Control of Steroid, Heme, and Carcinogen Metabolism by Nuclear Pregnane X Receptor and Constitutive Androstane Receptor

Control of steroid, heme, and carcinogen metabolism by nuclear pregnane X receptor and constitutive androstane receptor

Wen Xie*†‡, Mei-Fei Yeuh‡§, Anna Radominska-Pandya¶, Simrat P. S. Saini*, Yoichi Negishi*, Bobbie Sue Bottroff†, Geraldine Y. Cabrera†, Robert H. Tukey§, and Ronald M. Evans†ʈ

*Center for Pharmacogenetics and Department of Pharmaceutical Sciences, University of Pittsburgh, Pittsburgh, PA 15213; †Howard Hughes Medical Institute, Gene Expression Laboratory, The Salk Institute for Biological Studies, La Jolla, CA 92037; §Laboratory of Environmental Toxicology, Departments of Chemistry, and Biochemistry and Pharmacology, University of California at San Diego, La Jolla, CA 92093; and ¶Department of Biochemistry and Molecular Biology, University of Arkansas for Medical Sciences, Little Rock, AR 72205

Contributed by Ronald M. Evans, December 31, 2002 Through a multiplex promoter spanning 218 kb, the phase II UDP- ogens, estrogen, and thyroxin metabolism. Moreover, activation of glucuronosyltransferase 1A (UGT1) gene encodes at least eight dif- PXR in transgenic mice is sufficient to increase levels of UGT ferently regulated mRNAs whose protein products function as the expression and activity, as well as bilirubin and steroid clearance. principal means to eliminate a vast array of steroids, heme metabolites, environmental toxins, and drugs. The orphan nuclear receptors Materials and Methods pregnane X receptor (PXR) and constitutive androstane receptor Animals. The generation of VP-hPXR, hPXR (previously known as (CAR) were originally identified as sensors able to respond to numer- VPSXR and SXR) transgenic mice and PXR null mice has been ous environmentally derived foreign compounds (xenobiotics) to described (8). promote detoxification by phase I cytochrome P450 genes. In this report, we show that both receptors can induce specific UGT1A UGT1A1 Promoter Cloning and Site-Directed Mutagenesis. The hu- isoforms including those involved in estrogen, thyroxin, bilirubin, and man UGT1A1 promoter was amplified by using a BAC clone carcinogen metabolism. Transgenic mice expressing a constitutively containing the entire UGT1 locus template (2). All primers were active form of human PXR show markedly increased UGT activity designed based on the sequence encoding the human UGT1A locus toward steroid, heme, and carcinogens, enhanced bilirubin clearance, published in the National Center for Biotechnology Information as well as massively increased steroid clearance. The ability of PXR and GenBank (accession no. AF297093). Site-directed mutagenesis was constitutive androstane receptor and their ligands to transduce both performed by PCR and confirmed by DNA sequencing. the phase I and phase II adaptive hepatic response defines a unique transcriptional interface that bridges the ingestion and metabolism of DNA-Binding Analysis. Electrophoretic mobility-shift assays environmental compounds to body physiology. (EMSA) were performed by using in vitro-transcribed and -translated protein (TNT, Promega) as described (14). Oligonucleotides Ј he metabolism of steroid hormones, other endogenous com- were: UGT1A1, 5 -CTAACGGTTCATAAAGGGTATTAGGT- 3Ј; UGT1A6, 5Ј-CGAGTAGGTCATAAAGGTCACA-3Ј; and Tpounds, and xenobiotics occurs