Genotoxic effects of heterocyclic aromatic amines in human derived ...

Mutagenesis vol.14 no.6 pp.533–539, 1999

Genotoxic effects of heterocyclic aromatic amines in human derived hepatoma (HepG2) cells

Siegfried Knasmu¨ller1,6,*, Christina E. Schwab1,*, Susan J.Land2, Ching Y.Wang3, Ratna Sanyal1, Michael Kundi4, Wolfram Parzefall1 and Firouz Darroudi5 1Institute

of Cancer Research, University of Vienna, Borschkegasse 8a, A-1090 Vienna, Austria, 2Molecular and Chemical Carcinogenesis Program, Karmanos Cancer Institute, Detroit, MI, USA, 3Department of Urology, SUNY Health Science Center, Syracuse, NY, USA, 4Institute for Environmental Hygiene, University of Vienna, Vienna, Austria and 5MGC, Department of Radiation Genetics and Chemical Mutagenesis, State University of Leiden, Medical Centre, Leiden, The Netherlands

In order to study the mutagenic effects of heterocyclic aromatic amines (HAAs) in cells of human origin, five compounds, namely 2-amino-3-methyl-imidazo[4,5-f]quinoline (IQ), 2-amino-3,4-dimethyl-imidazo[4,5-f]quinoline (MeIQ), 2-amino-3,8-dimethyl-imidazo[4,5-f]quinoxaline (MeIQx), the pyridoimidazo derivative 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) and 3-amino-1,4-dimethyl5H-pyrido[4,3-b]indole (Trp-P-1), were tested in micronucleus (MN) assays with a human derived hepatoma (HepG2) cell line. All HAAs caused significant, dosedependent effects. The activities of IQ, MeIQ, MeIQx and PhIP were similar (lowest effective concentrations 25– 50 µM), whereas Trp-P-1 was effective at a dose of ù2.1 µM. In addition, the HAAs were tested in MN assays with Chinese hamster ovary (CHO) cells and in Salmonella strain YG1024 using HepG2 cell homogenates as an activation mix. In the CHO experiments, positive results were obtained with Trp-P-1 and PhIP, whereas the other compounds were devoid of activity under all experimental conditions. The discrepancy in the responsivity of the two cell lines is probably due to differences in their acetylation capacity: enzyme measurements with 2-aminofluorene as a substrate revealed that the cytosolic acetyltransferase activity in the HepG2 cells is ~40-fold higher than that of the CHO cells. In the bacterial assays all five HAAs gave positive results but the ranking order was completely different from that seen in the HepG2/MN experiments (IQ . MeIQ . Trp-P-1 ù MeIQx .. PhIP) and the mutagenic potencies of the various compounds varied over several orders of magnitude. The order obtained in bacterial tests with rat liver S9 mix was more or less identical to that seen in the tests with HepG2 cell homogenates but the concentrations of the amines required to give positive results were in general substantially lower (10–5– 10–1 µM). Overall, the results of the present study indicate that MN/HepG2 tests might reflect the mutagenic effects of HAAs more adequately than other in vitro mammalian cell systems due to the presence of enzymes involved in the metabolic conversion of the amines.

Introduction Heterocyclic aromatic amines (HAAs) were detected in the 1970s by Sugimura and co-workers in pyrolysates of amino acids of protein-rich foods (Sugimura et al., 1977). In recent years, considerable effort has been made to characterize, identify and quantitate these compounds in human foodstuffs and to elucidate their potential health risks in humans ¨ vervik and Gustavsson, 1990; (Frederick et al., 1988; O IARC, 1993). HAAs are mutagenic in pro- and eukaryotic organisms and experiments with laboratory rodents also showed that these food-derived compounds cause cancer. Furthermore, epidemiological studies indicated that they may be involved in the etiology of colon cancer in humans (IARC, 1993; Hatch et al., 1992; Felton et al., 1997). HAAs have been tested extensively in microbial in vitro assays and are potent mutagens in these test procedures (De Meester, 1989), however, the results obtained in genotoxicity tests with mammalian cell lines are highly divergent (see IARC, 1993). The primary aim of the present study was the investigation of the mutagenic effects of five different HAAs in human derived cells. So far, only a few studies have been carried out with HAAs in human cells and in most of these experiments the genotoxic potential of the amines was assessed by DNA adduct measurements (Totsuka et al., 1996; Turteltaub et al., 1997; Wakabayashi et al., 1997). The human HepG2 cell line we used was isolated originally from a primary human liver tumor biopsy from an 11-year-old boy and has retained the activities of a variety of xenobiotic metabolizing enzymes (for a review see Knasmu¨ller et al., 1998). It has been shown that representatives of different classes of promutagens give positive results in mutagenicity tests with HepG2 cells and evidence is accumulating that the predictive value of assays with HepG2 cells is higher than that of conventional assays with cells which require addition of exogenous cell homogenates for activation of HAAs and other classes of promutagens (Knasmu¨ller et al., 1998, 1999). In the present study, we used the induction of micronuclei (MN) as an endpoint. Darroudi and Natarajan (1993) have recently described a protocol for the determination of MN in HepG2 cells and the mechanisms which lead to the formation of MN are well understood (Natarajan and Obe, 1982). In order to determine to what extent the genotoxic effects and the order of genotoxic potencies of the different HAAs depends on the metabolic activation system and on the type of indicator cells used, additional experimental series were carried out in which induction of MN and/or gene mutations by HAAs was measured in Chinese hamster ovary (CHO) cells and in bacterial indicators (Salmonella typhimurium YG1024). The results obtained in the latter test system upon activation of HAAs with HepG2 derived cell homogenate were compared with those found under identical conditions with rat

whom correspondence should be addressed. Tel: 143 1 4277 65142; Fax: 143 1 4277 9651; Email: [email protected] *The first two authors contributed equally

6To

© UK Environmental Mutagen Society/Oxford University Press 1999

533

S.Knasmu¨ller et al.

