Abstract
Inflammatory processes are associated with compromised metabolism and elimination of drugs in the liver, largely mediated by proinflammatory cytokines, such as interleukin-6. The Hepa-RG cell line is an established surrogate for primary human hepatocytes (PHH) in drug metabolism and toxicity studies. However, the impact of inflammatory signaling on HepaRG cells has not been well characterized. In this study, the response of primary human hepatocytes and HepaRG cells to interleukin (IL)-6 was comparatively analyzed. For this purpose, broad-spectrum gene expression profiling, including acute-phase response genes and a large panel of drug-metabolizing enzyme and transporter (DMET) genes as well as their modifiers and regulators, was conducted in combination with cytochrome P450 (P450) activity measurements. Exposure of PHH and HepaRG cells to IL-6 resulted in highly similar coordinated reduction of DMET mRNA, including major ATP-binding cassette transporters (ABCs), P450s, glutathione S-transferases (GSTs), uridine diphosphate glucuronosyltransferases (UGTs), and solute carriers (SLCs). Enzyme activities of CYP1A2, CYP2B6, CYP2C8, CYP2C9, CYP2C19, and CYP3A4 were reduced upon 48–72 hours exposure to IL-6 in PHH and HepaRG. However, although these effects were not significant in PHH due to large interindividual donor variability, the impact on HepaRG was more pronounced and highly significant, thus emphasizing the advantage of HepaRG as a more reproducible model system. Exposure of HepaRG cells to interleukin-1β and tumor necrosis factor α resulted in similar effects on gene expression and enzyme activities. The present study emphasizes the role of proinflammatory cytokines in the regulation of drug metabolism and supports the use of HepaRG in lieu of PHH to minimize subject variability.
Introduction
It is well-known that proinflammatory cytokines such as interleukin (IL)-1β, IL-6, and tumor necrosis factor α (TNF-α) lead to compromised metabolism of drugs in the liver, primarily by affecting cytochrome P450 (P450) enzyme activities (Renton, 2005; Morgan et al., 2008; Dickmann et al., 2012). In particular, IL-6 was shown to regulate various P450s, including the important CYP3A4 (Jover et al., 2002; Aitken et al., 2006; Dickmann et al., 2011). Moreover, IL-6 downregulates the expression of several phase 2 drug-metabolizing enzymes (Shimada et al., 1999; Congiu et al., 2002; Richardson et al., 2006) and drug transporters (Sukhai et al., 2000; Siewert et al., 2004; Teng and Piquette-Miller, 2005; Yang et al., 2012). These changes largely contribute to complications in drug therapy, such as toxicity or decreased clearance, and alterations in physiologic function (Renton, 2005). The mechanisms that affect the regulation of drug-metabolizing enzymes and transporters (DMET) during the acute-phase response (APR) are still scarcely understood, although transcriptional suppression is considered to be the primary mechanism of regulation (Jover et al., 2002; Aitken et al., 2006). Moreover, most studies in this field have been conducted in rodent models (Aitken et al., 2006; Petrovic et al., 2007; Morgan et al., 2008). However, translation of results to humans is difficult due to species differences, particularly in the genomic response to inflammatory stimuli (Seok et al., 2013).
Liver tissue-derived primary human hepatocytes (PHH) are considered the “gold standard” model for the investigation of hepatic metabolism of drugs and other xenobiotics at the cellular level (Lecluyse and Alexandre, 2010; Godoy et al., 2013). However, PHH are restricted in availability and have a limited life span. Furthermore, PHH exhibit marked interindividual variations, including variability in expression and corresponding activities of many genes related to drug metabolism (Rogue et al., 2012). The HepaRG cell line has been proposed as an alternative model for PHH. These cells were isolated from a hepatocellular carcinoma of a female patient suffering from chronic hepatitis C infection (Gripon et al., 2002). HepaRG cells are bipotent progenitor cells and can differentiate into either biliary or hepatocyte lineages, representing the only example of complete differentiation in vitro (Cerec et al., 2007). As shown by genome-wide gene expression profiling studies, HepaRG cells are more similar to PHH and human liver tissue than any other liver cell line, particularly concerning the drug-processing genes (Hart et al., 2010; Rogue et al., 2012). In particular, HepaRG cells demonstrate stable expression of key phase 1 (e.g., P450s) and phase 2 (e.g., uridine diphosphate glucuronosyltransferases (UGTs) and glutathione S-transferases (GSTs)) enzymes, drug transporters (e.g., ATP-binding cassette transporters (ABCs) and solute carriers (SLCs)), and nuclear receptors (e.g., constitutive androstane receptor (CAR), pregnane X receptor (PXR), and peroxisome proliferator–activated receptors (PPARs)) (Aninat et al., 2006; Andersson et al., 2012). Moreover, major P450s were shown to be functionally expressed and selectively inhibited/induced by prototypical P450 inhibitors and inducers (Turpeinen et al., 2009). Therefore, HepaRG cells are increasingly being used as a surrogate in vitro model for drug metabolism and disposition studies (Andersson et al., 2012).
