Prediction of Aryl Hydrocarbon Receptor-Mediated Enzyme

Note that the increase in the RNA level by treatment with losartan and pantoprazole, which is only small on this scale, was highly significant with re...
0 downloads 0 Views 159KB Size
Chem. Res. Toxicol. 1998, 11, 1447-1452

1447

Prediction of Aryl Hydrocarbon Receptor-Mediated Enzyme Induction of Drugs and Chemicals by mRNA Quantification Roland Fro¨tschl,† Lubomir Chichmanov,† Ullrich Kleeberg,‡ Alfred G. Hildebrandt,‡ Ivar Roots,† and Ju¨rgen Brockmo¨ller*,† Institute of Clinical Pharmacology, University Hospital Charite´ , Humboldt-University of Berlin, Schumannstrasse 20/21, D-10098 Berlin, Germany, and Federal Institute of Drugs and Medical Devices, Seestrasse 10, D-13353 Berlin, Germany Received July 10, 1998

Enzyme-specific testing for drug interactions by in vitro techniques has become a routine practice in drug development. With many drugs, enzyme induction has similar importance for the prediction of drug-drug interactions. We developed a method for recognizing enzyme induction mediated via the aryl hydrocarbon receptor. This type of induction may be clinically important since experimental data suggest a higher rate of toxification in induced subjects. Twenty-four drugs and environmental chemicals, selected as prototype inducers or being chemically related to known inducers, including HIV protease inhibitors nelfinavir, saquinavir, ritonavir, and indinavir, were tested for their potency to induce cytochrome P450 1A1 mRNA in human Hela cell cultures by a quantitative reverse transcriptase polymerase chain reaction. Known prototype inducers such as β-naphthoflavone and 3-methylcholanthrene exhibited the highest inducing potency quantified with an Imax value (maximal induction of cytochrome P450 1A1 mRNA synthesis) of 5.48 and 10.7 × 106 mRNA molecules per 150 ng of total RNA, respectively. The enzyme-inducing efficacy of some compounds such as resveratrol (2.92 × 106) and the protease inhibitors was not much lower (2.23-3.08 × 106). All compounds that were structurally similar to benzimidazoles exhibited some extent of enzyme induction; e.g., Imax values were 0.86 × 106, 0.20 × 106, and 0.14 × 106 for omeprazole, lansoprazole, and losartan, respectively. To predict the clinical relevance of these inducing effects, the concentration at half-maximal induction IM was estimated; the plasma concentrations of these drug substances were within 1 order of magnitude of the IM values, upon usual dosage. In conclusion, cytochrome P450 1A1 enzyme induction by drugs is a common phenomenon, though there is a great range in the inducing efficacy. In vitro prediction of enzyme induction may be useful for explaining or foreseeing drug interactions, drug side effects, or toxicity by xenobiotics.

Introduction Cytochrome P450 (P450)1 enzymes are mixed-function oxidases. Enzyme-specific testing for drug interactions based on mutual inhibition of P450 enzymes by in vitro techniques is currently a routine practice during phases I-III of drug development. This is used to predict potentially dangerous drug interactions (1, 2). With many drugs, enzyme induction may have similar importance for prediction of drug-drug interactions. There are different types of enzyme induction, such as the P450 2B type triggered by phenobarbital, the P450 * To whom correspondence should be addressed: Institut fu¨r Klinische Pharmakologie, Universita¨tsklinikum Charite´, HumboldtUniversita¨t zu Berlin, Schumannstrasse 20/21, D-10098 Berlin, Germany. Phone: +49 30 2802 8912. Fax: +49 30 2802 5153. E-mail: [email protected]. † Humboldt-University of Berlin. ‡ Federal Institute of Drugs and Medical Devices. 1 Abbreviations: AhR, aryl hydrocarbon receptor; bp, base pairs; DMEM, Dulbecco’s Modified Eagle’s Medium; DMSO, dimethyl sulfoxide; GSTA1, glutathione S-transferase type A1; HIV, human immunodeficiency virus; IM, half-maxmal induction of P450 1A1 mRNA; Imax, maximal induction of P450 1A1 mRNA; P450, cytochrome P450; RT-PCR, reverse transcriptase polymerase chain reaction; TIC, transcription initiation complex; Ugt1*06, UDP glucuronyltransferase type 1*06.

