Abstract
The molecular mechanisms of regulation of theCYP3A4 gene have been examined in an in vitro reporter gene system, containing −1 kb of the CYP3A4 promoter, in a HepG2 cell line. This system allows for the separate and combined transfection of expression plasmids encoding the human glucocorticoid receptor (hGR) and the human pregnane X receptor (hPXR), and, therefore, the opportunity to assess the role of these receptors in the induction process. Hydrocortisone produces a dose-dependent increase in CYP3A4 activation, a response that is increased in the presence of either receptor. Moreover, transfection of the hPXR decreased the EC50 for hydrocortisone-dependent induction by a factor of 3.3, a response that was not changed by simultaneous cotransfection of the hGR. In addition, the hydrocortisone dose-response curve falls within the physiological blood level concentration of this steroid, implicating a regulatory role for hydrocortisone in the basal level of CYP3A4 expression. Although the responses to dexamethasone and rifampicin were both increased by both receptors, dexamethasone activation of CYP3A4 was similar for both the hGR and the hPXR, whereas rifampicin-dependent activation favored the hPXR. We conclude that regulation of the CYP3A4 gene is receptor-dependent and that hydrocortisone may function as a regulator of basal expression via the hPXR and the hGR. The implications of this latter conclusion for possible regulatory interactions between hydrocortisone and xenobiotic inducers remain to be clarified.
Members of the cytochrome P450 3A subfamily are highly expressed in human liver and intestine, and play a pivotal role in the metabolism of clinically used drugs including erythromycin, cyclosporine, nifedipine, midazolam, and certain toxic environmental chemicals (Shimada et al., 1994). CYP3A4 is predominantly expressed in human liver and intestine, where it comprises approximately 30 to 50% of the total cytochrome P450 population in these tissues (Watkins, 1994) and is highly inducible in humans by synthetic glucocorticoids [dexamethasone (DEX)1], macrolide antibiotics [rifampicin (RIF)], and phenobarbitone (Guengerich, 1999). In addition, the human response to inducers is variable, and 4- to 8-fold variations in the induction of enterocyte CYP3A4 mRNA and the corresponding catalytic activity in patients receiving RIF, a known CYP3A4 inducer, have been reported (Kolars et al., 1992). This isoform is also inhibited by structurally diverse drugs, including cimetidine, troleandomycin, propofol, ketoconazole, and erythromycin (Chang and Kam, 1999). Thus, taken collectively, this diverse range of substrates, inducers, and inhibitors of CYP3A4 activity creates the potential for clinically significant drug interactions involving this isoform, particularly in therapeutic areas where polypharmacy is common practice, for example, in the elderly (Lehmann et al., 1998), or in the long-term prophylaxis of psychiatric disorders (Ketter et al., 1995).
Recently, several laboratories have investigated the molecular basis for the induction of CYP3A4 reporter gene expression. The CYP3A4 promoter has been cloned, and a 20-bp region residing ∼150bp upstream of the transcription initiation site has been shown to confer responsiveness to DEX and RIF (Hashimoto et al., 1993; Barwick et al., 1996). There has been much discussion as to the potential role of the glucocorticoid receptor (GR) in the induction of the CYP3A subfamily, and this laboratory has recently reported that CYP3A4 inducers cannot be regarded as a homogeneous group, as some inducer responses are dependent on the GR whereas others are not (Ogg et al., 1999).
Computer sequence analysis of −1 kb of the CYP3A4 regulatory region isolated in our laboratory and as originally reported by Hashimoto et al. (1993) has demonstrated the presence of response elements including a glucocorticoid response element, which exhibits 58% sequence homology to the consensus glucocorticoid response element (GGTACA-nnn-TGTTCT; Beato, 1989).
