- Email: [email protected]
Genotoxicity assessment of the antiepileptic drug AMP397, an Ames-positive aromatic nitro compound
Mutation Research 518 (2002) 181–194
Genotoxicity assessment of the antiepileptic drug AMP397, an Ames-positive aromatic nitro compound Willi Suter a,∗ , Andreas Hartmann a , Franziska Poetter a , Peter Sagelsdorff b , Peter Hoffmann a , Hans-Jörg Martus a a
Toxicology/Pathology, Novartis Pharma AG, 4002 Basel, Switzerland b Syngenta AG, Basel, Switzerland
Received 29 January 2002; received in revised form 19 April 2002; accepted 22 April 2002
Abstract AMP397 is a novel antiepileptic agent and the first competitive AMPA antagonist with high receptor affinity, good in vivo potency, and oral activity. AMP397 has a structural alert (aromatic nitro group) and was mutagenic in Salmonella typhimurium strains TA97a, TA98 and TA100 without S9, but negative in the nitroreductase-deficient strains TA98NR and TA100NR. The amino derivative of AMP397 was negative in wild-type strains TA98 and TA100. AMP397 was negative in a mouse lymphoma tk assay, which included a 24 h treatment without S9. A weak micronucleus induction in vitro was found at the highest concentrations tested in V79 cells with S9. AMP397 was negative in the following in vivo studies, which included the maximum tolerated doses of 320 mg/kg in mice and 2000 mg/kg in rats: MutaTM Mouse assay in colon and liver (5 × 320 mg/kg) at three sampling times (3, 7 and 31 days after the last administration); DNA binding study in the liver of mice and rats after a single treatment with [14 C]-AMP397; comet assay (1 × 2000 mg/kg) in jejunum and liver of rats, sampling times 3 and 24 h after administration; micronucleus test (2 × 320 mg/kg) in the bone marrow of mice, sampling 24 h after the second administration. Based on these results, it was concluded that AMP397 has no genotoxic potential in vivo. In particular, no genotoxic metabolite is formed in mammalian cells, and, if formed by intestinal bacteria, is unable to exert any genotoxic activity in the adjacent intestinal tissue. These data were considered to provide sufficient safety to initiate clinical development of the compound. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Antiepileptic; Nitroaromatic compound; Nitroreductase; DNA binding; Comet assay; MutaTM Mouse
1. Introduction Epilepsy is one of the most common neurological disorders, with prevalence in excess of 0.5% of the world population. Despite a variety of antiepileptic drugs available, there is still a high medical need for improved treatments of epilepsy. About 30% of ∗ Corresponding author. Tel.: +41-61-3241341; fax: +41-61-3241274. E-mail address: [email protected] (W. Suter).
patients see their seizures inadequately controlled [1], and many of those who become free of seizures suffer from adverse effects, the most frequent complaints being cognitive impairment and decrease in overall energy level [2,3]. It is now well established that the activation of alpha-amino-3-hydroxy-5-methyl4-isoxazole-propionic acid (AMPA) receptors is involved in the initiation and propagation of seizures [4–6], therefore, AMPA receptor antagonism appears a logical strategy for the development of new anticonvulsant drugs. AMP397 was selected for the
1383-5718/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 3 - 5 7 1 8 ( 0 2 ) 0 0 1 0 5 - 5
182
W. Suter et al. / Mutation Research 518 (2002) 181–194
Fig. 1. Structures of AMP 397 (top) and AAE669 (bottom).
development as an antiepileptic drug because it combines high affinity for the AMPA receptor, good in vivo potency and oral anticonvulsant activity in rodent seizure models. The compound is a nitro-quinoxaline derivative (Fig. 1), which caused concern in toxicology because of the possible mutagenic and carcinogenic potential of the aromatic nitro group. The genotoxicity and carcinogenicity of nitroaromatic compounds has been studied intensively over the last 50 years beginning with 4-nitroquinoline-1-oxide, which was found to be a potent carcinogen [7]. Nitroaromatic compounds are found as air pollutants, in cigarette smoke and in grilled foods. They are important synthesis intermediates, e.g. for the production of azo dyes and are used as herbicides as well as human and veterinary medicine [8]. Among the latter, nitrofuran derivatives [9], mainly used to treat topical and urinary tract infection, metronidazole [10], used against Trichomonas vaginalis infections, and aristolochic acid, a mixture of naturally occurring carboxylic acids as a component of herbal medicines prepared from the Aristolochia plant [11], have been extensively studied. Many nitroaromatic compounds have been shown to
bind covalently to DNA [8,12]. The reactive forms are metabolically generated through nitroreduction and, in many cases, through oxidative pathways. Whereas, the oxidative pathways depend on the presence of the cytochrome P450 family of enzymes and occur, therefore, mainly in the liver, nitroreduction is found in bacterial as well as in mammalian cells [13,14]. This may explain why, nitro compounds are generally fairly strong mutagens in the Salmonella mutagenicity assay, whereas, mutagenicity in mammalian cells and carcinogenicity in the rodent is not always found for Salmonella-positive nitro compounds. This has resulted in labelling the Salmonella mutagenicity assay as ‘not predictive’ for this class of compounds [15]. The evaluation of the mutagenic potential of a nitro compound, therefore, relies strongly on the outcome of tests conducted in mammalian systems. The results of the standard in vitro mammalian assays have to be interpreted with caution, since nitroreduction is oxygen sensitive so that in vitro metabolism by rat liver S9 may produce a completely different spectrum of metabolites than observed in vivo [16,17]. In addition, metabolic activation, i.e. nitroreduction by the intestinal flora and the possible effect of bacterial metabolites on mammalian cells cannot be appropriately studied in vitro. It was, therefore, decided to study the genotoxic potential of AMP397 in an extensive battery of in vivo studies, focusing on well-exposed and metabolically active tissues, where the likelihood to detect genotoxic effects was considered to be highest.
2. Materials and methods 2.1. Test articles AMP397 (CAS registry number 188696-80-2) was synthesised in the laboratories of Nervous System Research, Novartis Pharma AG, Basel, Switzerland. The positive control chemicals were purchased from the following manufacturers: 9-Aminoacridine (Fluka), 2-aminoanthracene (Fluka), benzo(a)pyrene (Sigma), cyclophosphamide (ASTA medica AG), ethylmethane sulfonate (Aldrich), methylmethane sulfonate (Sigma), mitomycin C (Boehringer Ingelheim), 2nitrofluorene (Merck), N-nitroso-N -ethylurea (Sigma), sodium azide (Fluka).
W. Suter et al. / Mutation Research 518 (2002) 181–194
183
2.2. Rat liver S9
2.4. V79 micronucleus test
The S9 liver homogenate was prepared as described in the literature [18]. Seven to 9-week-old male Crl:Wist Han (Charles River, Germany) were injected intraperitoneally with 500 mg/kg Aroclor 1254 and sacrificed 5 days later. Their livers were homogenised, diluted 1:4 with 0.15 M KCl and centrifuged for 10 min at 9000 g. The supernatant was frozen in small aliquots and stored at −70 to −80 ◦ C until use. The S9-mix used for the Salmonella mutagenicity experiments was prepared according to Maron and Ames [18] with 10% rat liver S9-fraction. The composition and final concentrations of the S9-mix used for the V79 micronucleus was as follows: Glucose–6–phosphate, 4.4 mM; NADP, 0.84 mM; KCl, 30 mM; 0.032% NaHCO3 ; S9 fraction, 10%. For the mouse lymphoma assay the final concentrations of the S9-mix components in the treatment medium were as follows: Hanks BSS, Glucose–6–phosphate, 5.9 mM; NADP, 0.32 mM; KCl, 1.5 mM; S9 fraction, 2%.
V79 chinese hamster cells were obtained from M. Fox (Paterson Laboratories, Manchester, UK). The assay was conducted as described in detail by Miller et al. [21]. In brief, based on the data of a preliminary cytotoxicity test using (established by use of the MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide, Sigma) exclusion assay), 6 concentrations, leading to a reduction of MTT uptake of 25–80%, were selected for the main experiment. For the micronucleus experiment, the cells were treated 1 day after subculture with the test compound solution for 3 h (with and without S9) or for 20 h (without S9). After the treatment, cells were washed with PBS and further incubated with fresh medium for 45 (or, 28 in case of 20 h treatment) hours. Before sampling, the cells were examined qualitatively for morphology and growth (degree of confluence). Based on this assessment, the concentration that inhibited growth by approximately 50%, and two lower concentrations were chosen for analysis. A period of 48 h after the start of treatment, cells were removed with trypsin EDTA (Gibco BRL) and spread on glass slides by cytocentrifugation (Shandon Cytospin). Cells were fixed and stained with Schiff reagent (Merck). 4000 cells (2000 cells per culture) were analysed automatically using a LEITZ MIAS image analyser from at least three concentration groups of each test compound as well as from negative and positive controls. A result was considered positive if the micronucleus frequency was ≥1% and at least 0.6 above the concurrent solvent-control value. These criteria are based on an in house validation study of the assay. The historical mean solvent-control data were as follows: 3 h treatment without S9, 0.54 ± 0.17; 20 h treatment without S9, 0.60 ± 0.16; with S9, 3 h treatment with S9, 0.61 ± 0.19.
2.3. Ames test The method used followed the recommendations by Maron and Ames [18] and the OECD guideline [19]. The Salmonella typhimurium strains TA1535, TA97a, TA98, TA100 and TA102 were obtained from B.N. Ames, Biochemistry Department, University of California, Berkeley, CA, USA. The nitroreductase resistant strains were provided by Dr. P. Arni (Syngenta AG, Basel, Switzerland) who had obtained them from Dr. H.S. Rosenkranz (Pittsburgh, PA, USA). These bacteria, S. typhimurium strains TA98NR and TA100NR, are lacking the ‘classical’ nitroreductase and were isolated as niridazole resistant derivatives of S. typhimurium strains TA98 and TA100, respectively [20]. A test compound was judged to be mutagenic in the plate test if it produces, in at least one concentration and one strain, a response equal to twice (or more) the control incidence. The only exception is strain TA102, which has a relatively high spontaneous revertant number, where an increase by a factor of 1.5 above the control level is taken as an indication of a mutagenic effect.
