the same sample; this treatment (CAT/LUC) value was divided by the mean (CAT/LUC) value from 3–4 control wells. The control value was 0.974 (±) 0.087. The values represent means ± standard deviation from a representative experiment (repeated two to four times) with three to four replicates per treatment per experiment.
We first tested whether PCBs, previously shown to induce CYP3A23 in rat liver in vivo and in primary rat hepatocytes (Schuetz et al., 1986), could activate(CYP3A23) 2-tk-CAT in LLC-PK1 cells co-transfected with PXR. Treatment with various nonplanar polychlorinated biphenyls caused dose-dependent activation of(CYP3A23) 2-tk-CAT (Fig.2), but only in cells co-transfected with PXR (Fig. 4). Strikingly, except for PCB #47, the rank order of PCB congeners as inducers of (CYP3A23) 2-tk-CAT was directly associated with the extent of ortho-chlorination, a finding we had shown previously to be correlated with their hierarchy as inducers of the endogenous CYP3A23 gene in rat hepatocyte cultures (Schuetz et al., 1986). Treatment with 10 μm (data not shown) or 20 μm of another class of environmental contaminants, the organochlorine pesticides transnonachlor or chlordane (Fig. 4) induced the CYP3A23DR-3, but only in cells co-transfected with PXR (Fig. 4).
Polychlorinated biphenyls activate theCYP3A23 DR-3. LLC-PK1 cells co-transfected as in 1 were treated with 5–20 μm of the indicated PCBs with the position of chlorines as follows: (PCB 184 [▪], 2, 2′, 3′, 4, 4′, 6, 6′), (PCB 188 [•], 2, 2′, 3′, 4, 5′, 6, 6′), (PCB 200 [□], 2, 2′, 3, 3′, 4′, 5, 6, 6′), (PCB 196 [▴], 2, 2′, 3, 3′, 4, 4′, 5′, 6), (PCB 47 [▵], 2, 2′, 4, 4′). Data was graphed and analyzed as in Figure 1. Control, cells co-transfected with(CYP3A23) 2-tk-CAT, TK-LUC, and PXR in the absence of xenobiotic treatment. The control value was 0.998 ± 0.07. The student’s t test was used to determine that of the treated values, congener, #47, #196 were not significant at 5 μm with p < 0.05 as the limit of significance. All other congeners were significantly different from the control with p < 0.05 (∗),p < 0.01 (∗∗), or p < 0.005 (∗∗∗).
PXR is required for xenobiotic and steroid induction of CYP3A23. 200,000 LLC-PK1 cells were transfected with (CYP3A23) 2-tk-CAT and TK-LUC with or without co-transfected PXR and then treated 24 hr with steroids (100 μm) except dexamethasonet-butylacetate (50 μm), organochlorine pesticides (20 μm), or PCBs (50 μm). Data is graphed as -fold change from untreated control 1.00 ± 0.009 (variation between measurements was within the size of the symbol) determined as in Fig. 1. Values represent means ± standard deviation from three to four replicates per treatment per experiment (representative of two to three independent experiments).
We next determined whether other classes of antihormones besides antiglucocorticoids could use PXR to activate the CYP3A23DR-3. The antiandrogen cyproterone acetate dose-response curve forCYP3A23 DR-3 activation was similar in potency and efficacy to the glucocorticoid agonist dexamethasone-t-butylacetate (Fig. 3), whereas the antimineralocorticoid spironolactone was a less efficacious inducer.
Steroid antihormones activate PXR. LLC-PK1 cells co-transfected as in Fig. 1 were treated with various concentrations of dexamethasone-t-butylacetate (✙) or spironolactone (▴), or RU486 (⊞), or cyproterone acetate (•), or not treated with steroid (control). Cell extracts were assayed for CAT and Luciferase activities. Data was graphed as the percentage of maximal response (100 μmdexamethasone-t-butylacetate) and values represent means ± standard deviation from two to four independent experiments with three to four replicates per treatment per experiment. The student’s t test was used to determine whether the treated values were significantly different from the control, withp < 0.05 as the limit of significance. ∗,p < 0.05; ∗∗, p < 0.005.
For each of the agents not investigated previously (Kliewer et al., 1998) for PXR activation, we compared the activity of(CYP3A23) 2-tk-CAT in LLC-PK1 cells with or without co-transfected PXR. None of the agents transcriptionally activated (CYP3A23) 2-tk-CAT in the absence of co-transfected PXR (Fig. 4) or in the presence of co-transfected pSG5 expression vector (Stratagene, LaJolla, CA) substituted for mPXR (data not shown). Importantly, by transfecting cells at a high density (2 × 105/well), we minimized toxic effects of any of the xenobiotics or steroids on cell viability as judged by the recovery of protein per well and trypan blue exclusion. To further control for possible nonspecific effects on the TK promoter, we normalized CAT activity to co-transfected TK-LUC in all experiments (Figs. 1-4). Because the amount of co-transfected hormone receptor can affect the extent of gene activation (data not shown) we tested the possibility that some of the drugs might produce an increase in PXR expression by activating the simian virus 40 promoter driving PXR expression. However, none of the drugs activated simian virus 40-LUC (PGL2-promoter, Promega, Madison, WI) transfected into LLC-PK1 cells (data not shown). Thus, it is unlikely the level of expressed PXR is affected by these drugs.
