Reciprocal activation of xenobiotic response genes by nuclear receptors SXR/PXR and CAR.

The cytochrome P450 (CYP) gene products such as CYP3A and CYP2B are essential for the metabolism of steroid hormones and xenochemicals including prescription drugs. Nuclear receptor SXR/PXR (steroid and xenobiotic receptor/pregnenolone X receptor) has been shown both biochemically and genetically to activate CYP3A genes, while similar studies have established constitutive androstane receptor (CAR) as a CYP2B regulator. The response elements in these genes are also distinct, furthering the concept of independent regulation. Unexpectedly, we found that SXR can regulate CYP2B, both in cultured cells and in transgenic mice via adaptive recognition of the phenobarbital response element (PBRE). In a type of functional symmetry, orphan receptor CAR was also found to activate CYP3A through previously defined SXR/PXR response elements. These observations not only provide a rational explanation for the activation of multiple CYP gene classes by certain xenobiotics, but also reveal the existence of a metabolic safety net that confers a second layer of protection to the harmful effects of toxic compounds and at the same time increases the propensity for drug-drug interactions.

Humanized xenobiotic response in mice expressing nuclear receptor SXR.

The cytochrome CYP3A gene products, expressed in mammalian liver, are essential for the metabolism of lipophilic substrates, including endogenous steroid hormones and prescription drugs. CYP3A enzymes are extremely versatile and are inducible by many of their natural and xenobiotic substrates. Consequently, they form the molecular basis for many clinical drug-drug interactions. The induction of CYP3A enzymes is species-specific, and we have postulated that it involves one or more cellular factors, or receptor-like xeno-sensors. Here we identify one such factor unequivocally as the nuclear receptor pregnenolone X receptor (PXR) and its human homologue, steroid and xenobiotic receptor (SXR). We show that targeted disruption of the mouse PXR gene abolishes induction of CYP3A by prototypic inducers such as dexamethasone or pregnenolone-16alpha-carbonitrile. In transgenic mice, an activated form of SXR causes constitutive upregulation of CYP3A gene expression and enhanced protection against toxic xenobiotic compounds. Furthermore, we show that the species origin of the receptor, rather than the promoter structure of CYP3A genes, dictates the species-specific pattern of CYP3A inducibility. Thus, we can generate 'humanized' transgenic mice that are responsive to human-specific inducers such as the antibiotic rifampicin. We conclude that SXR/PXR genes encode the primary species-specific xeno-sensors that mediate the adaptive hepatic response, and may represent the critical biochemical mechanism of human xenoprotection.

Characterization of DNA-binding proteins required for glucocorticoid induction of CYP3A23.

Cytochrome P450 (CYP) 3A23 is transcriptionally regulated in rat liver by such glucocorticoids as dexamethasone (DEX) and by such antiglucocorticoids as pregnenolone 16 alpha-carbonitrile (PCN). Based on studies of CYP3A23 gene fragments expressed in primary cultures of adult rat hepatocytes and tested for DNA-protein interactions, we have proposed that the mechanism of CYP3A23 induction by these steroid hormones involves the glucocorticoid receptor or a protein induced by glucocorticoids indirectly interacting with proteins constitutively bound to an enhancer element consisting of a direct repeat of 7-bp separated by two nucleotides in the 5'-flanking region of the CYP3A23 gene (L. Quattrochi et al., J. Biol. Chem. 270, 28,917, 1995). In the present study, we prepared and transiently expressed in cultured rat hepatocytes 20-bp double-stranded (ds)-oligonucleotides containing this direct repeat or various mutations of this direct repeat inserted into a chloramphenicol acetyltransferase (CAT) reporter plasmid. We found that both repeats were necessary for induction of CAT by either DEX or PCN. Analysis of proteins bound to CYP3A23 enhancer through the use of uv cross-linking revealed two rat liver nuclear proteins with molecular masses of approximately 130 and 100 kDa, as well as several proteins of molecular masses between 45 and 60 kDa, that specifically bind to the 20-bp ds-oligonucleotide CYP3A23 enhancer. Methylation interference assays determined that all guanine residues within the direct repeats of this oligonucleotide are important for protein binding. Mutations of these guanine residues abolished binding of nuclear proteins and eliminated DEX or PCN inducibility of CAT. These data suggest that constitutively bound proteins, interacting with the CYP3A23 enhancer possibly as a heterodimeric complex, play a role in the glucocorticoid inducibility of CYP3A23.

Trans-species gene transfer for analysis of glucocorticoid-inducible transcriptional activation of transiently expressed human CYP3A4 and rabbit CYP3A6 in primary cultures of adult rat and rabbit hepatocytes.