2.5. Mouse lymphoma thymidine kinase assay L5178Y mouse lymphoma cells, tk+/− clone 3.7.2C, were obtained from American Type Culture Collection by D. Geleick, Ciba Geigy Ltd., Basle, and were kept as frozen stock in liquid nitrogen. The test compound was dissolved in 0.5 M Na2 HPO4 at the maximum soluble concentration of 22 mg/ml. The final concentration of the solvent in the treatment
184
W. Suter et al. / Mutation Research 518 (2002) 181–194
medium was 2% in Experiment 1 and part of Experiment 2. In part of Experiment 2 and in Experiment 3 the end concentration of the 0.5 M Na2 HPO4 was increased to 4% in order to reach higher test compound concentrations and achieve stronger cytotoxicity. The fluctuation protocol with treatment times of 3 h (with and without S9-mix) and 24 h (without S9-mix) was used as described by Cole et al. [22] and in the corresponding OECD guideline [23]. All calculations were performed with ‘MUTANT’ software for fluctuation assays (York Electronic Research, UK). These calculations are based on recommendations by UKEMS guidelines found in the literature [24]. The concentrations for the mutagenicity experiments were chosen based on the results from a preliminary dose-finding experiment using the following calculations: adapted survival (%S) was calculated by multiplying ‘cloning efficiency’ by ‘cell count day 0’ of the corresponding culture divided by ‘cell count day 0’ of the solvent-control cultures. Relative adapted survival (%RS) was calculated from %S by setting the solvent-control value to 100% and calculating values of treated groups relative to it. On the basis of these results, at least five concentrations between 10 and 20% and 100% RS were chosen for the mutagenicity experiment. 2.6. Animal husbandry Generally, mice were kept individually in type II Macrolon® cages and rats pairwise in Macrolon© Type 4 cages on sawdust bedding. For the DNA binding study the animals were housed in groups of two animals (rats) and four animals (mice). The animals were fed standard rodent chow NAFAG 8900 FOR GLP (Nafag, Gossau SG, Switzerland) and given tap water in bottles ad libitum. The animals were housed in a fully air-conditioned animal room, at 22 ± 3 ◦ C, at a relative humidity of 35–75% and a 12 h light/dark cycle. Young adult animals were used in all studies. 2.7. Bone marrow micronucleus test in mice treated by the oral route The study protocol was based on the method developed by Romagna and Staniforth [25] and was in compliance with the corresponding OECD guideline [26]. The test article was dissolved in 0.2 M NaHCO3 . The
dose-finding experiment with AMP397 showed that treatment of CD-1 mice (Charles River, Germany) by oral gavage with 450.5, 500 or 800 mg/kg, twice with an interval of 24 h, led to strong signs of toxicity such as laboured breathing, ataxia, and strong sedation. At 320 mg/kg, the same symptoms were visible, but with less severity, and no animals died. On the basis of these results, doses of 32, 100 and 320 mg/kg were chosen for this micronucleus test. In the main experiment five male and five female mice were treated as described above and bone marrow was sampled 48 h after the first application. Nucleated cells were removed from the bone marrow samples using cellulose columns. The cells were loaded on poly-l-lysine coated glass slides by cytocentrifugation using a Shandon Cytospin stained with May Grünwald stain (5%) and Giemsa (14%). The slides were automatically evaluated with a LEITZ MIAS image analyser [27]. No statistical analysis was performed since all values in the treated groups were ≤ the frequency of micronucleated polychromatic erythrocyte in the concurrent vehicle control group. 2.8. Gene mutation assay in MutaTM Mouse In a dose-finding experiment groups of three male CD-1 mice were treated by oral gavage on five consecutive days with 125, 200 or 320 mg/kg per day. Clear signs of toxicity such as ataxia, sedation, Straub fur or closed eyes were observed in all dose groups. At 320 mg/kg per day, these symptoms were very severe and this dose was, therefore, chosen for the mutagenicity test. A total of 320 mg/kg per day AMP397 was administered on five consecutive days in a volume of 20 ml/kg, and using 0.2 M NaHCO3 as the vehicle to groups of seven to nine male mice (MutaTM Mouse, Covance Research Products, USA). The AMP397- treated animals were sacrificed on 3, 10 and 31 days after the last application, the vehicle control animals on day 31 and the three positive control animals (50 mg/kg per day N-nitroso-N -ethylurea) on day 10. Colon and livers were sampled, flash frozen in liquid nitrogen and stored at −80 ◦ C until analysis. Genomic DNA was isolated using the phenol/chloroform method [28]. For in vitro packaging of the DNA into the lambda phage vector TranspackTM (Stratagene, USA) was used under the conditions recommended by the manufacturer. The mutant
W. Suter et al. / Mutation Research 518 (2002) 181–194
185
frequencies were determined using Escherichia coli galE− LacZ− , which allows selection for LacZdeficient mutants [29,30]. The test conditions and procedures were chosen in accordance with the recommendations by Covance [31]. In particular, 300’000 plaque forming units (pfu) derived from at least four independent packaging reactions were analysed per DNA sample. No statistical analysis of the data was conducted, because all mutant frequencies of the treated groups were lower than the concurrent vehicle control value.
of the inevitable bacterial contamination of jejunum samples, they were analysed semi-automatically with the Comet Assay II image analysis system (Perceptive Instruments). Fifty cells per slide were evaluated on each of the two replicate slides. The parameter measured was the tail moment as described by Olive et al. [35]. The statistical evaluation was conducted using the SAS software package (SAS version 6.12). An ANOVA was followed by a Dunn test (because of heterogeneity) or a Bartlett’s test in case of the 24 and 3 h experiment, respectively.
2.9. Comet assay with jejunum and liver cells of rats
2.10. DNA binding study in rats and mice
The alkaline comet assay protocol used was based on published recommendations and descriptions of methods [32,33]. Groups of five male Crl:Wist Han rats (Charles River, Germany) were administered once by oral gavage a dose of 2000 mg/kg (suspended in sodium carboxymethylcellulose solution [CMC] 1%, application volume 5 ml/kg). Vehicle control animals received CMC 1% (5 ml/kg) and positive control animals 300 mg ethylmethane sulfonate suspended in 1% CMC. Jejunum and liver were sampled 3 or 24 h after the application. The tissue was washed in HBSS/25 mM EDTA/10% DMSO and afterwards kept in buffer on ice until further preparation. For the preparation of single cell suspensions a small piece of the tissue was taken and a small amount of buffer was pipetted onto the tissue; then the tissue was minced with a scalpel into fine pieces to obtain a cell suspension. A volume of 15 l of the cell suspension was then mixed 1:10 with 0.5% LMP-agarose. A volume of 45 l of the agarose suspension were put on the slides; two samples were placed on one slide and each was covered with a 24 mm × 24 mm coverslip until the agarose hardened. The coverslip was then removed and the slide was immersed into a jar containing lysis buffer (2.5 mM NaCl, 100 mM EDTA, 10 mM Tris, 10% DMSO, 1% sodium lauroyl sarcosinate, 1% Triton X-100). Alkaline unwinding time was 15 min (jejunum) or 20 min (liver), electrophoresis 20 min. After neutralisation, slides were alcohol-dried and afterwards stained with 2.5 g/ml propidium iodide in Vectashield (Vector Laboratories, USA). Liver samples were analysed fully automatically using a Leitz MIAS image analysis system [34]. Because
Two rats and eight mice were treated with [14 C]AMP397 at a dose level of 13.2 mg/kg body weight (2 mCi/kg body weight). The animals received 10 ml/kg body weight of the dosing solution by single oral gavage and were killed after 24 h. The third rat remained untreated and was used to show that the work-up of the samples was performed without external contamination with radiolabel. A radioactivity dose of about 2 mCi/kg body weight was administered to achieve a limit of detection for the CBI in the liver of <0.1. Compounds with a CBI of <0.1 were defined to be non-genotoxic in this assay [36,37]. The chemical dose was based on the specific activity of the radiolabeled test compound. Individual livers of rats and two pools of livers, each prepared from four identically treated mice, were homogenised in 75 mM NaCl, 10 mM Tris/HCl, 10 mM EDTA, pH 7.8, and DNA was isolated as described by Sagelsdorff et al. [38,39]. Essentially, chromatin was precipitated and the pellet was blended in a denaturing lysing medium, deproteinated with chloroform/isoamyl alcohol/phenol (CIP), extracted with diethyl ether and the DNA was further purified by adsorption on a hydroxylapatite column, dialysis and precipitation with ethanol. The highly purified DNA was dissolved in 20 mM succinic acid containing 8 mM CaCl2 and an aliquot was used for UV-determination of the DNA content. Another aliquot was mixed with scintillation cocktail for the determination of radioactivity. To determine background radioactivity, DNA was isolated accordingly from the liver of the untreated rat. The total count—upon comparison with historical controls—was used to show that the work-up of the
186
W. Suter et al. / Mutation Research 518 (2002) 181–194
samples was performed without external contamination with radiolabel. On the basis of the background values compiled during the last years, the mean historical background and the respective S.D. was calculated. Assuming that a vial containing a DNA sample with low amount of radioactivity shows the same statistic variation as the background, a limit of detection for radioactivity in a vial was calculated on a level of 2 S.D. from the historical background. If no radioactivity was detectable in a DNA sample this limit of detection was used to calculate the maximum possible covalently bound DNA radioactivity. The covalently bound DNA radioactivity was converted to the units of the covalent binding index (CBI) according to the following equation [36]: CBI = =
µmol chemical bound/mol DNA nucleotides mmol chemical applied/kg body weight dpm/mg DNA × 309 × 106 dpm/kg body weight
The average weight of 1 mol of DNA nucleotides is 309 g.
3. Results and discussion The results are shown in detail in Tables 1–7 and in Fig. 2. A summarised overview is presented in Table 8. The standard Ames test showed, in the absence of rat liver S9, a clear dose dependent increase in rever-
tant numbers in strain TA98 and positive results also in strains TA100 and TA97a (Table 1). This result was not surprising, since many aromatic nitro compounds are positive in this assay [15]. The comparison with published Salmonella mutagenicity data shows that AMP397 is one of the weakest nitroaromatic mutagens known. It induced 0.077 revertants/nmol, which lies at the lower end of the potency ranges found by McCann et al. [40] (0.26–20800 revertants/nmol) and compiled by Purohit and Basu ([8] 0.0029–734000 revertants/nmol). This could be due to one or several of the following reasons: (i) lack of uptake of the test compound by the bacteria; (ii) low activation rate by the ‘classical’ nitroreductase; (iii) low affinity of the reactive metabolite(s) for the DNA; (iv) formation of weakly mutagenic types of DNA damage in bacteria. The absence of an effect in nitroreductase-deficient strains and the negative results obtained with the amino analogue AAE669 (Fig. 2) indicated that the nitro group is metabolised and activated to electrophilic intermediates by the ‘classical’ nitro reductase [20] in S. typhimurium. It is not clear, however, why mutagenicity occurred only in strain TA98 and only at higher concentrations in the preincubation assay as compared to the effects found in the plate incorporation assay, where mutagenicity was observed in S. typhimurium strains TA97a, TA98 and TA100. The plate incorporation assay was also found to be more sensitive than the preincubation assay for the detection of the mutagenicity of 4-nitro-phenylenediamine [41]. In the absence of direct measurements of metabolism on the agar plates it is also difficult to explain why
Fig. 2. Mutagenicity data in S. typhimurium strain TA98 without S9. The experiments were conducted under standard conditions and using the plate incorporation method, (䊉) EE699 (amino analogue of AMP397) standard test; (䊏) MP397 tested with TA98NR (nitroreductase deficient); (䉬) MP397 standard test.
W. Suter et al. / Mutation Research 518 (2002) 181–194
187
188
W. Suter et al. / Mutation Research 518 (2002) 181–194
Table 2 Micronucleus test in V79 chinese hamster fibroblasts Treatment (20 h, −S9)
Cells with MN (%)
Treatment (3 h, +S9)
Cells with MN (%)
0.50 0.68 0.28 0.78 Too toxic
S9 412.3 g/ml 833.0 g/ml 1682.9 g/ml 3400.0 g/ml
0.73 0.73 0.75 1.38 Too toxic
19.30
CP, 20 M
2.60
Cells with MN (%)
Treatment (3 h, +S9)
Cells with MN (%)
Experiment 2 MEM 2537.3 g/ml 2893.6 g/ml 3300.0 g/ml
0.43 0.55 0.75 0.60
S9 538.6 g/ml 790.6 g/ml 1160.4 g/ml 1703.2 g/ml
0.65 0.75 0.88 1.40 Too toxic
EMS, 12.5 mM
9.80
CP, 17.5 M
6.80
Experiment 1 MEM 412.3 g/ml 833.0 g/ml 1682.9 g/ml 3400.0 g/ml EMS, 6 mM Treatment (3 h, −S9)
The numbers (% cells with MN) represent the percentage of cells with one or more micronuclei. Historical negative control values: without S9, 3 h treatment 0.54 ± 0.17; without S9, 20 h treatment, 0.60 ± 0.16; without S9, 3 h treatment, 0.61 ± 0.19. Abbreviation, MN: micronucleus.