Although the influence of inflammatory mediators on drug detoxification functions has been studied in PHH and in HepaRG cells, the data are rather fragmentary, particularly for HepaRG. Studies investigating the effects of inflammatory signaling in PHH often focused on particular groups of drug-metabolizing enzymes (e.g., P450s) (Aitken et al., 2006) or drug transporters (Le Vee et al., 2008). In HepaRG cells, an induction of the APR was previously shown by increased C-reactive protein (CRP) expression upon lipopolysaccharide or IL-6 stimulation, simultaneously causing a suppression of CYP3A4 or CYP1A2 and CYP3A4 expression, respectively (Aninat et al., 2008; Bachour-El Azzi et al., 2014). Furthermore, it was shown that IL-1β leads to downregulation of organic anion transporter expression (Le Vee et al., 2008). However, the present data do not allow systematic comparisons of the effects of inflammatory signaling neither with respect to the differences between PHH and HepaRG, nor regarding quantitative changes for a wider selection of relevant genes. In this study, we used IL-6 as a prototypical cytokine to induce the APR in PHH and HepaRG. By simultaneously quantitating a large number of APR and DMET genes, we show that the IL-6–induced gene expression changes are highly similar in PHH and HepaRG, both quantitatively and qualitatively. Furthermore, the enzyme activities of major P450 enzymes were shown to be comparable between the two models. Additionally, the impact of the cytokines IL-1β and TNF-α was investigated in HepaRG cells. Our results confirm a coordinated response of major DMETs to inflammatory signaling and extend the current knowledge toward less well investigated genes. HepaRG cells broadly mirrored the results observed in PHH and appear to be advantageous as a more reproducible model system.
Materials and Methods
Reagents.
William’s E Medium was purchased from Invitrogen Life Technologies (Darmstadt, Germany). Fetal bovine serum (FBS) was from PAA Laboratories (Pasching, Austria), human insulin from Sanofi (Frankfurt, Germany), and hydrocortisone from Pfizer Pharma (Karlsruhe, Germany). HEPES, l-glutamine, minimum essential medium nonessential amino acids, penicillin/streptomycin (Pen/Strep), phosphate-buffered saline, and sodium pyruvate were purchased from GIBCO (Carlsbad, CA). Bovine serum albumin, dexamethasone, and dimethylsulfoxide (DMSO) were from Sigma-Aldrich (Steinheim, Germany). Human recombinant IL-6 was purchased from Promo Cell (Heidelberg, Germany). Human recombinant IL-1β and TNF-α were purchased from Sigma-Aldrich. All cytokines were reconstituted and stored as high concentration stocks, according to manufacturer specifications. All TaqMan assays were purchased from Applied Biosystems (Foster City, CA).
Human Hepatocyte Cultures.
With written informed consent from donors (11 females, 3 males) and approvals of the local ethics committees in Berlin, Munich, Regensburg, and Tuebingen, PHH were isolated from partial liver resections by collagenase digest, as described previously (Godoy et al., 2013; Lee et al., 2013). Donor data are shown in Table 1. Isolated cells with a viability of more than 70% as determined via trypan blue exclusion test were seeded at a density of 4 × 105 viable cells/well onto BioCoat Collagen I Cellware 12-well culture plates (BD Biosciences, Bedford, MA) in William’s E Medium, supplemented with 10% FBS, 100 U/ml Pen/Strep, 2 mM l-glutamine, 32 mU/ml human insulin, 1 mM sodium pyruvate, 1× nonessential amino acids, 15 mM HEPES, and 0.8 µg/ml hydrocortisone. After 24 hours, cells were equilibrated for another 24 hours in cultivation medium, containing William’s E Medium, supplemented with 10% FBS, 100 U/ml Pen/Strep, 2 mM l-glutamine, 32 mU/ml human insulin, 0.1% DMSO, and 0.1 µM dexamethasone. Cells were maintained at 37°C in 5% CO2 throughout the experiment with the exception of the shipping period. Media were changed every 24 hours. Concentrations of IL-6 used for the experiments ranged from 0.1 pg/ml to 50 ng/ml. Treatments were repeated every 24 hours.