Interindividual variation in the spontaneous and in the glucocorticoid-or rifampicin-inducible expression of the CYP3A cytochromes P450, the dominant froms of this supergene family that catalyze the oxidation of numerous drugs and environmental chemicals in human liver, remains largely unexplained, due in part to the lack of a validated animal model. We analyzed the 5'-flanking sequences of CYP3A genes from the rat (CYP3A23, CYP3A2), rabbit (CYP3A6), and human (CYP3A4, CYP3A5, CYP3A7) and found variable regions separated by three areas (consensus I, II, and III) of sequence homology immediately upstream of their respective promoters. We used trans-species gene transfer in cellulo as a new approach for determining the basis for qualitative differences among species in liver expression of different forms of CYP3A. When we transfected into cultured rat hepatocytes vectors containing 5'-flanking DNA from CYP3A23, CYP3A4, or CYP3A6 genes, we found that CAT activity was induced on treatment with dexamethasone or pregnenolone-16 alpha-carbonitrile only if consensus II sequences were included. Rifampicin treatment had no effect. When the same constructions containing consensus II were transfected into rabbit hepatocytes, increased activity was observed on treatment of the cells with dexamethasone or with rifampicin but not with pregnenolone-16 alpha-carbonitrile. These results suggest that the host cellular environment rather than the structure of the gene dictates the pattern of CYP3A inducibility. The application of this new model system will provide a unique technique for identifying mechanisms of induction and advancing the development of appropriate toxicological models for human safety assessment.

A novel cis-acting element in a liver cytochrome P450 3A gene confers synergistic induction by glucocorticoids plus antiglucocorticoids.

The induction by dexamethasone of rat liver CYP3A1 differs from classical glucocorticoid gene regulation in part because both glucocorticoids and antiglucocorticoids such as pregnenolone 16 alpha-carbonitrile (PCN) induce CYP3A1 through transcriptional gene activation. In the present study, we transiently expressed in primary cultures of rat hepatocytes plasmids consisting of CYP3A1 5'-flanking sequences fused to a chloramphenicol acetyltransferase reporter plasmid. Deletional analysis identified a 78-base pair (bp) element located approximately 135 bp upstream of the transcriptional start site that was inducible by treatment of the cultures with dexamethasone or PCN and was induced synergistically by dexamethasone plus PCN. Nuclear extract from control rat liver protected two regions within the 78-bp sequence against digestion with DNase I. The same two regions were protected when nuclear extracts from dexamethasone-treated animals were used. Analysis of both of the "footprints" (FP1 and FP2) failed to reveal a classical sequence for the glucocorticoid-responsive element. A 33-bp element that includes FP1 sequences inserted into the chloramphenicol acetyltransferase reporter plasmid and transiently expressed in rat hepatocytes conferred a profile of dexamethasone and PCN induction similar to that of the 78-bp element. However, an Escherichia coli expressed glucocorticoid receptor protein failed to protect sequences within FP1 in DNase I footprinting experiments and failed to change its mobility in gel shift assays. Moreover, as judged by the gel shift assay, the specific protein binding to this fragment was the same whether nuclear extracts from the liver of untreated or dexamethasone-treated rats were used. We conclude that the activation of CYP3A1 gene transcription by glucocorticoids may involve proteins already bound to the controlling element in the CYP3A1 gene through a mechanism in which GR in the presence of hormone does not bind directly to CYP3A1 DNA.

Induction of cytochrome P-450 by glucocorticoids in rat liver. I. Evidence that glucocorticoids and pregnenolone 16 alpha-carbonitrile regulate de novo synthesis of a common form of cytochrome P-450 in cultures of adult rat hepatocytes and in the liver in vivo.

We administered a series of steroid hormones to primary nonproliferating cultures of adult rat hepatocytes and found that dexamethasone and other glucocorticoids but not sex steroid hormones, mineralocorticoids, or derivatives of pregnenolone other than pregnenolone 16 alpha-carbonitrile (PCN) stimulated de novo synthesis of an immunoreactive protein, indistinguishable from the form of cytochrome P-450 (P450PCN) induced by PCN in rat liver. No difference were discerned among purified liver cytochromes from rats treated with dexamethasone, PCN or dexamethasone plus PCN, among proteolytic digests of these proteins, or among the immunoprecipitated cytochromes prepared from cultured hepatocytes treated with these steroids as judged by electrophoresis on polyacrylamide gels containing sodium dodecyl sulfate followed by immunoblot analysis. Of the steroids tested, dexamethasone proved to be the most efficacious inducer increasing the rate of synthesis of P450PCN from 0.05% of total cellular protein synthesis in incubated control cultures (measured as incorporation of [3H]leucine into immunoprecipitable P450PCN) to as much as 9.4% in cultures incubated for 5 days in medium containing dexamethasone (10(-5) M). As with traditional glucocorticoid-responsive liver functions, induction of immunoreactive P450PCN was dependent on the concentration of dexamethasone (10(-8) to 10(-5) M) and was promptly reversed upon withdrawal of the steroid. However, during the 24-h interval between 24 to 48 h of culture age the hepatocytes were refractory to either induction or de-induction of immunoreactive P450PCN even though continuous exposure of the cells to dexamethasone (including this interval) was mandatory for maximal induction of P450PCN at 120 h in culture. Unlike cultured rat hepatocytes, HTC hepatoma cultures failed to exhibit dexamethasone-responsive expression of immunoreactive P450PCN. We conclude that glucocorticoids and PCN constitute a specific "class" of synthetic and endogenous inducers of a single form of cytochrome P-450.