liver S9 mix, which is used in routine mutagenicity tests. In order to explain the discrepancies seen in the MN assays with HepG2 and CHO cells, acetyltransferase activities were measured comparatively in both cell types. On the basis of the results obtained in the various experimental series and in previous studies with rodents, attempts were made to draw conclusions on the usefulness of the HepG2/ MN assay for the investigation of the genotoxic properties of HAAs. Materials and methods Chemicals The test compounds 2-amino-3-methylimidazo[4,5-f]quinoline (IQ), 2-amino3,4-dimethyl-imidazo[4,5-f]quinoline (MeIQ), 2-amino-3,8-dimethyl-imidazo[4,5-f]quinoxaline (MeIQx), 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) and 3-amino-1,4-dimethyl-5H-pyrido[4,3-b]indole (Trp-P-1) were purchased from Toronto Research Chemicals (Toronto, Canada). Cyclophosphamide (CP), benzo[a]pyrene (B[a]P) and aflatoxin B1 (AFB1) were purchased from Sigma (St Louis, MO). For the enzyme measurements, paraoxon (Sigma), 2-aminofluorene (2-AF) (Aldrich Chemical Co., Milwaukee, MN) and [3H]acetyl coemzyme A ([3H]AcCoA; ICN Radiochemicals, Costa Mesa, CA) were used. Cells and culture conditions The human HepG2 cell line was originally established from a human liver tumor biopsy (Aden et al., 1979) and was obtained from G. Dalner (Stockholm, Sweden). The cells were grown in Dulbecco’s minimal essential medium (DMEM; Boehringer Mannheim, Mannheim, Germany), supplemented with 15% fetal calf serum (Gibco BRL Life Technologies, Paisley, UK) and antibiotics (penicillin, 100 U/ml, streptomycin 0.1 mg/ml; Sigma). CHO wild-type cells (CHO 9) were cultured in Ham’s F10 medium supplemented with 15% newborn calf serum (both from Gibco BRL Life Technologies) and antibiotics (penicillin, 100 U/ml, streptomycin 0.1 mg/ ml; Sigma). For the bacterial tests, the O-acetyltransferase-overproducing S.typhimurium strain YG1024 (Watanabe et al., 1990), as well as the parent strain TA98, were used. Strain YG1024 was a gift from T. Nohmi (National Institute of Hygienic Science, Division of Mutagenesis, Tokyo, Japan). Strain TA98 was obtained from B.N. Ames (University of California, Berkeley, CA). The cells were kept on master plates which were supplemented with antibiotics. Preparation of cell homogenates Two different types of HepG2 cell homogenates were used. For MN assays with CHO cells, the activation mix was composed according to the protocol of Darroudi and Natarajan (1993). Briefly, fresh HepG2 cells were trypsinized, centrifuged (800 g, 7 min) and transferred to Tris–sucrose buffer (53106 cells/ml), which was prepared according to Darroudi and Natarajan (1993). Subsequently, the suspension was sonicated (Sonifier 250; Branson, Danbury, CT) and centrifuged for 20 min at 4°C (13 200 g). The supernatant was stored in 200 µl portions at –70°C. Immediately before the start of the mutagenicity assays, cofactor solutions were added which were composed as described earlier (Darroudi and Natarajan, 1993). In the bacterial assays, a more concentrated cell homogenate was used, which was composed according to Duverger-van Bogaert et al. (1993, 1995). The cells were gently centrifuged. Subsequently, the volume of the cell pellet was measured and a 10-fold volume of sucrose–imidazole buffer (0.25 M, pH 7.4, 85.3 g/l sucrose, 3 ml 1 M imidazole) added. The cell pellet was homogenized and centrifuged for 10 min at 9000 g. The supernatants were stored in small aliquots and kept at –70°C. In order to compare the activation capacities of the HepG2 cell homogenates with that of rat S9 (used in standard tests), commercial rat liver S9 homogenate from Aroclor 1254-induced animals was purchased from ICN Biochemicals (Erscheule, Germany). The protein concentrations of the cell homogenates were determined by the method of Lowry et al. (1951) and, for reasons of comparison, the homogenates were diluted with 1.5 M KCl to establish identical protein concentrations. MN assay with HepG2 cells The assay was performed essentially as described by Natarajan and Darroudi (1991). Briefly, the cells were grown in culture flasks (25 cm2) in 7 ml complete medium, then they were treated with solutions of the test compounds for 1 h. Subsequently, the cells were washed twice with phosphate-buffered saline (PBS) free of Mg21 and Ca21 and cultivated for 28 h in complete growth medium supplemented with cytochalasin B1 (final concentration 3 µg/ml). Subsequently, the cells were trypsinized and treated with cold

534

hypotonic KCl (5.6 g/l) and air dried preparations were made which were stained with 2% aqueous Giemsa solution (Gurr R 66). For each experimental point, 1000 binucleated cells (BN) were evaluated. The survival rates were determined with tetrazolium salt according to Mosman (1983). Two independent experiments were carried out with the individual test compounds and in each experiment solvent controls and positive controls were included. MN assay with CHO cells The tests were performed as described earlier by Darroudi and Natarajan (1993). Cells in exponential growth phase were trypsinized and centrifuged. Subsequently the pellet (~13106 cells) was resuspended in 0.2 ml F10 medium. The HAAs and B[a]P were dissolved in sterile DMSO; subsequently these solutions were diluted 1:1 with distilled water to reach the final concentrations. The incubation mixtures consisted of 0.2 ml of the cell suspension, 0.7 ml of HepG2 S13 cell homogenate (see above) and 0.1 ml of different concentrations of the stock solutions of the HAAs. The mixtures were incubated for 2 h at 37°C with shaking. Subsequently, the cells were washed in PBS and seeded into culture flasks. To obtain BN, cytochalasin B1 (3 µg/ml) was added to the growth medium immediately after treatment. After 20 h the cells were trypsinized, treated with hypotonic KCl (5.6 g/l) and fixed in acetic acid:methanol (1:3). For each experimental point, 1000 BNC were evaluated. The survival rate was determined with tetrazolium salt according to Mosman (1983). With each test compound two independent experiments were performed and in each solvent controls and positive controls (CP) were included. To exclude the possibility that the DMSO concentration had a cytotoxic effect untreated cultures were additionally prepared in the first experimental series and viability and MN formation in untreated and DMSOexposed cells were compared. The results are described in the legend to Table II. It can be seen that the solvent did not affect the viability of the cells and had no impact on MN formation. Bacterial assay The bacterial mutagenicity assays were carried out essentially as described earlier by Maron and Ames (1984) as liquid holding tests. Stationary phase overnight cultures of strain YG1024 (0.1 ml, ~1–23109 cells/ml) were incubated for 30 min with 0.2 ml HepG2 cell homogenates or aliquot amounts of rat S9 mix (final protein concentration 1 mg/ml) and 0.01 ml of different concentrations of the various test compounds. Subsequently, the incubation mixtures were plated with 2 ml top agar on selective agar plates. After 2 days, the His1 revertants were counted. For each experimental point, three plates were evaluated. In addition, identical experimental series were carried out with the parental strain TA98. In all experimental series, AFB1 was used as a positive control. Determination of N-acetyltransferase activities in CHO and HepG2 cells The enzyme fractions were prepared as described by Glowinski et al. (1983). All steps were carried out at 4°C. HepG2 cells which had been cultivated on plates were washed several times with PBS, scraped gently from the surface and centrifuged. Approximately 23109 cells were resuspended in a 4-fold volume of sodium pyrophosphate buffer (50 mM, pH 7.0, supplemented with 1 mM dithiothreitol) and sonicated for 5315 s (Sonicator W-375; Ultrasonics Inc.). Homogenates were centrifuged (20 min, 9000 g, 4°C), followed by centrifugation of the supernatant at 105 000 g for 60 min to separate the cytosolic supernatant from the microsomal pellet. For preparation of the microsomal fractions, the pellets were washed twice and finally resuspended in an equivalent amount of buffer and homogenized (Sonicator W-375). Cytosols and microsomes were used immediately for the enzyme assays. As positive controls, enzyme fractions were prepared from rat liver homogenates. In order to establish identical experimental conditions, the protein concentrations of the various fractions were determined (Bio-Rad, Hemel Hempstead, UK) and identical protein concentrations established by appropriate dilution. N-acetyltransferases activities were determined according to Land et al. (1989). The enzyme measurements were carried out in the presence and absence of paraoxon to distinguish between cytosolic and microsomal acetylation enzymes (Glowinski et al., 1983). The reaction mix consisted of 12.5 nmol 2-AF in 5 µl DMSO, various concentrations of cytosolic or microsomal enzymes and 50 mM sodium pyrophosphate buffer for a total volume of 200 µl. The mixtures were preincubated for 3 min at 37°C before the reaction was started by addition of 20 nmol [3H]AcCoA in 50 µl sodium pyrophosphate buffer (sp. act. 4–7 mCi/mmol). The radiolabeled product was extracted into 1.5 ml chloroform, washed twice with water, evaporated to dryness and counted in 3 ml of a toluene-based scintillator. All samples were assayed at six to eight different enzyme concentrations. The activities were derived from an average of those concentrations that showed a linear response.

