Cytochrome P450 Enzyme Levels in Hepg2 Cells and Cryopreserved Primary Human Hepatocytes and Their Induction in Hepg2 Cells

Available online at www.sciencedirect.com

Toxicology in Vitro 21 (2007) 1581–1591 www.elsevier.com/locate/toxinvit

Cytochrome P450 enzyme levels in HepG2 cells and cryopreserved primary human hepatocytes and their induction in HepG2 cells

Walter M.A. Westerink *, Willem G.E.J. Schoonen

Department of Pharmacology, NV Organon, Molenstraat 110, 5340 BH Oss, The Netherlands

Received 12 January 2007; accepted 29 May 2007 Available online 8 June 2007

Abstract

Early in vitro toxicity screening might improve the success rate of new chemical entities in pharmaceutical development. In previous studies, the advantage of cytotoxicity screening with the HepG2 cell line was shown. Cytotoxicity could be identiﬁed for 70% of the compounds in these assays as compared with known toxicity in either in vitro assays in primary hepatocytes, in in vivo assays in rats, or in (pre-)clinical development in humans. The low Phase I and II enzyme levels in HepG2 cells might have been responsible for the fact that 30% of the compounds scored negative. Therefore, we performed two follow-up studies in which Cytochrome P450 (CYP) enzymes and Phase II metabolism were examined. In the present study, the transcript levels of CYP1A1, 1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, and 3A4 were measured with quantitative PCR. Results showed that transcripts of all CYPs were present in HepG2 cells, however, mRNA levels of most CYPs were dra- matically lower than in primary human hepatocytes. These results were conﬁrmed with luminometric assays which were used to measure the enzyme activities of CYP1A1, 1A2, 2C9, and 3A4. Regulation of CYP1A1, 1A2, 2B6, 2C8, 2D6, 2E1, and 3A4 by the aryl hydrocarbon receptor, pregnane X receptor and constitutive androstane receptor was studied in HepG2 cells at the mRNA and/or enzyme level. Regulation of CYP1A1, 1A2, 2B6, and 3A4 mRNA levels was similar to the regulation in primary human hepatocytes. In contrast, CYP2C8 mRNA levels are inducible in primary human hepatocytes, but not in HepG2 cells, after treatment with PXR/CAR activators. Consistent with other studies, CYP2D6 and 2E1 transcript levels were not changed after treatment with AhR, PXR, and CAR activators. Moreover, CYP1A1 and 1A2 enzyme levels could be induced by AhR agonists and CYP3A4 by PXR agonists. As a consequence of the low levels of CYPs in HepG2 cells, cytotoxicity of several compounds might have been missed or underes- timated as compared with cytotoxicity in primary human hepatocytes. Inducing HepG2 cells with particular receptor stimulators might lead to higher toxicity for several of the tested compounds. Compared to primary human hepatocytes, HepG2 cells are a relatively easy- to-handle tool to study the up-regulation of CYP1A1, 1A2, 2B6, and 3A4. 2007 Elsevier Ltd. All rights reserved.

Keywords: HepG2; Hepatocytes; Phase I metabolism; Cytochrome P450; Induction studies; Cytotoxicity

1. Introduction tion in an early phase of development of drugs may improve the success rate of new chemical entities (Caldwell Approximately, 40% of the new drug candidates fail in et al., 2001). However, this strategy implies a large number the developmental phase due to toxicological side eﬀects of compounds for which only a small amount of material is (Kola and Landis, 2004). Screening on toxicity and deselec- available. Therefore medium or high throughput screening methods are necessary. * Corresponding author. In two previous studies (Schoonen et al., 2005a,b) the E-mail address: [email protected] (W.M.A. Westerink). advantage of early toxicity medium throughput screening

0887-2333/$ - see front matter 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.tiv.2007.05.014 1582 W.M.A. Westerink, W.G.E.J. Schoonen / Toxicology in Vitro 21 (2007) 1581–1591 was shown with seven different fluorometric assays on four tative PCR was used to measure mRNA levels of CYP1A1, different cell lines. In these studies cytotoxicity could be 1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, and 3A4. shown for 70% of the compounds with known toxicity in Enzyme activities of CYP1A1, 1A2, 2C9, and 3A4 were either in vitro assays in primary hepatocytes, in in vivo measured using luminometric P450-Glo substrates (Cali assays in rats, or in (pre-)clinical development in humans. et al., 2006). Furthermore, the inducibility of CYPs in Toxicity of the remaining 30% of compounds could not HepG2 cells was studied after exposure of HepG2 cells to be established. The human hepatoma cell line HepG2 was agonists of the xenobiotic receptors AhR, CAR and chosen for further toxicity screening as detoxication and PXR. Indigo, indirubin, b-naphthoflavone (BNF), 3-meth- activation processes of compounds are studied most opti- ylcholanthrene (3MC), and 2,3,7,8-tetrachlorodibenzo-p- mally in liver cells. However, a discrepancy exists between dioxin (TCDD) were used as AhR agonists (Adachi Cytochrome P450 (CYP) and Phase II metabolism of pri- et al., 2001; Runge et al., 2000; Yueh et al., 2005). Tularik mary hepatocytes and HepG2 cells. Low CYP and Phase T0901317, CITCO, rifampicin (RIF), and phenobarbital II enzyme levels in HepG2 cells might have been responsible (PB) were used as PXR and/or CAR agonists. T0901317 for the fact that 30% of the compounds were falsely classi- is in principle a potent LXR ligand that is also capable fied as non-toxic (Hewitt and Hewitt, 2004; Rodriguez- of activating human PXR (Shenoy et al., 2004). CITCO Antona et al., 2002; Wilkening et al., 2003). However, a and RIF are potent CAR and PXR activators, respectively recent study with Cellulomics techniques on HepG2 cells (Maglich et al., 2003). PB is an activator of the PXR and counteracts the hypothesis that the metabolic competence although it does not appear to bind directly to CAR, it is of HepG2 cells is significantly limiting the production of also an activator of CAR (Chen et al., 2004). reactive metabolites (O’Brien et al., 2006). With this novel A parallel study deals with Phase II metabolism and com- method, cytotoxicity in HepG2 cells could be established pares the mRNA levels and enzyme activities of 10 Phase II for more than 90% of 243 drugs with varying degrees of tox- enzymes in HepG2 cells with those of cryopreserved icity including drugs that produce their toxicities by a reac- primary human hepatocytes (Westerink and Schoonen, tive metabolite. Nevertheless we performed two follow-up accepted for publication). Moreover, induction of Phase II studies in which CYP enzymes and Phase II metabolism enzymes by AhR, PXR and CAR activators was studied. were examined. In the present study the focus is on CYPs. The role of Phase I enzymes is to introduce a new func- 2. Materials and methods tional group (e.g., hydroxylation) or modify an existing functional group (e.g., O-dealkylation) so as to facilitate 2.1. Materials Phase II conjugation reactions. Whilst Phase I reactions generally result in a more polar metabolite, it is the conju- All compounds and reagents were of analytical grade. gation reactions (e.g., glucuronidation, sulphonation) that Indirubin was obtained from BIOMOL International (Exe- result in marked increases in water solubility and facilitate ter, UK), 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) excretion. from Promochem (Wesel, Germany), Tularik T0901317 This transformation in two steps facilitates excretion and from NV Organon (Oss, The Netherlands), and P450-Glo detoxicates a wide variety of xenobiotics. For example the substrates for CYP1A1, 1A2, 2C9, and 3A4 from Promega Phase II enzyme glutathione S-transferase P1-1 prevents (Madison, USA). All other compounds were obtained from 4-nitroquinoline 1-oxide (4-NQO) DNA-adduct formation Sigma–Aldrich (St. Louis, MO, USA). by the formation of 4-NQO glutathione conjugates (Mor- row et al., 1998; Townsend et al., 2002). However, many 2.2. Cell culture xenobiotics are also toxicated by CYPs and Phase II enzymes (Rooney et al., 2004). A well known example is HepG2 cells were obtained from the American Type the carcinogenicity of benzo[a]pyrene in which the activities Culture Collection (Rockville, MD, USA). Cells were cul- of CYP1A1 and epoxide hydrolase lead to the formation of tured in Dulbecco’s Modified Eagles medium and Nutrient the carcinogenic metabolite benzo[a]pyrene-7,8-diol-9,10- mixture F-12 (in vitro technologies inc., Baltimore, USA) epoxide (Bao et al., 2002; Huang et al., 2005). mixed in a ratio of 1:1 with 10% defined supplemented The CYP and Phase II enzymes are mainly regulated by bovine calf serum (dBCS) from Hyclone (Utah, USA) ligand-activated nuclear transcription factors, such as the and 1% Penicillin–Streptomycin (10,000 U/ml, Gibco). aryl hydrocarbon receptor (AhR; for the CYP1A family), Cultures were maintained in a humidified atmosphere with the constitutive androstane receptor (CAR; for the CYP2B 5% CO2 at 37 C and medium was refreshed every three or family) and the pregnane X receptor (PXR; for the CYP3A four days with subculturing. All experiments in HepG2 family) (Bock and Kohle, 2004; Maglich et al., 2002; Man- cells were performed between passage 8 and 16. During this kowski and Ekins, 2003; Xie et al., 2003). period, no significant change in the CYP expression was In the present study we characterized CYP metabolism observed. To compare the mRNA expression levels in in HepG2 cells. Transcript levels and enzyme activities in HepG2 with primary hepatocytes, cells from passage 8 the HepG2 cell line were compared with levels and activi- and 9 were used. For induction studies in HepG2, cells ties in cryopreserved primary human hepatocytes. Quanti- from passage 12, 14, and 16 were used. W.M.A. Westerink, W.G.E.J. Schoonen / Toxicology in Vitro 21 (2007) 1581–1591 1583