HepaRG Cultures.
HepaRG cells (batch HPR101007) were obtained from Biopredic International (Rennes, France) and expanded according to the provider’s instructions to set up a working bank. Working bank cell vials, containing 1.5 million cells, were thawed and cultured in HepaRG growth medium in 25-cm2 (T-25) tissue culture flasks (Sarstedt, Newton, NC) for 14 days. HepaRG growth medium was based on 500 ml William’s E Medium, supplemented with 10% FBS, 100 U/ml Pen/Strep, 2 mM l-glutamine, 32 mU/ml human insulin, and 20 µg/ml hydrocortisone. Medium was exchanged every 2 or 3 days. Cells were passaged and transferred to MULTIWELL 24-well plates (BD Biosciences) in a density of 50,000 cells/well and cultivated for 2 more weeks. Medium was replaced by HepaRG growth medium containing 1% DMSO for 2 days. Starting from the third day, cells were cultivated in HepaRG growth medium containing 2% DMSO (HepaRG differentiation medium) for another 12 days. At that stage, HepaRG cells reached a differentiated hepatocyte-like morphology and showed liver-specific functions. The cells were further maintained in HepaRG differentiation medium for the duration of the experiments with exchange of medium every 2 or 3 days. Cells were maintained at 37°C in 5% CO2 throughout the experiment. Concentrations of IL-6, IL-1β, and TNF-α used for the experiments ranged from 0.1 pg/ml to 50 ng/ml. Treatments were repeated every 24 hours.
Quantitative Real-Time Polymerase Chain Reaction.
Total RNA was isolated from PHH and HepaRG cells using the RNeasy Mini Kit, including on-column genomic DNA digestion with RNase-free DNase Set (Qiagen, Hilden, Germany). The RNA integrity and quantity were analyzed with the Agilent 2100 Bioanalyzer using the RNA 6000 Nano Kit (Agilent Technologies, Waldbronn, Germany). Synthesis of cDNA was performed with 500 ng RNA using Taqman Reverse Transcription Reagents (Applera, Darmstadt, Germany). Quantification of expression of 86 genes was performed using Fluidigm’s BioMark HD high-throughput quantitative chip platform (Fluidigm, San Francisco, CA), following the manufacturer’s instructions (Spurgeon et al., 2008). All corresponding TaqMan assays are listed in Supplemental Table 3. The mRNA expression levels were normalized to the glyceraldehyde-3-phosphate dehydrogenase mRNA expression. Relative gene expression changes were calculated using the delta delta Ct (ΔΔCt) method (Livak and Schmittgen, 2001).
Measurement of P450 Enzyme Activities.
Cytochrome P450 enzyme activities were determined in PHH and HepaRG cell culture supernatants using a liquid chromatography with tandem mass spectrometry–based substrate cocktail assay, as described previously (Feidt et al., 2010). CYP2D6 enzyme activity (propafenone 5-hydroxylation) was not quantifiable in HepaRG cells by liquid chromatography–tandem mass spectrometry cocktail assay and was therefore excluded from further analysis.
Statistical Analyses.
For demonstration of gene expressin changes, the mean fold changes as obtained from the ΔΔCT method and their S.D. are shown in bar graphs. Due to the considerably skewed symmetry of up- and downregulation in the linear fold change, all statistical analyses were carried out using the ΔCt values only. Statistical significance was determined by comparing the t ratio with the t distribution for the number of degrees of freedom calculated with a two-way analysis of variance and applying the Bonferroni correction for multiple comparisons, using GraphPad Prism 5.0.4 software (GraphPad Software, La Jolla, CA). For demonstration of changes in P450 enzyme activities, the mean activities are shown as pmol/min per 106 cells in bar graphs. Statistical significance was determined by log2 transformation of data and subsequent Bonferroni post hoc test as described above, using GraphPad Prism 5.0.4 software (GraphPad Software).