The complexity of CYP3A4 gene regulation is also underscored by the recent identification of a human orphan nuclear receptor, termed pregnane X receptor (PXR), which binds to a response element in the CYP3A4 promoter and is activated by a range of drugs and steroids known to induce CYP3A4 expression (Lehmann et al., 1998). The PXR response element is localized to a ∼20-bp sequence in the CYP3A4 promoter (at approximately −150 bp in the −1 kb CYP3A4 promoter fragment we have isolated; Ogg et al., 1999), containing two copies of the nuclear receptor half-site sequence AG(G/T)TCA organized as an everted repeat (ER) and separated by 6 bp, collectively termed an ER6 motif. Activation of the PXR by drug inducers appears to have a functional role in vivo as RIF concentrations required to activate CYP3A4 via the PXR/ER6 in vitro has an EC50 value of 800 nM, well below patient blood levels of 10 to 40 uM of RIF (Lehmann et al., 1998). Comparison of the human PXR (hPXR) with the recently cloned mouse PXR (Kliewer et al., 1998) reveals marked differences in their activation by certain drugs, which may account in part for the species-specific effects of inducers on CYP3A gene expression.
An in vitro reporter gene assay system (96-well plate format) to assess the molecular mechanisms of CYP3A4 activation has recently been reported by researchers in this laboratory (Ogg et al., 1999). This validated system consists of −1 kb of the CYP3A4 promoter, a cytomegalovirus (CMV) promoter, and a secretory placental alkaline phosphatase (SPAP) reporter gene, and when this construct is transfected into HepG2 cells, all of the known in vivo inducers of CYP3A4 in humans activate reporter gene activity. Accordingly, this system now affords us the opportunity to assess the role of single and combined transfections of the human glucocorticoid receptor (hGR) and hPXR on both xenobiotic and steroid activation of theCYP3A4 gene, and, therefore, to further our understanding of the molecular mechanisms of regulation of this clinically important human liver cytochrome P450 isoform.
Materials and Methods
Chemicals.
Hydrocortisone, DEX, and RIF were of cell culture quality and were purchased from Sigma Chemical Co. (St. Louis, MO). Solvents (methanol and dimethyl sulfoxide) used were of cell culture grade and were obtained from Fisher Scientific Co. (Fairlawn, NJ) and Sigma, respectively.
Plasmids.
The alkaline phosphatase reporter plasmid pCMV-cSPAP (termed pCMV) was a gift from Glaxo-Wellcome (Ware, UK) and contained an SPAP reporter gene. Secretory reporter gene activity could therefore be directly and rapidly assayed in the culture medium without the need for disruption of the HepG2 cells. The CYP3A4 reporter construct, p3A4-CMV-cSPAP (termed p3A4) was previously engineered in our laboratory (Williams, 1997; Ogg et al., 1999) by inserting ∼1 kb (−1087 to −57 bp) of the CYP3A4 upstream regulatory region (Hashimoto et al., 1993) into the pCMV-cSPAP plasmid. We chose to use the 1-kb promoter fragment of CYP3A4 as this has been characterized previously by Hashimoto et al. (1993) and contains several regulatory elements including the response elements for the hGR, hPXR (termed ER6) estrogen receptor, HNF-5, HNF-4, AP3, CAAT box, BTE, and TATA box. In this context, it should be noted that Lehman et al. (1999) reported that several xenobiotics and steroids can activate a minimal construct containing only the ER6 response element. However, in view of the additional regulatory elements present in our 1-kb fragment, which may modify any ER6 -mediated responses, we chose to use the larger 1-kb fragment in our studies. The pSG5-hGR plasmid containing the full-length hGR was a gift from Dr. Jonathan Tugwood, Zeneca Central Toxicology Laboratory, Macclesfield, UK. The latter three plasmids were as previously described in detail byOgg et al., 1999. The pSG5-hPXR expression plasmid (termed hPXR) was a gift from Dr Steven Kliewer (Glaxo Wellcome Research and Development, Research Triangle Park, NC) and was generated by polymerase chain reaction (PCR) amplification and subcloning of nucleotides 1–1608 of the hPXR clone into the pSG5 expression vector (Lehmann et al., 1998). Plasmids were purified using Qiagen Endo Free Maxi Preps according to manufacturer instructions, as the presence of endotoxin contamination of DNA can adversely affect the transfection efficiency (Weber et al., 1996).
Cell Culture and Transient Transfection.