3A type triggered by rifampin or dexamethasone, the P450 4 type triggered by peroxisome proliferators such as clofibrate, and the P450 1A type triggered by dioxins, methylcholanthrene, and omeprazole, which is mediated by the cytosolic aryl hydrocarbon receptor (AhR) (3). P450 1 enzymes are important metabolizers of drugs and some steroids, and they may act as activators of polycyclic aromatic hydrocarbons and toxic environmental pollutants (4, 5), capable of transforming some procarcinogens into ultimate carcinogens (6). In each type of induction, several P450 enzymes and phase II enzymes are induced in parallel. In the AhRmediated type, the following enzymes are induced: P450 1A1 (7), P450 1A2 (8), P450 1B1 (9), menadione oxidoreductase (10), aldehyde dehydrogenase (type 3) (11), glucuronyltransferase (Ugt1*06) (12), and glutathione S-transferase A1 (GSTA1) (13). Thus, enzyme induction via the AhR may result in increases in phase I and phase II metabolism. The caffeine test, which measures the P450 1A2 activity, is the most prominent in vivo probe for this type of enzyme induction (14); earlier, phenacetin was used for this purpose (15). These functional tests truly reflect the activity changes by induction. However, they are not

10.1021/tx980164h CCC: $15.00 © 1998 American Chemical Society Published on Web 11/25/1998

1448 Chem. Res. Toxicol., Vol. 11, No. 12, 1998

Fro¨ tschl et al.

Table 1. Tested Concentration Range and Imax and IM Values ( Standard Deviations

inducer losartan (39) dipropyl-5-carboxamidotryptamine (40) efaroxan (41) resveratrol (21) 2-phenylbenzthiazole

medical relevance

H2O 2% ethanol

2.6-258 2.5-124

0.14 ( 0.005 0.08 ( 0.005

37.5 ( 12 56.0 ( 27

R2-adrenoreceptor agonist food compound in grapes certain phenylbenzimidazoles (2-phenylbenzimidazole-5-sulfonic acid is used for a UV-B light protection)

H2O 1% DMSO 2% DMSO

0.79-79 5-250 4.7-473

0.53 ( 0.04 2.92 ( 0.15 1.42 ( 0.05

10.5 ( 5 22.4 ( 9 31.1 ( 8

1% DMSO 2% DMSO 2% DMSO 2% DMSO 2% DMSO 2% DMSO

5.1-515 5.1-512 3.7-74 4-120 0.74-37 0.74-37 6.2-617 4.7-470

0.54 ( 0.019 3.93 ( 0.23 10.70 ( 0.48 3.48 ( 0.22 5.48 ( 0.37 0.89 ( 0.03 0.12 ( 0.003 0.23 ( 0.025

20.5 ( 10 19.0 ( 4 29.0 ( 16 33.0 ( 29 11.0 ( 1.7 9.30 ( 3 275 ( 130 172 ( 120

2.2-222 2.1-209 2.4-235 1.6-162 169-8466 5.2-261 36-565 1.6-160

3.15 ( 0.10 0.16 ( 0.002 0.86 ( 0.03 0.20 ( 0.007 0.62 ( 0.008 1.60 ( 0.08 0.15 ( 0.006 2.32 ( 0.04

43.0 ( 14 41.2 ( 16 104 ( 32 109 ( 90 3820 ( 90

2% HCl (0.1 M) antimalarial drug H2O H+-, K+-ATPase inhibitors H2O 1% DMSO H2O anthelmintic drug, benzimidazole derivative H2O antifungal drug 1.5% DMSO anticonvulsant drug H2O HIV protease inhibitors, known to cause H2O drug interactions by P450 3A4 inhibition

ritonavir saquinavir (46) nelfinavir (47) a

IM ( SD (µM)

angiotensin ΙΙ receptor antagonist selective 5-HT1A receptor antagonist

phenylbenzimidazole (42) 2-phenylbenzoxazole 3-methylcholanthrene known Ah receptor ligands benzo[a]pyrene β-naphthoflavone (19) R-naphthoflavone (20) isosafrol (43) domperidone (27) antiemetic drug, benzimidazolone primaquine (22) pantoprazole (24) omeprazole (14, 23) lansoprazole mebendazole (44) miconazole (44) phenytoin (45) indinavir

vehiclea

tested Imax ( SD concentration (106 mRNA (µM) molecules/150 ng)

1.4-140 1.5-150 1.5-150

2.23 ( 0.07 3.08 ( 0.13 2.40 ( 0.05

174 ( 60 31.0 ( 22 29.0 ( 23 27.0 ( 17 37.0 ( 17

Final vehicle concentration in cell culture medium.

suitable for screening purposes; their precision and sensitivity are rather limited, and the effect on gene regulation may not be detectable in cases of simultaneous enzyme inhibition. For instance, resveratrol, which appears in our study as a strong inducer of the Ahassociated gene battery, is also a strong mechanismbased inhibitor of P450 1A1 and P450 1A2 (16); therefore, the increase in the amount of mRNA measured by our technique will probably not be reflected in increased activity of these enzymes, but this type of enzyme inhibition is not likely to occur with the other coregulated enzymes such as Ugt1*06. Thus, a universal in vivo probe for AhR-mediated enzyme induction is not conceivable. Different from toxicologically relevant chemicals, only a few drugs are known as inducers of P450 1A1 and 1A2, and the drug has attracted much attention when after approval of omeprazole, this drug became known as a P450 1A1 and 1A2 enzyme inducer (14, 17, 18). Therefore, we searched for methods for quantitatively predicting the P450 1A1 enzyme-inducing potency of drugs. Known prototype inducers of cytochrome P450 1A1 and 1A2 like benzo[a]pyrene, 3-methylcholanthrene, and β-naphthoflavone (19) were also tested to compare the inducing potency of the other substances as described in Table 1, including R-naphthoflavone (20), resveratrol (21), primaquine (22), human immunodeficiency virus (HIV) protease inhibitors, proton pump inhibitors omeprazole (14, 23), lansoprazole, pantoprazole (24), and other benzimidazoles. Benzimidazole derivatives were studied so structure-function relationships in enzyme induction could be investigated. Enzyme induction was measured with several inducer concentrations and expressed as the maximal effect (Imax) and as a concentration at which the half-maximal induc-

