2024 Volume 47 Issue 6 Pages 1218-1223
Unknown interactions between drugs remain the limiting factor for clinical application of drugs, and the induction and inhibition of drug-metabolizing CYP enzymes are considered the key to examining the drug–drug interaction (DDI). In this study, using human HepaRG cells as an in vitro model system, we analyzed the potential DDI based on the expression levels of CYP3A4 and CYP1A2. Rifampicin and omeprazole, the potent inducers for CYP3A4 and CYP1A2, respectively, induce expression of the corresponding CYP enzymes at both the mRNA and protein levels. We noticed that, in addition to inducing CYP1A2, omeprazole induced CYP3A4 mRNA expression in HepaRG cells. However, unexpectedly, CYP3A4 protein expression levels were not increased after omeprazole treatment. Concurrent administration of rifampicin and omeprazole showed an inhibitory effect of omeprazole on the CYP3A4 protein expression induced by rifampicin, while its mRNA induction remained intact. Cycloheximide chase assay revealed increased CYP3A4 protein degradation in the cells exposed to omeprazole. The data presented here suggest the potential importance of broadening the current DDI examination beyond conventional transcriptional induction and enzyme-activity inhibition tests to include post-translational regulation analysis of CYP enzyme expression.
Drug–drug interactions (DDIs), indicating the impact of one drug on the pharmacological effects of another, are a major cause of adverse drug effects that increase the economic burden and reduce the patient QOL, posing a serious public health concern.1) Many studies have investigated the roles of CYP enzymes in drug metabolism. Induction or inhibition of CYP enzymes is directly linked to drug-induced toxicity and DDIs that lead to treatment failure.2,3) Owing to the importance of DDIs in drug development, the U.S. Food and Drug Administration (FDA), European Medicines Agency (EMA), and Pharmaceuticals and Medical Devices Agency (PMDA) of Japan recommend the testing of the CYP induction/inhibition potential of all drugs.4–7)
In preclinical drug screening, the use of in vitro model systems has many advantages, including fewer animals required, smaller sample volumes for testing, shorter development periods, and increased throughput in the evaluation of multiple test compounds. Primary cultures of human hepatocytes are used as the gold standard model systems for CYP induction assays.8) However, primary human hepatocytes exhibit several limitations, such as isolation difficulty, high variability between different donors, quick failure in biological function, and limited proliferation capacity.9) Therefore, alternative approaches have been developed for evaluating CYP induction/inhibition. The human HepaRG cells, obtained from the liver tumor of a patient with hepatocarcinoma and hepatitis C infection, exhibit hepatocyte-like functions and are used as alternative in vitro models to primary human hepatocytes to evaluate CYP induction/inhibition.10) HepaRG 5F clone, generated by Sigma-Aldrich, exhibits improved growth characteristics and enhanced functionality, expressing various factors involved in drug metabolism, including a large panel of phase I and phase II drug-metabolizing enzymes and drug transporters.
In this study, we examined the drug-induced induction of the expression of CYP3A4 and CYP1A2, two major CYP enzymes accounting for approximately 30 and 12% of the total CYP enzyme expression in the liver, respectively,11) in HepaRG 5F cells. Notably, omeprazole, a well-known CYP1A2 inducer, induced both CYP1A2 and CYP3A4 mRNA expression but did not induce CYP3A4 expression. Subsequently, we investigated the mechanisms underlying the omeprazole-dependent suppression of CYP3A4 protein expression.
An undifferentiated human hepatoma cell line HepaRG 5F (MTOX1010) was purchased from Sigma-Aldrich (U.S.A.). Cells were seeded onto collagen-coated 24-well plates and cultured in William’s E medium (WEM) containing HepaRG Thawing and Plating Medium Supplement (ADD671C, Biopredic International, France), 2 mM alanine (Ala)-glutamine (Gln), and 1% penicillin–streptomycin solution (Nacalai Tesque, Japan). After reaching confluence, cells were differentiated by culturing in WEM containing HepaRG Maintenance and Metabolism Medium Supplement (ADD621C) for a minimum of 14 d, followed by 48 h in WEM supplemented with HepaRG Induction Medium Supplement (ADD641C). Drug-induced expression of CYP proteins and mRNA was measured in differentiated HepaRG 5F cells grown in WEM with HepaRG Serum-Free Induction Medium Supplement (ADD651C). Where specified, cells were treated with rifampicin (Nacalai) or omeprazole (Tokyo Chemical Industry, Japan).