2.3. Cryopreserved primary hepatocytes 10 ml culture medium. Next cells were collected by centri- fugation for 5 min at 272 g. The supernatant was removed Vials containing 5 · 106 cryopreserved primary human and total cellular RNA was isolated with Trizol reagent. hepatocytes were purchased from In Vitro Technologies. For cDNA synthesis 2.5 lg total RNA from HepG2 After thawing the viability of the hepatocytes was assessed cells or primary hepatocytes was used and 0.5 lg random by using trypan blue exclusion. The viability of all vials hexamer primer (GE Healthcare Bio-Sciences Corp, Piscat- used was >80%. Measured enzyme activities were corrected away, USA) was added. The mixture was heated at 70 C for viability. Hepatocytes from two donors were used: EG, for 10 min and thereafter quickly chilled on ice for 2 min. 68 years, white female (did not use tobacco, alcohol or cDNA was synthesized in a total volume of 25 ll contain- other substances; medical history of hypertension); HRU, ing 50 mM Tris–HCl (pH 8.3), 75 mM KCl, 3 mM MgCl2, 55 years, white male (did not use tobacco, alcohol or other 10 mM dithiothreitol, 0.5 mM dNTPs and 200 U Super- substances; medical history of seizures and hypertension). script II Rnase HÀ Reverse Transcriptase (Invitrogen). Furthermore, 10-donor pooled cryopreserved hepatocytes This mixture was incubated for 1 h at 42 C. Undiluted with lot number KDN were used. At In Vitro Technologies cDNA was used for the analysis of CYP2A6, 2C9 and the CYP and Phase II enzyme activities in hepatocytes were 2C19 expression in HepG2 cells. For all other measure- routinely checked after thawing and hepatocytes were only ments cDNA was diluted to a concentration equivalent released when activities were comparable to activities in of 10 ng/ll RNA. fresh primary hepatocytes. Appropriate substrates were used to check the CYP1A2, 2A6, 2C9, 2C19, 2D6, 2E1, 2.6. Quantitative PCR 3A4, SULT, and UGT activity. Quantitative PCR was performed using an ABI PRISM 2.4. Preparation of compound solutions 7900 HT Sequence Detection system (Applied Biosystems, Inc. Foster City, USA). Specific primers were designed Compounds were dissolved in 100% dimethyl sulphox- using Primer Express software (version 2.0, Applied Bio- ide (DMSO). From the stock, 10-fold dilution series were systems). To avoid the influence of DNA contamination, prepared in DMSO. DMSO solutions were added to cul- primer pairs were designed to span an intron–exon bound- ture medium leading to a final DMSO concentration of ary (Table 1). Quantitative PCR was performed using 0.1%. cDNA equivalent to 50 ng of total RNA in a volume of 25 ll. Due to the low expression in HepG2 cells, a cDNA 2.5. RNA isolation and cDNA synthesis concentration equivalent to 500 ng of total RNA was used for analysis of CYP2A6, 2C9, and 2C19. The total mixture HepG2 cells were seeded on petri-dishes with a density contained cDNA, 300 nM forward primer, 300 nM reverse of 4500 cells/cm2 in culture medium with 10% dBCS. After primer and 1· SYBR green PCR Master Mix (Applied Bio- 24 h this medium was replaced with medium containing systems). The program used was 10 min at 95 C, 40 cycles different concentrations of the inducers 3MC, BNF, indiru- of 15 s at 95 C and 1 min at 60 C 100%, followed by a dis- bin, indigo, TCDD, CITCO, RIF, PB, T0901317, or vehi- sociation curve step. Expression levels were normalized by cle alone (0.1% DMSO). After exposure for 24 h, total using the b-actin housekeeping gene. cellular RNA was isolated using Trizol reagent according to the manufacturer’s protocol (Invitrogen, Karlsruhe, 2.7. Specificity of luminometric P450-Glo substrates Germany). Cryopreserved primary hepatocytes were not cultured The luminometric P450-Glo substrates are derivates of but were directly used for RNA isolation. A vial containing D-luciferin and converted by CYP enzymes into D-luciferin 5 · 106 hepatocytes was thawed and cells were added to (Cali et al., 2006). The specificity of Luciferin(Luc)-CEE,

Table 1 Primers used for the analysis of gene expression by quantitative PCR Gene Accession number Forward primer Reverse primer CYP1A1 ENST00000140465 CGGCCCCGGCTCTCT CGGAAGGTCTCCAGGATGAA CYP1A2 NM_000761 AGCTTCTCCTGGCCTCTGC GGACTTTTCAGGCCTTTGGG CYP2A6 ENST00000301141 CTATGGCTTCTTGCTGCTCATG CTTGCCGATCACTCTGTCAATC CYP2B6 ENST00000324071 TTAGGGAAGCGGATTTGTCTTG GGAGGATGGTGGTGAAGAAGAG CYP2C8 ENST00000285985 CACCCAGAGGTCACAGCTAAAGT CATGTGGCTCCTATCCTGCAT CYP2C9 ENST00000260682 ACATCAGCAAATCCTTAACCAATCT GGGTTTCAGGCCAAAATACAGA CYP2C19 ENST00000285981 TGCTCTCCTTCTCCTGCTGAAG TGCCAACGACACGTTCAATC CYP2D6 ENST00000320777 CGCATCCCTAAGGGAACGA TTCCAGACGGCCTCATCCT CYP2E1 ENST00000252945 GCAAGAGATGCCCTACATGGA GGGCACGAGGGTGATGAA CYP3A4 NM_017460 CAGGAGGAAATTGATGCAGTTTT GTCAAGATACTCCATCTGTAGCACAGT b-Actin NM_001101 CTGGCACCCAGCACAATG GCCGATCCACACGGAGTACT 1584 W.M.A. Westerink, W.G.E.J. Schoonen / Toxicology in Vitro 21 (2007) 1581–1591