tion was achieved (IM). The IM value may be compared with concentrations seen in humans and allows estimation of the relevance of the induction.

Materials and Methods Tested Drugs and Chemicals. Hela cells were from Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (Braunschweig, Germany); dimethyl sulfoxide (DMSO) and ethanol of analytical grade were from Merck (Darmstadt Germany). Oligonucleotide primers were from TIB-Molbiol (Berlin Germany). Tested compounds were β-naphthoflavone, R-naphthoflavone, isosafrol, primaquine, 2-phenylbenzimidazole, 2-phenylbenzthiazole, 2-phenylbenzoxazole 3-methylcholanthrene, and benzo[a]pyrene (Aldrich, Steinheim, Germany), resveratrol and miconazole (Sigma, Deisenhofen, Germany), dipropyl-5-carboxamidotryptamine, phenytoin, Efaroxan, and domperidone (RBI, Natick, MA), omeprazole (Antra Astra, Ha¨ssle, Sweden), lansoprazole (Agopton, Takeda, Aachen, Germany), pantoprazole (Pantozol, Byk Gulden, Konstanz, Germany), mebendazole (Sigma, Deisenhofen, Germany), losartan (Lorzaar, MSD Chibropharm, Haar, Germany), nelfinavir (Viracept, Agouron, La Jolla, CA), indinavir (Crixivan, MSD, Hoddesdon, U.K.), ritonavir (Norvir, Abbott, Queenborough, U.K.), and Saquinavir (Invirase, Roche, Welwyn Garden City, U.K.). Treatment and Preparation of Hela Cells. Hela cells were cultured in 25 cm2 flasks in Dulbecco’s Modified Eagle’s Medium (DMEM) with 10% v/v newborn calf serum to near confluence, and the inducer substances dissolved in DMSO, ethanol, or water were added. The maximal concentration of solvent was 2%; in each series, the same amount of solvent was used. Control culture flasks were pretreated with the same amount of DMSO, ethanol, or water. Cells were incubated for 18 h at 37 °C in a 5% CO2 atmosphere, harvested by conventional trypsinization with trypsin/EDTA, pelleted by centrifugation, and washed twice with a Dulbecco’s phosphate-buffered saline [10 mM Na,K-phosphate (pH 7.4), 2.5 mM KCl, and 150