Results MN assays with HepG2 cells The results of the MN assay with HepG2 cells are summarized in Table I. All compounds gave clear positive results. The

Genotoxic effects of heterocyclic aromatic amines

Table I. MN induction by HAAsa in HepG2 cells Compound PhIP

Positive control Trp-P-1

Positive control IQ

Positive control MeIQ

Positive control MeIQx

Positive control

Table II. MN induction in CHO cells by HAAs in the presence of HepG2 cell homogenate

Concentration MN/103 (µM) BNCb

Survival MN/103 (%)c BNCb

Survival (%)c

0.0 25 50 100 300 600 900 900 0.0 0.2 0.7 2.1 6.3 12.6 25.2 900 0.0 25 50 100 300 600 900 900 0.0 25 50 100 300 600 900 900 0.0 25 50 100 300 600 900 900

100 106 105 97 94 70 82 n.d. 100 93 106 93 97 52 10 n.d. 100 79 105 93 84 45 52 n.d. 100 91 103 84 74 75 63 n.d. 100 93 83 85 74 74 82 n.d.

100 96 101 105 82 80 65 n.d. 100 101 86 94 72 65 35 n.d. 100 91 98 101 90 65 71 n.d. 100 110 91 92 68 75 71 n.d. 100 82 94 101 81 85 80 n.d.

21 38 49 50 57 64 37 73 21 22 22 26 34 40 Toxic 73 21 29 37 46 48 50 58 73 21 33 42 51 56 62 65 73 21 41 44 48 49 59 59 73

18 34 31 45 44 52 57 81 18 28 25 32 29 38 (42) 81 22 21 32 38 49 47 60 75 22 29 28 52 60 59 61 75 22 24 39 38 50 61 59 75

aThe

HAAs as well as B[a]P were dissolved in DMSO. The cells were exposed to the mutagens for 1 h. In all experiments, solvent controls were included and B[a]P (900 µM) was used as a common positive control b1000 binucleated cells (BNC) were scored per experimental point. cCell survival was determined according to Mosmann (1983).

dose–response effects obtained with the various amines were similar and the potencies of PhIP, IQ, MeIQx and MeIQ were more or less identical, whereas Trp-P-1 caused a significant effect at concentrations ù2 µM. The other compounds were also tested in preliminary range-finding experiments at concentrations below 25 µM (10 and 5 µM), but no positive results were obtained under these conditions (data not shown). Cell survival was much more affected by Trp-P-1 than by the other compounds and reduced to 50% at a concentration of 12.6 µM (Table I). B[a]P was tested in all experiments under identical experimental conditions as IQ, MeIQ and MeIQx. The numbers of MN formed after exposure of the cells to 900 µM B[a]P are listed in Table I and it can be seen that its mutagenic activity was similar to that of the quinolines and quinoxalines. MN assays with CHO cells The results obtained in MN assays with CHO cells are summarized in Table II. With Trp-P-1 and PhIP, clear positive results were measured, whereas IQ, MeIQ and MeIQx did not induce MN even when tested at relatively high concentrations

Test compounda

Concentration Experiment 1 Experiment 2 (µM) MN/103 Survival MN/103 Survival (%)c BNCb (%)c BNCb

PhIP

0.0 134 268 401 5000 0.0 11 22 33 5000 0.0 2350 4700 5000 0.0 2350 4700 5000 0.0 2350 4700 5000


Positive control IQ Positive control MeIQ Positive control MeIQx Positive control

10 38 171 188 179 10 76 78 48 179 18 24 24 177 18 27 21 177 18 25 22 177

100 87 43 17 67 100 86 54 9 67 100 96 83 76 100 94 87 76 100 86 81 76

14 29 150 169 157 15 66 62 61 186 17 14 19 191 17 17 22 191 17 19 20 191

100 89 51 25 73 100 101 86 9 89 100 104 98 82 100 99 101 82 100 82 84 82

aThe HAAs as well as CP were dissolved in DMSO. In the first experiment, the effect of DMSO (which was used as a solvent in all experiments) on MN formation and survival was investigated. Exposure of the cells to the solvent (5% in the medium for 120 min) had no effect. In untreated cultures, the MN frequency was 14 MN/1000 BNC and in DMSO-exposed cells 15 MN/1000 BNC. The survival rate in the untreated cells was 100% and in the DMSO-treated cells 101%. The cells were exposed to the mutagens for 2 h. CP (5000 µM) was used as a positive control in all experiments. b1000 binucleated cells were scored per experimental point. cCell survival was determined according to Mosmann (1983).