Results
Impact of IL-6 on PHH Gene Expression.
Our broad-spectrum gene expression panel included 86 genes coding for major acute-phase (AP) proteins, important DMETs, as well as their regulators and modifiers. PHH from one individual donor were exposed to IL-6 for 24 hours in a concentration range of 0.1 pg/ml–50 ng/ml. IL-6 caused a marked increase in mRNA expression of acute-phase proteins (e.g., CRP and SAA1/SAA2) with maximum fold increase at concentrations of 10 ng/ml and higher (Fig. 1A). Within the same concentration range, the mRNA expression of several major DMETs declined to a maximum fold decrease (Fig. 1, B–I). Thus, to comparatively investigate large-scale gene expression changes in PHH, a concentration of 10 ng/ml IL-6 for maximum effectiveness was applied. Changes in mRNA expression upon IL-6 stimulation were investigated after 8 and 24 hours in hepatocytes from 8 and 14 individual donors, respectively. The suppressor of cytokine signaling-coding SOCS3, a negative feedback regulator of the Janus kinase/signal transducer and activator of transcription signaling pathway, was >sixfold induced 8 hours after the IL-6 stimulus and thus confirmed the activation of the IL-6 receptor complex. Activation of the systemic inflammatory response was reflected by almost 100-fold induced CRP mRNA expression after 8 hours, increasing up to approximately 200-fold after 24 hours (Fig. 2A). These effects were statistically highly significant. Among the drug metabolism genes, a coordinated downregulation of P450 isoforms in PHH in response to IL-6 stimulation was largely confirmed (Fig. 2B). Except for CYP2E1 (see below), P450s of major importance in drug metabolism were downregulated by at least 40%. On average, expression of CYP1A2 was decreased by >75%, CYP3A4 by >80%, CYP2C9 by >60%, and CYP2D6 by ∼50% after 24 hours of IL-6 stimulation. CYP7A1 was significantly downregulated by >80% as early as 8 hours after the IL-6 challenge. Interestingly, CYP2E1 was the only phase 1 metabolism gene that was significantly upregulated, showing nearly twofold induction at 24 hours. Of note, among the tested phase 1 metabolism genes, the mRNA expression of ADH1A, ALDH2, and DPYD did not display any significant effects.
The effects on phase 2 metabolism genes were rather diverse (Fig. 2C). NAT1 and NAT2 were significantly downregulated by approximately 50% each as early as 8 hours after the IL-6 stimulus. A significant downregulation of GSTA2, GSTM1, NAT2, and UGT2B7 was observed after 24 hours. Although highly variable, the expression of SULT1B1 was significantly induced (>twofold) at 24 hours. All major ABC and SLC drug transporter genes tested were also significantly downregulated by more than 40% after challenging the cells with IL-6 for 24 hours (Fig. 2E). Remarkably, ABCC2 (MRP2) was downregulated by >50% and SLC10A1 (NTCP) by >75%. Among the major DMET modifiers, AHR, NR1I2 (PXR), and NR1I3 (CAR) were identified as significantly downregulated by approximately 50%, whereas HNF4A, PPARG, and RXRA appeared to be less influenced. In contrast, SOD2 was the only gene in this panel that was significantly upregulated (Fig. 2D). Taken together, these data demonstrate that the most significant impact of IL-6 was found for genes that code for P450s as well as ABC and SLC drug transporters. Almost all of their major isoforms were transcriptionally downregulated. The average phase 2 metabolism gene expression, however, appeared to be not as strongly affected by IL-6.
Impact of IL-6 on HepaRG Gene Expression.