AMP397 was not mutagenic for Salmonella in the presence of S9. In rats and dogs it was found that metabolism does not play an important role in the elimination of AMP397 and that binding to plasma proteins was low. It appears unlikely, therefore, that the compound was rapidly degraded or bound by proteins of the S9-mix. Bacteriotoxicity did not provide information to answer this question, because the test compound was not bacteriotoxic under all test conditions up to the highest soluble concentrations. One is, therefore, left with the speculation that the bacteria might absorb nucleophilic products from the S9-mix, which react with the reactive intermediates formed by the bacterial nitroreductase. The absence of a mutagenic effect in the presence of S9 is important, because it shows that mammalian liver enzymes in general and mammalian nitroreductases in particular, do not metabolise AMP397 into mutagenic intermediates under the conditions of the Ames test. Other nitroaromatic compounds have been found to be activated to mutagenic intermediates by mammalian nitroreductases [12] and by nitroreductase-independent pathways [42].
The two in vitro mammalian assays conducted, i.e. the V79 micronucleus test and the mouse lymphoma thymidine kinase assay, gave contradictory results. The V79 micronucleus test showed a weak increase in the frequency of micronucleated cells in the presence of S9 at the highest concentration that could be evaluated (1682.9 and 1160.4 g/ml in the first and second experiment, respectively (Table 2)). The highest non-mutagenic concentrations (833 and 790.6 g/ml, in the first and second experiment, respectively) were about 400 times higher than the peak plasma levels measured in man after a single oral treatment with up to 800 mg (about 2 g/ml). The mouse lymphoma assay provided negative results up to the precipitating concentration of 880 g/ml. There are at least three possible explanations for the contradictory results obtained in the V79 micronucleus test and the mouse lymphoma assay: (i) the test conditions used for the two studies were different with regard to cell type, genetic endpoint, S9 concentration, etc.; (ii) the V79 micronucleus test and the mouse lymphoma assay were performed with different batches of test compound. The data of the chemical analysis showed that the batch used for the V79 micronucleus test (purity 97%) was slightly less pure than the material used for all other studies (purity 98.3%). Since the impurities have not been identified, it is not clear whether the contradictory results shown in Tables 2 and 3 are due to the presence of mutagenic impurities in the batch used for the V79 micronucleus test; (iii) for unknown reasons the test material used for the mouse lymphoma assay was less soluble and precipitated at concentrations ≥440 g/ml, which prevented the use of the clearly cytotoxic concentrations, required to induce an increased micronucleus frequency in V79 cells. In conclusion, it is not entirely clear whether AMP397 can be activated to mutagenic intermediates by S9-mix from Aroclor 1254 pre-treated rats under some test conditions at very high concentrations. However, as shown in Table 1, under the conditions of the Ames test, no induction of revertants was found and, thus, an activation of AMP397 can be excluded with certainty. It is important to mention in this context that pre-treatment of rats with Aroclor 1254 leads to a significant increase in cytosolic nitroreductase activity [43], which indicates that the S9-mix used for the in vitro experiments had a comparatively high nitroreductase activity.
W. Suter et al. / Mutation Research 518 (2002) 181–194
189
Table 3 Mouse lymphoma thymidine kinase assay 24 h Treatment (g/ml)
Experiment 1 0 55 110 220 440 P Linear trend MMS 5 3 h Treatment (g/ml)
Experiment 2 0 55 110 220 440 $$,P Linear trend MMS 10 0 880
P
Linear trend
−S-9
3 h treatment (g/ml)
%RS
RTG
MF§
100.00 97.06 78.51 92.51 77.55
1.00 1.20 0.96 1.12 0.97
278.49 211.19 NS 260.08 NS 291.52 NS 181.47 NS
0 55 110 220 440 P
NS
Linear trend B(a)P 1.5
34.09
0.22
2303.34
−S-9
3 h treatment (g/ml)
%RS
RTG
MF§
100.00 89.20 83.85 79.42 79.07
1.00 1.00 0.97 0.99 (0.81)
142.75 150.63 NS 173.62 NS 168.58 NS (221.49)
0 55 110 220 440 P
NS 60.02
0.42
1163.09
Linear trend B(a)P 1.5
100.00 77.24
1.00 0.92
200.70 156.10 NS
0 880
NS
Linear trend
24 h Treatment (g/ml)
Experiment 3 0 110 220 440 P 880 P Linear trend MMS 5
P
+S-9 %RsS
RTG
MF§
100.00 113.57 107.18 104.27 112.92
1.00 0.98 1.23 1.46 1.48
236.63 216.95 321.78 182.61 205.89
NS NS NS NS
NS 8.49
0.07
14704.81
%RS
RTG
MF§
100.00 104.46 100.25 86.99 94.40
1.00 1.13 1.16 1.29 1.15
215.66 170.58 199.00 191.02 184.75
+S-9
NS NS NS NS
NS 1.50
0.01
3634.17
100.00 94.42
1.00 1.09
213.97 195.01 NS NS
−S-9 %RS
RTG
MF§
100.00 98.59 83.51 72.36 70.75
1.00 1.04 1.35 0.89 0.82
119.27 122.28 132.81 144.26 132.07
NS NS NS NS
NS 47.95
0.50
1259.99
Solvent 0.5 M Na2 HPO4 . Experiment 1 and Experiment 2 (upper table), end concentration of the solvent 2%; Experiment 2 lower table and Experiment 3, end concentration of the solvent 4%. Abbreviations: §, 5-TFT resistant mutants/106 viable cells 2 days after treatment; %RS, percent relative survival adjusted by post treatment cell count factor (cytotoxicity parameter); RTG, relative total growth (cytotoxicity parameter); MF, mutant frequency; P, precipitating dose. NS, not significant; $, treatment has high heterogeneity, but is included in analysis; $$, treatment excluded due to excessive heterogeneity.
Negative and positive mutagenicity results in mammalian cells in vitro have been reported for several nitroaromatic compounds. In all cases the genotoxic potential was found to be much lower in mammalian
cells as compared to bacteria, which can be explained by the high nitroreductase activity in the bacterial tester strains. Metronidazole has been particularly well studied [8,10,44–47] and it appears that in mammalian
190
W. Suter et al. / Mutation Research 518 (2002) 181–194
Table 4 Bone marrow micronucleus test in mice Group
Dose (mg/kg)
Sex
Number of mice
Frequency (%) MPE
Control AMP397 AMP397 AMP397 Cyclophosphamide
Vehicle 32 100 320 40
m m m m m
+f +f +f +f +f
5 5 5 4 5
+5 +5 +5 +5 +5
0.16 0.16 0.16 0.14 1.46
PE ± ± ± ± ±
0.04 0.06 0.05 0.03 0.23
30.6 33.0 37.3 35.5 29.1
± ± ± ± ±
7.5 9.8 13.2 5.7 9.6
Each value represents mean ± S.D. of a treatment group consisting of 10 (highest test compound dose group: 9) animals (4000 PE/animal). Abbreviations: m, male; f, female; PE, polychromatic erythrocytes; MPE, micronucleated polychromatic erythrocytes; frequency (%), percentage of polychromatic erythrocytes carrying micronuclei (MPE) or percentage of polychromatic erythrocytes in the total number of erythrocytes (PE).
Table 5 MutaTM Mouse LacZ assay Group
Treatment (five doses/5 days)
Dose (mg/kg)
Sampling time (days after last treatment)
Animals per group
Mutant frequency (mean ± SD)
Mean mutant frequencies in colon 0 Vehicle I AMP397 II AMP397 III AMP397 ENU N-Nitroso-N -Ethylurea
0 320 320 320 50
31 3 10 31 10
Seven Males Seven Males Seven Males Nine Males Three Males
9.58 7.03 9.26 9.22 197.65
± ± ± ± ±
2.67 1.09 3.10 2.31 41.06
Mean mutant frequencies in liver 0 Vehicle I AMP397 II AMP397 III AMP397 ENU N-Nitroso-N -Ethylurea
0 320 320 320 50
31 3 10 31 10
Seven Males Seven Males Seven Males Nine males Three Males
7.25 6.26 6.98 6.64 22.75
± ± ± ± ±
1.42 1.56 1.39 1.69 4.42
Mean mutant frequencies ± S.D. are shown for each treatment group.
Table 6 Comet assay in rats Treatment
Dose (mg/kg)
Organ
Number of animals
Tail moment (mean ± S.D.) 3 h Sampling
24 h Sampling
Sampling 3 h after administration Control Vehicle AMP397 2000 EMS 300
Jejunum Jejunum Jejunum
Five Males Five Males Five Males
0.48 ± 0.16 0.50 ± 0.31 1.17 ± 0.46
0.34 ± 0.15 0.65 ± 0.31 0.94 ± 0.32
Control AMP397 EMS
Liver Liver Liver
Five Males Five Males Three Males
1.44 ± 0.34 1.37 ± 0.37 11.50 ± 3.08
1.30 ± 0.19 1.22 ± 0.13 10.17 ± 1.80
Vehicle 2000 300
Abbreviation: S.D. is standard deviation.
W. Suter et al. / Mutation Research 518 (2002) 181–194
191
Table 7 Radioactivity in liver DNA isolated 24 h after oral administration of [14 C]AMP397 to rats and mice Animal No.a
Rat number 1
Rat number 2
Mice number 3–6
Mice number 7–10
Effective dose (mg/kg BW) (dpm/kg BW)·10−9
10.6 3.55
10.6 3.57
11.0 3.68
11.1 3.71
DNA Amount in vial (mg) Gross activity (cpm) Net activity (cpm)b Net activity (dpm)c Specific activity (dpm/mg)
3.49 14.1 <3.13 <3.91 <1.12
3.28 13.9 <3.13 <3.91 <1.19
3.64 13.3 <3.13 <3.91 <1.07
3.82 14.2 <3.13 <3.91 <1.02
Binding to DNA CBI unitsd
<0.10
<0.10
<0.09
<0.09
Rat number 11 0 0 3.09 12.5 – – – –
a
The livers of four identically treated mice were pooled for DNA isolation. On the basis of 38 background values compiled during the last 5 years, a respective value of 11.4 cpm with a S.D. of 1.11 cpm was calculated. The S.D. of the difference (S.D.difference ) between a sample and the background can be calculated as the square root of the sum of the squares of the S.D. for the sample (S.D.sample ) and the S.D. of the background (S.D.background ): b
S.D.difference =
(S.D.2sample + S.D.2background )
Assuming that a vial containing a DNA sample from a treated animal with low amount of radioactivity shows the same statistic variation as the background (1 S.D. = 1.11 cpm), the limit of detection (LoD) for radioactivity in a vial can be calculated on a level of 2 S.D.: LoD = 2 × (1.112 + 1.112 ) = 3.13 cpm c d
The counting efficiency was 80%. For the calculation of the CBI, see Section 2.10.