in the hepatogastrointestinal Ј Ј tract. This catabolic process is mediated by phase I enzymes, such UGT1A6m, 5 -CGAGTAGAACATAAAGAACACA-3 . as the monooxygenase CYP enzymes (1), as well as the phase II ͞ conjugating enzymes, such as UDP-glucuronosyltransferase Plasmid Construction and Transfection. The tk-UGT1A6 DR3-Luc (UGTs), sulfotransferases, and GSTs (2). The UGT 1A (UGT1A) and its mutant variants were generated by insertion of correspond- locus is controlled by 13 promoters (A1–A12), spanning Ͼ200 kb ing annealed oligonucleotides into the tk-Luc vector. The expres- of upstream sequence encoding overlapping but distinct mRNAs sion vectors for hPXR, VP-hPXR, and VP-CAR have been (3). Using UDP-glucuronic acid as a cosubstrate, UGT enzymes described (14). CV-1 cell transfection using N-[1-(2,3-dioleoy- convert a diverse set of lipophilic substances to water-soluble loxy)propyl]-N,N,N-trimethylammonium methylsulfate (Roche glucuronides and function as the principal means to eliminate Biochemicals) and HepG2 cell transfections using Lipofectamine steroids, heme metabolites, environmental toxins, and drugs from Plus (Invitrogen) were carried out as before (14, 15). the body (2). Crigler–Najjar syndrome type I results in a lethal Northern and Western Blot Analysis. accumulation of bilirubin due to a defect in the 1A1 promoter, Total RNA was prepared from liver tissues by using TRIzol reagent (Invitrogen). Northern hy- whereas Gilbert’s disease (7% of the population) is a more mild 1A1 bridization was carried out as described (8). Liver microsomes were promoter defect typified by increased sensitivity to certain drugs prepared and analyzed for UGT1A expression. Western blot pro- like Tylenol (2). files were conducted by using an anti-UGT1A antibody, which Nuclear receptor (NR) pregnane X receptor (PXR, also known recognizes only the UGT1A family of proteins, and an anti- as the steroid and xenobiotic receptor or SXR) and the constitutive UGT1A1 antibody that is specific for UGT1A1. androstane receptor (CAR) were originally shown to act through NR response elements localized in the promoters of the target UGT Analysis. The UGT analysis was similar to that described before CYP3A and CYP2B genes (4–10). Pharmaceutical compounds (16). In brief, mouse liver microsomes were prepared as described such as phenobarbital and rifampicin (RIF) have been reported as (16). The [14C]UDP-GlcUA was used as the sugar donor, and TLC UGT inducers, although the molecular basis remains to be defined was used to separate glucuronidation products. The products were (11, 12). Both phenobarbital and RIF have been empirically used to treat hyperbilirubinemia, a clinical accumulation of serum bilirubin due to insufficient glucuronidation (11, 13). These com- Abbreviations: CAR, constitutive androstane receptor; PXR, pregnane X receptor; RIF, pounds are activators of PXR and CAR, suggesting a plausible rifampicin; TCPOBOP, 1,4-bis-[2-(3,5-dichloropyridyloxy)]benzene; UGT, UDP-glucurono- signaling pathway for UGT transcription. syltransferase; NR, nuclear receptor; EMSA, electrophoretic mobility-shift assay. In this report, we show that both receptors can induce specific ‡W.X. and M.-F.Y. contributed equally to this work. UGT1A isoforms including those specialized for bilirubin, carcin- ʈTo whom correspondence should be addressed. E-mail: [email protected].