(ù4 mM). Trp-P-1 was clearly more active than PhIP and caused a 7-fold enhancement over the spontaneous frequency at 11 µM. Both compounds affected cell survival. Again TrpP-1 was more effective than PhIP in this regard. Bacterial mutagenicity tests In preliminary experiments with strain TA98 only marginal effects were induced by the HAAs upon activation with HepG2 cell homogenate (data not shown). Therefore, the derivative YG1024 was used in all further experiments. This strain has enhanced O-acetyltransferase activity and is highly sensitive to amines (Watanabe et al., 1990). In order to compare the results of the bacterial mutagenicity assays with those generated with mammalian cells, the tests were carried out as preincubation assays. With all amines, a significant, dose-dependent increase in His1 mutant numbers was observed. It can be seen in Table III that the mutagenic potencies of the different compounds vary over several orders of magnitude: the quinolines IQ and MeIQ were the most potent mutagens. With both amines, positive effects were seen at 0.003 µM, followed by Trp-P-1 and MeIQx. PhIP was the weakest mutagen. The concentrations required to cause a significant increase over the spontaneous background values were ~3310–9 M in the case of MeIQ and IQ, 1310–7 M for Trp-P-1, 5310–5 M for MeIQx and 4310–4 M for PhIP. When Aroclor 1254-induced rat S9 mix was used under identical conditions, more pronounced effects were seen with all compounds (Table IV). The concen535


Table III. Induction of His1revertants in Salmonella strain YG1024 upon activation with HepG2 cell homogenatea

Table IV. Induction of His1 revertants in Salmonella strain YG1024 upon activation with liver S9 mix from Aroclor-induced ratsa

Test compound

Concentration (µM)

His1 revertants/plateb (mean 6 SD)c

Test compound

Concentration (µM)

His1 revertants/plateb (mean 6 SD)c

PhIP

0 448 900 1797 10 µg/plate 0 0.15 0.30 0.60 1.20 10 µg/plate 0 0.0031 0.0063 0.0127 0.0254 10 µg/plate 0 0.0031 0.0063 0.0127 0.0254 10 µg/plate 0 58 117 233 468 10 µg/plate

30 6 2.3 46 6 4.5* 59 6 1.0* 66 6 6.5* 366 6 48.7* 30 6 2.3 119 6 15.3* 178 6 1.5* 231 6 7.8* 377 6 32.5* 366 6 48.7* 30 6 2.3 78 6 5.6* 97 6 4.5* 131 6 4.7* 263 6 26.1* 366 6 48.7* 30 6 2.3 61 6 2.5* 74 6 6.4* 121 6 7.0* 200 6 2.6* 366 6 48.7* 30 6 2.3 98 6 3.0* 185 6 21.3* 239 6 15.6* 467 6 39.5* 366 6 48.7*

PhIP

0 0.22 0.44 0.89 1.79 3.59 7.19 14.38 4 µg/plate 0.0000 0.0012 0.0022 0.0046 0.0092 0.0184 0.0368 0.0737 4 µg/plate 0.000000 0.000012 0.000025 0.000049 0.000099 0.000198 0.000397 0.000794 0.001587 4 µg/plate 0.00000 0.00023 0.00047 0.00094 0.00190 0.00377 0.00758 0.01519 4 µg/plate 0.00000 0.00164 0.00329 0.00654 0.01309 0.02622 0.05251 0.10483 4 µg/plate

17 6 2.0 23 6 4.0 29 6 2.3* 34 6 4.0* 50 6 5.5* 68 6 6.6* 134 6 36.5* 196 6 58.9* 471 6 46.7* 17 6 2.0 36 6 3.7* 39 6 7.6* 65 6 3.0* 183 6 41.0* 205 6 45.6* 514 6 46.0* 994 6 220.0* 471 6 46.7* 17 6 2.0 41 6 8.0* 38 6 2.0* 65 6 10.2* 55 6 13.6* 92 6 17.9* 190 6 60.8* 322 6 96.5* 548 6 140.4* 471 6 46.7* 17 6 2.0 41 6 11* 43 6 3.6* 70 6 5.8* 145 6 35.5* 244 6 33.0* 348 6 62.4* 881 6 146.0* 471 6 46.7* 17 6 2.0 36 6 4.0* 45 6 6.8* 115 6 12.0* 148 6 39.0* 309 6 63.6* 819 6 143.0* 870 6 113.,1* 471 6 46.7*


Positive control IQ



Positive control

aVarious concentrations of the test compounds (dissolved in DMSO) were preincubated with 0.1 ml of an overnight culture of strain YG1024 (1– 23108 cells) and 0.5 ml activation mix prepared from HepG2 cells (see Materials and Methods). The final protein concentration in the incubation mixtures was 1.0 mg protein/ml. The mixtures were incubated for 30 min at 37°C and subsequently plated on histidine-free selective agar plates and the number of His1 revertant colonies evaluated after 2 days. Aflatoxin B1 was used as a positive control. bThree plates were counted per experimental point. cAsterisks indicate statistical significance (Dunnett’s test, P ø 0.05)

tration of MeIQ required to cause a significant increase in His1 revertants in assays with rat S9 was 13-fold lower than that which caused a clear effect with HepG2 activation. The corresponding values for Trp-P-1, IQ, PhIP and MeIQx were 125, 258, 1018 and 35 400, respectively. Determination of the N-acetyltransferase activities in HepG2 and CHO cells The most striking discrepancy seen in the present experiments was that MeIQ, MeIQx and IQ were potent genotoxins in HepG2 cells, but devoid of activity in the CHO test system. We asked whether this phenomenon might be due to differences in the metabolic activation capacities of the two cell lines. The results of a series of experiments in which N-acetyltransferase activities with 2-AF as substrate were comparatively measured in microsomal and cytosolic preparations of CHO and HepG2 cells are shown in Table V. In the microsomal fraction, the activities in the cell lines were below the detection limit. In the cytosolic fractions, the N-acetyltransferase activity in the human derived cells was ~40-fold higher compared with that found in CHO cells. The highest 2-AF N-acetyltransferase activity was measured in the cytosol from rat livers (~4-fold higher than the activity in the human HepG2 cytosol). 536


Positive control IQ



Positive control aVarious

concentrations of the test compounds (dissolved in DMSO) were preincubated with 0.1 ml of an overnight culture of strain YG1024 (1– 23108 cells) and 0.5 ml activation mix prepared with liver S9 mix (Maron and Ames, 1983) prepared from Aroclor 1254-induced male Fischer 344 rats. The final protein concentration in the incubation mixtures were 1.0 mg protein/ml. The mixtures were incubated for 30 min at 37°C and subsequently plated on histidine-free selective agar plates and the number of His1 revertant colonies counted after 2 days. As a positive control, aflatoxin B1 was used. bThree plates were evaluated per experimental point. cAsterisks indicate statistical significance (Dunnett’s test, P ø 0.05).

