Hepatocyte nuclear factor 4{alpha} regulates rifampicin-mediated induction of CYP2C genes in primary cultures of human hepatocytes

Abstract

CYP2C enzymes are expressed constitutively and comprise approximately 20% of the total cytochrome P450 in human liver. However, the factors influencing the transcriptional regulation of the CYP2C subfamily have only been addressed recently. In the present study, we used primary cultures of human hepatocytes to investigate the role of HNF4alpha in the pregnane X receptor (PXR)/rifampicin-mediated up-regulation of CYP2C8, CYP2C9, and CYP2C19 gene expression. We first identified new proximal cis-acting HNF4alpha sites in the proximal CYP2C8 promoter [at -181 base pairs (bp) from the translation start site] and the CYP2C9 promoter (at -211 bp). Both sites bound HNF4alpha in gel shift assays. Thus, these and recent studies identified a total of three HNF4alpha sites in the CYP2C9 promoter and two in the CYP2C8 promoter. Mutational studies showed that the HNF4alpha sites are needed for up-regulation of the CYP2C8 and CYP2C9 promoters by rifampicin. Furthermore, silencing of HNF4alpha abolished transactivation of the CYP2C8 and CYP2C9 promoters by rifampicin. Constitutive promoter activity was also decreased. Quantitative polymerase chain reaction analysis demonstrated that silencing HNF4alpha reduced the constitutive expression of CYP2C8 (53%), CYP2C9 (55%), and CYP2C19 (43%) mRNAs and significantly decreased the magnitude of the rifampicin-mediated induction of CYP2C8 (6.6- versus 2.7-fold), CYP2C9 (3- versus 1.5-fold), and CYP2C19 (1.8- versus 1.1-fold). These results provide clear evidence that HNF4alpha contributes to the constitutive expression of the human CYP2C genes and is also important for up-regulation by the PXR agonist rifampicin.

Fig. 1.

Identification of a new HNF4α-binding site in the CYP2C8 promoter region. EMSA demonstrates binding of the new putative HNF4α-binding site of CYP2C8 at −181 bp to HNF4α. The ³²P-labeled probe containing the new putative HNF4α-binding site of CYP2C8 was incubated with in vitro synthesized HNF4α at room temperature for 20 min. Excess (5× or 50×) wild-type (wt) or mutant (mut) cold competitors (CC) were added to the binding reactions for competition analysis. Antibody (Ab) against HNF4α (last lane) resulted in a supershifted band. S, shifted complex; SS, supershifted band.

Fig. 2.

EMSA demonstrates binding of HNF4α to the new putative HNF4α-binding site of the CYP2C9 promoter at −211 bp. A ³²P-labeled oligonucleotide probe containing the new putative HNF4α-binding site of CYP2C9 was incubated with in vitro synthesized HNF4α at room temperature for 20 min. Excess (5× or 50×) wild-type (wt) or mutant (mut) cold competitors (CC) (CYP2C9 or APF-1 oligonucleotides containing an HNF4α site) was added to binding reactions for competition analysis. Antibody (Ab) against HNF4α resulted in a supershifted band (last lane) with the CYP2C9 oligomer. S, shifted complex; SS, supershifted band.

Fig. 3.

Mutations in the HNF4α-binding sites at −152 or −181 bp significantly reduced transactivation of the CYP2C8 promoter by exogenous HNF4α (A) and induction by rifampicin (RIF) in primary human hepatocytes (B). A, primary cultures of human hepatocytes from donor Hu-0861 were transfected with wild-type CYP2C8 (−8.9 to −8.5)-3 kb or mutant constructs containing mutations at −152 bp (2C8/152-mut), −181 bp (2C8/181-mut), or both sites (2C8/dmut). After 12 h, cells were infected with 2.5 × 10⁹ viral particles/ml AdHNF4α or LacZ as a control as described under Materials and Methods. After 48 h, luciferase activity was measured and normalized to the internal control pRL. Transfections were performed in triplicate, and values represent the mean ± S.E. of fold activation relative to that of the wild-type CYP2C8 (−8.9 to −8.5)-3 kb promoter infected with LacZ. AdHNF4α significantly activated the wild-type and single mutant CYP2C8 reporter constructs (*, p < 0.05) and the single mutants 2C8/152-mut, and 2C8/181-mut but not the double mutant. Activation of the CYP2C8 mutant constructs (2C8/152-mut, 2C8/181-mut, or 2C8/HNF4αdmut) was significantly lower than of the wild-type CYP2C8 reporter construct (#, p < 0.05, whereas the basal activity of those transfected with lacZ was significantly lower for the 2C8/152-mut and 2C8/HNF4αdmut than those of the wild-type construct (†, p < 0.05). B, effect of mutation of the HNF4α sites in the CYP2C8 promoter on activation of CYP2C8 by rifampicin. Primary cultures of human hepatocytes transfected with the wild-type CYP2C8 (−8.9 to −8.5)-3 kb promoter or its HNF4α mutants were treated with 0.2% DMSO or 10 μM rifampicin for 24 h. Values for rifampicin-mediated transactivation are expressed relative to those of the wild-type CYP2C8 promoter treated with the vehicle DMSO. Rifampicin significantly increased activity of the wild-type CYP2C8 and 2C8/152-mut reporter activity compared with the vehicle control (*, p < 0.05), whereas mutation of each of the HNF4α sites significantly decreased constitutive activity (†, p < 0.05) or the induction by rifampicin (#, p < 0.05).