HepaRG cells were exposed to IL-6 for 24 hours in a concentration range of 0.1 pg/ml–50 ng/ml. Similar to the dose-response curves shown for PHH (Fig. 1), IL-6 concentrations of 10 ng/ml and higher showed maximum effectiveness toward induction and repression of mRNA expression of AP proteins and most DMETs, respectively (Fig. 3). Notably, a somewhat reduced sensitivity of HepaRG cells compared with PHH toward IL-6 was noted for concentrations between 0.1 pg/ml and 100 pg/ml, as indicated in a shift of the dose-response curves toward higher concentrations. However, this shift was generally less than one decimal. Using our large gene expression panel, we then determined the changes in mRNA expression at two time points (8 and 24 hours) upon stimulation with 10 ng/ml IL-6. The impact of IL-6 on major AP marker genes in HepaRG was significant and highly similar to PHH (Fig. 4A). Statistically highly significant transcriptional downregulation by at least 60% was observed for almost all P450 isoforms (Fig. 4B). The magnitudes of regulation were comparable to those observed in IL-6–stimulated PHH. In contrast to PHH, the expression of CYP2E1 was marginally—but not significantly—downregulated at 24 hours. The expression of the CYP2D6 gene was not quantifiable in HepaRG cells, confirming previous findings (Kanebratt and Andersson, 2008). The IL-6–induced effects on expression of phase 2 metabolism genes were less pronounced than on phase 1 metabolism genes, comparable to the findings in PHH (Fig. 4C). In contrast to PHH, however, UGT1A1 was significantly downregulated by >50% upon IL-6 stimulation for 24 hours. Furthermore, SULT1B1 mRNA expression was slightly decreased, but still highly variable. GSTP1 could not be detected in HepaRG cells, as previously shown in the literature (Rogue et al., 2012). Also, in contrast to findings in PHH, the mRNA expression of ABCB1 and ABCC2 was not significantly downregulated, while SLC22A7 was much stronger downregulated (Fig. 4E). Interestingly, there was no detectable response of the SLC genes to a short IL-6 stimulation (8 hours). Among the major DMET modifiers, NR1I2 (PXR) and NR1I3 (CAR) were identified as significantly downregulated on the transcriptional level by >50%, in agreement with the findings in PHH (Fig. 4D). Notably, the mRNA expression of PPARA was significantly decreased after IL-6 stimulation for 24 hours, which was not observed in PHH. Overall, HepaRG showed a coordinated downregulation of many important DMET genes upon IL-6 stimulation, which was quantitatively and qualitatively similar to the effects observed in PHH. This was further demonstrated by correlation analysis (Fig. 5). Only a few DMET genes (e.g., CYP2E1, UGT1A1, ABCB1, and ABCC2) appeared to be differentially expressed. The detailed expression data of all tested genes are presented in Supplemental Table 1.
Impact of IL-6 on Cytochrome P450 Activities in PHH and HepaRG Cells.
IL-6 stimulation in PHH was shown to cause many transcriptional changes of DMET genes, in particular of P450 isoforms. However, lacking correlations between gene expression and protein level or activity are frequently observed. For instance, factors such as the half-life of proteins play an important role. Hence, it was investigated whether IL-6 also had an impact on the activities of major P450s. Metabolite formation rates of CYP1A2, 2B6, 2C19, 2C8, 2C9, and 3A4 were determined in PHH from three individual donors that were exposed to IL-6 for up to 72 hours (Fig. 6). The formation rates of acetaminophen (Fig. 6A), OH-bupropion (Fig. 6B), N-DE-amodiaquine (Fig. 6D), and o-OH-atorvastatin (Fig. 6F) were most strongly affected by IL-6 treatment; however, none of the effects was statistically significant, most likely due to the pronounced interindividual variability between the three donors. Moreover, basal activities of most P450s declined rapidly within the first 24 hours of culture time (Supplemental Fig. 1).
The enzyme activities of P450s 1A2, 2B6, 2C19, 2C8, 2C9, and 3A4 could be determined in IL-6–challenged HepaRG cells, being very stable in the control-treated cells over a time span of at least 72 hours. Fig. 7, summarizes the treatment effects, as determined in four independent experiments. After 48 hours of IL-6 treatment, the activities of all examined P450s were statistically significantly reduced by at least 60%. After 72 hours, the formation rates of, for example, o-OH-atorvastatin (Fig. 7F), OH-bupropion (Fig. 7B), and N-DE-amodiaquine (Fig. 7D) were decreased by >80%. Taken together, the results to date demonstrated highly comparable cellular responses of PHH and HepaRG cells. Transcriptional downregulation of the analyzed P450s resulted in decreased enzyme activities. The lack of interindividual variability in the HepaRG cells appeared to facilitate experimental reproducibility.