ImmunoblottingCells were lysed in Laemmli buffer containing 1× cOmplete Protease Inhibitor cocktail (Roche Diagnostics, Switzerland), as described previously.12) Immunoblots were performed using our standard method13) with antibodies against CYP1A2 (sc-53614, 1 : 200, Santa Cruz, U.S.A.), CYP3A4 (sc-53850, 1 : 200, Santa Cruz), and α-tubulin (T6199, 1 : 1,000, Sigma-Aldrich). The cycloheximide (CHX) chase assay was performed by adding CHX (Wako, 100 µg/mL) to the medium.14)
Quantitative RT-PCR (qRT-PCR)Total cell RNA was isolated with the RNeasy mini kit (Qiagen, the Netherlands). cDNA was synthesized using SuperScript VILO cDNA synthesis kit (Invitrogen, U.S.A.) and quantitative real-time PCR was performed using THUNDERBIRD SYBR qPCR mix (TOYOBO, Japan) and StepOnePlus (Applied Biosystems, U.S.A.).15) 18S rRNA was used as an internal control. The primer sets used in this study included: CYP3A4, Fw: 5′-TGTAAAGAAACACAGATCCCCCTGA-3′, Rv: 5′-CAGGCTCCACTTACGGTGCC-3′, CYP1A2, Fw: 5′-TACCTGCCTAACCCTGCCCT-3′, Rv: 5′-ACACTGTTCTTGTCAAAGTCCTGA-3′, TWIST1, Fw: 5′-GGCCAGGTACATCGACTTCC-3′, Rv: 5′-CTCCATCCTCCAGACCGAGA-3′, ALB, Fw: 5′-GAGACCAGAGGTTGATGTGATG-3′, Rv: 5′-AGTTCCGGGGCATAAAAGTAAG-3′, TF, Fw: 5′-GTCAACTGTGTCCAGGGTGTGG-3′, Rv: 5′-TCAGACACTTGAAGGCTCCCG-3′, and 18S rRNA, Fw: 5′-CGGACAGGATTGACAGATTG-3′, Rv: 5′-GGCATCACAGACCTGTTATTG-3′.
Quantification and Statistical AnalysisWestern blot band intensities were quantified using Image J software. Statistical analysis was performed using GraphPad Prism 8 with the statistical tests shown in the figure legends.
Undifferentiated HepaRG 5F cells showed elongated cell morphology and appeared to be actively dividing (Fig. 1A). After 14 d of culture in the differentiation medium (see Materials and Methods), the cells formed clusters of granular epithelial cells resembling hepatocytes. Subsequent culture of the cells in an induction medium promoted their differentiation into hepatocytes (Fig. 1A). qRT-PCR analysis verified reduced expression levels of the mesenchymal marker, TWIST family bHLH transcription Factor 1 (TWIST1), and increased expression levels of the hepatocyte markers, albumin (ALB) and transferrin (TF), in differentiated HepaRG cells (Fig. 1B), thereby confirming the successful differentiation of the cells.16,17)
(A) Representative images of undifferentiated and differentiated HepaRG cells. To obtain differentiated HepaRG cells, cells were initially cultured in the HepaRG maintenance and metabolism medium (MMM) for 14 d and then in the HepaRG induction medium (IM) for two days. Scale bar, 50 µm. (B) Expression levels of hepatocyte markers in undifferentiated (Undiff.) and differentiated (Diff.) HepaRG cells. Bar graphs show the mRNA expression levels of ALB and TF determined via qRT-PCR. TWIST1 was determined as an undifferentiated marker gene. Values were normalized to the 18S rRNA expression level and plotted relative to the expression levels in undifferentiated HepaRG 5F cells. n = 5–6 biological replicates. (C) Schematic experimental design for drug treatment in differentiated HepaRG 5F cells. Rif, rifampicin. Ome, omeprazole. (D, E) Protein and mRNA expression levels of CYP3A4 and CYP1A2 in differentiated HepaRG 5F cells treated with rifampicin (50 µM) or omeprazole (50 µM) for 48 or 72 h. Values were plotted relative to the expression levels in vehicle (Veh)-treated cells. For a direct comparison of mRNA and protein responses, the samples were aliquoted in duplicate and used for qRT-PCR and Western blotting. n = 3 biological replicates. Values are means ± standard error of the mean (S.E.M.). Statistical analyses were conducted via one-way ANOVA followed by Tukey’s multiple comparison test in (D) and (E) and via an unpaired t-test in (B). ** p < 0.01, *** p < 0.001, **** p < 0.0001, n.s., not significant.