Luc-ME, Luc-H, and Luc-BE for CYP enzymes, was Rockford, USA) (Westerink and Schoonen, accepted for assessed for CYP1A1, 1A2, 2A6, 2B6, 2C8, 2C9, 2C19, publication). 2D6, 2E1, and 3A4. Firstly, a NADPH regenerating system was prepared in phosphate buffered saline (pH 7.4). This 2.9. Measurement of induction of CYP1A1, 1A2 and 3A4 mixture contained 2.6 mM NADP+, 6.6 mM glucose-6- enzyme activities in HepG2 cells phosphate, 6.6 mM MgCl2 and 0.8 U/ml glucose-6-phosphate dehydrogenase. The cofactor solution was pre- HepG2 cells were trypsinized, counted and resuspended warmed at 37 C for 10 min. Supersomes (BD Biosciences, in culture medium to a final concentration of 3 · 104 cells/ USA) expressing one CYP were added to this cofactor mix- well for the CYP1A1 and CYP1A2 assay and to 1.5 · 105 ture. The final concentration of supersomes in the cofactor cells/well for the CYP3A4 assay. 190 ll cell suspension mixture was 10 pmol/ml. The cofactor/supersome mixture was added to a well of a white 96 well culture plate (Per- was kept at 37 C. Subsequently, 50 ll of either 60 lM kin–Elmer). The 96 well microtiter plates were incubated Luc-CEE, 200 lM Luc-Me, 200 lM Luc-H, or 200 lM for 24 h in a humidified atmosphere at 37 C under 5% Luc-BE in PBS was added to a well of a white 96 well plate CO2. Next day 10 ll medium containing different concen- (Perkin–Elmer, Groningen, The Netherlands). The plates trations of 3MC, BNF, indirubin, indigo, TCDD, CITCO, with substrates were pre-heated at 37 C for 5 min. Next, RIF, PB, T0901317, or vehicle alone (0.1% DMSO) was 50 ll from the cofactor/supersome mixture was added to added. Subsequently, plates were incubated for another one of the substrates. Substrates were incubated with the 24 h in a humidified atmosphere at 37 C under 5% CO2. enzymes for 30 min and the luminescence was measured After exposure, plates were washed twice with PBS and on a luminometer (TopCountNT, Perkin–Elmer). the CYP1A1, 1A2, and 3A4 activities were assessed by using 50 ll Luc-CEE (30 lM), Luc-ME (100 lM), or 2.8. Comparison of CYP1A1, 1A2, 2C9 and 3A4 enzyme Luc-BE (100 lM). In each assay the substrate was present activities in primary human hepatocytes and HepG2 cells at a concentration equal to the Km value. Under these test with P450-Glo substrates conditions, substrate conversion remained linear in time with all enzymes. Next 50 ll luciferin detection reagent The P450-Glo assays were performed according to the was added and plates were shaken for 10 min. Then the manufacturer’s protocol (Promega). Hepatocytes or luminescence signal was measured with a TopCountNT 4 HepG2 cells (5 · 10 ) were washed with PBS and resus- luminometer. For calculation of the amount of D-luciferin pended in 50 ll PBS. Next 50 ll of either 60 lM Luc- formed, a D-luciferin standard curve was made. CEE (CYP1A1), 200 lM Luc-ME (CYP1A2), 200 lM Luc-H (CYP2C9), or 200 lM Luc-BE (CYP3A4) was 2.10. Statistical analysis added. In each assay the substrate was present at a concentration equal to the Km value (Promega). HepG2 cells and Each experiment was performed in duplicate in three primary hepatocytes were incubated with P450-Glo sub- independent experiments except where indicated otherwise. strates for up to 1 h. Under these test conditions, substrate Data are expressed as mean ± SEM. Data obtained from conversion remained linear in time with all enzymes. Due mRNA and enzyme activity induction studies were com- to the aspecificity of Luc-ME, CYP1A2 activity was always pared for statistically significant differences using the Stu- determined in the presence of 10À6 M sulphaphenazole. dent’s unpaired t test (p < 0.05). For the mRNA This concentration blocks the CYP2C9 activity completely induction studies, only induction levels above 2-fold with and the CYP2C8 activity for 60% in human supersomes statistically significant differences were seen as relevant. (data not shown). In some experiments CYP activity was blocked by coincubation with inhibitors to show the spec- 3. Results ificity of the P450-Glo substrates. CYP1A1 activity was blocked with 10À6 M ketoconazole, CYP1A2 activity with 3.1. Comparison of CYP mRNA expression in HepG2 cells 10À6 M furafylline, CYP2C9 activity with 10À6 M sulpha- and primary human hepatocytes phenazole, and CYP3A4 activity with 10À6 M ketoconazole. Subsequently, 50 ll luciferin detection reagent was Transcripts of CYP1A1, 1A2, 2A6, 2B6, 2C8, 2C9, added. Plates were shaken for 10 min and the luminescence 2C19, 2D6, 2E1, and 3A4 were present in both cryopre- signal was measured on a TopCountNT luminometer. The served hepatocytes and HepG2 cells, although mRNA lev- conversion rate expressed as the increase in luminescence els of most CYP enzymes were much lower in HepG2 cells per minute per 105 cells (Dluminescence/min/105 cells) (Fig. 1). was calculated in the linear part of the reaction time. For CYP1A1 levels were 3- and 37-fold higher in hepato- calculation of the amount of D-luciferin formed, a D-lucif- cytes from donor EG and HRU. In contrast CYP1A1 lev- erin standard curve was made. Furthermore, cell lysates els in the donor pool were 10 times lower than those in were made and the protein content of HepG2 cells and HepG2 cells. Levels of the second CYP1A family member, primary human hepatocytes was measured with the BCA CYP1A2, were 24-, 386-, and 55-fold higher in hepatocytes protein assay kit from Pierce (Pierce Biotechnologies, from donor EG, HRU, and the pool. Likewise, Cyp2B6 W.M.A. Westerink, W.G.E.J. Schoonen / Toxicology in Vitro 21 (2007) 1581–1591 1585

3.2. Speciﬁcity of P450-Glo substrates

The specificity of the luminometric substrates was assessed for 11 human CYP isoforms (Fig. 2). Luc-CEE showed the greatest specificity for CYP1A1 and showed very little reactivity with the other CYPs. The second luminometric substrate Luc-ME showed the greatest reactivity with CYP1A2 as expected. However, this substrate also reacted substantially with CYP2C8 and 2C9. CYP2C8 and 2C9 reactivity was 24% and 32% of the reactivity with CYP1A2. Thus to assess CYP1A2 activity, CYP2C8 and 2C9 were inhibited with 10À6 M sulphaphenazole. This concentration blocked CYP2C9 completely and 60% of Fig. 1. Cytochrome P450 (CYP) 1A1, 1A2, 2A6, 2B6, 2C8, 2C9, 2C19, CYP2C8 activity but had no effect on CYP1A2 activity 2D6, 2E1 and 3A4 mRNA expression levels in cryopreserved primary (data not shown). Luc-H was very specific and reacted only human hepatocytes (donor EG, donor HRU, 10-donor pool) and HepG2 cells. Results are presented as the mean mRNA levels expressed in copy with CYP2C9. Luc-BE was quite specific for CYP3A4 and number ratio of the CYP and b-actin transcripts of two independent showed only little reactivity with CYP1A1 (16%) and 2C8 measurements. (14%).