4150–4155 ͉ PNAS ͉ April 1, 2003 ͉ vol. 100 ͉ no. 7 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0438010100 Downloaded by guest on September 26, 2021 MEDICAL SCIENCES

Fig. 1. Induction of UGT1A expression and glucuronidation by PXR activation. (A) Liver microsomes were prepared from WT and VP-hPXR transgenic mice and subject to Western blot proﬁling by using a pan anti-UGT1A antibody and a speciﬁc anti-UGT1A1 antibody. (B) Mouse liver total RNAs were subjected to Northern blot analysis. The membranes were probed for UGT1A6, UGT1As, CYP3A11, and GAPDH as a loading control. (C) Induction of UGT1A mRNA by RIF in hPXR transgenic mice. WT or transgenic males were gavaged with a single dose of solvent or RIF (50 mg͞kg) 24 h before death. (D) The parental HepG2 cells or the HepG2-hPXR stable cells were treated with DMSO or 25 ␮M RIF for 24 h and subsequently harvested for Northern blot analysis to detect UGT1A1 mRNA. (E) Glucuronidation activity toward ␤-estradiol. The result is shown as the autoradiograph of a TLC plate. (F) Glucuronidation toward thyroid hormones (rT3 and T4) and xenobiotics (4-nitrophenol and 4-OH-PhIP). Results are presented as fold increase in glucuronidation activity over WTs and represent the mean and standard error.

visualized with a Molecular Dynamics Storm 820 PhosphorImager. at 4 p.m. by retroorbital eye bleeding within 1 min of initial Alternatively, silica gel in zones corresponding to the glucuronide disturbance. The 24-h urine was collected by using mouse metabolic bands were visualized by autoradiography, or corresponding areas cages (Nalge). Corticosterone levels in plasma or urine were from control lanes were scraped into scintillation vials, and radio- measured with a Corticosterone [125I]RIA kit from ICN. The activity was measured by liquid scintillation counting. statistical analysis was performed with INSTAT 2.03 software.