Table 8 Summary of genetic toxicology results Test systems
Test conditions
Result
Ames test
Strains TA1535, TA 97a, TA98, TA100, TA102; +/−S9; 17.6–4400 g/plate
Dose dependent increase −S9 with strains TA98 and TA100; lowest mutagenic concentration 88 g/plate
Ames test
Strains TA98NR, TA100NR, +/−S9; 500–5000 g/plate
No increases in revertant rates
V79 micronucleus test
V79 chinese hamster fibroblasts, +S9, 3 h treatment; −S9, 3 and 20 h treatment; 412.3–3400 g/ml;
Weak increase in frequency of micronucleated cells at the highest, cytotoxic concentration, i.e. at ≥1160.4 g/ml;
Mouse lymphoma tk assay
Fluctuation protocol, +S9, 3 h treatment; −S9, 3 and 24 h treatment; 55–880 g/ml;
No increases in mutant frequencies
Bone marrow micronucleus test
CD-1 mice, 2 oral treatments with a 24 h interval, sampling 48 h after 1st treatment; 32, 100 and 320 mg/kg
No bone marrow toxicity, no effects on percentage of micronucleated polychromatic erythrocytes
MutaTM Mouse LacZ assay
MutaTM Mouse, 5 oral treatments, 320 mg/kg, sacrifice 3, 7 or 21 days after last treatment
No effects on frequencies of mutated LacZ genes in colon and liver
Comet assay
Han:Wist rats, single oral treatment, 2000 mg/kg, sampling after 3 and 24 h
No effects on DNA migration (tail moment) in jejunum and liver
DNA binding
Han:Wist rats and CD-1 mice, single oral treatment with 10.3–11.5 mg/kg, i.e. 1.55–1.74 mCi/kg; sacrifice 24 h after application
No evidence for DNA binding in the liver, CBI < 0.1
Abbreviations: CBI, covalent binding index (definition see Section 2.10); NR, nitro reductase deficient.
192
W. Suter et al. / Mutation Research 518 (2002) 181–194
cells effects were found mainly under hypoxic test conditions [16]. Since it is very difficult to simulate in vitro the oxygen pressure prevailing in mammalian tissues, relevant information about the safety of a nitro compound in patients cannot be expected from in vitro experiments. Aromatic nitro compounds have been tested in vivo and found to be positive in BigBlueTM transgenic mice [48] and DNA binding [49] as well as comet formation [50] was observed in mice treated with such compounds. The genotoxic potential of AMP397 was, therefore, studied in a series of in vivo studies with animals treated by the oral route. Oral treatment of animals also allows studying in the intestines possible effects of metabolites formed by the intestinal flora, which cannot, or can only with great difficulty, be studied in mammalian cells in vitro. Apart from the standard bone marrow micronucleus test in mice, which provided clear negative results (Table 4), the assays focused on the tissues that were thought to be most at risk. The liver is known to have nitroreductase activity [43,44,51] and could, therefore, be expected to produce reactive and/or genotoxic intermediates. Distribution studies showed relatively high tissue concentrations in the liver of rats, i.e. 4830 ng-eq/g of tissue, 1 h after treatment with a single oral dose of 60 mg/kg [14 C]AMP397. This was clearly >1754 ng-eq/g of tissue found in the bone marrow or the 1750 ng-eq/g of tissue measured in the plasma under the same test conditions. There was no evidence for the induction of gene mutations in MutaTM Mouse (Table 5), for DNA damage in rats, as measured in the comet assay (Table 6), or for DNA binding in rats or mice (Table 7). The intestinal tract was considered to be the appropriate target to investigate, because of the possibility that intestinal bacteria could form and release reactive intermediates into the lumen of the gut. The negative data from the MutaTM Mouse study (Table 5) and the comet assay with cells of rats clearly showed that there was no genotoxic insult detectable in the cells of the jejunum and the colon. The rat was used to conduct the comet assay since higher doses could be administered without strong systemic toxicity, which can be explained by the lower absorption of the test compound in rats as compared to mice. Rats could be treated with 2000 mg/kg, whereas, mice showed strong signs of systemic toxicity at 320 mg/kg. Con-
sequently, a higher exposure of the gut tissue to AMP397 and its potential reactive metabolites produced by the intestinal flora was achievable in rats. Assuming a comparable ratio between intestinal volume and body weight, the local intestinal exposure of a human volunteer/patient treated with 800 mg (highest dose used in a clinical tolerability study) would be >125 times lower than the exposure of the rats in the comet assay. In addition, there is evidence that the nitroreductase activity is much lower in human faecal samples than that found in samples from laboratory rodents [52]. Since, as shown by the data presented, reactive intermediates are formed exclusively by nitroreduction, it can be expected that the rodent models used to investigate the occurrence of genotoxic effects in the gut represent worst case scenarios. In conclusion, the unequivocally negative results from four in vivo mutagenicity assays give a high degree of confidence that AMP397 can be developed and used for the treatment of epilepsy without exposing the patients to an undue genotoxic hazard. References [1] P. Kwan, M.J. Brodie, Early identification of refractory epilepsy, N. Engl. J. Med. 342 (2000) 314–319. [2] R.S. Fischer, B.G. Vickrey, P. Gibson, B. Hermann, P. Penovich, A. Scherer, S. Walker, The impact of epilepsy from the patient’s perspective I. Descriptions and subjective perceptions, Epilepsy Res. 41 (2000) 39–51. [3] R.S. Fischer, B.G. Vickrey, P. Gibson, B. Hermann, P. Penovich, A. Scherer, S. Walker, The impact of epilepsy from the patient’s perspective II: views about therapy and health care, Epilepsy Res. 41 (2000) 53–61. [4] A. Tortorella, T. Halonen, N. Sahibzada, K. Gale, A crucial role of the alpha-amino-3-hydroxy-5-methylisoxazole-4propionic acid subtype of glutamate receptors in piriform and perirhinal cortex for the initiation and propagation of limbic motor seizures, J. Pharmacol. Exp. Ther. 280 (1997) 1401– 1405. [5] H. Katsumori, Y. Mibabe, M. Osawa, R. Ashby Jr., Acute effect of various GABA receptor agonists and glutamate antagonists on focal hippocampal seizures in freely moving rats elicited by low-frequency stimulation, Synapse 28 (1998) 103–109. [6] W. Löscher, Pharmacology of glutamate receptor antagonists in the kindling model of epilepsy, Prog. Neurobiol. 54 (1998) 721–741. [7] W. Nakahara, F. Fukuoka, T. Sugimuira, Carcinogenic action of 4-nitro-quinoline-1-oxide, Gann 48 (1957) 129–137. [8] V. Purohit, A.K. Basu, Mutagenicity of nitroaromatic compounds, Chem. Res. Toxicol. 13 (2000) 673–692.
W. Suter et al. / Mutation Research 518 (2002) 181–194 [9] S.M. Cohen, Toxicity and carcinogenicity of nitrofurans, in: G.T. Bryan (Ed.), Carcinogenesis, Vol 4, Nitrofurans, Raven Press, New York, 1978, pp. 171–231. [10] L. Dobias, M. Cerna, P. Roessner, R. Sram, Genotoxicity and carcinogenicity of metronidazole, Mutation Res. 317 (1994) 177–194. [11] W. Pfau, B.L. Pool-Zobel, C.-W. von der Lieth, M. Wiessler, The structural basis for the mutagenicity of aristolochia acid, Cancer Lett. 55 (1990) 7–11. [12] P.C. Howard, R.H. Heflich, F.E. Evans, F.A. Beland, Formation of DNA adducts in vitro and in Salmonella typhimurium upon metabolic reduction of the environmental mutagen 1-nitropyrene, Cancer Res. 43 (1983) 2052–2058. [13] T.C. Pederson, J.S. Siak, The activation of mutagens in diesel particle extract with rat liver S9 enzymes, J. Appl. Toxicol. 1 (1981) 61–66. [14] J.H. Tocher, Reductive activation of nitroheterocyclic compounds, Gen. Pharmac. 28 (1997) 485–487. [15] J. Ashby, R.W. Tennant, Chemical structure, Salmonella mutagenicity and extent of carcinogenicity as indicators of genotoxic carcinogenesis among 222 chemicals tested in rodents by the U.S. NCI/NTP, Mutation Res. 204 (1988) 17– 115. [16] M. Korbelik, D. Horvat, The mutagenicity of nitroaromatic drugs. Effect of metronidazole after incubation in hypoxia in vitro, Mutation Res. 78 (1980) 201–207. [17] J.R. Thornton-Manning, B.A. Smith, F.A. Beland, F.L. Heflich, S9-mediated metabolism of 1-nitropyrene to a mutagen in chinese hamster ovary cells by ring-oxidation under aerobic conditions and by nitroreduction under anaerobic condition, Carcinogenesis 12 (1991) 43–51. [18] M. Maron, B.N. Ames, Revised methods for the Salmonella mutagenicity tests, Mutation Res. 113 (1983) 173–215. [19] OECD guideline for testing of chemicals, Nr. 471, genetic toxicology: bacterial reverse mutation test, adopted 21 July 1997. [20] E.J. Rosenkranz, E.C. McCoy, R. Mermelstein, H.S. Rosenkranz, Evidence for the existence of distinct nitroreductases in S. typhimurium: roles in mutagenesis, Carcinogenesis 3 (1982) 121–123. [21] B. Miller, S. Albertini, F. Locher, V. Thybaud, E. Lorge, Comparative evaluation of the in vitro micronucleus test and the in vitro chromosome aberration test: industrial experience, Mutation Res. 392 (1997) 45–59. [22] J. Cole, D.B. Mc Gregor, M. Fox, R.C. Garner, J. Thacker, Gene mutation assay in cultured mammalian cells, in: D.J. Kirkland (Ed.), Basic mutagenicity tests UKEMS recommended procedures, Cambridge University Press, Cambridge, 1990, pp. 87–114. [23] OECD. Guideline for testing of chemicals, 476, genetic toxicology: in vitro mammalian cell gene mutation tests, adopted 21 July 1997. [24] W.D. Robinson, M.H.L. Green, J. Cole, R.C. Garner, M.J.R. Healy, D. Gatehouse, Statistical evaluation of bacterial/mammalian fluctuation tests, in: D.J. Kirkland (Ed.), Statistical evaluation of mutagenicity test data, Cambridge University Press, Cambridge, 1990, pp. 102–140.
193
[25] F. Romagna, C.D. Staniforth, The automated bone marrow micronucleus test, Mutation Res. 213 (1989) 91–104. [26] OECD guideline for the testing of chemicals, 474, genetic toxicology: mammalian erythrocyte micronucleus test, adopted 21 July 1997. [27] W. Frieauff, F. Romagna, Technical aspects of automatic micronucleus analysis in rodent bone marrow assays, Cell Biol. Toxicol. 10 (1994) 283–289. [28] BigBlueTM transgenic mouse mutagenesis assay system (stratagene, 1 May 1994). [29] J.A. Gossen, A.C. Molijn, G.R. Douglas, J. Vijg, Application of galactose sensitive E. coli strains as selective hosts for lacZ− plasmids, Nucl. Acid Res. 20 (1992) 3254. [30] J.A. Gossen, J. Vijg, A selective system for lacZ− phage using a galactose-sensitive E. coli host, Biotechniques 14 (1993) 326–330. [31] MutaTM mouse: mutation assay instruction manual (Covance Hazleton, version obtained on 15 July 1996). [32] G. Speit, A. Hartmann, The comet assay (single-cell gel test)—A sensitive genotoxicity test for the detection of DNA damage and repair, DNA Repair Protocols 113 (1999) 203– 212. [33] R.R. Tice, E. Agurell, D. Anderson, B. Burlinson, A. Hartmann, H. Kobayashi, Y. Miyamae, E. Rojas, J.-C. Ryu, Y.F. Sasaki, The single cell gel/comet assay: guidelines for in vitro and in vivo genetic toxicology testing, Environ. Mol. Mutagen. 35 (2000) 206–221. [34] W. Frieauff, A. Hartmann, W. Suter, Automatic analysis of slides processed in the comet assay, Mutagenesis 16 (2001) 133–137. [35] P.L. Olive, Cell proliferation as a requirement for development of the contact effect in chinese hamster V79 spheroids, Radiation Res. 117 (1989) 79–92. [36] W.K. Lutz, In vivo covalent binding of organic chemicals to DNA as a quantitative indicator in the process of chemical carcinogenesis, Mutation Res. 65 (1979) 289–356. [37] W.K. Lutz, Quantitative evaluation of DNA binding data for risk estimation and for classification of direct and indirect carcinogens, J. Cancer Res. Clin. Oncol. 112 (1986) 85–91. [38] P. Sagelsdorff, W.K. Lutz, C. Schlatter, The relevance of covalent binding to mouse liver DNA to the carcinogenic action of hexachlorocyclohexane isomers, Carcinogenesis 4 (1983) 1267–1273. [39] P. Sagelsdorff, W.K. Lutz, C. Schlatter, DNA methylation in rat liver by daminozide, 2,2-dimethylhydrazine, and dimethylnitrosamine, Fundamental Appl. Toxicol. 11 (1988) 723–730. [40] J. McCann, E. Choi, E. Yamasaki, B.N. Ames, Detection of carcinogens as mutagens in the Salmonella/microsome test: assay of 300 chemicals, Proc. Natl. Acad. Sci. U.S.A. 72 (1975) 5135–5139. [41] J.M. Gentile, G.J. Gentile, M.J. Plewa, Mutagenicity of selected aniline derivatives to Salmonella following plant activation and mammalian hepatic activation, Mutation Res. 188 (1987) 185–196. [42] M.W. Chou, R.H. Heflich, P.P. Fu, Metabolism of 1-nitrobenzo(a)pyrene by rat liver microsomes to potent mutagenic metabolites, Carcinogenesis 7 (1986) 1837–1844.