3.3. Comparison of CYP enzyme activities in HepG2 cells transcript levels were 10-fold higher in the donor pool and and primary human hepatocytes more than 100-fold higher in donor EG and HRU. Expres- sion levels of CYP2D6 diﬀered less than 100-fold between The CYP activities in HepG2 cells and donor EG the hepatocyte samples and HepG2 cells. Levels of expressed in Dluminescence/min/105 cells are shown in CYP2D6 were 92-, 44-, and 7-fold higher in hepatocytes Fig. 3. Results were in concordance with the mRNA levels from donor EG, HRU, and the pool. Furthermore, levels measured. CYP1A1 activity was higher in donor EG as of CYP2A6, 2C8, 2C9, 2C19, 2E1, and 3A4 were generally compared with activity in HepG2 cells and could be inhib- between 100- and 1000-fold higher in the hepatocytes. ited with 10À6 M ketoconazole. CYP1A2 activity was

Fig. 2. Speciﬁcity of luciferin(Luc)-CEE (a), Luc-ME (b), Luc-H (c), and Luc-BE (d) for the human cytochrome P450 isotypes 1A1, 1A2, 2A6, 2B6, 2C8, 2C9, 2C18, 2C19, 2D6, 2E1, and 3A4. Results are presented as the mean activity of three independent measurements ± SEM. 1586 W.M.A. Westerink, W.G.E.J. Schoonen / Toxicology in Vitro 21 (2007) 1581–1591

in HepG2 cells was detectable, but 15-times lower than the activity in donor EG. Activities could be inhibited substan- À6 tially with 10 M ketoconazole. A D-luciferin standard curve showed that an increase of 10,000 light units is equal to the formation of 0.62 pmol D-luciferin (data not shown). Moreover, the protein content of 105 HepG2 cells and hepatocytes was 0.28 and 0.16 mg, respectively.

3.4. Eﬀects of AhR activators on CYP mRNA expression

The transcript levels of CYP1A1, 1A2, 2B6, 2C8, 2D6, 2E1, and 3A4 were measured after 24 h exposure to AhR agonists. The very low expression and consequently the Fig. 3. Cytochrome P450 (CYP) 1A1, 1A2, 2C9, and 3A4 activity large amount of cDNA needed made it impossible to study (Dluminescence/min/105 cells) in cryopreserved primary human hepato- the up-regulation of CYP2A6, 2C9, and 2C19. None of the cytes (donor EG) and HepG2 cells. CYP1A1, 1A2, 2C9, and 3A4 enzyme AhR agonists changed CYP2C8, 2D6 and 2E1 transcript activities were blocked with 10À6 M ketoconazole, furafylline, sulphaphenazole, and ketoconazole, respectively. An increase of 10,000 light units/ levels (data not shown). The effects on the transcript levels 5 5 min/10 cells equals to the formation of 0.62 pmol D-luciferin/min/10 of CYP1A1, 1A2, 2B6, and 3A4 are presented in Fig. 4. cells and the protein content of 105 HepG2 cells and hepatocytes was 0.28 As expected, transcript levels of CYP1A1 were highly and 0.16 mg, respectively. Results are presented as the mean activity of induced after exposure to the five AhR agonists. The lowest two independent measurements (À = basal activity, + = activity with activation doses for BNF, indirubin, indigo, TCDD, and inhibitor). The dashed line indicates the background luminescence. 3MC were 10À7 M, 10À8 M, 10À8 M, 10À10 M, and 10À8 M, respectively. The highest activation doses for detectable in both primary hepatocytes and HepG2 cells, BNF, indirubin, indigo, TCDD and 3MC (with fold but was 80-times lower in HepG2 cells. The CYP1A2 activ- changes in parentheses) were 10À4 M (239), 10À4 M ity could be blocked completely with 10À6 M furafylline. (1359), 10À4 M (85), 10À7 M (452), and 10À5 M (487), The activity of CYP2C9 was high in donor EG but not respectively. detectable in HepG2 cells. CYP2C9 activity in the hepato- Likewise, CYP1A2 was induced after exposure to the cytes could be blocked completely with 10À6 M sulphaphe- five AhR agonists. The lowest activation doses for BNF, nazole. Furthermore, results showed that CYP3A4 activity indirubin, indigo, TCDD and 3MC were 10À7 M,

Fig. 4. Induction of Cytochrome P450 (CYP) 1A1, 1A2, 2B6, and 3A4 mRNA expression in HepG2 cells after 24 h treatment with the AhR activators b- naphthoflavone (BNF), indirubin, indigo, TCDD, and 3-methylcholanthrene (3MC). The level in uninduced HepG2 cells is set at 1. This level represents a CYP/b-actin transcript copy number ratio of 2.65 · 10À5, 9.06 · 10À6, 4.72 · 10À5, and 3.44 · 10À5 for CYP1A1, 1A2, 2B6, and 3A4, respectively. Data are presented as the mean expression ± SEM (n = 3). *Significant difference (p < 0.05). W.M.A. Westerink, W.G.E.J. Schoonen / Toxicology in Vitro 21 (2007) 1581–1591 1587

10À8 M, 10À8 M, 10À10 M, and 10À8 M. The highest activa- CYP1A1, 1A2, 2B6, and 3A4 mRNA levels were signifi- tion doses for BNF, indirubin, indigo, TCDD and 3MC cantly induced by one or more of the tested activators (with fold changes in parentheses) were 10À4 M (51), (Fig. 5). 10À4 M (244), 10À4 M (444), 10À7 M (487), and 10À5 M CYP1A1 mRNA levels were enhanced after exposure to (276), respectively. 10À5 M T0901317 and CITCO, and 10À4 M PB. However, Surprisingly, CYP2B6 and CYP3A4 transcript levels induction was only very small (2- to 4-fold) in comparison were also enhanced after treatment with some of the to inductions of between 100- and more than 1000-fold AhR agonists, but fold changes were 10- to 100-fold lower observed after treatment with the AhR agonists. CYP1A2 than for the CYP1A isoforms. transcript levels were only slightly increased after exposure CYP2B6 transcript levels were induced after treatment to 10À5 M T0901317. with BNF, indirubin and 3MC. However, indigo and CYP2B6 was induced 3.5-fold after treatment with TCDD treatment did not change the CYP2B6 mRNA lev- 10À7 M CITCO. No effect of T0901317, RIF, and PB on els. The lowest activation dose for BNF, indirubin, and CYP2B6 expression was observed. 3MC was 10À5 M. The maximum induction of 3- to 4-fold CYP3A4 transcript levels were induced after treatment was observed at 10À4 M for BNF and indirubin, and at with T0901317 and RIF. No effects were seen with PB 10À5 M for 3MC. and CITCO. The lowest activation doses for T0901317 CYP3A4 was induced after exposure to indirubin, and RIF were 10À7 M and 10À5 M. The highest activation indigo and 3MC. No effect of BNF and TCDD was doses for T0901317 and CITCO were 10À7 Mand10À4 M, observed. After exposure to 10À4 M indirubin and indigo leading to 2.6- and 4.7-fold induction, respectively. an induction of 3.5- and 2-fold was measured. A higher induction of 10-fold was observed at 10À5 M 3MC. 3.6. Effect of activators on enzyme activities of CYP1A1, 1A2, and 3A4 3.5. Effects of PXR and CAR activators on CYP mRNA expression The CYP1A1 and 1A2 activities were assessed with Luc- CEE and Luc-ME for BNF, indirubin, indigo, TCDD, and The expression levels of CYP enzymes in HepG2 cells 3MC. The induction of CYP1A1 and 1A2 enzyme activities were measured after 24 h exposure to several PXR/CAR by AhR agonists are shown in Fig. 6. The AhR agonists agonists. Transcript levels of CYP2C8, 2D6, and 2E1 were tested induced the CYP1A1 enzyme activities above the not enhanced by the CAR/PXR agonists (data not shown). basal CYP1A1 activity in HepG2 cells. TCDD was the