Bilirubin Clearance. Adult males were given a single does of bilirubin Results (10 mg͞kg body weight). Blood samples were collected in untreated tubes1hafterinjection.Serumwasprepared by centrifugation at Activation of PXR Induces UGT Expression and Glucuronidation in 4,000 rpm for 10 min. Total and conjugated bilirubin levels were Transgenic Mice. The regulatory locus of the UGT1A gene is highly measured by Antech Diagnostic (Lake Success, NY). The statistical complex. It harbors 12 promoters that control the independent analysis was performed with INSTAT 2.03 software. expression of a specific UGT isoform with distinctive substrate specificity (2). The isoforms vary at the amino terminus and Blood and Urine Collection and Hormone Analysis. Mouse blood share a common body coded by exons 2–5. For example, the 1A1 samples were collected in EDTA-coated tubes (Becton Dickinson) isoform metabolizes estrogens but not androgens or other

Xie et al. PNAS ͉ April 1, 2003 ͉ vol. 100 ͉ no. 7 ͉ 4151 Downloaded by guest on September 26, 2021 Fig. 2. Cloning of the UGT1A1 gene promoter and its activation by PXR and CAR. (A) Schematic representation of the UGT1 gene locus. The 5Ј regulatory sequences for UGT1A1 isoform are indicated. (B) Analysis of the UGT1A1 5Ј regulatory sequences. The Enhancer 1 (Ϫ11,001 to Ϫ8,311), Enhancer 2 (Ϫ8,537 to Ϫ4,636), Enhancer 3 (Ϫ3,714 to Ϫ2,068), and Promoter (Ϫ2,574 to Ϫ5), relative to the transcription start site, were ampliﬁed by PCR and cloned into pGL3-Luc reporter vector. The reporter genes were transfected into HepG2 cells in the presence of an expression vector for VP-hPXR. The fold induction was calculated from the luciferase activity of VP-hPXR cotransfected cells over empty vector transfected cells. (C) Enhancer 3 was dissected into RE-a and RE-b and tested for transactivation in the presence of VP-hPXR or VP-CAR. (D) Deletion mutants used in transfection experiments are shown diagrammatically. (E) EMSA of the putative DR-3 element using in vitro-translated proteins of hPXR, CAR, and RXR␣. The competition was performed by using a 50-fold excess of unlabeled probe. (F) Mutation of the DR-3 in the context of RE-a blocks response to VP-hPXR.

steroids. It also metabolizes bilirubin, carcinogens, and Tylenol. As shown in Fig. 1A, the VP-hPXR mice have increased levels of To explore whether individual isoforms of UGTs are transcrip- both general UGT1A protein and the UGT1A1 isoform. In additional targets for PXR, we examined profiles of hepatic proteins tion to UGT1A1, Northern blot analysis revealed that UGT1A6 and mRNAs isolated from transgenic mice harboring an acti- mRNA is up-regulated (Fig. 1B). As expected, up-regulation of vated hPXR in the liver (8). Liver microsomal proteins were general UGT1A mRNAs was observed when a probe containing prepared and subjected to Western blot analysis using a pan- the common 2–5 exons was used. This probe encompasses exons UGT1A antibody or a UGT1A1 isoform-specific antibody. that are shared by all UGT1A family members and thus measures