Effects of IL-1β and TNF-α in HepaRG Cells.
To obtain additional data for the HepaRG model, we exposed the cells to IL-1β and TNF-α. Dose-response experiments demonstrated that 5 ng/ml IL-1β are sufficient to induce maximum mRNA expression fold changes of the genes of interest, whereas a concentration of 10 ng/ml TNF-α was necessary to induce maximum effects on gene expression (Fig. 3). Supplemental Table 1 summarizes the mRNA expression changes of all tested genes. The mean expression changes of AP and DMET genes in IL-1β– and TNF-α–challenged HepaRG cells, as determined in three independent experiments, are illustrated in a heat map, including also IL-6–mediated changes in PHH and HepaRG cells (Fig. 8). As observed with IL-6, AP markers were highly increased upon stimulation with IL-1β or TNF-α. Fewer phase 1 metabolism genes were downregulated 8 hours after both stimulations, compared with IL-6. In particular, the effects elicited by TNF-α were not as pronounced at this earlier time point. After 24 hours, impairment of mRNA expression of almost all phase 1/2 metabolism, transporter, and modifier genes was similar to IL-6, in particular in IL-1β–challenged HepaRG cells. Notably, dose-response curves showed much higher sensitivity of cells toward IL-1β compared with IL-6 and TNF-α (Fig. 3). In summary, exposure of HepaRG cells to IL-1β and TNF-α for 24 hours caused a coordinated downregulation of important DMET genes, in particular of P450 isoforms. However, the sensitivity and the potency of these two inflammatory mediators as well as the extent of suppression varied among the examined genes. The activities of CYP1A2, 2B6, 2C19, 2C8, 2C9, and 3A4 were also determined in HepaRG exposed to IL-1β and TNF-α for up to 72 hours. Supplemental Table 2 summarizes the metabolite formation rates of all six tested P450s, as determined in two independent experiments. All formation rates were reduced as early as 48 hours after exposure of cells to IL-1β and TNF-α. After 72 hours, both treatments caused suppressions of all examined P450 activities by more than 80%. Thus, both IL-1β and TNF-α demonstrated similar repressive properties toward P450 activities.
Discussion
This work was aimed at evaluating the robustness and the suitability of the highly differentiated human HepaRG cell line for the investigation of the impact of inflammatory signaling on drug metabolism gene expression and drug detoxification capacity. We compared first relative expression changes of over 80 relevant APR marker genes and major drug metabolism genes in IL-6–challenged PHH and HepaRG cells. Dose-response curves were obtained in PHH and HepaRG cells for dose finding and to assure that our experimental conditions were comparable to those of previous studies. An IL-6 concentration of 10 ng/ml was determined to be sufficient in both PHH and HepaRG cells to induce maximum effects on gene expression, in agreement with previous studies using a similar concentration of this particular cytokine for comparative analyses (Aitken and Morgan, 2007; Dickmann et al., 2011). Although being very low (∼pg) in healthy individuals, physiologic IL-6 concentrations well above 10 ng/ml serum were previously reported in patients with severe illness (Rose-John, 2012). Because the inflammatory response is of local character, considerably higher cytokine concentration may be possible in the liver (Jones et al., 2011; Rose-John, 2012). In this study, the APR was strongly induced by 10 ng/ml IL-6 in both cell models, PHH and HepaRG cells, with a rapid increase in SOCS3 expression, a negative feedback inhibitor of Janus kinase/signal transducer and activator of transcription signaling, as well as a several hundred-fold increase in CRP mRNA expression, confirming efficient activation of the APR. Among the DMET genes studied, coordinated downregulation of almost all major P450 genes occurred in both cell types, which has not been shown before by such a systematic approach. The extent of P450 downregulation was generally stronger after 24 hours and reached over 80% for some P450s, including CYP3A4 and CYP7A1, in both cell models. Although the downregulating effects of proinflammatory cytokines on CYP3A4 have been repeatedly reported (Jover et al., 2002; Aitken and Morgan, 2007), CYP7A1 has been less well studied in this regard. Downregulation of CYP7A1 may be explained by repression of farnesoid X receptor (Kim et al., 2003).