Fig. 5. Induction of Cytochrome P450 (CYP) 1A1, 1A2, 2B6, and 3A4 mRNA expression in HepG2 after 24 h treatment with the PXR/LXR activator T0901317, PXR activator rifampicin (RIF), the CAR/PXR activator phenobarbital (PB), and the CAR activator CITCO. The level in uninduced HepG2 cells is set at 1. This level represents a CYP/b-actin transcript copy number ratio of 2.65 · 10À5, 9.06 · 10À6, 4.72 · 10À5, and 3.44 · 10À5 for CYP1A1, 1A2, 2B6, and 3A4, respectively. Data are presented as the mean expression ± SEM (n = 3). *Signiﬁcant diﬀerence (p < 0.05). 1588 W.M.A. Westerink, W.G.E.J. Schoonen / Toxicology in Vitro 21 (2007) 1581–1591

Fig. 6. Dose dependent induction of Cytochrome P450 (CYP) 1A1 and 1A2 enzyme activities (Dluminescence/min/105 cells) after 24 h treatment with b- naphthoflavone (BNF), indirubin, indigo, TCDD, and 3-methylcholanthrene (3MC). The dashed lines indicate the basal enzyme activities in HepG2 cells. 5 5 5 An increase of 10,000 light units/min/10 cells equals to the formation of 0.62 pmol D-luciferin/min/10 cells and the protein content of 10 HepG2 cells was 0.28 mg. Data are presented as the mean activity ± SEM (n = 3). most potent CYP1A1 inducer with a maximum induction lower in HepG2 cells as compared to levels in primary of almost 1000-fold at 10À8 M and a significant induction human hepatocytes. The mRNA levels of CYP1A1, 1A2, of almost 10-fold at 10À11 M. Indirubin, indigo and 3MC 2B6, 2C8, 2C9, and 2C19 in the three primary hepatocyte were also potent inducers (>100-fold). Indirubin, indigo sources showed a variation of more than 10-fold. This and 3MC induced CYP1A1 activity at concentrations as might have been due to genotype or food intake. low as 10À8 M. BNF was the weakest inducer of CYP1A1 CYPs play a pivotal role in the toxication of some com- (10-fold). pounds (Lewis et al., 1998; Vignati et al., 2005). Thus the Moreover, the five AhR agonists induced CYP1A2 low enzyme levels in HepG2 cells might result in an under- enzyme activities above the basal CYP1A2 activity. TCDD estimation of the toxicity of compounds as compared with induced CYP1A2 activity at concentrations as low as toxicity in primary human hepatocytes. On the other hand, 10À10 M and at 10À9 M the maximum induction was 15- cytotoxicity was observed in HepG2 for several com- fold. Indirubin enhanced CYP1A2 enzyme activity at all pounds like iproniazid, dacarbazine, nitrofurantoin, and tested concentrations, induction being observed at 10À8 M benzo[a]pyrene that need metabolic activation via and the maximal induction of 25-fold was observed at CYP3A4, 1A1, and/or 1A2 (Schoonen et al., 2005a,b). 10À6 M. BNF, indigo and 3MC were weaker inducers of Similar observations were made by O’Brien et al. (2006). the CYP1A2 enzyme activity than TCDD and indirubin. This implies that for several compounds, the CYP activities These compounds showed a maximum induction of between in HepG2 cells are high enough or that the compounds 5- and almost 10-fold. The lowest activation dose for BNF themselves activate a xenobiotic receptor, which induces and indigo was 10À6 M, while for 3MC it was 10À7 M. CYP expression. CYP3A4 activity was assessed with Luc-BE for T0901317 Rodriguez-Antona et al. (2002) showed that the low and PB. RIF could not be tested because the dark red color CYP expression in HepG2 and other permanent hepatoma of this compound disturbs the luminescence measurement. cells is caused by a decreased CYP transcription. A change Treatment with T0901713 induced CYP3A4 activity signif- in hepatic transcription factors might be involved in this icantly above the basal activity (Fig. 7). A maximum induc- process. Support for this hypothesis comes from transfec- tion just above 2-fold was measured at 10À5 M. No effect of tion studies with C/EBPa in HepG2 cells, leading to higher PB on CYP3A4 expression was observed. levels of CYP2B6, 2C9, and 2D6 (Jover et al., 1998). Another way to induce CYP enzyme levels is by treatment 4. Discussion with AhR, CAR, and PXR agonists. CYP1A1 contributes to the toxicity of many carcino- In the present study the CYP mRNA expression and gens, especially PAHs (Conney, 1982; Shimada et al., enzyme activities in HepG2 cells and cryopreserved pri- 1989, 1996). The main pathway for CYP1A1 induction is mary human hepatocytes were compared. The HepG2 cell through activation of the AhR pathway (Whitlock, 1999). line is derived from a Caucasian human hepatoblastoma In the present study all five AhR activators used highly and has both Phase I and II metabolizing enzymes (Hewitt induced CYP1A1 mRNA levels and enzyme activities. and Hewitt, 2004; Knasmuller et al., 1999, 2004). Hewitt TCDD was the most potent inducer followed by indirubin and Hewitt (2004) showed that the enzyme activities and indigo. Indirubin and indigo were identified in human depend on the source and culture conditions and that con- urine in normal donors and can be endogenous ligands of sequently characterization of HepG2 cells is essential. the AhR as both compounds are present in the human Results from the present study showed that the tran- serum at a concentration of 10À8 M(Adachi et al., 2001). script levels and enzyme activities of most CYPs were much In yeast AhR activation assays, these compounds were W.M.A. Westerink, W.G.E.J. Schoonen / Toxicology in Vitro 21 (2007) 1581–1591 1589