Prediction of Ah Receptor-Mediated Enzyme Induction mM NaCl] solution. Cells were counted using a “Neubauer improved” counting chamber and stored in liquid nitrogen until RNA extraction. Isolation of RNA. Total RNA was isolated from 106 cells by following the method of Chomczynski and Sacchi (25). The RNA preparations were digested with 10 units of RNAse-free RQ1 DNAse (Promega, Madison, WI) for 15 min at 37 °C, followed by phenol/chloroform extraction and ethanol precipitation. The RNA concentration was determined spectrophotometrically. Quantification of P450 1A1 mRNA. The quantitative P450 1A1 reverse transcriptase polymerase chain reaction (RTPCR) assay was performed as described previously (26, 27), except that the total RNA used in each reaction was 150 ng. In brief, an artificial standard RNA (aRNA) was constructed from genomic P450 1A1, spanning intron 6 from nucleotides 64356870 (28). An EcoRI restriction site was introduced at position 6724 by primer-directed mutagenesis to distinguish between amplified contaminating genomic DNA and aRNA. RT reactions were set up according to the cDNA synthesis system manual of Life Technologies (Gaithersburg, MD) using moloney murine leukemia virus RT and a temperature-sensitive primer RT1 (5′-CAGGAAGAGAAAGAC-3′, positions 68336847) (29). Reaction mixtures were adjusted to a final volume of 20 µL containing 8 units of RNAsin (Promega) and 150 ng of total cellular RNA. Assays were overlaid with 25 µL of mineral oil (Sigma, St. Louis, MO) and incubated at 37 °C for 1 h. Tubes were heated to 95 °C for 10 min, and 30 µL of PCR mix was added at 95 °C, containing 50 mM KCl, 10 mM Tris (pH 8.3), 10 pmol of primers 1A8 (5′-GGCTTTTACATCCCCAAGGGGCG3′, positions 6435-6457) and 1A9 (5′-ATACACTTCCGCTTGCCCATGCC-3′, positions 6786-6808), 0.5 mM MgCl2, and dNTP and 1.25 units of AmpliTaq (both at 67 µM). Cycling conditions were 1 min at 95 °C, 40 s at 66 °C, and 90 s at 72 °C for 30 cycles followed by 7 min at 72 °C. After EcoRI digestion, the PCR products were separated on a 2% agarose gel. mRNAderived amplificates show a 182 bp band, genomic DNA-derived ones a 373 bp band (28), and artificial RNA-derived ones two bands of 288 and 85 bp. Equality of the 373 and 182 bp fragments in PCR amplification was ensured by amplifying both in one PCR and taking samples at different cycle numbers. Quotients of band intensities were the same at all cycles. Tests for intercalation of ethidium bromide showed that the same amounts of DNA (micrograms) of the different bands resulted in densitometric measurements not differing by more than 5% (data not shown). Serial dilutions of aRNA were added to a series of RNA samples extracted from cell cultures with a constant RNA concentration, and the resulting amplificates were recorded with a digital image system. A typical gel image is shown in Figure 1. Bands were integrated with the Quantiscan densitometry software from Biosoft (Cambridge, U.K.), and density values of 182 bp bands were corrected for size by addition of log a (a ) 288/182 ) 1.58) before correlation with 288 bp values. The concentration of mRNA was determined as the intercept of the linear regression curves of all aRNA intensities versus the mRNA intensities. Inter- and Intra-Assay Coefficients of Variation Were Determined. (1) Interassay. One cell sample (induced with 1.5 µM Nelfinavir as described above) was divided into equal aliquots, and isolation and mearsurement of total RNA were carried out on different days as described above. The mean value of this series was 2.6 × 105 P450 1A1 mRNA molecules/ 150 ng of total RNA with a coefficient of variation of 12.7%. (2) Intra-Assay. One cell sample (induced with 40 µM Efaroxan as described above) was divided into equal aliquots, and isolation and measurement of total RNA were carried out on the same day as described above. The mean of this series was 3.2 × 105 with a coefficient of variation of 4.65%. Basal levels of P450 1A1 mRNA differed depending on the vehicle that was used, and the mean and the percent coefficient of variation were as follows: water, 24 630 (12%); 1% DMSO,

Chem. Res. Toxicol., Vol. 11, No. 12, 1998 1449

Figure 1. Gel electrophoreses for measurement of mRNA after treatment with vehicle (control) and 37.5 and 150 µM saquinavir. Each lane represents one P450 1A1-specific competitive RT-PCR from 150 ng of total RNA and the indicated copy number of internal standard aRNA (27). The numbers on top indicate the copy numbers of the internal standard aRNA. The marker is a 100 bp DNA ladder from Fermantas (St. Leon-Rot, Germany). 67 793 (8.4%); 1.5% DMSO, 97 465 (5.4%); 2.0% DMSO, 136 161 (11.4%); 2% Ethanol, 2949 (10.8%); and 0.002 M HCl, 54 957 (15.8%) Data Analysis. IM represents the concentration of the inducer which results in half-maximum induction. Imax, which is the maximum amount of P450 1A1 mRNA formed, was estimated by nonlinear regression analysis using the following equation:

mRNA ) I0 +

ImaxC IM + C

where mRNA is the total amount of P450 1A1 mRNA synthesized, I0 is the basal amount of P450 1A1 mRNA synthesized, and C is the inducer concentration. For nonlinear regression analysis, the program Origin 4.1 (Microcal, Northampton, MA) was used.

Results We evaluated an in vitro system for quantification of enzyme-inducing properties using known and suspected inducers in comparison with water, DMSO, or ethanol as vehicles. All data were equalized by subtraction of the P450 1A1 mRNA levels of the respective vehicle incubations. In the incubations with drugs, the lowest concentrations tested were below the clinical plasma levels and the upper concentrations were limited by solubility or eventual toxicity. Toxicity was limiting in

1450 Chem. Res. Toxicol., Vol. 11, No. 12, 1998

Fro¨ tschl et al.

mebendazole also induced P450 1A1 mRNA expression, but its IM value was very high. This drug is normally ingested in high doses but only poorly absorbed, resulting in plasma levels of around 0.1 µM. The drugs losartan, phenytoin, and domperidone had only a low inducing effect on P450 1A1 mRNA expression but nevertheless increased the expression level 4-7-fold above the basal level. This increase was statistically significant in all substances tested here (p < 0.001, Kruskal-Wallis-Test).

Discussion

Figure 2. Concentration-dependent enzyme induction in Hela cell culture. The lines were obtained by fitting the data to a Michaelis-Menten model, and Imax and IM values were calculated (corresponding to Vmax and KM). Note that the increase in the RNA level by treatment with losartan and pantoprazole, which is only small on this scale, was highly significant with respect to the measurements with vehicle alone.