Measurement of the metabolism of cytochrome P-450 in cultured hepatocytes by a quantitative and specific immunochemical method.

We have defined conditions that permit quantitative and specific measurement of the metabolism of the major phenobarbital-inducible form of cytochrome P-450 protein in primary non-proliferating monolayer cultures of adult rat hepatocytes. Isolated antibodies specifically directed against phenobarbital cytochrome P-450 are used to immunoprecipitate the cytochrome from lysates of cultured hepatocytes pulse-labelled with [(3)H]leucine. Phenobarbital cytochrome P-450 protein is then isolated from the immunoprecipitate by electrophoresis on polyacrylamide gradient slab gels. Specificity of the assay for phenobarbital cytochrome P-450 was established by competition experiments involving other forms of purified cytochrome P-450 as well as by testing antibodies directed against these other forms of the cytochrome. Using purified phenobarbital cytochrome P-450, radiolabelled in both its haem and apoprotein portions, as an internal standard, we demonstrated that, with this immunoassay, recovery of cytochrome P-450 from microsomal samples is nearly complete. Basal rates of synthesis of phenobarbital cytochrome P-450 representing as little as 0.02-0.05% of total cellular protein synthesis were reliably and reproducibly detected in hepatocyte culture maintained in serum-free medium for 72h. Moreover, inclusion of phenobarbital in the culture medium for 96h stimulated not only synthesis de novo of phenobarbital cytochrome P-450 protein, but also accumulation of spectrally and catalytically active cytochrome P-450. Advantages of this immunoassay are that metabolism (synthesis or degradation) of the haem or protein of this important form of the cytochrome can be measured conveniently in the small samples available from cultured cells without the necessity of preparing subcellular fractions.

Inhibition by cycloheximide of degradation of cytochrome P-450 in primary cultures of adult rat liver parenchymal cells and in vivo.

Degradation of cytochrome P-450 was studied in adult rat liver parenchymal cells in primary monolayer culture. In cells incubated in standard culture medium, the amount of cytochrome P-450 decreased at an accelerated rate relative to either the rate of degradation of total protein in the cells or the turnover of cytochrome P-450 in vivo. This change was succeeded by a spontaneous increase in the activity of haem oxygenase, an enzyme system that converts haem into bilirubin in vitro, measured in extracts from the cultured cells. This finding suggests that the rate of cytochrome P-450 breakdown may be controlled by factor(s) other than the activity of haem oxygenase. The decline in cytochrome P-450 and the subsequent increase in haem oxygenase activity was prevented by incubation of hepatocytes in medium containing an inhibitor of protein synthesis such as cycloheximide, puromycin, actinomycin D, or azaserine. The effect of cycloheximide appeared to be due to decreased breakdown of microsomal (14)C-labelled haem. By contrast, cycloheximide was without effect on the degradation of total protein, measured either in homogenates or in microsomal fractions prepared from the cultured cells. These results suggest that the conditions of cell culture stimulate selective degradation of cytochrome P-450 by a process that is inhibited by cycloheximide and hence may require protein synthesis. The findings in culture were verified in parallel studies of cytochrome P-450 degradation in vivo. After administration of bromobenzene, the degradation of the haem moiety of cytochrome P-450 was accelerated in vivo in a manner resembling that observed in cultured hepatocytes. Administration of cycloheximide to either bromobenzene-treated rats or to untreated rats decreased the degradation of the haem moiety of cytochrome P-450. However, the drug failed to affect degradation of haem not associated with cytochrome P-450, suggesting that cycloheximide is not a general inhibitor of haem oxidation in the liver. These findings confirm that the catabolism of hepatic cytochrome P-450 haem is controlled by similar cycloheximide-sensitive processes in the basal steady state in vivo, as stimulated by bromobenzene in vivo, or in hepatocytes under the conditions of cell culture. We conclude that the rate-limiting step in this process appears to require protein synthesis and precedes cleavage of the haem ring.