194
W. Suter et al. / Mutation Research 518 (2002) 181–194
[43] A.K.D. Hajos, G.W. Winston, Dinitropyrene nitroreductase activity of purified NAD(H)-quinone oxidoreductase: role in rat liver cytosol and induction by aroclor-1254 pretreatment, Carcinogenesis 12 (1991) 697–792. [44] D.B. McGregor, I. Edwards, C.R. Wolf, L.M. Forrester, W.J. Caspary, Endogenous xenobiotic enzyme levels in mammalian cells, Mutation Res. 261 (1991) 29–39. [45] R. Fahrig, M. Engelke, Reinvestigation of in vivo genotoxicity studies in man. Part I: no induction of DNA strand breaks in peripheral lymphocytes after metronidazole therapy, Mutation Res. 395 (1997) 215–221. [46] E. Touati, M. Hofnung, J.M. Thiberge, V. Michel, A. Labigne, P.J. Jenks, Short-term infection with Helicobacter pylori and 1 week exposure to metronidazole does not enhance gastric mutation frequency in transgenic mice, J. Antimicrobic. Chemother. 46 (2000) 987–992. [47] D. Menéndez, E. Rojas, L.A. Herrera, M.C. Lopez, M. Sordo, G. Elizondo, P. Ostrosky-Wegman, DNA breakage due to metronidazole treatment, Mutation Res. 478 (2001) 153–158.
[48] W. Suter, R. Ahiabor, B. Blanco, F. Locher, F. Mantovani, M. Robinson, G. Sreenan, F. Staedtler, T. Swinger, A. Vignutelli, E. Perentes, Evaluation of the in vivo genotoxic potential of three carcinogenic aromatic amines using the BigBlueTM transgenic mouse mutation assay, Environ. Mol. Mutagen. 28 (1996) 354–362. [49] K.B. Declos, C.E. Cerniglia, K.L. Dooley, W.L. Campbell, W. Franklin, R.P. Walker, The role of intestinal microflora in the metabolic activation of 6-nitrochrysene to DNA-binding derivatives in mice, Toxicology 60 (1990) 137–150. [50] Y.F. Sasaki, K. Sekihashi, F. Izumiyama, E. Nishidate, A. Saga, K. Ishida, S. Tsuda, The comet assay with multiple mouse organs: comparison of comet assay results and carcinogenicity with 208 chemicals selected from IARC monographs and U.S. NTP carcinogenicity database, Crit. Rev. Toxicol. 30 (2000) 629–799. [51] D.E. Rickert, Metabolism of Nitroaromatic compounds, Drug Metabol. Rev. 18 (1987) 23–53. [52] R. Rowland, Factors affecting metabolic activity of the intestinal microflora, Drug Metabol. Rev. 19 (1988) 243–261.
Genotoxicity assessment of the antiepileptic drug AMP397, an Ames-positive aromatic nitro compound Willi Suter a,∗ , Andreas Hartmann a , Franziska Poetter a , Peter Sagelsdorff b , Peter Hoffmann a , Hans-Jörg Martus a a
Toxicology/Pathology, Novartis Pharma AG, 4002 Basel, Switzerland b Syngenta AG, Basel, Switzerland
Received 29 January 2002; received in revised form 19 April 2002; accepted 22 April 2002
Abstract AMP397 is a novel antiepileptic agent and the first competitive AMPA antagonist with high receptor affinity, good in vivo potency, and oral activity. AMP397 has a structural alert (aromatic nitro group) and was mutagenic in Salmonella typhimurium strains TA97a, TA98 and TA100 without S9, but negative in the nitroreductase-deficient strains TA98NR and TA100NR. The amino derivative of AMP397 was negative in wild-type strains TA98 and TA100. AMP397 was negative in a mouse lymphoma tk assay, which included a 24 h treatment without S9. A weak micronucleus induction in vitro was found at the highest concentrations tested in V79 cells with S9. AMP397 was negative in the following in vivo studies, which included the maximum tolerated doses of 320 mg/kg in mice and 2000 mg/kg in rats: MutaTM Mouse assay in colon and liver (5 × 320 mg/kg) at three sampling times (3, 7 and 31 days after the last administration); DNA binding study in the liver of mice and rats after a single treatment with [14 C]-AMP397; comet assay (1 × 2000 mg/kg) in jejunum and liver of rats, sampling times 3 and 24 h after administration; micronucleus test (2 × 320 mg/kg) in the bone marrow of mice, sampling 24 h after the second administration. Based on these results, it was concluded that AMP397 has no genotoxic potential in vivo. In particular, no genotoxic metabolite is formed in mammalian cells, and, if formed by intestinal bacteria, is unable to exert any genotoxic activity in the adjacent intestinal tissue. These data were considered to provide sufficient safety to initiate clinical development of the compound. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Antiepileptic; Nitroaromatic compound; Nitroreductase; DNA binding; Comet assay; MutaTM Mouse
1. Introduction Epilepsy is one of the most common neurological disorders, with prevalence in excess of 0.5% of the world population. Despite a variety of antiepileptic drugs available, there is still a high medical need for improved treatments of epilepsy. About 30% of ∗ Corresponding author. Tel.: +41-61-3241341; fax: +41-61-3241274. E-mail address: [email protected] (W. Suter).
patients see their seizures inadequately controlled [1], and many of those who become free of seizures suffer from adverse effects, the most frequent complaints being cognitive impairment and decrease in overall energy level [2,3]. It is now well established that the activation of alpha-amino-3-hydroxy-5-methyl4-isoxazole-propionic acid (AMPA) receptors is involved in the initiation and propagation of seizures [4–6], therefore, AMPA receptor antagonism appears a logical strategy for the development of new anticonvulsant drugs. AMP397 was selected for the
1383-5718/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 3 - 5 7 1 8 ( 0 2 ) 0 0 1 0 5 - 5
182
W. Suter et al. / Mutation Research 518 (2002) 181–194
Fig. 1. Structures of AMP 397 (top) and AAE669 (bottom).
development as an antiepileptic drug because it combines high affinity for the AMPA receptor, good in vivo potency and oral anticonvulsant activity in rodent seizure models. The compound is a nitro-quinoxaline derivative (Fig. 1), which caused concern in toxicology because of the possible mutagenic and carcinogenic potential of the aromatic nitro group. The genotoxicity and carcinogenicity of nitroaromatic compounds has been studied intensively over the last 50 years beginning with 4-nitroquinoline-1-oxide, which was found to be a potent carcinogen [7]. Nitroaromatic compounds are found as air pollutants, in cigarette smoke and in grilled foods. They are important synthesis intermediates, e.g. for the production of azo dyes and are used as herbicides as well as human and veterinary medicine [8]. Among the latter, nitrofuran derivatives [9], mainly used to treat topical and urinary tract infection, metronidazole [10], used against Trichomonas vaginalis infections, and aristolochic acid, a mixture of naturally occurring carboxylic acids as a component of herbal medicines prepared from the Aristolochia plant [11], have been extensively studied. Many nitroaromatic compounds have been shown to
bind covalently to DNA [8,12]. The reactive forms are metabolically generated through nitroreduction and, in many cases, through oxidative pathways. Whereas, the oxidative pathways depend on the presence of the cytochrome P450 family of enzymes and occur, therefore, mainly in the liver, nitroreduction is found in bacterial as well as in mammalian cells [13,14]. This may explain why, nitro compounds are generally fairly strong mutagens in the Salmonella mutagenicity assay, whereas, mutagenicity in mammalian cells and carcinogenicity in the rodent is not always found for Salmonella-positive nitro compounds. This has resulted in labelling the Salmonella mutagenicity assay as ‘not predictive’ for this class of compounds [15]. The evaluation of the mutagenic potential of a nitro compound, therefore, relies strongly on the outcome of tests conducted in mammalian systems. The results of the standard in vitro mammalian assays have to be interpreted with caution, since nitroreduction is oxygen sensitive so that in vitro metabolism by rat liver S9 may produce a completely different spectrum of metabolites than observed in vivo [16,17]. In addition, metabolic activation, i.e. nitroreduction by the intestinal flora and the possible effect of bacterial metabolites on mammalian cells cannot be appropriately studied in vitro. It was, therefore, decided to study the genotoxic potential of AMP397 in an extensive battery of in vivo studies, focusing on well-exposed and metabolically active tissues, where the likelihood to detect genotoxic effects was considered to be highest.
2. Materials and methods 2.1. Test articles AMP397 (CAS registry number 188696-80-2) was synthesised in the laboratories of Nervous System Research, Novartis Pharma AG, Basel, Switzerland. The positive control chemicals were purchased from the following manufacturers: 9-Aminoacridine (Fluka), 2-aminoanthracene (Fluka), benzo(a)pyrene (Sigma), cyclophosphamide (ASTA medica AG), ethylmethane sulfonate (Aldrich), methylmethane sulfonate (Sigma), mitomycin C (Boehringer Ingelheim), 2nitrofluorene (Merck), N-nitroso-N -ethylurea (Sigma), sodium azide (Fluka).
W. Suter et al. / Mutation Research 518 (2002) 181–194
183
2.2. Rat liver S9
2.4. V79 micronucleus test
The S9 liver homogenate was prepared as described in the literature [18]. Seven to 9-week-old male Crl:Wist Han (Charles River, Germany) were injected intraperitoneally with 500 mg/kg Aroclor 1254 and sacrificed 5 days later. Their livers were homogenised, diluted 1:4 with 0.15 M KCl and centrifuged for 10 min at 9000 g. The supernatant was frozen in small aliquots and stored at −70 to −80 ◦ C until use. The S9-mix used for the Salmonella mutagenicity experiments was prepared according to Maron and Ames [18] with 10% rat liver S9-fraction. The composition and final concentrations of the S9-mix used for the V79 micronucleus was as follows: Glucose–6–phosphate, 4.4 mM; NADP, 0.84 mM; KCl, 30 mM; 0.032% NaHCO3 ; S9 fraction, 10%. For the mouse lymphoma assay the final concentrations of the S9-mix components in the treatment medium were as follows: Hanks BSS, Glucose–6–phosphate, 5.9 mM; NADP, 0.32 mM; KCl, 1.5 mM; S9 fraction, 2%.