All cell culture media and supplements were purchased from Life Technologies (Paisley, Scotland, UK). HepG2 cells are a human hepatocyte carcinoma line obtained from European Collection of Animal Cell Cultures (ECACC; no. 85011430; Porton Down, UK). HepG2 cells were cultured in minimum essential medium with Earle's salts supplemented with 1% nonessential amino acids, 2 mM l-glutamine, 100 ug/ml gentamycin, and 10% fetal bovine serum.
The transient transfection protocol used was that previously used in our laboratory (Ogg et al., 1999), which takes place over a 5-day period and is based on the calcium phosphate precipitation method previously described by Jordan et al., 1996. To ensure reproducibility, HepG2 cells only up to passage 13 after receipt by ECACC were used.
Day 1: Seeding of HepG2 Cells.
Cells were diluted to a concentration of 6 × 105 cells/ml in minimum essential medium, seeded in 96-well plates, and placed in a humidified container in a cell culture incubator at 37C° in 5% CO2 for 24 h.
Day 2: Transfection.
One hour before transfection, the growth medium was removed and replaced by fresh medium, and the cells were returned to the incubator. During this time, the transfection mixture was prepared on ice for optimum precipitate formation. DNA precipitate suspension (2 ml) was prepared at a standard DNA concentration of 12.5 ug/ml. Where indicated, the hGR and/or the hPXR plasmids were added at a concentration of 12.5 ug/ml. The transfection mix (120 μl) was added per well (equivalant to 12 μl precipitate/well). The plates were placed in a humidified container and returned to the cell culture incubator.
Day 3: Xenobiotic Dosing.
The growth media were aspirated and assayed for SPAP activity. Subsequently, hydrocortisone was dissolved in methanol and used in final concentrations of 0.05, 0.1, 0.5, 1, 5, 10, 25, and 50 uM. DEX and RIF were dissolved in dimethyl sulfoxide and were used in concentrations of 50 uM, concentrations previously determined to give a maximum induction response (Ogg et, 1999). Solvent controls were included in the corresponding transfection experiments. Xenobiotic solutions and solvent controls were added to the cell culture media in a final concentration of 0.1% solvent, of which 120 μl was added per well. All solutions were freshly prepared on the day of dosing.
Day 4: Cells Incubated.
Day 5: Alkaline Phosphatase Activity Determination.
Secreted alkaline phosphatase activity was assayed by transferring aliquots of the cell culture medium (25 μl/well) into 96-well Optiplates (polystyrene microplates; Packard, Pangbourne, UK). Samples were heat-treated for 1 h at 65°C in an oven to deactivate endogenous cellular alkaline phosphatase activity. SPAP activity was assayed using the Aurora alkaline phosphatase chemiluminescent assay kit (ICN, Thame, UK) according to the manufacturer's protocol in a Packard LumiCount automated plate reader (Packard). SPAP activity was assayed both at 24 h post-transfection (before inducer addition) and 72 h post-transfection (48 h after inducer addition). The transient transfection induction assay is summarized in Fig.1.
Data Analysis.
A pCMV control was included for every dose of inducer, and the relative changes for the difference in reporter gene activity between 72 and 24 h were calculated to take into account any variation in cell seeding, transfection efficiency, cytotoxicity of inducers, or cell proliferation effects.
In addition, to calculate accurately the fold induction due to inducer treatment, it is essential to compare the change in the relative expression of SPAP from the test (p3A4) to the control (pCMV) plasmids in the presence of inducers compared with the solvent control, as follows:
Results and Discussion
In the absence of any added receptor, the inducibility of CYP3A4 by different concentrations of hydrocortisone showed a concentrationdependent increase in CYP3A4 reporter gene expression (Fig. 2), indicating the probable presence of constitutively expressed receptors in the HepG2 cell line. This induction response was enhanced by single cotransfection with either the hGR or hPXR expression plasmids, and simultaneous cotransfection with both receptors caused a similar induction to that observed with either receptor alone (Fig. 2). Duel receptor transfection was clearly not additive or synergistic, probably indicating competition by the two receptors for available hydrocortisone. The shape of the dose-response curves indicated that the influence of both receptors was to decrease the threshold value for CYP activation. In this context, it is relevant to note that the normal serum level of hydrocortisone in humans ranges from 0.2 to 0.75 μM (Williams and Marks, 1994), thereby inferring a possible physiological role for hydrocortisone in regulating the basal level of transcription of the CYP3A4 gene. It should be noted that we have not determined the constitutive levels of hydrocortisone in the HepG2 cells; therefore, the data shown in Fig. 2 are uncorrected for this. The fact that CYP3A4 was trans-activated in the absence of cotransfected hGR and hPXR receptors (albeit at a lower level) suggests to us that native HepG2 cells express sufficient amounts of these receptors to contribute to the overall response. Currently, we are investigating this possibility by determining the basal level of mRNA receptor expression using an reverse transcriptase-PCR-based method.