Interestingly, CYP2E1 expression was almost twofold induced at 24-hour IL-6 treatment in PHH, whereas in HepaRG the observed minor decrease was not significant. Induced expression of CYP2E1 occurs in a range of inflammatory pathologic conditions, including nonalcoholic steatohepatitis and related diseases, in which it is discussed to contribute to oxidative stress, lipid peroxidation, and hepatocyte injury (Aubert et al., 2011). However, it should be noted that proinflammatory cytokines in rodent models usually cause downregulation of CYP2E1, emphasizing species-dependent differences (Siewert et al., 2000; Hakkola et al., 2003). The lack of inducibility of CYP2E1 in HepaRG cells is probably related to the generally low expression and may also depend on the culture conditions (Kanebratt and Andersson, 2008; Le Hegarat et al., 2010). Among further phase 1 genes, we found that mRNA expression of ADH1A, ALDH2, and DPYD was only moderately affected in both cell systems.
Regarding P450 enzyme activities, the six P450s, 1A2, 2B6, 2C8, 2C9, 2C19, and 3A4, were all reduced in the PHH cells, although none of these results achieved statistical significance. This was mainly due to the enormous variability between PHH donors. We observed a decline in basal P450 activities during the first day of culture period, which was previously reported in primary hepatocytes (Feidt et al., 2010; Lübberstedt et al., 2011). In the HepaRG cells, much better preservation of the six measured P450 activities over the investigated culture period of 72 hours was observed, in agreement with previous studies (Kanebratt and Andersson, 2008; Turpeinen et al., 2009; Lübberstedt et al., 2011). A limitation of HepaRG is the very low expression of CYP2D6, which was not quantifiable by either quantitative polymerase chain reaction or our liquid chromatography–tandem mass spectrometry cocktail, and was thus excluded from further analysis, as previously noted (Kanebratt and Andersson, 2008). IL-6 treatment led to consistent and highly significant decreases in all six enzyme activities in HepaRG cells. These data emphasize the lack of interindividual variability as a distinct advantage of the HepaRG cell model.
Less attention has been credited to the effects of proinflammatory cytokines on phase 2 drug metabolism. Whereas major GSTs and UGTs were shown to be expressed at similar levels in PHH and HepaRG (Aninat et al., 2006; Kanebratt and Andersson, 2008), the impact of cytokines has not been addressed to date in HepaRG cells to our knowledge. The expression changes in HepaRG cells exposed to IL-6 were qualitatively similar to those observed in PHH. Dose-response experiments indicated a differential quantitative response of the two cell models. Hence, for major DMET genes, HepaRG cells demonstrated a delayed response toward lower concentrations of IL-6. The morphologic heterogeneity of HepaRG cells, displaying both hepatocyte- and biliary-like epithelial phenotypes (Parent et al., 2004), may be one reason for this observation. The mRNA expression of GSTs, NATs (N-acetyltransferases), and UGTs was comparably suppressed, whereas TPMT was not significantly affected in both systems. Measurement of SULT1B1 mRNA levels proved to be highly variable in both PHH and HepaRG, but appeared to be rather strongly (∼twofold) upregulated by IL-6 in PHH, although slightly downregulated in HepaRG. The major drug transporters ABCG2 (BCRP), SLC10A1 (NTCP), SLC22A7, and SLCO1B1 (OATP2) were consistently transcriptionally repressed by IL-6 in PHH and HepaRG. ABCB1 (MDR1) and ABCC2 (MRP2) were less affected in HepaRG compared with PHH. We also determined gene expression changes for a panel of gene regulators. Principal functionality of major nuclear receptors in HepaRG cells is supported by similar inducibility of P450 gene expression with prototypical inducers compared with PHH (Aninat et al., 2006; Turpeinen et al., 2009). IL-6 stimulation of HepaRG cells profoundly and significantly suppressed mRNA expression of CAR and PXR in both cell models. Considering the magnitude and duration of the effect, it appears likely that the downregulation of these two important xenosensors contributes at least in part to the downregulation of P450s and other DMET genes during inflammation. By contrast, the Ah receptor was significantly downregulated by IL-6 in PHH, whereas the slight decreases observed in HepaRG were not significant. Interestingly, mRNA levels of HNF4A (HNF4-α) were affected in both cell types only at the earlier 8-hour time point, but appeared to be normalized at 24 hours. A participation of HNF4-α in IL-6–induced gene expression changes may thus be less likely. Taken together, these data demonstrated highly comparable effects of IL-6 on gene expression in PHH and HepaRG, suggesting HepaRG as a valid model to investigate inflammatory responses. The reader is referred to a recent publication (Rubin et al., 2015) that appeared during revision of our paper that also described similar suppression by IL-6 of CYP1A2, CYP2B6, and CYP3A4 mRNA levels and catalytic activities in cryopreserved human hepatocytes and in HepaRG cells.