4152 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0438010100 Xie et al. Downloaded by guest on September 26, 2021 combined regulation of the entire locus (17). Whereas the induction of UGT in VP-hPXR mice is clear, it is not as dramatic as that of CYP3A11, probably because not all UGT isoforms are PXR targets (Fig. 1B). The up-regulation is liver-specific, as no change in UGT expression was seen in tissues that do not express the transgene, such as the small intestine (data not shown). A ligand-dependent induction of UGT1A was also observed in the humanized mice that express a full-length hPXR transgene (8). As expected, treatment of these mice with the prescription antibiotic RIF, a potent hPXR- specific activator, elevated UGT1As mRNA within 24 h of a single oral dose (Fig. 1C). This establishes that human PXR and its hormone-specific ligands can regulate the UGT1A locus in vivo. The sufficiency of PXR to mediate transactivation of the 1A1 ‘‘Gilbert’s disease’’ isoform was examined in cultured cells. For these studies, HepG2 cells stably expressing hPXR were generated, and the expression of the transduced hPXR was confirmed by Northern blot analysis (data not shown). In contrast to the parental HepG2 line, the HepG2-hPXR stable cells are readily responsive to RIF, inducing UGT1A1 to a level that is compa- rable to the PXR-independent inducer ␤-naphthoflavone (Fig. 1D). ␤-Naphthoflavone is a prototypical aromatic hydrocarbon standard that induces UGT1A1 in PXR-deficient HepG2 cells. A sensitive cotransfection screen confirms that ␤-naphthoflavone does not activate PXR even at concentrations exceeding its maximal activation of UGT1A1 (data not shown). Together, these results suggest that at least two UGT1A isoforms (A1 and A6) are potential transcriptional targets of PXR. This is consistent with a recent quantitative PCR analysis that UGT1A mRNAs are regulated in vivo by PXR-specific ligands (18). Presumably, increased UGT expression would be associated with increased enzymatic activity. Compared with WT littermates, liver microsomes prepared from VP-hPXR mice exhibit profoundly higher glucuronidation activity toward ␤-estradiol, a known 1A1-selective substrate (Fig. 1E), thyroid hormones (rT3 and T4) classified as 1A1-preferred substrates (19), as well as the

xenobiotic carcinogens 4-nitrophenol and 4-OH-PhIP, which are MEDICAL SCIENCES substrates for multiple 1A isoforms (20) (Fig. 1F). Microsomes Fig. 3. Characterization of PXR and CAR binding in UGT1A6 promoter. (A) The prepared from RIF-treated HepG2-hPXR cells also exhibited sequence of the DR-3 NR response element found in the UGT1A6 gene. A mutant higher glucuronidation activity (data not shown). variant (mutated nucleotides underlined) and a previously identiﬁed DR-3 type of PXR response element found in the rat CYP3A23 gene are also shown. (B) ͞ ͞ ͞ Regulation of UGT1A Gene Promoters by PXR and CAR. PXR RXR and CAR RXR heterodimers bind to the UGT1A6 DR-3. EMSA was To delineate ␣ the molecular basis of UGT1A regulation, we have cloned the performed by using in vitro-translated hPXR, VP-hPXR, CAR, and RXR proteins and radiolabeled oligonucleotides of UGT1A6 (lanes 1–10) and its mutant human UGT1 gene locus and analyzed its potential regulation by UGT1A6m (lanes 11 and 12). The binding of PXR͞RXR heterodimers to CYP3A23 PXR and CAR. As mentioned above, the 200-kb human UGT1 was also included as positive controls (lanes 13 and 14). (C) The tk-UGT1A6͞DR- gene locus has a highly specialized genomic organization, in that 3-Luc reporter gene was transfected into CV-1 cells in the presence of the empty each of the UGT1A family members has a unique first exon but vector or the expression vectors for hPXR, mPXR, or CAR. The transfected cells shares common exons 2–5 (Fig. 2A and ref. 17). Each first exon is were subsequently mock-treated or treated with indicated compounds for 24 h proximal to its own unique promoter, which, in turn, is expressed before luciferase assays. Results are shown as fold induction over solvent controls in a characteristic tissue pattern and each UGT isoform has its and represent the average and standard error from triplicate assays. The concentrations are 10 ␮M for PCN, RIF, and clotrimazole; 300 ␮g͞ml for St. John’s wort; substrate preferences. For example, expressed in liver, gut, and ␮ kidney, UGT1A1 metabolizes estrogen, bilirubin, thyroid hormone, 5 M for androstenol; and 250 nM for TCPOBOP. and many other substrates. An 11-kb sequence of the 5Ј regulatory region of the human UGT1A1 gene was cloned by PCR. Four is required for PXR transactivation. Inspection of this 47-bp fragments of regulatory sequences, a promoter and three enhancer enhancer element revealed a DR-3-like NR response element elements (E1–3), that encompass the 11-kb sequence were sub- (GGTTCATAAAGGGTA). EMSA revealed that the PXR͞ cloned into the pGL vector to generate a series of luciferase 3 RXR␣ heterodimers could bind to this element (Fig. 2E). Inter- reporter constructs. The reporter genes were cotransfected with expression vectors for constitutively active forms of the receptors, estingly, a low level of binding was also seen when only PXR was VP-hPXR and VP-CAR, to test their ability to mediate transacti- present, and the migration was more consistent with a homodimer vation by these receptors. As shown in Fig. 2 B and C, the E3 than a monomer (Fig. 2E). The binding is specific as efficient sequence that flanks Ϫ3,714 to Ϫ2,068 relative to the transcription competition can be achieved by excessive unlabeled oligonucleo- ͞ ␣ start site is readily and profoundly activated in the presence of tides. The CAR RXR heterodimers, but not CAR by itself, also VP-hPXR and VP-CAR. A further dissection of the E3 fragment exhibit measurable binding to the same element (Fig. 2E and data revealed that the PXR-responsive region is localized to the 386-bp not shown), consistent with the transactivation of RE-a by CAR RE-a (nucleotides Ϫ3,529 to Ϫ3,143) region (Fig. 2C). The 5Ј and our previous observation that CAR can share the DR-3 deletion mutant 1 of RE-a does not affect the inducibility, whereas element with PXR (14). Mutation of the DR-3 in the context of deletion mutant 2 completely abolishes the activation (Fig. 2D). RE-a abolishes VP-hPXR-dependent induction (Fig. 2F), indicat- Therefore, the 47-bp sequence between deletion mutants 1 and 2 ing that this element is necessary for PXR transactivation. In