Next, we examined drug treatment-induced expression of CYP3A4 and CYP1A2 in differentiated HepaRG cells. The cells were treated with the pregnane X receptor (PXR) activator, rifampicin, or aryl hydrocarbon receptor (AhR) activator, omeprazole, every 24 h for two or three days (48 or 72 h in total). Then, qRT-PCR and immunoblotting analyses were conducted to determine the mRNA and protein expression levels of the CYP enzymes in cells with or without drug treatment (Fig. 1C). As expected, CYP1A2 expression was significantly induced by omeprazole but not rifampicin at both the mRNA and protein levels (Figs. 1D, E). Both mRNA and protein expression levels of CYP3A4 were also expectedly increased in cells treated with rifampicin (Figs. 1D, E). However, at both time points examined, we noticed a differential response between mRNA and protein in CYP3A4/CYP3A4 expression in omeprazole treatment; although it did not induce CYP3A4 protein expression (Fig. 1D), omeprazole significantly upregulated the CYP3A4 mRNA expression levels comparable to those in rifampicin-treated cells (Fig. 1E). These results indicate that CYP3A4 protein expression is limited in omeprazole-treated cells as compared to rifampicin-treated cells.
Co-application of Omeprazole Abrogates Rifampicin-Induced CYP3A4 ExpressionWe, next, examined whether omeprazole could interfere with the rifampicin-induced CYP3A4 expression. Cells were treated with rifampicin with or without omeprazole for three days. As shown in Fig. 2A, cells treated with rifampicin and omeprazole exhibited higher CYP3A4 mRNA expression levels than the vehicle controls (14.89-fold change vs. Ctrl, p = 0.0019, Fig. 2A) that were comparable to those in cells treated with only rifampicin (0.9118-fold change vs. Rif, p = 0.8687, Fig. 2A). Even with the presence of comparable amounts of induced CYP3A4 mRNA, CYP3A4 protein expression levels were found significantly lower in the rifampicin plus omeprazole-treated group than in the rifampicin only-treated group (0.5730-fold change vs. Rif, p = 0.0019, Fig. 2A).
(A) Protein and mRNA expression levels of CYP3A4 in differentiated HepaRG cells treated with rifampicin (50 µM) in the presence or absence of omeprazole (50 µM) for 72 h. Values were plotted relative to the expression levels in Rif(–)/Ome(–) cells. n = 3–4 biological replicates. (B) CYP3A4 protein stability in HepaRG cells with or without omeprazole (50 µM), determined via the cycloheximide (CHX) chase assay. Cells were treated with rifampicin (20 µM) 24 h before CHX treatment to prevent degradation of omeprazole by CYP3A4 during the experiment.18) n = 3 biological replicates. Values are means ± S.E.M. Statistical analyses were conducted via one-way ANOVA followed by Tukey’s multiple comparison test in (A) and via two-way ANOVA followed by Sidak’s post-hoc test in (B). ##p < 0.01 vs. CHX in (B). * p < 0.05, ** p < 0.01, **** p < 0.0001, n.s., not significant.
The data above prompted us to test whether omeprazole promotes degradation of CYP3A4 protein. To monitor the half-life of CYP3A4 protein in the presence or absence of omeprazole, we performed a chase assay using CHX, a de novo protein synthesis inhibitor (Fig. 2B). To express CYP3A4, HepaRG cells were treated with a relatively low concentration of rifampicin (20 µM) for 24 h, as high CYP3A4 expression counteracts the effects of omeprazole by metabolizing it.18) After removing rifampicin, the cells were treated with CHX with or without omeprazole and cultured for the indicated time periods. Proteins were extracted from cells and subjected to immunoblotting analysis. Consistent with previous reports,19,20) CYP3A4 protein levels remained stable until 6 h, and the levels at 3 or 6 h in the CHX-treated group were not significantly different from those prior to CHX treatment (Fig. 2B). We found that the cells treated with CHX and omeprazole together exhibit significantly decreased CYP3A4 expression levels compared to those in the time-matched CHX only group, which reached statistical significance at 6 h in comparison to pre-CHX values (Fig. 2B). These results indicate that omeprazole triggers a decrease in CYP3A4 protein stability, thereby contributing to the limited accumulation of CYP3A4 protein in HepaRG cells.
In this study, we observed a previously undescribed post-translational regulation of CYP3A4 protein accumulation by omeprazole. Despite its ability to induce CYP3A4 mRNA expression, omeprazole failed to increase the CYP3A4 protein expression in HepaRG cells. Concurrent administration of omeprazole and rifampicin reduced the rifampicin-induced CYP3A4 protein expression, without affecting its mRNA expression, suggesting a post-transcriptional regulation mechanism. Using a CHX chase assay, we observed an accelerated CYP3A4 protein degradation in omeprazole-treated cells. Although further investigation is warranted to elucidate the specific effects of omeprazole on CYP3A4, our data provide evidence for the effect of omeprazole on the degradation of CYP3A4 protein, which affects the CYP3A4 protein accumulation levels in HepaRG cells.