Discussion In the present investigation, the effects of various HAAs (TrpP-1, IQ, MeIQ, MeQx and PhIP) were studied in MN assays with HepG2 cells and attempts were made to compare the effects of the amines in the human derived cells with those caused in other indicator cells (CHO cells and bacteria). On the basis of the results obtained in the individual test systems and of in vivo data, conclusions were drawn on the predictive

Genotoxic effects of heterocyclic aromatic amines

Table V. Determination of N-acetyltransferase activities in cytosolic fractions of CHO cells, HepG2 cells and rat livera CHO cytosol

HepG2 cytosol

Rat cytosol

mg protein

nmol/min

mg protein

nmol/min

mg protein

nmol/min

0.000 0.231 0.461 0.553 0.600 0.646 0.692 0.738 0.784

0.000 0.000 0.003 0.004 0.005* 0.007* 0.007* 0.008* 0.009*

0.000 0.032 0.065 0.130 0.195 0.325 0.487 0.649 0.812 0.974

0.000 0.019 0.034 0.073 0.093 0.159 0.223 0.270 0.329 0.400

0.000 0.125 0.250 0.375 0.626 1.252 1.878

0.000 0.296 0.430 0.636 0.926 1.514 1.808

Average 0.010 nmol/min/mg



aThe

measurements were carried out as described in Materials and methods. Data which were used to calculate the enzyme activities are printed in bold.

value and usefulness of the MN/HepG2 system for the investigation of the mutagenic effects of the HAAs. We found clear positive results with all five HAAs in the MN/ HepG2 assays and also demonstrated that cell homogenates of HepG2 cells are capable of activating the amines so that positive results are obtained with indicator cells lacking enzymes which are required for their metabolic activation. At present, data on the effects of HAAs in human derived cell lines and primary human cells are scarce: Morgenthaler et al. (1995, 1998) tested PhIP and IQ in a human lymphoblastoid (TK6) cell line for induction of gene mutations; data from experiments with human lymphocytes are also available (Aeschbacher and Ruch, 1989), but in contrast to the present experiments with HepG2 cells, addition of rat liver homogenate was required for metabolic activation of the amines in both experimental systems. We earlier reported results from experiments in which we used the MN/HepG2 system to investigate protective effects of dietary constituents towards HAAs, but no dose–response effects and comparisons with other test systems were included in this study (Sanyal et al., 1997). Furthermore, data from a few DNA adduct measurements with primary human cells from different organs are also available (Totsuka et al., 1996; Turteltaub et al., 1997; Wakabayashi et al., 1997). The mutagenic potencies as well as the ranking orders of mutagenic activities of the various HAAs differ strongly in the individual test systems. In the HepG2/MN assay the order of activity was Trp-P-1 .. MeIQ ù PhIP ù MeIQ ù IQ. The tryptophan derivative was active at concentrations ù2 µM. The potencies of the other four amines were similar and positive effects were found at doses ù25 µM. In the CHO/ MN assay with HepG2 cell homogenates, MN induction and cytotoxic effects were only seen with Trp-P-1 and PhIP (Table II), whereas the quinolines (IQ and MeIQ) as well as MeIQx were devoid of activity under all test conditions. These findings are in agreement with earlier results obtained in CHO cells and with other mammalian cell lines in the presence of rodent derived activation mixtures: the quinolines (IQ and MeIQ) and the quinoxaline (MeIQx) caused only moderate or negative effects (Thompson et al., 1983, 1987; Holme et al., 1987; Loprieno et al., 1991; IARC, 1993), whereas Trp-P-1 and PhIP induced chromosomal aberrations and SCEs in CHO cells

(Thompson et al., 1987; Tucker et al., 1989; Otsuka et al., 1996a,b). It is well documented that acetylation reactions play a crucial role in the metabolic conversion of HAAs (for reviews ¨ vervik and Gustavsson, see Aeschbacher and Turesky, 1990; O 1990; IARC, 1993). Wild et al. (1995) demonstrated that IQ, MeIQ and MeIQx induced His1 revertants only in a S.typhimurium TA98 strain which encodes for NAT2, whereas PhIP and Trp-P-2 were positive in the acetyltransferasedeficient counterpart strain as well. Similar results were recently reported from experiments with genetically engineered repairdeficient CHO cells (Wu et al., 1997): IQ caused induction of mutations in the aprt locus as well as cytotoxicity in a line expressing NAT2 activity, whereas PhIP was also active in acetyltransferase-deficient cells. The authors postulated that pathways other than acetylation, such as sulfation, might be involved in the activation of PhIP. The results of our enzyme measurements indicate that the N-acetyltransferase activity in HepG2 cells is ~40-fold higher than in CHO cells (Table V) and similar to that found in primary human hepatocytes (Land et al., 1989). Coroneos and Sim (1993) recently genotyped the HepG2 cell line and found it to be a NAT2*5B/*6A slow acetylator. In line with the results of the latter studies (Hein et al., 1994; Otsuka et al., 1996a,b; Wu et al., 1997), it is likely that the negative results seen with quinolines and quinoxalines in CHO cells are due to their low acetyltransferase activity. In contrast, the acetyltransferase activity in HepG2 cells is apparently sufficiently high to catalyze the generation of DNA-reactive metabolites from IQ, MeIQ and MeIQx. With all five HAAs, positive results were obtained in the Salmonella mutagenicity assays upon activation with HepG2 cell homogenate (Table III), but in contrast to the results seen in the HepG2/MN tests, a totally different ranking order of mutagenic activity was seen (IQ . MeIQ . Trp-P-1 ù MeIQx .. PhIP). While in HepG2 cells the mutagenic potencies of the individual compounds were of the same order of magnitude, pronounced differences were found in the bacterial assays. For example, PhIP was approximately five orders of magnitude less active than IQ. The ranking orders of mutagenic actvities of the amines seen under identical experimental conditions with rat liver S9 mix and HepG2 homogenate were more or less identical, but the concentrations required to produce positive results were much lower. Since the cell homogenates are devoid of acetyltransferase activity (Wiebel, 1993), it is conceivable that these differences are due to activities of enzymes which catalyze the first activation step of the HAAs, namely N-hydroxylation by cytochromes CYP1A1 and CYP1A2. However, the dramatic dissimilarity in the ranking orders seen in the HepG2/MN-tests and in the bacterial assays is likely due to differences in the subsequent activation steps (e.g. differences of the substrate specificities of human and bacterial actelytransferases). Watanabe et al. (1994) compared the induction of MN by IQ in genetically engineered Chinese hamster CHL cells expressing either bacterial (Salmonella) O-acetyltransferase or human NAT1 or NAT2 acetyltransferase activity. Since a much more pronounced effect was seen in cells expressing the bacterial enzyme, they concluded that the microbial transferase has a much higher activation capacity for IQ than the human. Several attempts have been made to establish structure– activity relationships for the genotoxic effects of HAAs which are all based on data obtained in bacterial mutagenicity assays 537