Discussion
More than 15 years have passed since it was determined that organochlorine pesticides and PCBs induce CYP3A (Schuetz et al., 1986). However, the mechanism by which these environmental contaminants up-regulate hepatic CYP3A genes has remained elusive. We had demonstrated previously that these agents inducede novo synthesis of CYP3A23 in primary rat hepatocyte cultures (Schuetz et al., 1986) and postulated that, like dexamethasone and PCN, these agents might induce CYP3A23 synthesis by interacting with the “PCN receptor” (Schuetz et al., 1986). Our current study identifies PXR, the recently identified receptor activated by PCN, as indispensable in mediating transcriptional activation of the CYP3A23 DR-3 by environmental contaminants such as the nonplanar PCBs. This study also sheds light on the heretofore puzzling finding that many classes of antihormones, in addition to antiglucocorticoids, induce CYP3A (Kocareket al., 1995) by demonstrating that these steroids activate PXR. Given the high concentrations of steroids and xenobiotics required to activate PXR and the structural and pharmacological diversity among the activators, it is likely that PXR-mediated steroid and xenobiotic activation of CYP3A is primarily for the purpose of metabolizing these foreign molecules.
Both the organochlorine pesticides and the polychlorinated biphenyls are ubiquitous environmental contaminants. We demonstrated previously that nonplanar PCB congeners induced de novo synthesis of CYP3A and induced CYP2B in primary rat hepatocyte cultures and in rat liver in vivo (Schuetz et al., 1986). Our data herein provide the mechanistic connection demonstrating that these same PCBs use PXR to transcriptionally activate the CYP3A23 DR-3 and, in general, parallel our previous findings. Indeed, it is possible that, in addition to activation of CYP3A, some biological activities of the nonplanar congeners, including toxicities, are mediated by interactions of these PCBs with PXR. PCBs cause liver hypertrophy, are tumor promoters, and cause neurotoxicity and disruptions in calcium regulation (Hansen, 1998). PXR-mediated PCB activation of theCYP3A23 DR-3 may be a predictive biomarker correlated with some PCB toxicities. Thus, dose-response analysis of PCB congeners or mixtures of PCB for PXR-mediated activation of the CYP3A23DR-3 may provide an assay to rapidly segregate toxic and nontoxic nonplanar PCBs. Interestingly, the antiandrogen cyproterone acetate, identified herein as a PXR activator, also includes liver cell proliferation, and previous attempts to determine the signaling pathway for proliferation have been futile (Menegazzi et al., 1997). Cyproterone acetate also causes DNA damage in liver (Werner et al., 1997), is a rodent liver tumor promoter (Duivenvoordenet al., 1995), and increases sensitivity of hepatocytes to undergo apoptosis (Oberhammer et al., 1996). It remains to be determined whether ligand activated PXR is involved in mediating any of these other biological activities associated with nonplanar PCBs or cyproterone acetate.
We conclude that PXR mediates induction of the CYP3A23 DR-3 by antihormones and environmental xenobiotics. Thus, PXR joins the aryl hydrocarbon receptor in being a transcription factor ligand-activated by multiple classes of agents (Nebert, 1989; Dzeletovic et al., 1997). By analogy to the aryl hydrocarbon receptor paradigm, it will be of interest in the future to determine whether individual differences in PXR ligand affinity among humans correlates with CYP3A inducibility and, like P-glycoprotein (Schuetz et al., 1996), also contributes to the wide variation in the CYP3A inductive response (Watkins et al., 1989; Kolars et al., 1992).
Acknowledgments
We gratefully acknowledge Dr. Steve Kliewer (Glaxo Wellcome Research and Development, Research Triangle Park, NC) for the mouse PXR.1 and (CYP3A23) 2-tk-CAT plasmids. We thank Nancy Wright and Kazuto Yasuda for their technical assistance.
Footnotes
- Received June 22, 1998.
- Accepted September 15, 1998.
-
Send reprint requests to: Dr. Erin Schuetz, Dept. of Pharmaceutical Sciences, St. Jude Children’s Research Hospital, 332 N. Lauderdale, Memphis TN 38105. E-mail: erin.schuetz{at}stjude.org
-
This work was supported by National Institute of Health Research Grants ES08658 (E.G.S.), ES05851 (J.D.S.), P30-CA21765 (E.G.S., J.D.S.) and by the American Lebanese Syrian Associated Charities (ALSAC).
-
1 CYP3A23 refers to the major dexamethasone, PCN and phenobarbital inducible form of CYP3A in rat liver (Komori and Oda, 1994) which is now recognized to be CYP3A23, not CYP3A1. TheCYP3A 5′flanking regulatory sequence previously identified as CYP3A1 (Burger et al., 1992; Quattrochi et al., 1995; Kliewer et al., 1998) is in fact CYP3A23 (Nelson et al., 1993; Barwick et al., 1996). Likewise, because the CYP3A1 cDNA probe used in all previous publications of CYP3A regulation cannot distinguish between CYP3A1 and CYP3A23, we refer to the CYP3A in these publications as CYP3A23.
Abbreviations
- PCN
- pregnenolone 16α-carbonitrile
- PXR
- pregnane X receptor (mouse PXR.1), GenBank accession no. AF031814
- PCB
- polychlorinated biphenyl
- CAT
- chloramphenicol acetyltransferase
- RXR
- retinoid X receptor
- The American Society for Pharmacology and Experimental Therapeutics