Whereas transfection of either receptor (either singly or in combination) increased the maximum induction response (Imax) to hydrocortisone, cotransfection of the hGR apparently did not alter the EC50 value for this steroid (Table 1). In contrast, transfection of the hPXR (either alone or in combination with the hGR) resulted in a 3- to 4-fold decrease in the EC50value for hydrocortisone-dependent induction of CYP3A4. Whereas we are aware that steroid binding to their cognate receptors is usually characterized by Kd values in the nanomolar range, we must emphasize that the EC50 values for hydrocortisone reported herein (micromolar range) relate to the transactivation reaction. Accordingly, as the overall transactivation reaction is multifactorial, and not solely a simple function of steroid-receptor binding, it is not surprising that these two values differ. The EC50values reported by us for CYP3A4 transactivation by hydrocortisone in the presence of cotransfected hPXR (0.35 μM, Table 1) are somewhat lower than those reported by Lehman et al. (1999), namely, in the range of 3 to 30 μM. We should point out that our reporter construct contains approximately the first 1 kb (−57 to −1087) of the CYP3A4 promoter whereas the transactivation data for hydrocortisone reported by Lehman et al. (1999) only used the hPXR response element (termed the ER6-RE). Although our 1-kb promoter fragment does indeed contain the ER6-RE (at −150 bp), there are additional response elements in our construct (Hashimoto et al., 1993) that very likely modify the influence of the effect of the ER6-RE in the overall transactivation process.
Taken collectively, these data indicate: 1) the importance of cellular receptors in hydrocortisone-dependent regulation of CYP3A4; and 2) that the hPXR very likely plays a more predominant role than the hGR, a hypothesis that requires additional examination. Table2 shows the DEX- and RIF-dependent activation of CYP3A4, and again, hGR and hPXR transfection, appear to increase the induction response, particularly for hPXR-mediated induction by RIF.
These studies have confirmed that the regulation of theCYP3A4 gene is receptor-dependent and multifactorial for the three inducers studied herein. Although we have separately demonstrated CYP3A4 activation by both an endogenous steroid and two xenobiotics, all three of which appear to activate both the hGR and the hPXR, the likelihood exists that there is mutual competition for these two receptors between endobiotics and xenobiotics; we are examining this hypothesis in our ongoing studies. In addition, our receptor cotransfection approach to investigate the roles of the hPXR and hGR in regulating the CYP3A4 gene may be further explored by mutating the cognate response elements in the CYP3A4 promoter and is also the subject of our on-going studies.
Footnotes
-
Send reprint requests to: G. Gordon Gibson, Ph.D., Molecular Toxicology Group, School of Biological Sciences, University of Surrey, Guildford, Surrey, GU2 5XH UK. E-mail:g.gibson{at}surrey.ac.uk
- Abbreviations used are::
- DEX
- dexamethasone
- GR
- glucocorticoid receptor
- hGR
- human glucocorticoid receptor
- PXR
- pregnane X receptor
- hPXR
- human pregnane X receptor
- RIF
- rifampicin
- ER
- everted repeat
- CMV
- cytomegalovirus
- SPAP
- secretory placental alkaline phosphatase
- Received August 10, 1999.
- Accepted February 16, 2000.
- The American Society for Pharmacology and Experimental Therapeutics