(Hatch et al., 1991; Sabbioni and Wild, 1992; Zhang et al., 1993; Vikse et al., 1995). On the basis of the present findings it is apparent that the effects caused by the HAAs in bacteria differ strongly from those seen in human derived cells. For example, PhIP, which is the most abundant HAA in fried meats, was in the present as well as in earlier bacterial studies (Wakabayashi et al., 1993) more than two orders of magnitude less mutagenic than IQ and MeIQ, whereas it was equally active in the HepG2/MN assays. Data from in vivo clastogenicity studies with HAAs are scarce: MeIQ and IQ were inactive in bone marrow MN assays with mice (Wild et al., 1985; Loprieno et al., 1991); the results obtained with PhIP in this test system are conflicting (Tucker et al., 1989; IARC, 1993; Director et al., 1996); in SCE tests, no positive effects were seen in bone marrow cells of mice with the latter compound (Director et al., 1996). For MeIQx and Trp-P-1 no data from in vivo MN assays with rodents are available, but i.p. administration of Trp-P-1 to rats induced chromsomal aberrations in hepatocytes (Sawada et al., 1991). In contrast to the moderate and negative effects seen in these in vivo mutagenicity tests with HAAs, it is well established that they form DNA adducts in a variety of organs and are potent carcinogens in rodents (Ohgaki et al., 1984, 1991; Takayama et al., 1984; Ito et al., 1991; Wakabayashi et al., 1992). The findings obtained in the MN tests with HepG2 cells correlate better with the carcinogenic effects of the HAAs in rodents than those obtained in conventional in vitro mutagenicity and certain in vivo tests: (i) compounds like quinolines/quinoxalines which give negative results in bone marrow MN assays in vivo and/or in CHO cells, but are clearly positive in HepG2/MN tests are carcinogenic in laboratory animals; (ii) the genotoxic effects of the five HAAs in HepG2 cells are similar, likewise also their carcinogenic activities do not differ substantially (range of TD50 in rats 0.57–5.7, in mice 11–41 mg/kg/day; Hatch et al., 1992), whereas their mutagenic activities differ over many orders of magnitude in bacterial tests. In this context it is also notable that Trp-P-1 is a more potent carcinogen in rats than MeIQ, MeIQx, IQ and PhIP and was also the most potent genotoxin in HepG2/MN assays, in mice however the TD50 values of Trp-P-1 are 2- to 3-fold higher than those of IQ and MeIQ (Hatch et al., 1992). Overall these comparisons indicate that the MN/HepG2 assay is an appropriate approach for the investigation of the mutagenic effects of HAAs. We anticipate that this model might also be useful for metabolic activation studies and for the detection of HAA protective dietary constituents (see also Knasmu¨ller et al., 1998, 1999; Schwab et al., 1999). Acknowledgements The authors thank R. Schulte-Hermann (University of Vienna) and A.T. Natarajan (University of Leiden) for encouragement and discussions. Part of the experimental work was supported by an EC grant (to S.K.).

References Aden,D.P., Fogel,A., Plotkin,S., Damjanov,I. and Knowles,B.B. (1979) Controlled synthesis of HBsAg in a differentiated human liver carcinomaderived cell line. Nature, 282, 615–616. Aeschbacher,H.U. and Ruch,E. (1989) Effects of heterocyclic amines and beef extract on chomosome aberrations and sister chromatid exchanges in cultured human lymphocytes. Carcinogenesis, 10, 429–433. Aeschbacher,H.U. and Turesky,R.J. (1990) Mammalian cell mutagenicity and metabolism of heterocyclic aromatic amines. Mutat. Res., 259, 235–250. Coroneos,E. and Sim,E. (1993) Arylamine N-acetyltransferase activity in human cultured cell lines. Biochem. J., 294, 481–486.