higher than those for CYP1A1 and CYP1A2. Classical AhR ligands and atypical CYP1A inducers were recently profiled in rats by microarray experiments (Coe et al., 2006). These authors also found that some of the AhR activators including flutamide, omeprazole and indole-3-carbi- nol showed cross-talk at high concentrations with PXR/ CAR pathways leading to CYP2B and CYP3A induction. Similar data were found for CYP2B6 in human hepatocytes with benzo[a]pyrene (Wilkening et al., 2003). These findings suggest that at high concentrations, some AhR agonists give cross-talk with the CAR/PXR pathway. Enzyme levels of CYP2B6 have not yet been measured Fig. 7. Dose dependent induction of Cytochrome P450 (CYP) 3A4 since a proper specific fluorophore or luminophore is still activity (Dluminescence/min/105 cells) after 24 h treatment with T0901317. The dashed line indicates the basal CYP3A4 activity in HepG2 cells. An unavailable. increase of 10,000 light units/min/105 cells equals to the formation of CYP3A4 metabolizes approximately 50% of the drugs 5 5 0.62 pmol D-luciferin/min/10 cells and the protein content of 10 HepG2 that are currently on the market and plays an important cells was 0.28 mg. Data are presented as the mean activity ± SEM (n = 3). role in the toxication of compounds (Vignati et al., 2005). Examples are CYP3A4 metabolites of flutamide, troglitazone, and isoniazid which cause hepatotoxicity comparable or even more potent than TCDD (Sugihara (Berson et al., 1993; Yamamoto et al., 2002; Walubo et al., 2004). Remarkably also the PXR/CAR activators et al., 1998). CYP3A4 is regulated by the CAR and PB, CITCO, and T0901317 induced CYP1A1 transcript PXR receptor (Moore et al., 2003; Waxman, 1999; Wei levels, although to much lower levels than real AhR ago- et al., 2002). In this study CYP3A4 mRNA regulation nists. CYP1A1 and/or CYP1A2 induction by PXR/CAR was shown with the PXR agonists RIF and T0901317, activators has been observed in other studies as well the latter being more potent. CYP3A4 was not induced (Gross-Steinmeyer et al., 2005; Rae et al., 2001). by the CAR agonists CITCO and PB. It has recently been CYP1A2 activates PAHs, nitrosamines and aryl amines observed in human hepatocytes that although PXR and into DNA-binding forms (Hecht, 1998). CYP1A2 induc- CAR both regulate CYP3A4 and CYP2B6 like in rodents, tion on transcriptional level can occur through AhR depen- human CAR preferentially induces CYP2B6 relative to dent and independent pathways (Corcos et al., 1998)or CYP3A4 because of its weak binding and activation of through post-transcriptional pathways (Silver et al., CYP3A4 ER6 (Faucette et al., 2006). In the study of 1990). Here all five AhR activators induced the CYP1A2 Faucette et al. (2006) treatment of human hepatocytes transcript levels by 10- to 100-fold, which was only slightly with 0.5–1 lM CITCO resulted in a more than 2-fold weaker than observed for CYP1A1. Of the PXR/CAR acti- mRNA induction of CYP3A4 in only one out of six vators only T0901317 showed statistically significant induc- donors. Cross-talk of the AhR agonists indirubin, indigo tion. Before induction, CYP1A2 levels were between 24- and 3MC with the PXR/CAR pathways leading to induc- and 386-fold lower. The induction leads to comparable or tion of CYP3A4 was also demonstrated. even higher mRNA and enzyme levels than in cryopre- The CYP2C subfamily contains four members of which served hepatocytes. CYP2C8, CYP2C9, and CYP2C19 are expressed in the CYP2B6 metabolizes endogenous substances such as liver (Goldstein and de Morais, 1994). Here, regulation testosterone (Gervot et al., 1999), serotonin (Fradette of CYP2C8 was studied. CYP2C8 plays a role in the oxida- et al., 2004), and xenobiotics of which clinical drugs such tive metabolism of taxol, arachidonic acid, and retinoids as cyclophosphamide (Huang et al., 2000) and bupropion (Nadin and Murray, 1999). In human hepatocytes (Faucette et al., 2000) are examples. In the present study, CYP2C8 is inducible with PXR/CAR activators such as CYP2B6 was induced by the specific CAR agonist CITCO, CITCO, dexamethasone, RIF, and PB (Ferguson et al., while RIF showed only a tendency for induction. Regula- 2005; Gerbal-Chaloin et al., 2001). None of the AhR, tion of CYP2B6 by these ligands via CAR and PXR has PXR, and CAR activators used in the present study chan- also been documented by others (Madan et al., 2003; Mag- ged the transcript levels of CYP2C8. These results are not lich et al., 2002; Faucette et al., 2006). T0901317 and PB consistent with findings in primary human hepatocytes. did not induce CYP2B6 levels although CYP2B6 induction Consistent with other reports, CYP2D6 was a non- by PB was described in primary human hepatocytes by oth- inducible enzyme (Li et al., 1997; Rae et al., 2001). Regula- ers (Faucette et al., 2004; Madan et al., 2003). Madan et al. tion of CYP2E1 can occur by exposure to xenobiotics, such (2003) showed a large variation in CYP2B6 induction from as alcohol, isoniazid, and pyridine (Carroccio et al., 1994; 1.3-fold induction to a 17-fold induction in 13 different Catania et al., 2007; Raucy et al., 2004; Wu et al., 1997). donors. Surprisingly, the AhR agonists BNF, indirubin, However, as for CYP2D6, no involvement of the AhR, and 3MC also induced CYP2B6 expression in HepG2 cells. PXR, CAR or other nuclear receptors in the regulation The concentrations needed for CYP2B6 activation were of CYP2E1 has been observed in this study or by others. 1590 W.M.A. Westerink, W.G.E.J. Schoonen / Toxicology in Vitro 21 (2007) 1581–1591