V79 chinese hamster cells were obtained from M. Fox (Paterson Laboratories, Manchester, UK). The assay was conducted as described in detail by Miller et al. [21]. In brief, based on the data of a preliminary cytotoxicity test using (established by use of the MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide, Sigma) exclusion assay), 6 concentrations, leading to a reduction of MTT uptake of 25–80%, were selected for the main experiment. For the micronucleus experiment, the cells were treated 1 day after subculture with the test compound solution for 3 h (with and without S9) or for 20 h (without S9). After the treatment, cells were washed with PBS and further incubated with fresh medium for 45 (or, 28 in case of 20 h treatment) hours. Before sampling, the cells were examined qualitatively for morphology and growth (degree of confluence). Based on this assessment, the concentration that inhibited growth by approximately 50%, and two lower concentrations were chosen for analysis. A period of 48 h after the start of treatment, cells were removed with trypsin EDTA (Gibco BRL) and spread on glass slides by cytocentrifugation (Shandon Cytospin). Cells were fixed and stained with Schiff reagent (Merck). 4000 cells (2000 cells per culture) were analysed automatically using a LEITZ MIAS image analyser from at least three concentration groups of each test compound as well as from negative and positive controls. A result was considered positive if the micronucleus frequency was ≥1% and at least 0.6 above the concurrent solvent-control value. These criteria are based on an in house validation study of the assay. The historical mean solvent-control data were as follows: 3 h treatment without S9, 0.54 ± 0.17; 20 h treatment without S9, 0.60 ± 0.16; with S9, 3 h treatment with S9, 0.61 ± 0.19.
2.3. Ames test The method used followed the recommendations by Maron and Ames [18] and the OECD guideline [19]. The Salmonella typhimurium strains TA1535, TA97a, TA98, TA100 and TA102 were obtained from B.N. Ames, Biochemistry Department, University of California, Berkeley, CA, USA. The nitroreductase resistant strains were provided by Dr. P. Arni (Syngenta AG, Basel, Switzerland) who had obtained them from Dr. H.S. Rosenkranz (Pittsburgh, PA, USA). These bacteria, S. typhimurium strains TA98NR and TA100NR, are lacking the ‘classical’ nitroreductase and were isolated as niridazole resistant derivatives of S. typhimurium strains TA98 and TA100, respectively [20]. A test compound was judged to be mutagenic in the plate test if it produces, in at least one concentration and one strain, a response equal to twice (or more) the control incidence. The only exception is strain TA102, which has a relatively high spontaneous revertant number, where an increase by a factor of 1.5 above the control level is taken as an indication of a mutagenic effect.
2.5. Mouse lymphoma thymidine kinase assay L5178Y mouse lymphoma cells, tk+/− clone 3.7.2C, were obtained from American Type Culture Collection by D. Geleick, Ciba Geigy Ltd., Basle, and were kept as frozen stock in liquid nitrogen. The test compound was dissolved in 0.5 M Na2 HPO4 at the maximum soluble concentration of 22 mg/ml. The final concentration of the solvent in the treatment
184
W. Suter et al. / Mutation Research 518 (2002) 181–194
medium was 2% in Experiment 1 and part of Experiment 2. In part of Experiment 2 and in Experiment 3 the end concentration of the 0.5 M Na2 HPO4 was increased to 4% in order to reach higher test compound concentrations and achieve stronger cytotoxicity. The fluctuation protocol with treatment times of 3 h (with and without S9-mix) and 24 h (without S9-mix) was used as described by Cole et al. [22] and in the corresponding OECD guideline [23]. All calculations were performed with ‘MUTANT’ software for fluctuation assays (York Electronic Research, UK). These calculations are based on recommendations by UKEMS guidelines found in the literature [24]. The concentrations for the mutagenicity experiments were chosen based on the results from a preliminary dose-finding experiment using the following calculations: adapted survival (%S) was calculated by multiplying ‘cloning efficiency’ by ‘cell count day 0’ of the corresponding culture divided by ‘cell count day 0’ of the solvent-control cultures. Relative adapted survival (%RS) was calculated from %S by setting the solvent-control value to 100% and calculating values of treated groups relative to it. On the basis of these results, at least five concentrations between 10 and 20% and 100% RS were chosen for the mutagenicity experiment. 2.6. Animal husbandry Generally, mice were kept individually in type II Macrolon® cages and rats pairwise in Macrolon© Type 4 cages on sawdust bedding. For the DNA binding study the animals were housed in groups of two animals (rats) and four animals (mice). The animals were fed standard rodent chow NAFAG 8900 FOR GLP (Nafag, Gossau SG, Switzerland) and given tap water in bottles ad libitum. The animals were housed in a fully air-conditioned animal room, at 22 ± 3 ◦ C, at a relative humidity of 35–75% and a 12 h light/dark cycle. Young adult animals were used in all studies. 2.7. Bone marrow micronucleus test in mice treated by the oral route The study protocol was based on the method developed by Romagna and Staniforth [25] and was in compliance with the corresponding OECD guideline [26]. The test article was dissolved in 0.2 M NaHCO3 . The
dose-finding experiment with AMP397 showed that treatment of CD-1 mice (Charles River, Germany) by oral gavage with 450.5, 500 or 800 mg/kg, twice with an interval of 24 h, led to strong signs of toxicity such as laboured breathing, ataxia, and strong sedation. At 320 mg/kg, the same symptoms were visible, but with less severity, and no animals died. On the basis of these results, doses of 32, 100 and 320 mg/kg were chosen for this micronucleus test. In the main experiment five male and five female mice were treated as described above and bone marrow was sampled 48 h after the first application. Nucleated cells were removed from the bone marrow samples using cellulose columns. The cells were loaded on poly-l-lysine coated glass slides by cytocentrifugation using a Shandon Cytospin stained with May Grünwald stain (5%) and Giemsa (14%). The slides were automatically evaluated with a LEITZ MIAS image analyser [27]. No statistical analysis was performed since all values in the treated groups were ≤ the frequency of micronucleated polychromatic erythrocyte in the concurrent vehicle control group. 2.8. Gene mutation assay in MutaTM Mouse In a dose-finding experiment groups of three male CD-1 mice were treated by oral gavage on five consecutive days with 125, 200 or 320 mg/kg per day. Clear signs of toxicity such as ataxia, sedation, Straub fur or closed eyes were observed in all dose groups. At 320 mg/kg per day, these symptoms were very severe and this dose was, therefore, chosen for the mutagenicity test. A total of 320 mg/kg per day AMP397 was administered on five consecutive days in a volume of 20 ml/kg, and using 0.2 M NaHCO3 as the vehicle to groups of seven to nine male mice (MutaTM Mouse, Covance Research Products, USA). The AMP397- treated animals were sacrificed on 3, 10 and 31 days after the last application, the vehicle control animals on day 31 and the three positive control animals (50 mg/kg per day N-nitroso-N -ethylurea) on day 10. Colon and livers were sampled, flash frozen in liquid nitrogen and stored at −80 ◦ C until analysis. Genomic DNA was isolated using the phenol/chloroform method [28]. For in vitro packaging of the DNA into the lambda phage vector TranspackTM (Stratagene, USA) was used under the conditions recommended by the manufacturer. The mutant
W. Suter et al. / Mutation Research 518 (2002) 181–194
185
frequencies were determined using Escherichia coli galE− LacZ− , which allows selection for LacZdeficient mutants [29,30]. The test conditions and procedures were chosen in accordance with the recommendations by Covance [31]. In particular, 300’000 plaque forming units (pfu) derived from at least four independent packaging reactions were analysed per DNA sample. No statistical analysis of the data was conducted, because all mutant frequencies of the treated groups were lower than the concurrent vehicle control value.
of the inevitable bacterial contamination of jejunum samples, they were analysed semi-automatically with the Comet Assay II image analysis system (Perceptive Instruments). Fifty cells per slide were evaluated on each of the two replicate slides. The parameter measured was the tail moment as described by Olive et al. [35]. The statistical evaluation was conducted using the SAS software package (SAS version 6.12). An ANOVA was followed by a Dunn test (because of heterogeneity) or a Bartlett’s test in case of the 24 and 3 h experiment, respectively.
2.9. Comet assay with jejunum and liver cells of rats
2.10. DNA binding study in rats and mice
The alkaline comet assay protocol used was based on published recommendations and descriptions of methods [32,33]. Groups of five male Crl:Wist Han rats (Charles River, Germany) were administered once by oral gavage a dose of 2000 mg/kg (suspended in sodium carboxymethylcellulose solution [CMC] 1%, application volume 5 ml/kg). Vehicle control animals received CMC 1% (5 ml/kg) and positive control animals 300 mg ethylmethane sulfonate suspended in 1% CMC. Jejunum and liver were sampled 3 or 24 h after the application. The tissue was washed in HBSS/25 mM EDTA/10% DMSO and afterwards kept in buffer on ice until further preparation. For the preparation of single cell suspensions a small piece of the tissue was taken and a small amount of buffer was pipetted onto the tissue; then the tissue was minced with a scalpel into fine pieces to obtain a cell suspension. A volume of 15 l of the cell suspension was then mixed 1:10 with 0.5% LMP-agarose. A volume of 45 l of the agarose suspension were put on the slides; two samples were placed on one slide and each was covered with a 24 mm × 24 mm coverslip until the agarose hardened. The coverslip was then removed and the slide was immersed into a jar containing lysis buffer (2.5 mM NaCl, 100 mM EDTA, 10 mM Tris, 10% DMSO, 1% sodium lauroyl sarcosinate, 1% Triton X-100). Alkaline unwinding time was 15 min (jejunum) or 20 min (liver), electrophoresis 20 min. After neutralisation, slides were alcohol-dried and afterwards stained with 2.5 g/ml propidium iodide in Vectashield (Vector Laboratories, USA). Liver samples were analysed fully automatically using a Leitz MIAS image analysis system [34]. Because
Two rats and eight mice were treated with [14 C]AMP397 at a dose level of 13.2 mg/kg body weight (2 mCi/kg body weight). The animals received 10 ml/kg body weight of the dosing solution by single oral gavage and were killed after 24 h. The third rat remained untreated and was used to show that the work-up of the samples was performed without external contamination with radiolabel. A radioactivity dose of about 2 mCi/kg body weight was administered to achieve a limit of detection for the CBI in the liver of <0.1. Compounds with a CBI of <0.1 were defined to be non-genotoxic in this assay [36,37]. The chemical dose was based on the specific activity of the radiolabeled test compound. Individual livers of rats and two pools of livers, each prepared from four identically treated mice, were homogenised in 75 mM NaCl, 10 mM Tris/HCl, 10 mM EDTA, pH 7.8, and DNA was isolated as described by Sagelsdorff et al. [38,39]. Essentially, chromatin was precipitated and the pellet was blended in a denaturing lysing medium, deproteinated with chloroform/isoamyl alcohol/phenol (CIP), extracted with diethyl ether and the DNA was further purified by adsorption on a hydroxylapatite column, dialysis and precipitation with ethanol. The highly purified DNA was dissolved in 20 mM succinic acid containing 8 mM CaCl2 and an aliquot was used for UV-determination of the DNA content. Another aliquot was mixed with scintillation cocktail for the determination of radioactivity. To determine background radioactivity, DNA was isolated accordingly from the liver of the untreated rat. The total count—upon comparison with historical controls—was used to show that the work-up of the
186
W. Suter et al. / Mutation Research 518 (2002) 181–194
samples was performed without external contamination with radiolabel. On the basis of the background values compiled during the last years, the mean historical background and the respective S.D. was calculated. Assuming that a vial containing a DNA sample with low amount of radioactivity shows the same statistic variation as the background, a limit of detection for radioactivity in a vial was calculated on a level of 2 S.D. from the historical background. If no radioactivity was detectable in a DNA sample this limit of detection was used to calculate the maximum possible covalently bound DNA radioactivity. The covalently bound DNA radioactivity was converted to the units of the covalent binding index (CBI) according to the following equation [36]: CBI = =
µmol chemical bound/mol DNA nucleotides mmol chemical applied/kg body weight dpm/mg DNA × 309 × 106 dpm/kg body weight
The average weight of 1 mol of DNA nucleotides is 309 g.