538

Darroudi,F. and Natarajan,A.T. (1993) Metabolic activation of chemicals to mutagenic carcinogens by human hepatoma microsomal extracts in Chinese hamster ovary cells (in vitro). Mutagenesis, 8, 11–15. De Meester,C. (1989) Bacterial mutagenicity of heterocyclic amines found in heat-processed food. Mutat. Res., 221, 235–262. Director,A.E., Nath,J., Ramsey,M.J., Swiger,R.R. and Tucker,J.D. (1996) Cytogenetic analysis of mice chronically fed the food mutagen 2-amino-1methyl-6-phenylimidazo[4,5-b]pyridine. Mutat. Res., 359, 53–61. Duverger-van Bogaert,M., Dierickx,P.J., Stecca,C. and Crutzen,M.-C. (1993) Metabolic activation by a supernatant from human hepatoma cells: a possible alternative in mutagenic tests. Mutat. Res., 292, 199–204. Duverger-van Bogaert,M., Dierickx,P.J. and Crutzen,M.-C. (1995) Metabolic activation of aromatic amines by a human hepatoma cell (Hep G2) supernatant tested by means of Salmonella typhimurium strains with different acetyltransferase activities. Mutat. Res., 335, 219–227. Felton,J.S., Malfatti,M.A., Knize,M.G., Salmon,C.P., Hopmans,E.C. and Wu,R.W. (1997) Health risks of heterocyclic amines. Mutat. Res., 376, 37–41. Frederick,T.H., Knize,M.G., Healy,S.K., Slezak,T.S. and Felton,J.S. (1988) Cooked food mutagens reference list. Environ. Mutagen., 12 (suppl. 4), 1–83. Glowinski,I.B., Savage,L., Lee,M.S. and King,C.M. (1983) Relationship between nucleic acid adduct formation and deacylation of arylhydroxamic acids. Carcinogenesis, 4, 67–75. Hatch,F.T., Knize,M.G. and Felton,J.S. (1991) Quantitative structure activity relationships of heterocyclic amine mutagens. Environ. Mol. Mutat., 17, 4–19. Hatch,F.T., Knize,M.G., Moore,D.H. and Felton,J.S. (1992) Quantitative correlation of mutagenic and carcinogenic potencies for heterocyclic amines from cooked foods and additional aromatic amines. Mutat. Res., 271, 269–287. Hein,D.W., Rustan,D.T., Ferguson,R.J., Doll,M.A. and Gray,K. (1994) Metabolic activation of aromatic and heterocyclic arylamines by wild type and mutant recombinant human NAT1 and NAT2 acteyltransferases. Arch. Toxicol., 68, 129–133. Holme,J.A., Hongslo,J.K., Soderlund,E., Bunborg,G., Christensen,T., Alexander,J. and Dybing,E. (1987) Comparative genotoxic effects of IQ and MeIQ in Salmonella typhimurium and cultured mammalian cells. Mutat. Res., 187, 181–190. IARC (1993) Some Naturally Occuring Substances: Food Items and Constituents, Heterocyclic Aromatic Amines and Mycotoxins. IARC Scientific Publications no. 56, IARC, Lyon. Ito,N., Hasegawa,R., Sano,M., Tamano,S., Esumi,H., Takayama,S. and Sugimura,T. (1991) A new colon and mammary carcinogen in cooked food, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP). Carcinogenesis, 12, 1503–1506. Knasmu¨ller,S., Parzefall,W., Sanyal,R., Ecker,S., Schwab,C., Uhl,M., MerschSundermann,V., Williamson,G., Hietsch,G., Langer,T., Darroudi,F. and Natarajan,A.T. (1998) Use of metabolically competent human hepatoma cells for the detection of mutagens and antimutagens. Mutat. Res., 402, 185–200. Knasmu¨ller,S., Uhl,M., Fahrig,R., Darroudi,F., Mersch-Sundermann,V., Bader,A., Sanyal,R., Schwab,C., Hietsch,G., Parzefall,W. and Natarajan,A.T. (1999) Development and application of tests with human derived cells for the detection of environmental genotoxins. J. Indian Genet. Soc., in press. Land,S.J., Zukowski,K., Lee,M.-S., Debiec-Rychter,M., King,C.M. and. Wang,C.Y. (1989) Metabolism of aromatic amines: relationships of N-acetylation, O-acetylation, N,O-acetyltransfer and deacetylation in human liver and urinary bladder. Carcinogenesis, 10, 727–731. Loprieno,N., Boncristiani,G. and Loprieno,G. (1991) An experimental approach to identify the genetic risk from cooked food mutagens. Food Chem. Toxicol., 29, 377–386. Lowry,O.H., Rosebrough,N.J., Farr,A.L. and Randall, F.J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem., 62, 315–323. Maron,D.M. and Ames B.N. (1984) Revised methods for the Salmonella mutagenicity test. In Kilbey,B.J., Legator,M., Nichols,W. and Ramel,C. (eds), Handbook of Mutagenicity Test Procedures, 2nd Edn. Elsevier, Amsterdam, The Netherlands, pp. 93–141. Morgenthaler,P.M.L. and Holzha¨user D. (1995) Analysis of mutations induced by 2-amino-1-methyl-6-phenylimidazo [4,5-b]pyridine in human lymphoblastoid cells. Carcinogenesis, 16, 713–718. Morgenthaler,P.M.L., OpHetVelt,C., Jaccaud,E. and Turesky,R.J. (1998) Mutagenicity of 2-amino-3-methylimidazo[4,5-f]quinoline in human lymphoblastoid cells. Carcinogenesis, 19, 1749–1754. Mosman,T. (1983) Rapid colorimetric assays for cellular growth and survival— application to proliferation and cytotoxicity assays. J. Immunol. Methods, 65, 55–63. Natarajan,A.T. and Darroudi,F. (1991) Use of human hepatoma cells for in vitro