Figure 3. Maximal cytochrome P450 1A1 inducing effect of 24 drugs and chemicals.

the case of miconazole. Figure 2 illustrates the concentration-dependent enzyme induction. Obviously, there is a wide range from weakly efficient inducers to strong inducers, with saquinavir and resveratrol being the strongest inducers among drugs or food compounds tested. The inducing effect could be described by a Michaelis-Menten equation. As shown in Figure 3, the most potent inducers were the experimentally used polycyclic aromatic hydrocarbons and environmental pollutants such as benzo[a]pyrene, present in tobacco smoke and diesel exhausts. Other compounds in this class were 2-phenylbenzoxazole, β-naphthoflavone, and 3-methylcholanthrene. Their IM values were also in the lower range (Table 1). All HIV protease inhibitors exhibited a relatively strong induction of P450 1A1 (Figure 3), and saquinavir seemed to be the most potent inducer (Figures 2 and 3). Within this range of potency were also primaquine and resveratrol. We tested several benzimidazole derivatives and related compounds. Among all benzimidazoles, omeprazole was the most potent inducer of AhR-regulated enzymes, and the inducing potency of lansoprazole and pantoprazole was almost 5-fold lower than that of omeprazole; however, pantoprazole exhibited a lower IM value than omeprazole and lansoprazole. The anthelmintic drug

Obviously, enzyme induction is a relatively common phenomenon. Often, such effects of drugs may not be known or may only become apparent during wide clinical use as it was observed some years ago with omeprazole (14, 23). Sometimes, such inducing effects may be difficult to detect. For instance, protease inhibitors are strong enzyme inhibitors, and the reason patients receiving these drugs did not develop toxicity due to overdose of methadone was obscure. In fact, they requested higher methadone doses, and this effect may be explained by AhR-mediated enzyme induction of phase I and phase II enzymes. The chemicals tested here were selected by having some structural similarity to known enzyme inducers. Classical inducers binding to the Ah receptor are planar polycyclic compounds, but the others are rather diverse substances having only at least one planar aromatic structure in common. Some substances such as 2-phenylbenzimidazole, 2-phenylbenzthiazole, and 2-phenylbenzoxazole were selected because of their structural similarities, but even here it is difficult to explain why there is a significant increase in inducing potency in the given order. Obviously, the inducing potency increases with increasing electronegativity from nitrogen to oxygen. The final aim to resolve the structurefunction relationships in Ah receptor-mediated enzyme induction requires many studies beyond the scope of our investigation Earlier toxicological data might suggest that drugs with strong P450 1A1-inducing potency may increase cancer risk (30). Furthermore, P450 1A1 induction is an indicator of the activation of the AhR-regulated gene battery. Multiple drug interaction would result, and therefore, inducing potency of drugs should be known. As shown in Figure 3, there is a broad range from nearly no inducing potency to very strong inducing potency. The P450 1A1-inducing activity of benzo[a]pyrene, 3-methylcholanthrene, β-naphthoflavone, omeprazole, lansoprazole, and domperidone was already described (27, 3133). The other substances all cause statistically highly significant induction of P450 1A1 but with great differences in their inductive strength. These results indicate that there must be differences in the affinity of the inducer substances for the AhR and also in the ability to form a highly active transcription initiation complex (TIC). So, naphthoflavones seem to have nearly the same affinity for AhR but are leading to TICs with different activities. It was previously described (34) that R-naphthoflavone is a weak AhR agonist and P450 1A1 inducer but antagonizes the P450 1A1-inducing potency of 2,3,7,8tetrachlorodibenzodioxin. This indicates that R-naphthoflavone is a potent ligand of the AhR but forms a weakly active TIC (35) compared to β-naphthoflavone. This effect may also be present in the case of the

Prediction of Ah Receptor-Mediated Enzyme Induction

Figure 4. Comparison of half-maximal inducing concentrations (IM ( 1 order of magnitude) in vitro with the concentrations of these compounds found in humans. No concentration data were available for the other tested substances.

phenylbenzoles where 2-phenylbenzoxazole is forming the most active TIC, whereas 2-phenylbenzimidazole has nearly the same IM but leads to a TIC being much less active. The ulcer drugs omeprazole, lansoprazole, and pantoprazole also show such differences. Lansoprazole causes a 1.5-fold and omeprazole even a 5-fold higher induction than pantoprazole. The naphthoflavones, 3methylcholanthrene, and benzo[a]pyrene are known ligands of the AhR, but this is questionable for the ulcer drugs and unknown for all others (36). Obviously, there is a dose-dependent and saturable mechanism by which the tested chemical produces enzyme induction. The most likely explanation is that the chemicals or their metabolites generated in this assay caused transcriptional activation, but other mechanisms causing an increased intracellular residence time of the mRNA cannot be ruled out. Surprisingly, resveratrol, a phytoalexine (3,5,4′-trihydroxy-trans-stilbene) found in several fruits and suggested to be a useful chemopreventive agent because of its cyclooxygenase-1 inhibition and antitumorigenic activity (21), was a potent inducer of P450 1A1 mRNA transcription. The levels for primaquine and the HIV protease inhibitors were in the same range. Saquinavir was the most potent among the protease inhibitors, being 1.3-fold more potent than the others, and also had the lowest IM. As with all in vitro systems, how representative this is for the situation in humans should be discussed. Intracellular concentrations in vivo may differ from the in vitro concentrations because of differences in plasma protein binding; our assays included only 10% fetal bovine serum. Figure 4 illustrates the fact that the concentration causing half-maximal induction was within 1 order or magnitude of plasma concentrations achieved during therapy. The intracellular concentrations might be even considerably higher due to intracellular accumulation of the mostly lipophilic compounds. One might argue on the other hand that the active concentrations are lower in humans because of a higher level of plasma protein binding, but direct transfer from a cellor lipoprotein-bound state into the cells might also be possible, making the question about the free fraction less important. We tested several systems for measuring enzyme induction, including gel shift assays, reporter gene assays (37), and cell culture systems using HepG2 cells. Green fluorescence protein was used as a reporter molecule because it allows expression detection in living