Xie et al. PNAS ͉ April 1, 2003 ͉ vol. 100 ͉ no. 7 ͉ 4153 Downloaded by guest on September 26, 2021 Fig. 4. Increased bilirubin clearance in VP-hPXR mice. Adult males of WT and VP-hPXR mice were treated with a single dose of bilirubin (10 mg͞kg), and serum levels of total (A) and conjugated (B) bilirubin were measured 1 h after injection. The P value for total bilirubin is 0.001 between WT and VP-hPXR.

addition to the 47-bp enhancer element, RE-a contains three additional previously reported putative NR binding sites, one of which binds to CAR (21). Our EMSA revealed that two of the DR-4-type elements can also bind to PXR (data not shown). Therefore, these additional sites may also contribute to 1A1 transactivation by PXR and CAR. Consistent with the induction of UGT1A6 mRNA in VP- hPXR mice, we have previously identified a putative DR-3- binding site in the UGT1A6 gene (Fig. 3A and ref. 4). As expected, the WT hPXR (Fig. 3A, lane 5), CAR (data not shown), and their respective activated variants VP-hPXR (Fig. 3A, lane 7) and VP-CAR (Fig. 3A, lane 9) bound to this element in an RXR-dependent manner (Fig. 3B). The specific binding was abrogated when this DR-3 was disrupted (Fig. 3A, lanes 11 and 12). To examine whether this DR-3 element can mediate transactivation by PXR and CAR, a luciferase reporter gene containing three copies of this element was placed upstream of a minimal thymidine kinase promoter and transfected into monkey kidney CV-1 cells together with expression vectors for hPXR, mPXR, or CAR. Significant activation of this UGT1A6 reporter by hPXR was seen when RIF, clotrimazole, and an extract of St. John’s wort, but not androstenol or 1,4-bis-[2-(3,5-dichloropyridyloxy)]benzene (TCPOBOP), were added to the culture medium (Fig. 3C). mPXR-mediated UGT activation was seen in the presence of St. John’s wort, clotrimazole, as well as PCN, an mPXR-specific ligand. As predicted, CAR activates this UGT reporter in a ligand- independent manner. The activation by CAR is inhibited or Fig. 5. Increased glucuronidation and output of corticosterone in VP-hPXR mice. (A) Increased glucuronidation of three corticosteroids, corticosterone, potentiated by androstenol or TCPOBOP, respectively. The cortisone, and cortisol by liver microsomes of VP-hPXR mice. (B) Blood samples inhibitory effect of androstenol was compromised when from 6-week-old males were collected via retroorbital eye bleeding, and TCPOBOP was coadded. plasma were prepared and subjected to corticosterone RIA. Nonparametric Mann–Whitney test P Ͻ 0.0001. (C) Twenty-four-hour urine samples were Activation of PXR Enhances Bilirubin Clearance. UGT1A1 is the collected from 8- to 10-week-old males and subjected to corticosterone principal UGT isoform that facilitates bilirubin glucuronidation measurement. The P value is 0.5739 between WT and hPXR (not signiﬁcant), and subsequent clearance. The identification of UGT1A1 as 0.0004 between WT and VP-hPXR, and 0.0026 between hPXR and VP-hPXR. direct target of PXR prompted us to examine the effect of PXR activation on bilirubin conjugation and clearance. The VP- hPXR mice were subjected to a single dose of bilirubin, and the their WT littermates. Therefore, activation of PXR in transgenic serum bilirubin levels were monitored. As shown in Fig. 4, 1 h mice is sufficient to promote bilirubin conjugation and clear- after injection, the remaining serum levels of both total and ance, consistent with the up-regulation of UGT1A1 activity in conjugated bilirubin in transgenic mice were less than half of these animals.