Fig. 4.

Effect of mutation of different HNF4α-binding sites on CYP2C9 promoter transactivation by HNF4α and rifampicin (RIF) in primary human hepatocytes. A, primary cultures of human hepatocytes from donor Hu-1125 were transfected with the wild-type CYP2C9-3 kb promoter or CYP2C9 promoters containing mutations of the individual HNF4α-binding sites at −150 (2C9/150-mut), −185 (2C9/185-mut), or −211 bp (2C9/211-mut), or mutation of all three sites (2C9/HNF4αtmut) followed by infection with AdHNF4α or LacZ as a control. Luciferase activity assays were performed 48 h later. All transfections were performed in quadruplicate, and values represent the mean ± S.E. of the fold activation relative to that of the CYP2C9-3 kb promoter infected with LacZ. AdHNF4α significantly enhanced the activation of wild-type CYP2C9 and single HNF4α mutants (*, p < 0.05) but not the triple mutation. Luciferase activity of the CYP2C9 mutants transfected with LacZ (†, p <0.05) or adenoviral HNF4α was significantly lower than that of the wild-type CYP2C9 reporter construct (#, p < 0.05). B, effect of mutation of the HNF4α-binding sites on transactivation of the CYP2C9 promoter by rifampicin (RIF). Cultures of primary human hepatocytes were transfected with the CYP2C9-3 kb promoter or its HNF4α mutants followed by treatment with 10 μM rifampicin for 24 h. Values for rifampicin-mediated transactivation are expressed as fold relative to the activity of the wild-type CYP2C9 promoter construct treated with the vehicle DMSO. Rifampicin significantly increased activity of the wild-type CYP2C9 promoter and promoter constructs containing single mutants compared with the vehicle control (*, p < 0.05). However, activation of the CYP2C9 HNF4α promoter mutants by rifampicin was significantly lower than that of the wild-type (#, p < 0.05;), and the triple mutation abolished rifampicin activation. Constitutive activity of the mutants in the absence of rifampicin was significantly lower than that of wild type (†, p < 0.05).

Fig. 5.

Silencing HNF4α decreases basal CYP2C8 and CYP2C9 promoter activity and essentially abolishes rifampicin (RIF)-mediated transactivation in primary cultures of human hepatocytes. A, primary cultures of human hepatocytes were transfected with the wild-type CYP2C8 (−8.9 to −8.5)-3 kb promoter reporter construct. After 12 h, cells were infected with siHNF4-I or scrambled (SCR) adenovirus for 1.5 h and then incubated at 37°C for 24 h. Transfected cells were then treated with 10 μM rifampicin or 0.2% DMSO for 24 h. Transfections were performed in triplicate, and values represent the mean ± S.E. in two donors (Hu-0747 and Hu-0808). Treatment with rifampicin resulted in a 2-fold increase in wild-type CYP2C8 (−8.9 to −8.5)-3 kb promoter reporter activity (*, p < 0.05; **, p < 0.01). Silencing HNF4α resulted in significant down-regulation of the basal wild-type CYP2C8 (−8.9 to −8.5)-3 kb reporter activity (†, p < 0.05; ††, p < 0.01) and abolished rifampicin-mediated transactivation (##, p < 0.01). B, silencing HNF4α reduced both the basal and rifampicin-mediated transactivation of the wild-type CYP2C9-3 kb reporter construct. All transfections were performed in triplicate samples from donors Hu-0798 and Hu-0813, and values represent the mean ± S.E. Treatment with rifampicin resulted in a significant 2-fold increase in wild-type CYP2C9 promoter reporter activity compared with DMSO treatment and a small but significant increase after silencing HNF4α (*, p < 0.01; **, p < 0.001). Silencing HNF4α resulted in significant down-regulation of the constitutive wild-type CYP2C9 reporter activity in the absence of rifampicin (†, p < 0.05; ††, p < 0.01) and significantly decreased transactivation by rifampicin (##, p < 0.01).