3. Results and discussion The results are shown in detail in Tables 1–7 and in Fig. 2. A summarised overview is presented in Table 8. The standard Ames test showed, in the absence of rat liver S9, a clear dose dependent increase in rever-
tant numbers in strain TA98 and positive results also in strains TA100 and TA97a (Table 1). This result was not surprising, since many aromatic nitro compounds are positive in this assay [15]. The comparison with published Salmonella mutagenicity data shows that AMP397 is one of the weakest nitroaromatic mutagens known. It induced 0.077 revertants/nmol, which lies at the lower end of the potency ranges found by McCann et al. [40] (0.26–20800 revertants/nmol) and compiled by Purohit and Basu ([8] 0.0029–734000 revertants/nmol). This could be due to one or several of the following reasons: (i) lack of uptake of the test compound by the bacteria; (ii) low activation rate by the ‘classical’ nitroreductase; (iii) low affinity of the reactive metabolite(s) for the DNA; (iv) formation of weakly mutagenic types of DNA damage in bacteria. The absence of an effect in nitroreductase-deficient strains and the negative results obtained with the amino analogue AAE669 (Fig. 2) indicated that the nitro group is metabolised and activated to electrophilic intermediates by the ‘classical’ nitro reductase [20] in S. typhimurium. It is not clear, however, why mutagenicity occurred only in strain TA98 and only at higher concentrations in the preincubation assay as compared to the effects found in the plate incorporation assay, where mutagenicity was observed in S. typhimurium strains TA97a, TA98 and TA100. The plate incorporation assay was also found to be more sensitive than the preincubation assay for the detection of the mutagenicity of 4-nitro-phenylenediamine [41]. In the absence of direct measurements of metabolism on the agar plates it is also difficult to explain why
Fig. 2. Mutagenicity data in S. typhimurium strain TA98 without S9. The experiments were conducted under standard conditions and using the plate incorporation method, (䊉) EE699 (amino analogue of AMP397) standard test; (䊏) MP397 tested with TA98NR (nitroreductase deficient); (䉬) MP397 standard test.
W. Suter et al. / Mutation Research 518 (2002) 181–194
187
188
W. Suter et al. / Mutation Research 518 (2002) 181–194
Table 2 Micronucleus test in V79 chinese hamster fibroblasts Treatment (20 h, −S9)
Cells with MN (%)
Treatment (3 h, +S9)
Cells with MN (%)
0.50 0.68 0.28 0.78 Too toxic
S9 412.3 g/ml 833.0 g/ml 1682.9 g/ml 3400.0 g/ml
0.73 0.73 0.75 1.38 Too toxic
19.30
CP, 20 M
2.60
Cells with MN (%)
Treatment (3 h, +S9)
Cells with MN (%)
Experiment 2 MEM 2537.3 g/ml 2893.6 g/ml 3300.0 g/ml
0.43 0.55 0.75 0.60
S9 538.6 g/ml 790.6 g/ml 1160.4 g/ml 1703.2 g/ml
0.65 0.75 0.88 1.40 Too toxic
EMS, 12.5 mM
9.80
CP, 17.5 M
6.80
Experiment 1 MEM 412.3 g/ml 833.0 g/ml 1682.9 g/ml 3400.0 g/ml EMS, 6 mM Treatment (3 h, −S9)
The numbers (% cells with MN) represent the percentage of cells with one or more micronuclei. Historical negative control values: without S9, 3 h treatment 0.54 ± 0.17; without S9, 20 h treatment, 0.60 ± 0.16; without S9, 3 h treatment, 0.61 ± 0.19. Abbreviation, MN: micronucleus.
AMP397 was not mutagenic for Salmonella in the presence of S9. In rats and dogs it was found that metabolism does not play an important role in the elimination of AMP397 and that binding to plasma proteins was low. It appears unlikely, therefore, that the compound was rapidly degraded or bound by proteins of the S9-mix. Bacteriotoxicity did not provide information to answer this question, because the test compound was not bacteriotoxic under all test conditions up to the highest soluble concentrations. One is, therefore, left with the speculation that the bacteria might absorb nucleophilic products from the S9-mix, which react with the reactive intermediates formed by the bacterial nitroreductase. The absence of a mutagenic effect in the presence of S9 is important, because it shows that mammalian liver enzymes in general and mammalian nitroreductases in particular, do not metabolise AMP397 into mutagenic intermediates under the conditions of the Ames test. Other nitroaromatic compounds have been found to be activated to mutagenic intermediates by mammalian nitroreductases [12] and by nitroreductase-independent pathways [42].
The two in vitro mammalian assays conducted, i.e. the V79 micronucleus test and the mouse lymphoma thymidine kinase assay, gave contradictory results. The V79 micronucleus test showed a weak increase in the frequency of micronucleated cells in the presence of S9 at the highest concentration that could be evaluated (1682.9 and 1160.4 g/ml in the first and second experiment, respectively (Table 2)). The highest non-mutagenic concentrations (833 and 790.6 g/ml, in the first and second experiment, respectively) were about 400 times higher than the peak plasma levels measured in man after a single oral treatment with up to 800 mg (about 2 g/ml). The mouse lymphoma assay provided negative results up to the precipitating concentration of 880 g/ml. There are at least three possible explanations for the contradictory results obtained in the V79 micronucleus test and the mouse lymphoma assay: (i) the test conditions used for the two studies were different with regard to cell type, genetic endpoint, S9 concentration, etc.; (ii) the V79 micronucleus test and the mouse lymphoma assay were performed with different batches of test compound. The data of the chemical analysis showed that the batch used for the V79 micronucleus test (purity 97%) was slightly less pure than the material used for all other studies (purity 98.3%). Since the impurities have not been identified, it is not clear whether the contradictory results shown in Tables 2 and 3 are due to the presence of mutagenic impurities in the batch used for the V79 micronucleus test; (iii) for unknown reasons the test material used for the mouse lymphoma assay was less soluble and precipitated at concentrations ≥440 g/ml, which prevented the use of the clearly cytotoxic concentrations, required to induce an increased micronucleus frequency in V79 cells. In conclusion, it is not entirely clear whether AMP397 can be activated to mutagenic intermediates by S9-mix from Aroclor 1254 pre-treated rats under some test conditions at very high concentrations. However, as shown in Table 1, under the conditions of the Ames test, no induction of revertants was found and, thus, an activation of AMP397 can be excluded with certainty. It is important to mention in this context that pre-treatment of rats with Aroclor 1254 leads to a significant increase in cytosolic nitroreductase activity [43], which indicates that the S9-mix used for the in vitro experiments had a comparatively high nitroreductase activity.
W. Suter et al. / Mutation Research 518 (2002) 181–194
189
Table 3 Mouse lymphoma thymidine kinase assay 24 h Treatment (g/ml)
Experiment 1 0 55 110 220 440 P Linear trend MMS 5 3 h Treatment (g/ml)
Experiment 2 0 55 110 220 440 $$,P Linear trend MMS 10 0 880
P
Linear trend
−S-9
3 h treatment (g/ml)
%RS
RTG
MF§
100.00 97.06 78.51 92.51 77.55
1.00 1.20 0.96 1.12 0.97
278.49 211.19 NS 260.08 NS 291.52 NS 181.47 NS
0 55 110 220 440 P
NS
Linear trend B(a)P 1.5
34.09
0.22
2303.34
−S-9
3 h treatment (g/ml)
%RS
RTG
MF§
100.00 89.20 83.85 79.42 79.07
1.00 1.00 0.97 0.99 (0.81)
142.75 150.63 NS 173.62 NS 168.58 NS (221.49)
0 55 110 220 440 P
NS 60.02
0.42
1163.09
Linear trend B(a)P 1.5
100.00 77.24
1.00 0.92
200.70 156.10 NS
0 880
NS
Linear trend
24 h Treatment (g/ml)
Experiment 3 0 110 220 440 P 880 P Linear trend MMS 5
P
+S-9 %RsS
RTG
MF§
100.00 113.57 107.18 104.27 112.92
1.00 0.98 1.23 1.46 1.48
236.63 216.95 321.78 182.61 205.89
NS NS NS NS
NS 8.49
0.07
14704.81
%RS
RTG
MF§
100.00 104.46 100.25 86.99 94.40
1.00 1.13 1.16 1.29 1.15
215.66 170.58 199.00 191.02 184.75
+S-9
NS NS NS NS
NS 1.50
0.01
3634.17
100.00 94.42
1.00 1.09
213.97 195.01 NS NS
−S-9 %RS
RTG
MF§
100.00 98.59 83.51 72.36 70.75
1.00 1.04 1.35 0.89 0.82
119.27 122.28 132.81 144.26 132.07
NS NS NS NS
NS 47.95
0.50
1259.99
Solvent 0.5 M Na2 HPO4 . Experiment 1 and Experiment 2 (upper table), end concentration of the solvent 2%; Experiment 2 lower table and Experiment 3, end concentration of the solvent 4%. Abbreviations: §, 5-TFT resistant mutants/106 viable cells 2 days after treatment; %RS, percent relative survival adjusted by post treatment cell count factor (cytotoxicity parameter); RTG, relative total growth (cytotoxicity parameter); MF, mutant frequency; P, precipitating dose. NS, not significant; $, treatment has high heterogeneity, but is included in analysis; $$, treatment excluded due to excessive heterogeneity.
Negative and positive mutagenicity results in mammalian cells in vitro have been reported for several nitroaromatic compounds. In all cases the genotoxic potential was found to be much lower in mammalian
cells as compared to bacteria, which can be explained by the high nitroreductase activity in the bacterial tester strains. Metronidazole has been particularly well studied [8,10,44–47] and it appears that in mammalian
190
W. Suter et al. / Mutation Research 518 (2002) 181–194
Table 4 Bone marrow micronucleus test in mice Group
Dose (mg/kg)
Sex
Number of mice
Frequency (%) MPE
Control AMP397 AMP397 AMP397 Cyclophosphamide
Vehicle 32 100 320 40
m m m m m
+f +f +f +f +f
5 5 5 4 5
+5 +5 +5 +5 +5
0.16 0.16 0.16 0.14 1.46
PE ± ± ± ± ±
0.04 0.06 0.05 0.03 0.23
30.6 33.0 37.3 35.5 29.1
± ± ± ± ±
7.5 9.8 13.2 5.7 9.6
Each value represents mean ± S.D. of a treatment group consisting of 10 (highest test compound dose group: 9) animals (4000 PE/animal). Abbreviations: m, male; f, female; PE, polychromatic erythrocytes; MPE, micronucleated polychromatic erythrocytes; frequency (%), percentage of polychromatic erythrocytes carrying micronuclei (MPE) or percentage of polychromatic erythrocytes in the total number of erythrocytes (PE).
Table 5 MutaTM Mouse LacZ assay Group
Treatment (five doses/5 days)
Dose (mg/kg)
Sampling time (days after last treatment)
Animals per group
Mutant frequency (mean ± SD)
Mean mutant frequencies in colon 0 Vehicle I AMP397 II AMP397 III AMP397 ENU N-Nitroso-N -Ethylurea
0 320 320 320 50
31 3 10 31 10
Seven Males Seven Males Seven Males Nine Males Three Males
9.58 7.03 9.26 9.22 197.65
± ± ± ± ±
2.67 1.09 3.10 2.31 41.06
Mean mutant frequencies in liver 0 Vehicle I AMP397 II AMP397 III AMP397 ENU N-Nitroso-N -Ethylurea
0 320 320 320 50
31 3 10 31 10
Seven Males Seven Males Seven Males Nine males Three Males
7.25 6.26 6.98 6.64 22.75
± ± ± ± ±
1.42 1.56 1.39 1.69 4.42
Mean mutant frequencies ± S.D. are shown for each treatment group.