The last part of this study was devoted to the analysis of the effects of two further important proinflammatory cytokines, IL-1β and TNF-α, in HepaRG cells, to extend the available systematic data. The maximum efficient concentration, as determined by dose-response experiments, was 5 and 10 ng/ml for IL-1β and TNF-α, respectively. These concentrations are in agreement with previous reports in PHH (Aitken and Morgan, 2007). Both treatments led to remarkably similar and drastic decreases in the transcripts of all major P450s, and measurement of enzyme activities confirmed the negative influence for all six measured P450 isoenzymes. Regarding phase 2 enzymes, the influence of IL-1β was somewhat more potent compared with IL-6, whereas TNF-α appeared to be less potent, although not for all genes. IL-1β also efficiently suppressed mRNA expression of all investigated transporters, confirming previous findings (Le Vee et al., 2008). The effects of TNF-α on drug transporters differed with respect to the investigated ABC transporters, which proved to be refractory toward this cytokine. A slight upregulation of MDR1 (ABCB1), as observed previously (Poller et al., 2010), was not significant. Transcriptional downregulation of the hepatic bile acid transporter NTCP (SLC10A1) is in agreement with a former study (Cherrington et al., 2013). Somewhat more extended effects of IL-1β and TNF-α were also observed for HNF4A, PPARA, and POR.
In conclusion, we report striking similarities in DMET gene expression changes upon stimulation with the proinflammatory cytokine IL-6 in HepaRG cells compared with PHH. These data emphasize the coordinated response of hepatocytes to inflammatory conditions, which appears to result in an almost complete shutdown of the drug metabolism system. Patients treated with any drug that has a narrow therapeutic index are thus at considerable risk to develop adverse drug reactions. The physiologic reason of this dramatic reorganization of gene expression might be to ensure availability of sufficient hepatic resources for the APR.
Acknowledgments
The authors thank Stefan Winter for statistical advice, and Igor Liebermann, Britta Klumpp, and Sonja Seefried for excellent technical assistance. The authors also thank Biopredic International, and especially Christophe Chesné, for providing HepaRG cells and technical support. Furthermore, Wolfgang Thasler (Munich) and Thomas Weiss (Regensburg) are gratefully acknowledged for contribution of human hepatocytes. Finally, the authors are indebted to the charitable foundation Human Tissue and Cell Research, Regensburg, for making human tissue available for research.
Authorship Contributions
Participated in research design: Klein, Thomas, Zanger.
Conducted experiments: Klein.
Contributed new reagents or analytic tools: Seehofer, Damm.
Performed data analysis: Klein, Hofmann.
Wrote or contributed to the writing of the manuscript: Klein, Zanger, Thomas, Damm.
Footnotes
- Received September 8, 2014.
- Accepted December 5, 2014.
This work was supported by the German Federal Ministry of Education and Research [Grants 0315755 and 0315741] and by the Robert Bosch Foundation, Stuttgart, Germany.
↵This article has supplemental material available at dmd.aspetjournals.org.
Abbreviations
- ABC
- ATP-binding cassette
- AP
- acute phase
- APR
- acute-phase response
- CAR
- constitutive androstane receptor
- CRP
- C-reactive protein
- DMET
- drug-metabolizing enzyme and transporter
- DMSO
- dimethylsulfoxide
- FBS
- fetal bovine serum
- GST
- glutathione S-transferase
- IL
- interleukin
- P450
- cytochrome P450
- Pen/Strep
- penicillin/streptomycin
- PHH
- primary human hepatocyte
- PPAR
- peroxisome proliferator–activated receptor
- PXR
- pregnane X receptor
- SLC
- solute carrier
- TNF
- tumor necrosis factor
- UGT
- uridine diphosphate glucuronosyltransferase
- Copyright © 2014 by The American Society for Pharmacology and Experimental Therapeutics