Genotoxic effects of heterocyclic aromatic amines metabolic activation of chemical mutagens/carcinogens. Mutagenesis, 6, 399–403. Natarajan,A.T. and Obe,G. (1982) Mutagenicity testing with cultured mammalian cells: cytogenetic assays. In Heddle J.A. (ed.), Mutagenicity: New Horizons in Genetic Toxicology. Academic Press, New York, NY, pp. 171–313. Ohgaki,H., Kusama,K., Matsukura,N., Morino,K., Hasegawa,H., Sato,S., Sugimura,T. and Takayama,S. (1984) Carcinogenicity in mice of a mutagenic compound, 2-amino-3-methylimidazo[4,5-f]quinoline, from broiled sardine, cooked beef and beef extract. Carcinogenesis, 5, 921–924. Ohgaki,H., Takayama,S. and Sugimura,T. (1991) Carcinogenicities of heterocyclic amines in cooked food. Mutat. Res., 259, 399–410. Otsuka,C., Miura,K.F. and Ishidate,M. (1996a) The possible role of acetyltransferase in the induction of cytogenetic effects induced by mutagen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine in cultured Chinese hamster cells. Mutat. Res., 371, 23–28. Otsuka,C., Miura,K.F., Sathoh,T., Hatanaku,M., Wakabayashi,K. and Ishidate,M. (1996b) Cytogenetic effects of a cooked food mutagen 2-amino1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) and its metabolite 2hydroxyamino-1-methyl-6-phenylimidazo[4,5-b]pyridine (N-OH-PhIP) on human and Chinese hamster cells in vitro. Mutat. Res., 367, 115–121. ¨ vervik,E. and Gustavsson,J.-A. (1990) Cooked-food mutagens: current O knowledge of formation and biological significance. Mutagenesis, 5, 437– 446. Sabbioni,G. and Wild,D. (1992) Quantitative structure activity relationships of mutagenic aromatic and heteroaromatic azides and amines. Carcinogenesis, 13, 709–714. Sanyal,R., Darroudi,F., Parzefall,W., Nagao,M. and Knasmu¨ller,S. (1997) Inhibition of the genotoxic effects of heterocyclic amines in human derived hepatoma cells by dietary bioantimutagens. Mutagenesis, 12, 297–303. Sawada,S., Yamanaka,T., Yamatsu,K., Furihata,C. and Matsushima,T. (1991) Chromosome aberrations, micronuclei and sister chromatid exchanges (SCEs) in rat liver induced in vivo by hepatocarcinogens including heterocyclic amines. Mutat. Res., 251, 59–69. Schwab,C., Fekadu,K., Quin M.M., Sanyal,R., Uhl,M., Hietsch,G., Rabot,S., Darroudi,F. and Knasmu¨ller,S. (1999) Development of test systems for the detection of compounds which prevent the genotoxic effects of heterocyclic aromatic amines: preliminary results with constituents of cruciferous vegetables and other dietary constituents. J. Environ. Pathol. Toxicol. Oncol., 18, 109–118. Sugimura,T., Nagao,M., Kawachi,T., Honda,M., Yahagi,T., Seino,Y., Sato,S., Matsukura,N., Matsushima,T., Shirai,A., Sawamura,M. and Matsumoto,H. (1977) Mutagen-carcinogens in foods, with special reference to highly mutagenic pyrolytic products in broiled foods. In Hiatt,H.H., Watson,J.D. and Winsten,J.A. (eds.), Origins of Human Cancer. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 1561–1577. Takayama,S., Nakatsuru,Y., Masuda,M., Ohgaki,H., Sato,S. and Sugimura,T. (1984) Demonstration of carcinogenicity in F344 rats of 2-amino-3methylimidazo[4,5-f]quinoline from broiled sardine, fried beef and beef extract. Gann, 75, 467–470. Thompson,L.H., Carrano,A.V., Salazar,E., Felton,J.S. and Hatch,F.T. (1983) Comparative genotoxic effects of the cooked-food-related mutagens Trp-P2 and IQ in bacteria and cultured mammalian cells. Mutat. Res., 117, 243–257. Thompson,L.H., Tucker,J.D., Steward,S.A., Christensen,M.L., Salazar,E.P., Carrano,A.V. and Felton,J.S. (1987) Genotoxicity of compounds from cooked beef in repair deficient CHO cells versus Salmonella mutagenicity. Mutagenesis, 12, 483–487. Totsuka,Y., Fukutome,K., Takahashi,M., Takahashi,S., Tada,A., Sugimura,T. and Wakabayashi,K. (1996) Presence of N2-(deoxyguanosin-8-yl)-2-amino3,8-dimethylimidazo[4,5-f]quinoxaline (dG-C8-MeIQx) in human tissues. Carcinogenesis, 17, 1029–1034. Tucker,J.D., Carrano,A.V., Allen,N.A., Christensen,M.L., Knize,M.G., Strout,C.L. and Felton,J.S. (1989) In vivo cytogenetic effects of cooked food mutagens. Mutat. Res., 224, 105–113. Turteltaub,K.W., Mauthe,R.J., Dingley,K.H., Vogel,J.S., Frantz,C.E., Garner, R.C and Shen,N. (1997) MeIQx–DNA adduct formation in rodent and human tissues at low doses. Mutat. Res., 376, 243–252. Vikse,R., Hatch,F.T., Winter,N.W., Knize,M.G., Grivas,S. and Felton,J.S. (1995) Structure–mutagenicity relationships of four amino-imidazonaphthyridines and imidazoquinolines. Environ. Mol. Mutagen., 26, 79–85. Wakabayashi,K., Nagao,M., Esumi,H. and Sugimura,T. (1992) Food derived mutagens and carcinogens. Cancer Res., 52, 2092–2098. Wakabayashi,K., Ushiyama,H., Takahashi,M., Nukaya,H., Kim,S.-B., Hirose,M., Ochiai,M., Sugimura,T. and Nagao,M. (1993) Exposure to heterocyclic amines. Environ. Health Perspect., 99, 129–133. Wakabayashi,K., Totsuka,Y., Tukutome,K., Oguri,A., Ushiyama,H. and

Sugimura,T. (1997) Human exposure to mutagenic/carcinogenic heterocyclic amines and comutagenic β-carbolines. Mutat. Res., 376, 253–259. Watanabe,M., Ishidate,M.,Jr and Nohmi,T. (1990) Sensitive method for the detection of mutagenic nitroarenes and aromatic amines: new derivatives of Salmonella typhimurium tester strains possessing elevated Oacetyltransferase levels. Mutat. Res., 234, 337–348. Watanabe,M., Matuoka,A., Yamazaki,N., Hayxashi,M., Deguchi,T., Nohmi,T. and Sofuni,T. (1994) New sublines of Chinese hamster CHL stably expressing human NAT1 or NAT2 acetyltransferase or Salmonella typhimurium O-acetyltransferase: comparison of the sensitivities to nitroarenes and aromatic amines using the in vitro micronucleus test. Cancer Res., 54, 1672–1677. Wiebel,F.J. (1993) Do we need in vivo assays in case of negative in vitro results? In Madle,S. and Mu¨ller,L. (eds), Current Issues in Genetic Toxicology. MMV Press, Munich, Germany, pp. 25–23. Wild,D., Gocke,E., Harnasch,D., Kaiser,G. and King,M.T. (1985) Differential mutagenic activity of IQ (2-amino-3-methylimidazo[4,5-f]quinoline) in Salmonella typhimurium strains in vitro and in vivo, in Drosophila, and in mice. Mutat. Res., 156, 93–102. Wild,D., FeserW., Michel,S., Lord,H.L. and Josephy,P.D. (1995) Metabolic activation of heterocyclic aromatic amines catalyzed by human arylamine N-acetyltransferase isozymes (NAT1 and NAT2) expressed in Salmonella typhimurium. Carcinogenesis, 16, 643–648. Wu,R.W., Tucker,J.D., Sorensen,K.J., Thompson,L.H. and Felton,J.S. (1997) Differential effect of acetyltransferase expression on the genotoxicity of heterocyclic amines in CHO cells. Mutat. Res., 390, 93–103. Zhang,Y.P., Kopman,G. and Rosenkranz,H.S. (1993) Structural basis of the mutagenicity of heterocyclic amines formed during the cooking process. Environ. Mol. Mutat., 21, 100–115. Received on December 16, 1998; accepted on June 28, 1999

539