In conclusion, HepG2 cells have low levels of CYPs. The hydroxylation as a selective marker of human cytochrome P450 2B6 results regarding CYP1A1, 1A2, 2B6, and 3A4 induction catalytic activity. Drug Metab. Dispos. 28, 1222–1230. show that there is a good correlation with earlier results Faucette, S.R., Wang, H., Hamilton, G.A., Jolley, S.L., Gilbert, D., Lindley, C., Yan, B., Negishi, M., LeCluyse, E.L., 2004. Regulation of observed in primary human hepatocytes. Furthermore, CYP2B6 in primary human hepatocytes by prototypical inducers. CYP2D6 and CYP2E1 could, like in primary hepatocytes, Drug Metab. Dispos. 32, 348–358. not be induced with AhR, PXR, and CAR agonists. In Faucette, S.R., Sueyoshi, T., Smith, C.M., Negishi, M., LeCluyse, E.L., contrast to results observed in primary human hepatocytes Wang, H., 2006. Differential regulation of hepatic CYP2B6 and no regulation of CYP2C8 with PXR/CAR activators was CYP3A4 genes by constitutive androstane receptor but not pregnane X receptor. J. Pharmacol. Exp. Ther. 317, 1200–1209. found in HepG2 cells. Ferguson, S.S., Chen, Y., LeCluyse, E.L., Negishi, M., Goldstein, J.A., Still, HepG2 cells are an easy to handle tool in contra- 2005. Human CYP2C8 is transcriptionally regulated by the nuclear diction to primary hepatocytes to study the regulation of receptors constitutive androstane receptor, pregnane X receptor, CYP1A1, 1A2, 2B6, and 3A4. Pre-stimulating HepG2 cells glucocorticoid receptor, and hepatic nuclear factor 4alpha. Mol. with AhR, PXR, and CAR activators before performing Pharmacol. 68, 747–757. Fradette, C., Yamaguchi, N., Du, S.P., 2004. 5-Hydroxytryptamine is cytotoxicity assays might lead to a better predictivity of biotransformed by CYP2C9, 2C19 and 2B6 to hydroxylamine, which is toxicity. converted into nitric oxide. Br. J. Pharmacol. 141, 407–414. Gerbal-Chaloin, S., Pascussi, J.M., Pichard-Garcia, L., Daujat, M., Waechter, F., Fabre, J.M., Carrere, N., Maurel, P., 2001. Induction of References CYP2C genes in human hepatocytes in primary culture. Drug Metab. Dispos. 29, 242–251. Adachi, J., Mori, Y., Matsui, S., Takigami, H., Fujino, J., Kitagawa, H., Gervot, L., Rochat, B., Gautier, J.C., Bohnenstengel, F., Kroemer, H., de Miller III, C.A., Kato, T., Saeki, K., Matsuda, T., 2001. Indirubin and Berardinis, V., Martin, H., Beaune, P., de Waziers, I., 1999. Human indigo are potent aryl hydrocarbon receptor ligands present in human CYP2B6: expression, inducibility and catalytic activities. Pharmaco- urine. J. Biol. Chem. 276, 31475–31478. genetics 9, 295–306. Bao, H., Vepakomma, M., Sarkar, M.A., 2002. Benzo(a)pyrene exposure Goldstein, J.A., de Morais, S.M., 1994. Biochemistry and molecular induces CYP1A1 activity and expression in human endometrial cells. J. biology of the human CYP2C subfamily. Pharmacogenetics 4, 285– Steroid Biochem. Mol. Biol. 81, 37–45. 299. Berson, A., Wolf, C., Chachaty, C., Fisch, C., Fau, D., Eugene, D., Gross-Steinmeyer, K., Stapleton, P.L., Tracy, J.H., Bammler, T.K., Loeper, J., Gauthier, J.C., Beaune, P., Pompon, D., 1993. Metabolic Lehman, T., Strom, S.C., Eaton, D.L., 2005. Influence of Matrigel- activation of the nitroaromatic antiandrogen flutamide by rat and overlay on constitutive and inducible expression of nine genes human cytochromes P-450, including forms belonging to the 3A and encoding drug-metabolizing enzymes in primary human hepatocytes. 1A subfamilies. J. Pharmacol. Exp. Ther. 265, 366–372. Xenobiotica 35, 419–438. Bock, K.W., Kohle, C., 2004. Coordinate regulation of drug metabolism Hecht, S.S., 1998. Biochemistry, biology, and carcinogenicity of tobacco- by xenobiotic nuclear receptors: UGTs acting together with CYPs and specific N-nitrosamines. Chem. Res. Toxicol. 11, 559–603. glucuronide transporters. Drug Metab. Rev. 36, 595–615. Hewitt, N.J., Hewitt, P., 2004. Phase I and II enzyme characterization of Caldwell, G.W., Ritchie, D.M., Masucci, J.A., Hageman, W., Yan, Z., two sources of HepG2 cell lines. Xenobiotica 34, 243–256. 2001. The new pre-preclinical paradigm: compound optimization in Huang, W.Y., Chatterjee, N., Chanock, S., Dean, M., Yeager, M., early and late phase drug discovery. Curr. Top. Med. Chem. 1, 353– Schoen, R.E., Hou, L.F., Berndt, S.I., Yadavalli, S., Johnson, C.C., 366. Hayes, R.B., 2005. Microsomal epoxide hydrolase polymorphisms and Cali, J.J., Ma, D., Sobol, M., Simpson, D.J., Frackman, S., Good, T.D., risk for advanced colorectal adenoma. Cancer Epidemiol. Biomarkers Daily, W.J., Liu, D., 2006. Luminogenic cytochrome P450 assays. Prev. 14, 152–157. Expert. Opin. Drug Metab. Toxicol. 2, 629–645. Huang, Z., Roy, P., Waxman, D.J., 2000. Role of human liver Carroccio, A., Wu, D., Cederbaum, A.I., 1994. Ethanol increases content microsomal CYP3A4 and CYP2B6 in catalyzing N-dechloroethylation and activity of human cytochrome P4502E1 in a transduced HepG2 of cyclophosphamide and ifosfamide. Biochem. Pharmacol. 59, 961– cell line. Biochem. Biophys. Res. Commun. 203, 727–733. 972. Catania, J.R., McGarrigle, B.P., Rittenhouse-Olson, K., Olson, J.R., Jover, R., Bort, R., Gomez-Lechon, M.J., Castell, J.V., 1998. Re- 2007. Induction of CYP2B and CYP2E1 in precision-cut rat liver slices expression of C/EBP alpha induces CYP2B6, CYP2C9 and CYP2D6 cultured in defined medium. Toxicol. in Vitro 21, 109–115. genes in HepG2 cells. FEBS Lett. 431, 227–230. Chen, Y., Ferguson, S.S., Negishi, M., Goldstein, J.A., 2004. Induction of Knasmuller, S., Mersch-Sundermann, V., Kevekordes, S., Darroudi, F., human CYP2C9 by rifampicin, hyperforin, and phenobarbital is Huber, W.W., Hoelzl, C., Bichler, J., Majer, B.J., 2004. Use of human- mediated by the pregnane X receptor. J. Pharmacol. Exp. Ther. 308, derived liver cell lines for the detection of environmental and dietary 495–501. genotoxicants; current state of knowledge. Toxicology 198, 315–328. Coe, K.J., Nelson, S.D., Ulrich, R.G., He, Y., Dai, X., Cheng, O., Knasmuller, S., Schwab, C.E., Land, S.J., Wang, C.Y., Sanyal, R., Kundi, Caguyong, M., Roberts, C.J., Slatter, J.G., 2006. Profiling the hepatic M., Parzefall, W., Darroudi, F., 1999. Genotoxic effects of heterocyclic effects of flutamide in rats: a microarray comparison with classical aryl aromatic amines in human derived hepatoma (HepG2) cells. Muta- hydrocarbon receptor ligands and atypical CYP1A inducers. Drug genesis 14, 533–540. Metab. Dispos. 34, 1266–1275. Kola, I., Landis, J., 2004. Can the pharmaceutical industry reduce Conney, A.H., 1982. Induction of microsomal enzymes by foreign attrition rates? Nat. Rev. Drug Discov. 3, 711–715. chemicals and carcinogenesis by polycyclic aromatic hydrocarbons: Lewis, D.F., Ioannides, C., Parke, D.V., 1998. Cytochromes P450 and G.H.A. Clowes Memorial Lecture. Cancer Res. 42, 4875–4917. species differences in xenobiotic metabolism and activation of carcin- Corcos, L., Marc, N., Wein, S., Fautrel, A., Guillouzo, A., Pineau, T., ogen. Environ. Health Perspect. 106, 633–641. 1998. Phenobarbital induces cytochrome P4501A2 hnRNA, mRNA Li, A.P., Reith, M.K., Rasmussen, A., Gorski, J.C., Hall, S.D., Xu, L., and protein in the liver of C57BL/6J wild type and aryl hydrocarbon Kaminski, D.L., Cheng, L.K., 1997. Primary human hepatocytes as a receptor knock-out mice. FEBS Lett. 425, 293–297. tool for the evaluation of structure–activity relationship in cytochrome Faucette, S.R., Hawke, R.L., LeCluyse, E.L., Shord, S.S., Yan, B., P450 induction potential of xenobiotics: evaluation of rifampin, Laethem, R.M., Lindley, C.M., 2000. Validation of bupropion rifapentine and rifabutin. Chem. Biol. Interact. 107, 17–30. W.M.A. Westerink, W.G.E.J. Schoonen / Toxicology in Vitro 21 (2007) 1581–1591 1591