4154 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0438010100 Xie et al. Downloaded by guest on September 26, 2021 Increased Corticosterone Clearance in VP-hPXR Mice. In addition to absent in CAR null mice. The carcinogenic potential of this gene xenobiotics, UGTs are essential for the metabolism and elimi- was suggested in the Gunn rat, where a mutation in the UGT1 allele nation of steroid hormones. The up-regulation of UGT enzyme leads to a decrease glucuronidation of the carcinogenic benzo- levels prompted us to assess a potential role of PXR in promoting (a)pyrene, leading to elevated levels of DNA adducts (24). More- steroid elimination and up-regulating the adrenal axis. As shown over, down-regulation of UGT1 mRNA is observed in the early in Fig. 5A, liver microsomes of the VP-hPXR mice exhibited stages of cancer but not in benign tumorogenesis (25). Environ- increased glucuronidation of several glucocorticoids, such as mental mutagens, such as PhIP and benzo(a)pyrenes, have been corticosterone, cortisol, and cortisone. To assess whether in- identified as substrates for several UGT1 proteins (15, 20, 26). creased glucuronidation is associated with increased output of Indeed, PhIP is glucuronidated at a higher rate in VP-hPXR mice glucocorticoid, mouse plasma and urine samples were collected (Fig. 1F). Thus, the creation of these transgenic mice not only and measured for corticosterone, the principal glucocorticoid in demonstrates a role for the receptor in its regulation, it provides a rodents. The average plasma corticosterone concentration in potential in vivo model to assess the molecular dynamics of carci- WT males is 88.78 ng͞ml, similar to published results (22), nogenesis and the contribution of glucuronidation to this process. whereas the levels were elevated to 322.4 ng͞ml in VP-hPXR Finally, the increased levels of corticosterone in VP-hPXR mice provide evidence that PXR activation may directly contribute to the mice (Fig. 5B), indicating an activation of the pituitary adrenal ͞ axis. A corresponding increase of urine corticosterone levels was increased metabolism and or elimination of steroids. Of note, also seen in VP-hPXR mice (Fig. 5C), whereas no significant increased glucuronidation was also seen for many other steroid changes in urinary (Fig. 5C) or blood (data not shown) corti- hormones, including additional glucocorticoids, thyroid hormones, costerone levels were observed in the untreated hPXR trans- and estradiol, suggesting a broader role for xenobiotic nuclear genic mice. Corticosterone is produced in the zona fasciculate of receptors in hormonal homeostasis. the adrenal gland cortex and subjected to the regulation of Although activation of PXR in transgenic mice is sufficient to pituitary–adrenal axis. Histological analysis of the adrenal induce the UGT locus, loss of PXR in knockout mice does not glands revealed no significant differences between WT and suppress basal expression levels (data not shown). The maintenance of basal expression in PXR null mice was also previously seen for transgenic mice at 2.5 months (data not shown). Therefore, CYP3A (8). It is possible that the sustained basal expression is whereas both elimination and synthesis of corticosterone were mediated, in part, by CAR, as it is capable of binding to and increased, this had not yet led to an obvious morphologic change, activating through the DR-3 response element. Therefore, this suggesting that the increased demand was in the normal capacity study provides another example of the proposed molecular fail-safe of the adrenal gland. The measurement of ACTH in plasma and model in xenobiotic regulation (14). immunostaining detection of ACTH producing pituitary corti- In conclusion, the xenobiotic receptors PXR and CAR function cotropes also revealed no significant changes between WT and as master sensors to control the phase I and II adaptive hepatic VP-hPXR transgenic mice (data not shown). responses by nature of their ability to coordinate the expression of In summary, activation of PXR results in increased levels of a defined network of target genes. Moreover, such coordinated steroid in both plasma and urine, consistent with the premise regulatory mechanisms may also include control of drug-effluxing that PXR activation can promote steroid clearance, presumably transporters (27). Therefore, by residing at the biochemical inter- via its induction of cytochrome P450 (CYP) and UGT enzyme face between mammals and their chemical environment, these synthesis. xenosensors represent a key protective mechanism to ensure elim- MEDICAL SCIENCES ination of steroid hormones along with a plethora of endotoxins and Discussion xenotoxins. The identification of the UGT locus as a direct target for hPXR and CAR has implications in both xenobiotic metabolism and human We thank Susan Nowell for Phip glucuronidation assay; Joanna Little, Ruth diseases. UGTs have been implicated in the etiology of human Yu, and Yanhong Shi for comments on the manuscript; Zheng Ma and genetic diseases and carcinogenesis (2). The mutations in the Lingyun Zhao for technical advice; Henry Juguilon and Jing Xu for UGT1A1 promoter are linked to inheritable hyperbilirubinemia as technical assistance; Joe Ritter for mouse anti-UGT1A1 antibody; and a result of decreased glucuronidation and, therefore, clearance of Elaine Stevens, Lita Ong, and Li Xu for administrative assistance. This work serum bilirubin (23). Our observation is also consistent with a was conducted in part by U.S. Public Health Service Grant GM49139 (to R.H.T.). W.X. is supported by the Competitive Medical Research Fund of recent study showing the regulation of bilirubin metabolism by the University of Pittsburgh Medical Center Health System and the Susan CAR. In a related paper, Huang et al. (28) showed that activation G. Komen Breast Cancer Foundation. R.M.E. is an Investigator of the of CAR increases hepatic expression of genes known to be involved Howard Hughes Medical Institute at the Salk Institute for Biological Studies in bilirubin metabolism including UGT1A1, and this induction is and March of Dimes Chair in Molecular and Developmental Biology.