Chem. Res. Toxicol., Vol. 11, No. 12, 1998 1451

cells. But the fluorescence signal of the system was difficult to quantify (38). This might also be due to differences in the copy number of plasmids per transfected cell which might be a problem in all reporter gene assays. Thus far, the Hela cell system was most stable and reproducible over time. Of course, these cells do not resemble very well the hepatocytes; however, the molecular structures of the human AhR and many transcription factors involved in that type of enzyme induction are the same. The quantitative extent of induction varies in different tissues and in the different coregulated enzymes. This situation is similar with in vitro test systems for metabolic inhibition; namely, if a strong effect was exhibited, it should be evaluated for clinical relevance in studies in volunteers or patients. Thus, the Hela system may be used in drug development to initially identify inducing drugs; in all strongly positive reactions, further clinical studies are advisable. It would be surprising if the inducing potency of drugs like primaquine, saquinavir, or resveratrol would not be reflected in the metabolism or pharmacokinetics of certain drugs in humans, especially since the IM values were in the range expected in humans.

Acknowledgment. This work was supported by Grants Fo2.1-68502-130 of the German Institute of Drugs and Medical Devices and 01EC9408 of the German Ministry of Education, Science, Research and Technology.

References (1) Ha, H. R., Chen, J., Leuenberger, P. M., Freiburghaus, A. U., and Follath, F. (1995) In vitro inhibition of midazolam and quinidine metabolism by flavonoids. Eur. J. Clin. Pharmacol. 48, 367-71. (2) Rost, K. L., and Roots, I. (1994) Accelerated caffeine metabolism after omeprazole treatment is indicated by urinary metabolite ratios: coincidence with plasma clearance and breath test. Clin. Pharmacol. Ther. 55, 402-11. (3) Swanson, H. I., and Bradfield, C. A. (1993) The AH-receptor: genetics, structure and function. Pharmacogenetics 3, 213-30. (4) Chaloupka, K., Harper, N., Krishnan, V., Santostefano, M., Rodriguez, L. V., and Safe, S. (1993) Synergistic activity of polynuclear aromatic hydrocarbon mixtures as aryl hydrocarbon (Ah) receptor agonists. Chem.-Biol. Interact. 89, 141-58. (5) Catteau, A., Douriez, E., Beaune, P., Poisson, N., Bonaiti Pellie, C., and Laurent, P. (1995) Genetic polymorphism of induction of CYP1A1 (EROD) activity. Pharmacogenetics 5, 110-9. (6) McManus, M. E., Burgess, W. M., Veronese, M. E., Huggett, A., Quattrochi, L. C., and Tukey, R. H. (1990) Metabolism of 2-acetylaminofluorene and benzo[a]pyrene and activation of foodderived heterocyclic amine mutagens by human cytochromes P-450. Cancer Res. 50, 3367-76. (7) Lusska, A., Wu, L., and Whitlock, J. P., Jr. (1992) Superinduction of CYP1A1 transcription by cycloheximide. Role of the DNA binding site for the liganded Ah receptor. J. Biol. Chem. 267, 15146-51. (8) Quattrochi, L. C., Vu, T., and Tukey, R. H. (1994) The human CYP1A2 gene and induction by 3-methylcholanthrene. A region of DNA that supports AH-receptor binding and promoter-specific induction. J. Biol. Chem. 269, 6949-54. (9) Alexander, D. L., Eltom, S. E., and Jefcoate, C. R. (1997) Ah receptor regulation of CYP1B1 expression in primary mouse embryo-derived cells. Cancer Res. 57, 4498-506. (10) Vasiliou, V., Theurer, M. J., Puga, A., Reuter, S. F., and Nebert, D. W. (1994) Mouse dioxin-inducible NAD(P)H: menadione oxidoreductase: NMO1 cDNA sequence and genetic differences in mRNA levels. Pharmacogenetics 4, 341-8. (11) Vasiliou, V., Kozak, C. A., Lindahl, R., and Nebert, D. W. (1996) Mouse microsomal Class 3 aldehyde dehydrogenase: AHD3 cDNA sequence, inducibility by dioxin and clofibrate, and genetic mapping. DNA Cell Biol. 15, 235-45. (12) Vasiliou, V., Shertzer, H. G., Liu, R. M., Sainsbury, M., and Nebert, D. W. (1995) Response of [Ah] battery genes to compounds that protect against menadione toxicity. Biochem. Pharmacol. 50, 1885-91.