Upon activation by rifampicin and omeprazole, PXR and AhR heterodimerize with the retinoic acid receptor α (RXRα)21,22) or AhR nuclear translocator (ARNT),23,24) forming a transcriptionally active complex. Previous chromatin immunoprecipitation assay demonstrated enrichment of RXRα25) and AhR26) at the CYP3A4 promoter region up to −10 kb upstream of the transcription start site (GSE127399 and GSE206029), which supports our observation of CYP3A4 mRNA induction by rifampicin and omeprazole. CYP3A4 induction by rifampicin and omeprazole is further demonstrated by established experimental models of hepatocytes, including HepG2 cells and human primary hepatocytes.27,28) The molecular mechanism(s) underlying omeprazole-induced enhancement of CYP3A4 protein degradation, on the other hand, remains totally unknown. Lansoprazole, another AhR-dependent CYP1A2 inducer, has also been reported to induce CYP3A4 mRNA without a concurrent increase in its protein expression,27) pointing to a plausible mechanism involving AhR-dependent protein degradation. Apart from its role as a transcriptional factor, AhR mediates proteolysis in a ligand-dependent manner by assembling an E3 ubiquitin ligase complex.29) Within this complex, AhR recruits target proteins for proteolysis, such as the peroxisome proliferator-activated receptor-γ and estrogen receptor, facilitating protein ubiquitination and subsequent degradation.29,30) As E3 ubiquitin ligase-mediated ubiquitylation is involved in CYP3A4 protein degradation,31) omeprazole possibly promotes proteasomal degradation of CYP3A4 via AhR-dependent E3 ubiquitin ligase-mediated ubiquitination. However, other mechanisms are equally possible. For example, acetaminophen appears to promote CYP3A4 protein degradation via a lysosome-dependent mechniasm.32) It is also reported that ticagrelor and pazopanib increase CYP3A4 mRNA expression but not its protein abundance; however, the underlying mechanism is still not understood.33) Further investigations are warranted to elucidate the mechanisms underlying drug-induced degradation of CYP3A4 protein.
CYP3A induction-mediated DDI is one of the major concerns in drug development and clinical practice.34) In preclinical safety assessments, effects of drugs on CYP induction are routinely investigated via the quantification of mRNA expression and enzyme activity.35) However, if the test compound acts as both an inducer (for mRNA expression) and inhibitor (for enzymatic activity), enzyme activity tests can provide mixed results. For example, ritonavir, a potent peptide-like human immunodeficiency virus protease inhibitor, is reported to induce CYP3A4 mRNA and protein expression in primary human hepatocytes. However, because ritonavir also acts as a CYP3A4 enzymatic inhibitor, induction of human CYP3A4 by ritonavir is masked when CYP3A4 enzyme activity is examined in cells.1) To minimize such confounding effects, the FDA recommends assessment of CYP induction via transcriptional analysis.4) In this study, we have added another exception to the discordance between mRNA expression and enzyme activity (protein expression) in drug-treated cells. Importantly, omeprazole inhibited rifampicin-induced CYP3A4 protein expression. A standard CYP3A4 enzyme activity assay can be employed to detect the omeprazole-induced inhibition of CYP protein expression; however, it does not allow us to discern whether this inhibition arises from proteolytic degradation or enzyme inhibition. The failure to distinguish between these mechanisms could result in clinical complications, as drug-promoted proteolysis may lead to sustained reduction in enzyme activity even after drug clearance, owing to its dependence on protein synthesis for recovery. Thus, our data suggest the necessity of assessment of drug-induced CYP protein degradation for a more precise prediction and evaluation of DDI, a clinically important task which could help establish appropriate pharmacological treatment.
Limitation of our study: We used only a single concentration of omeprazole (50 µM) and examined CYP3A4 mRNA/protein expression at a few timepoints throughout the experiments. Detailed analysis of temporal changes in CYP3A4 mRNA and CYP3A4 protein abundance after treatment of different concentrations of omeprazole will be informative for further understanding the mode of omeprazole’s action in DDI.
This work was supported in part by research Grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan (22H04987 to M.D.; 22K15274 to T.M.), the Basis for Supporting Innovative Drug Discovery and Life Science Research program of the Japan Agency for Medical Research and Development (JP21am0101092), the Kusunoki 125 of Kyoto University 125th Anniversary Fund, the Kobayashi Foundation, Astellas Foundation for Research on Metabolic Disorders, and SRF (to M.D.) as well as the Takeda Science Foundation and the Kao Foundation for Arts and Sciences (to T.M.). We thank Yukari Doi (Kyoto University) for performing pilot experiments for this project.
T.M. and M.D. conceived the project and designed the research; Y.F. and T.M. performed experiments and analyzed the data in collaboration with X.S., Y.A. and E.H.; Y.F., T.M., and M.D. wrote the paper with input from all authors. M.D. supervised the entire project.
The authors declare no conflict of interest.