1. Guengerich, F. P. (1989) Annu. Rev. Pharmacol. Toxicol. 29, 241–264. 15. Yueh, M. F., Nguyen, N., Famourzadeh, M., Strassburg, C. P., Oda, Y., Guengerich, F. P. 2. Tukey, R. H. & Strassburg, C. P. (2000) Annu. Rev. Pharmacol. Toxicol. 40, 581–616. & Tukey, R. H. (2001) Carcinogenesis 22, 943–950. 3. Gong, Q. H., Cho, J. W., Huang, T., Potter, C., Gholami, N., Basu, N. K., Kubota, S., 16. Radominska-Pyrek, A., Zimniak, P., Irshaid, Y. M., Lester, R., Tephly, T. R. & Pyrek, J. S. Carvalho, S., Pennington, M. W., Owens, I. S. & Popescu, N. C. (2001) Pharmacogenetics (1987) J. Clin. Invest. 80, 234–241. 11, 357–368. 17. Ritter, J. K., Chen, F., Sheen, Y. Y., Tran, H. M., Kimura, S., Yeatman, M. T. & Owens, 4. Blumberg, B., Sabbagh, W., Juguilon, H., Bolado, J., Jr., Ong, E. S. & Evans, R. M. (1998) I. S. (1992) J. Biol. Chem. 267, 3257–3261. Genes Dev. 12, 3195–3205. 18. Maglich, J. M., Stoltz, C. M., Goodwin, B., Hawkins-Brown, D., Moore, J. T. & Kliewer, S. A. 5. Kliewer, S. A., Moore, J. T., Wade, L., Staudinger, J. L., Jones, M. A., McKee, D. D., Oliver, (2002) Mol. Pharmacol. 62, 638–646. B. M., Willson, T. M., Zetterstrom, R. H., Perlmann, T. & Lehmann, J. (1998) Cell 92, 73–82. 19. Findlay, K. A., Kaptein, E., Visser, T. J. & Burchell, B. (2000) J. Clin. Endocrinol. Metab. 6. Bertilsson, G., Heidrich, J., Svensson, K., Asman, M., Jendeberg, L., Sydow-Backman, M., 85, 2879–2883. 20. Nowell, S., Massengill, J., Williams, S., Radominska-Pandya, A., Tephly, T. R., Cheng, Z., Ohlsson, R., Postlind, H., Blomquist, P. & Berkenstam, A. (1998) Proc. Natl. Acad. Sci. USA Strassburg, C. P., Tukey, R. H., MacLeod, S. L., Lang, N. P. & Kadlubar, F. F. (1999) 95, 12208–12213. Carcinogenesis 20, 101–108. 7. Honkakoski, P., Zelko, I., Sueyoshi, T. & Negishi, M. (1998) Mol. Cell. Biol. 18, 5652–5658. 21. Sugatani, J., Kojima, H., Ueda, A., Kakizaki, S., Yoshinari, K., Gong, Q. H., Owens, I. S., 8. Xie, W., Barwick, J. L., Downes, M., Blumberg, B., Simon, C. M., Nelson, M. C., Neuschwander- Negishi, M. & Sueyoshi, T. (2001) Hepatology 33, 1232–1238. Tetri, B. A., Brunt, E. M., Guzelian, P. S. & Evans, R. M. (2000) Nature 406, 435–439. 22. Smith, G. W., Aubry, J. M., Dellu, F., Contarino, A., Bilezikjian, L. M., Gold, L. H., Chen, 9. Wei, P., Zhang, J., Egan-Hafley, M., Liang, S. & Moore, D. D. (2000) Nature 407, 920–923. R., Marchuk, Y., Hauser, C., Bentley, C. A., et al. (1998) Neuron 20, 1093–1102. 10. Xie, W. & Evans, R. M. (2001) J. Biol. Chem. 276, 37739–37742. 23. Bosma, P., Chowdhury, J. R. & Jansen, P. H. (1995) Lancet 346, 314–315. 11. Yaffe, S. J., Levy, G., Matsuzawa, T. & Balish, T. (1996) N. Engl. J. Med. 275, 1461–1466. 24. Hu, Z. & Wells, P. G. (1992) J. Pharmacol. Exp. Ther. 263, 334–342. 12. Jemnitz, K., Lengyel, G. & Vereczkey, L. (2002) Biochem. Biophys. Res. Commun. 291, 29–33. 25. Strassburg, C. P., Manns, M. P. & Tukey, R. H. (1997) Cancer Res. 57, 2979–2985. 13. Cancado, E. L., Leitao, R. M., Carrilho, F. J. & Laudanna, A. A. (1998) Am. J. Gastroenterol. 26. Malfatti, M. A. & Felton, J. S. (2001) Carcinogenesis 22, 1087–1093. 93, 1510–1517. 27. Bock, K. W., Eckle, T., Ouzzine, M. & Fournel-Gigleux, S. (2000) Biochem. Pharmacol. 59, 467–470. 14. Xie, W., Barwick, J. L., Simon, C. M., Pierce, A. M., Safe, S., Blumberg, B., Guzelian, P. S. 28. Huang, W., Zhang, J., Chua, S. S., Qatanani, M., Han, Y., Granata, R. & Moore, D. D. & Evans, R. M. (2000) Genes Dev. 14, 3014–3023. (2003) Proc. Natl. Acad. Sci. USA 100, 4156–4161.

Xie et al. PNAS ͉ April 1, 2003 ͉ vol. 100 ͉ no. 7 ͉ 4155 Downloaded by guest on September 26, 2021