1452 Chem. Res. Toxicol., Vol. 11, No. 12, 1998 (13) Schrenk, D., Stuven, T., Gohl, G., Viebahn, R., and Bock, K. W. (1995) Induction of CYP1A and glutathione S-transferase activities by 2,3,7,8-tetrachlorodibenzo-p-dioxin in human hepatocyte cultures. Carcinogenesis 16, 943-6. (14) Rost, K. L., Brosicke, H., Brockmo¨ller, J., Scheffler, M., Helge, H., and Roots, I. (1992) Increase of cytochrome P450IA2 activity by omeprazole: evidence by the 13C-[N-3-methyl]-caffeine breath test in poor and extensive metabolizers of S-mephenytoin. Clin. Pharmacol. Ther. 52, 170-80. (15) Butler, M. A., Iwasaki, M., Guengerich, F. P., and Kadlubar, F. F. (1989) Human cytochrome P-450PA (P-450IA2), the phenacetin O-deethylase, is primarily responsible for the hepatic 3-demethylation of caffeine and N-oxidation of carcinogenic arylamines. Proc. Natl. Acad. Sci. U.S.A. 86, 7696-700. (16) Delaporte, E., and Wilkinson, G. R. (1998) Modulation of cytochrome P450 enzyme activity by resveratrol. Clin. Pharmacol. Ther. 63, 149. (17) Rost, K. L., Brosicke, H., Heinemeyer, G., and Roots, I. (1994) Specific and dose-dependent enzyme induction by omeprazole in human beings. Hepatology 20, 1204-12. (18) McDonnell, W. M., Scheiman, J. M., and Traber, P. G. (1992) Induction of cytochrome P450IA genes (CYP1A) by omeprazole in the human alimentary tract [see comments]. Gastroenterology 103, 1509-16. (19) Conney, A. H. (1982) Induction of microsomal enzymes by foreign chemicals and carcinogenesis by polycyclic aromatic hydrocarbons: G. H. A. Clowes Memorial Lecture. Cancer Res. 42, 4875917. (20) Gasiewicz, T. A., and Rucci, G. (1991) R-Naphthoflavone acts as an antagonist of 2,3,7,8-tetrachlorodibenzo-p-dioxin by forming an inactive complex with the Ah receptor. Mol. Pharmacol. 40, 607-12. (21) Jang, M., Cai, L., Udeani, G. O., Slowing, K. V., Thomas, C. F., Beecher, C. W., Fong, H. H., Farnsworth, N. R., Kinghorn, A. D., Mehta, R. G., Moon, R. C., and Pezzuto, J. M. (1997) Cancer chemopreventive activity of resveratrol, a natural product derived from grapes. Science 275, 218-20. (22) Zhao, X. J., and Ishizaki, T. (1997) The in vitro hepatic metabolism of quinine in mice, rats and dogs: comparison with human liver microsomes. J. Pharmacol. Exp. Ther. 283, 1168-76. (23) Diaz, D., Fabre, I., Daujat, M., Saint Aubert, B., Bories, P., Michel, H., and Maurel, P. (1990) Omeprazole is an aryl hydrocarbonlike inducer of human hepatic cytochrome P450. Gastroenterology 99, 737-47. (24) Masubuchi, N., Hakusui, H., and Okazaki, O. (1997) Effects of pantoprazole on xenobiotic metabolizing enzymes in rat liver microsomes: a comparison with other proton pump inhibitors. Drug Metab. Dispos. 25, 584-9. (25) Chomczynski, P., and Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156-9. (26) Gilliland, G., Perrin, S., Blanchard, K., and Bunn, H. F. (1990) Analysis of cytokine mRNA and DNA: detection and quantitation by competitive polymerase chain reaction. Proc. Natl. Acad. Sci. U.S.A. 87, 2725-9. (27) Fro¨tschl, R., Kleeberg, U., Hildebrandt, A. G., Roots, I., and Brockmo¨ller, J. (1996) Polymerase chain reaction (PCR)-based construction of a competitor target for quantitative reverse transcriptase PCR measurement of cytochrome P450 1A1 mRNA. Anal. Biochem. 242, 280-2. (28) Jaiswal, A. K., Gonzalez, F. J., and Nebert, D. W. (1985) Human P1-450 gene sequence and correlation of mRNA with genetic differences in benzo[a]pyrene metabolism. Nucleic Acids Res. 13, 4503-20. (29) Pfeffer, U., Fecarotta, E., and Vidali, G. (1995) Efficient one-tube RT-PCR amplification of rare transcripts using short sequencespecific reverse transcription primers. BioTechniques 18, 2046. (30) Kellermann, G., Shaw, C. R., and Luyten Kellerman, M. (1973) Aryl hydrocarbon hydroxylase inducibility and bronchogenic carcinoma. N. Engl. J. Med. 289, 934-7.