Madan, A., Graham, R.A., Carroll, K.M., Mudra, D.R., Burton, L.A., assays on ROS, glutathione depletion and calcein uptake. Toxicol. in Krueger, L.A., Downey, A.D., Czerwinski, M., Forster, J., Ribadene- Vitro 19, 505–516. ira, M.D., Gan, L.S., LeCluyse, E.L., Zech, K., Robertson Jr., P., Shenoy, S.D., Spencer, T.A., Mercer-Haines, N.A., Alipour, M., Garg- Koch, P., Antonian, L., Wagner, G., Yu, L., Parkinson, A., 2003. ano, M.D., Runge-Morris, M., Kocarek, T.A., 2004. CYP3A induc- Effects of prototypical microsomal enzyme inducers on cytochrome tion by liver x receptor ligands in primary cultured rat and mouse P450 expression in cultured human hepatocytes. Drug Metab. Dispos. hepatocytes is mediated by the pregnane X receptor. Drug Metab. 31, 421–431. Dispos. 32, 66–71. Maglich, J.M., Parks, D.J., Moore, L.B., Collins, J.L., Goodwin, B., Shimada, T., Iwasaki, M., Martin, M.V., Guengerich, F.P., 1989. Human Billin, A.N., Stoltz, C.A., Kliewer, S.A., Lambert, M.H., Willson, liver microsomal cytochrome P-450 enzymes involved in the bioacti- T.M., Moore, J.T., 2003. Identification of a novel human constitutive vation of procarcinogens detected by umu gene response in Salmonella androstane receptor (CAR) agonist and its use in the identification of typhimurium TA 1535/pSK1002. Cancer Res. 49, 3218–3228. CAR target genes. J. Biol. Chem. 278, 17277–17283. Shimada, T., Hayes, C.L., Yamazaki, H., Amin, S., Hecht, S.S., Maglich, J.M., Stoltz, C.M., Goodwin, B., Hawkins-Brown, D., Moore, Guengerich, F.P., Sutter, T.R., 1996. Activation of chemically diverse J.T., Kliewer, S.A., 2002. Nuclear pregnane x receptor and constitutive procarcinogens by human cytochrome P-450 1B1. Cancer Res. 56, androstane receptor regulate overlapping but distinct sets of genes 2979–2984. involved in xenobiotic detoxification. Mol. Pharmacol. 62, 638–646. Silver, G., Reid, L.M., Krauter, K.S., 1990. Dexamethasone-mediated Mankowski, D.C., Ekins, S., 2003. Prediction of human drug metaboliz- regulation of 3-methylcholanthrene-induced cytochrome P-450d ing enzyme induction. Curr. Drug Metab. 4, 381–391. mRNA accumulation in primary rat hepatocyte cultures. J. Biol. Moore, J.T., Moore, L.B., Maglich, J.M., Kliewer, S.A., 2003. Functional Chem. 265, 3134–3138. and structural comparison of PXR and CAR. Biochim. Biophys. Acta Sugihara, K., Kitamura, S., Yamada, T., Okayama, T., Ohta, S., 1619, 235–238. Yamashita, K., Yasuda, M., Fujii-Kuriyama, Y., Saeki, K., Matsui, Morrow, C.S., Diah, S., Smitherman, P.K., Schneider, E., Townsend, S., Matsuda, T., 2004. Aryl hydrocarbon receptor-mediated induction A.J., 1998. Multidrug resistance protein and glutathione S-transferase of microsomal drug-metabolizing enzyme activity by indirubin and P1-1 act in synergy to confer protection from 4-nitroquinoline 1-oxide indigo. Biochem. Biophys. Res. Commun. 318, 571–578. toxicity. Carcinogenesis 19, 109–115. Townsend, A.J., Kabler, S.L., Doehmer, J., Morrow, C.S., 2002. Nadin, L., Murray, M., 1999. Participation of CYP2C8 in retinoic acid 4- Modeling the metabolic competency of glutathione S-transferases hydroxylation in human hepatic microsomes. Biochem. Pharmacol. 58, using genetically modified cell lines. Toxicology 181–182, 265–269. 1201–1208. Vignati, L., Turlizzi, E., Monaci, S., Grossi, P., Kanter, R., Monshouwer, O’Brien, P.J., Irwin, W., Diaz, D., Howard-Cofield, E., Krejsa, C.M., M., 2005. An in vitro approach to detect metabolite toxicity due to Slaughter, M.R., Gao, B., Kaludercic, N., Angeline, A., Bernardi, P., CYP3A4-dependent bioactivation of xenobiotics. Toxicology 216, Brain, P., Hougham, C., 2006. High concordance of drug-induced 154–167. human hepatotoxicity with in vitro cytotoxicity measured in a novel Walubo, A., Smith, P., Folb, P.I., 1998. The role of oxygen free radicals in cell-based model using high content screening. Arch. Toxicol. 80, 580– isoniazid-induced hepatotoxicity. Methods Find. Exp. Clin. Pharma- 604. col. 20, 649–655. Rae, J.M., Johnson, M.D., Lippman, M.E., Flockhart, D.A., 2001. Waxman, D.J., 1999. P450 gene induction by structurally diverse Rifampin is a selective, pleiotropic inducer of drug metabolism genes xenochemicals: central role of nuclear receptors CAR, PXR, and in human hepatocytes: studies with cDNA and oligonucleotide PPAR. Arch. Biochem. Biophys. 369, 11–23. expression arrays. J. Pharmacol. Exp. Ther. 299, 849–857. Wei, P., Zhang, J., Dowhan, D.H., Han, Y., Moore, D.D., 2002. Specific Raucy, J.L., Lasker, J., Ozaki, K., Zoleta, V., 2004. Regulation of and overlapping functions of the nuclear hormone receptors CAR and CYP2E1 by ethanol and palmitic acid and CYP4A11 by clofibrate PXR in xenobiotic response. Pharmacogen. J. 2, 117–126. in primary cultures of human hepatocytes. Toxicol. Sci. 79, 233– Westerink, W.M., Schoonen, G.E.J., accepted for publication. Phase II 241. enzyme levels in HepG2 cells and cryopreserved human primary Rodriguez-Antona, C., Donato, M.T., Boobis, A., Edwards, R.J., Watts, human hepatocytes and their induction in HepG2 cells. Toxicol. in P.S., Castell, J.V., Gomez-Lechon, M.J., 2002. Cytochrome P450 Vitro. expression in human hepatocytes and hepatoma cell lines: molecular Whitlock Jr., J.P., 1999. Induction of cytochrome P4501A1. Annu. Rev. mechanisms that determine lower expression in cultured cells. Xeno- Pharmacol. Toxicol. 39, 103–125. biotica 32, 505–520. Wilkening, S., Stahl, F., Bader, A., 2003. Comparison of primary human Rooney, P.H., Telfer, C., McFadyen, M.C., Melvin, W.T., Murray, G.I., hepatocytes and hepatoma cell line Hepg2 with regard to their 2004. The role of cytochrome P450 in cytotoxic bioactivation: future biotransformation properties. Drug Metab. Dispos. 31, 1035–1042. therapeutic directions. Curr. Cancer Drug Targ. 4, 257–265. Wu, D., Ramin, S.A., Cederbaum, A.I., 1997. Effect of pyridine on the Runge, D., Kohler, C., Kostrubsky, V.E., Jager, D., Lehmann, T., Runge, expression of cytochrome P450 isozymes in primary rat hepatocyte D.M., May, U., Stolz, D.B., Strom, S.C., Fleig, W.E., Michalopoulos, culture. Mol. Cell Biochem. 173, 103–111. G.K., 2000. Induction of cytochrome P450 (CYP)1A1, CYP1A2, and Xie, W., Yeuh, M.F., Radominska-Pandya, A., Saini, S.P., Negishi, Y., CYP3A4 but not of CYP2C9, CYP2C19, multidrug resistance (MDR- Bottroff, B.S., Cabrera, G.Y., Tukey, R.H., Evans, R.M., 2003. 1) and multidrug resistance associated protein (MRP-1) by prototyp- Control of steroid, heme, and carcinogen metabolism by nuclear ical inducers in human hepatocytes. Biochem. Biophys. Res. Commun. pregnane X receptor and constitutive androstane receptor. Proc. Natl. 273, 333–341. Acad. Sci. USA 100, 4150–4155. Schoonen, W.G., de Roos, J.A., Westerink, W.M., Debiton, E., 2005a. Yamamoto, Y., Yamazaki, H., Ikeda, T., Watanabe, T., Iwabuchi, H., Cytotoxic effects of 110 reference compounds on HepG2 cells and for Nakajima, M., Yokoi, T., 2002. Formation of a novel quinone epoxide 60 compounds on HeLa, ECC-1 and CHO cells. II mechanistic assays metabolite of troglitazone with cytotoxicity to HepG2 cells. Drug on NAD(P)H, ATP and DNA contents. Toxicol. in Vitro 19, 491–503. Metab. Dispos. 30, 155–160. Schoonen, W.G., Westerink, W.M., de Roos, J.A., Debiton, E., 2005b. Yueh, M.F., Kawahara, M., Raucy, J., 2005. Cell-based high-throughput Cytotoxic effects of 100 reference compounds on Hep G2 and HeLa bioassays to assess induction and inhibition of CYP1A enzymes. cells and of 60 compounds on ECC-1 and CHO cells. I mechanistic Toxicol. in Vitro 19, 275–287.