Fro¨ tschl et al. (31) Nakama, A., Kuroda, K., and Yamada, A. (1995) Induction of cytochrome P450-dependent monooxygenase in serum-free cultured Hep G2 cells. Biochem. Pharmacol. 50, 1407-12. (32) Daujat, M., Charrasse, S., Fabre, I., Lesca, P., Jounaidi, Y., Larroque, C., Poellinger, L., and Maurel, P. (1996) Induction of CYP1A1 gene by benzimidazole derivatives during Caco-2 cell differentiation. Evidence for an aryl-hydrocarbon receptor-mediated mechanism. Eur. J. Biochem. 237, 642-52. (33) Krusekopf, S., Kleeberg, U., Hildebrandt, A. G., and Ruckpaul, K. (1997) Effects of benzimidazole derivatives on cytochrome P450 1A1 expression in a human hepatoma cell line. Xenobiotica 27, 1-9. (34) Santostefano, M., Merchant, M., Arellano, L., Morrison, V., Denison, M. S., and Safe, S. (1993) R-Naphthoflavone-induced Cyp1a1 gene expression and cytosolic aryl hydrocarbon receptor transformation. Mol. Pharmacol. 43, 200-6. (35) Merchant, M., Morrison, V., Santostefano, M., and Safe, S. (1992) Mechanism of action of aryl hydrocarbon receptor antagonists: inhibition of 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced CYP1A1 gene expression. Arch. Biochem. Biophys. 298, 389-94. (36) Daujat, M., Peryt, B., Lesca, P., Fourtanier, G., Domergue, J., and Maurel, P. (1992) Omeprazole, an inducer of human CYP1A1 and 1A2, is not a ligand for the Ah receptor. Biochem. Biophys. Res. Commun. 188, 820-5. (37) Brockmo¨ller, J., Fro¨tschl, R., Kleeberg, U., Hildebrandt, A. G., and Roots, I. (1997) Reporter gene assays as a tool to assess the enzyme inducing potency of drugs. Eur. Clin. Pharmacol. 52, A186. (38) Misteli, T., and Spector, D. L. (1997) Applications of the green fluorescent protein in cell biology and biotechnology. Nat. Biotechnol. 15, 961-4. (39) Stearns, R. A., Chakravarty, P. K., Chen, R., and Chiu, S. H. (1995) Biotransformation of losartan to its active carboxylic acid metabolite in human liver microsomes. Role of cytochrome P4502C and 3A subfamily members. Drug Metab. Dispos. 23, 207-15. (40) Watts, S. W., and Cohen, M. L. (1993) Further evidence that the guinea pig tracheal contractile serotonergic receptor is a 5-hydroxytryptamine 2 receptor: use of 5-methyltryptamine and dipropyl-5-carboxamidotryptamine. J. Pharmacol. Exp. Ther. 264, 271-5. (41) Tellez, S., Colpaert, F., and Marien, M. (1995) The R-2-adrenoceptor antagonist, (+)-efaroxan, enhances acetylcholine release in the rat cortex in vivo. Eur. J. Pharmacol. 277, 113-6. (42) Goin, C. J., and Mayer, V. W. (1995) Induction of chromosome loss in Saccharomyces cerevisiae strain D61.M by selected benzimidazole compounds. Mutat. Res. 343, 185-99. (43) Donato, M. T., Castell, J. V., and Gomez Lechon, M. J. (1995) Effect of model inducers on cytochrome P450 activities of human hepatocytes in primary culture. Drug Metab. Dispos. 23, 553-8. (44) Maurice, M., Pichard, L., Daujat, M., Fabre, I., Joyeux, H., Domergue, J., and Maurel, P. (1992) Effects of imidazole derivatives on cytochromes P450 from human hepatocytes in primary culture. FASEB J. 6, 752-8. (45) Kim, N. D., Yoo, J. K., Won, S. M., Park, S. S., and Gelboin, H. V. (1993) Phenytoin induction of cytochrome P4502B in mice: effects on hexobarbital hydroxylase activity. Xenobiotica 23, 21725. (46) Van Cleef, G. F., Fisher, E. J., and Polk, R. E. (1997) Drug interaction potential with inhibitors of HIV protease. Pharmacotherapy 17, 774-8. (47) Patick, A. K., Boritzki, T. J., and Bloom, L. A. (1997) Activities of the human immunodeficiency virus type 1 (HIV-1) protease inhibitor nelfinavir mesylate in combination with reverse transcriptase and protease inhibitors against acute HIV-1 infection in vitro. Antimicrob. Agents Chemother. 41, 2159-64.

TX980164H