Table 6 Comet assay in rats Treatment
Dose (mg/kg)
Organ
Number of animals
Tail moment (mean ± S.D.) 3 h Sampling
24 h Sampling
Sampling 3 h after administration Control Vehicle AMP397 2000 EMS 300
Jejunum Jejunum Jejunum
Five Males Five Males Five Males
0.48 ± 0.16 0.50 ± 0.31 1.17 ± 0.46
0.34 ± 0.15 0.65 ± 0.31 0.94 ± 0.32
Control AMP397 EMS
Liver Liver Liver
Five Males Five Males Three Males
1.44 ± 0.34 1.37 ± 0.37 11.50 ± 3.08
1.30 ± 0.19 1.22 ± 0.13 10.17 ± 1.80
Vehicle 2000 300
Abbreviation: S.D. is standard deviation.
W. Suter et al. / Mutation Research 518 (2002) 181–194
191
Table 7 Radioactivity in liver DNA isolated 24 h after oral administration of [14 C]AMP397 to rats and mice Animal No.a
Rat number 1
Rat number 2
Mice number 3–6
Mice number 7–10
Effective dose (mg/kg BW) (dpm/kg BW)·10−9
10.6 3.55
10.6 3.57
11.0 3.68
11.1 3.71
DNA Amount in vial (mg) Gross activity (cpm) Net activity (cpm)b Net activity (dpm)c Specific activity (dpm/mg)
3.49 14.1 <3.13 <3.91 <1.12
3.28 13.9 <3.13 <3.91 <1.19
3.64 13.3 <3.13 <3.91 <1.07
3.82 14.2 <3.13 <3.91 <1.02
Binding to DNA CBI unitsd
<0.10
<0.10
<0.09
<0.09
Rat number 11 0 0 3.09 12.5 – – – –
a
The livers of four identically treated mice were pooled for DNA isolation. On the basis of 38 background values compiled during the last 5 years, a respective value of 11.4 cpm with a S.D. of 1.11 cpm was calculated. The S.D. of the difference (S.D.difference ) between a sample and the background can be calculated as the square root of the sum of the squares of the S.D. for the sample (S.D.sample ) and the S.D. of the background (S.D.background ): b
S.D.difference =
(S.D.2sample + S.D.2background )
Assuming that a vial containing a DNA sample from a treated animal with low amount of radioactivity shows the same statistic variation as the background (1 S.D. = 1.11 cpm), the limit of detection (LoD) for radioactivity in a vial can be calculated on a level of 2 S.D.: LoD = 2 × (1.112 + 1.112 ) = 3.13 cpm c d
The counting efficiency was 80%. For the calculation of the CBI, see Section 2.10.
Table 8 Summary of genetic toxicology results Test systems
Test conditions
Result
Ames test
Strains TA1535, TA 97a, TA98, TA100, TA102; +/−S9; 17.6–4400 g/plate
Dose dependent increase −S9 with strains TA98 and TA100; lowest mutagenic concentration 88 g/plate
Ames test
Strains TA98NR, TA100NR, +/−S9; 500–5000 g/plate
No increases in revertant rates
V79 micronucleus test
V79 chinese hamster fibroblasts, +S9, 3 h treatment; −S9, 3 and 20 h treatment; 412.3–3400 g/ml;
Weak increase in frequency of micronucleated cells at the highest, cytotoxic concentration, i.e. at ≥1160.4 g/ml;
Mouse lymphoma tk assay
Fluctuation protocol, +S9, 3 h treatment; −S9, 3 and 24 h treatment; 55–880 g/ml;
No increases in mutant frequencies
Bone marrow micronucleus test
CD-1 mice, 2 oral treatments with a 24 h interval, sampling 48 h after 1st treatment; 32, 100 and 320 mg/kg
No bone marrow toxicity, no effects on percentage of micronucleated polychromatic erythrocytes
MutaTM Mouse LacZ assay
MutaTM Mouse, 5 oral treatments, 320 mg/kg, sacrifice 3, 7 or 21 days after last treatment
No effects on frequencies of mutated LacZ genes in colon and liver
Comet assay
Han:Wist rats, single oral treatment, 2000 mg/kg, sampling after 3 and 24 h
No effects on DNA migration (tail moment) in jejunum and liver
DNA binding
Han:Wist rats and CD-1 mice, single oral treatment with 10.3–11.5 mg/kg, i.e. 1.55–1.74 mCi/kg; sacrifice 24 h after application
No evidence for DNA binding in the liver, CBI < 0.1
Abbreviations: CBI, covalent binding index (definition see Section 2.10); NR, nitro reductase deficient.
192
W. Suter et al. / Mutation Research 518 (2002) 181–194
cells effects were found mainly under hypoxic test conditions [16]. Since it is very difficult to simulate in vitro the oxygen pressure prevailing in mammalian tissues, relevant information about the safety of a nitro compound in patients cannot be expected from in vitro experiments. Aromatic nitro compounds have been tested in vivo and found to be positive in BigBlueTM transgenic mice [48] and DNA binding [49] as well as comet formation [50] was observed in mice treated with such compounds. The genotoxic potential of AMP397 was, therefore, studied in a series of in vivo studies with animals treated by the oral route. Oral treatment of animals also allows studying in the intestines possible effects of metabolites formed by the intestinal flora, which cannot, or can only with great difficulty, be studied in mammalian cells in vitro. Apart from the standard bone marrow micronucleus test in mice, which provided clear negative results (Table 4), the assays focused on the tissues that were thought to be most at risk. The liver is known to have nitroreductase activity [43,44,51] and could, therefore, be expected to produce reactive and/or genotoxic intermediates. Distribution studies showed relatively high tissue concentrations in the liver of rats, i.e. 4830 ng-eq/g of tissue, 1 h after treatment with a single oral dose of 60 mg/kg [14 C]AMP397. This was clearly >1754 ng-eq/g of tissue found in the bone marrow or the 1750 ng-eq/g of tissue measured in the plasma under the same test conditions. There was no evidence for the induction of gene mutations in MutaTM Mouse (Table 5), for DNA damage in rats, as measured in the comet assay (Table 6), or for DNA binding in rats or mice (Table 7). The intestinal tract was considered to be the appropriate target to investigate, because of the possibility that intestinal bacteria could form and release reactive intermediates into the lumen of the gut. The negative data from the MutaTM Mouse study (Table 5) and the comet assay with cells of rats clearly showed that there was no genotoxic insult detectable in the cells of the jejunum and the colon. The rat was used to conduct the comet assay since higher doses could be administered without strong systemic toxicity, which can be explained by the lower absorption of the test compound in rats as compared to mice. Rats could be treated with 2000 mg/kg, whereas, mice showed strong signs of systemic toxicity at 320 mg/kg. Con-
sequently, a higher exposure of the gut tissue to AMP397 and its potential reactive metabolites produced by the intestinal flora was achievable in rats. Assuming a comparable ratio between intestinal volume and body weight, the local intestinal exposure of a human volunteer/patient treated with 800 mg (highest dose used in a clinical tolerability study) would be >125 times lower than the exposure of the rats in the comet assay. In addition, there is evidence that the nitroreductase activity is much lower in human faecal samples than that found in samples from laboratory rodents [52]. Since, as shown by the data presented, reactive intermediates are formed exclusively by nitroreduction, it can be expected that the rodent models used to investigate the occurrence of genotoxic effects in the gut represent worst case scenarios. In conclusion, the unequivocally negative results from four in vivo mutagenicity assays give a high degree of confidence that AMP397 can be developed and used for the treatment of epilepsy without exposing the patients to an undue genotoxic hazard. References [1] P. Kwan, M.J. Brodie, Early identification of refractory epilepsy, N. Engl. J. Med. 342 (2000) 314–319. [2] R.S. Fischer, B.G. Vickrey, P. Gibson, B. Hermann, P. Penovich, A. Scherer, S. Walker, The impact of epilepsy from the patient’s perspective I. Descriptions and subjective perceptions, Epilepsy Res. 41 (2000) 39–51. [3] R.S. Fischer, B.G. Vickrey, P. Gibson, B. Hermann, P. Penovich, A. Scherer, S. Walker, The impact of epilepsy from the patient’s perspective II: views about therapy and health care, Epilepsy Res. 41 (2000) 53–61. [4] A. Tortorella, T. Halonen, N. Sahibzada, K. Gale, A crucial role of the alpha-amino-3-hydroxy-5-methylisoxazole-4propionic acid subtype of glutamate receptors in piriform and perirhinal cortex for the initiation and propagation of limbic motor seizures, J. Pharmacol. Exp. Ther. 280 (1997) 1401– 1405. [5] H. Katsumori, Y. Mibabe, M. Osawa, R. Ashby Jr., Acute effect of various GABA receptor agonists and glutamate antagonists on focal hippocampal seizures in freely moving rats elicited by low-frequency stimulation, Synapse 28 (1998) 103–109. [6] W. Löscher, Pharmacology of glutamate receptor antagonists in the kindling model of epilepsy, Prog. Neurobiol. 54 (1998) 721–741. [7] W. Nakahara, F. Fukuoka, T. Sugimuira, Carcinogenic action of 4-nitro-quinoline-1-oxide, Gann 48 (1957) 129–137. [8] V. Purohit, A.K. Basu, Mutagenicity of nitroaromatic compounds, Chem. Res. Toxicol. 13 (2000) 673–692.
W. Suter et al. / Mutation Research 518 (2002) 181–194 [9] S.M. Cohen, Toxicity and carcinogenicity of nitrofurans, in: G.T. Bryan (Ed.), Carcinogenesis, Vol 4, Nitrofurans, Raven Press, New York, 1978, pp. 171–231. [10] L. Dobias, M. Cerna, P. Roessner, R. Sram, Genotoxicity and carcinogenicity of metronidazole, Mutation Res. 317 (1994) 177–194. [11] W. Pfau, B.L. Pool-Zobel, C.-W. von der Lieth, M. Wiessler, The structural basis for the mutagenicity of aristolochia acid, Cancer Lett. 55 (1990) 7–11. [12] P.C. Howard, R.H. Heflich, F.E. Evans, F.A. Beland, Formation of DNA adducts in vitro and in Salmonella typhimurium upon metabolic reduction of the environmental mutagen 1-nitropyrene, Cancer Res. 43 (1983) 2052–2058. [13] T.C. Pederson, J.S. Siak, The activation of mutagens in diesel particle extract with rat liver S9 enzymes, J. Appl. Toxicol. 1 (1981) 61–66. [14] J.H. Tocher, Reductive activation of nitroheterocyclic compounds, Gen. Pharmac. 28 (1997) 485–487. [15] J. Ashby, R.W. Tennant, Chemical structure, Salmonella mutagenicity and extent of carcinogenicity as indicators of genotoxic carcinogenesis among 222 chemicals tested in rodents by the U.S. NCI/NTP, Mutation Res. 204 (1988) 17– 115. [16] M. Korbelik, D. Horvat, The mutagenicity of nitroaromatic drugs. Effect of metronidazole after incubation in hypoxia in vitro, Mutation Res. 78 (1980) 201–207. [17] J.R. Thornton-Manning, B.A. Smith, F.A. Beland, F.L. Heflich, S9-mediated metabolism of 1-nitropyrene to a mutagen in chinese hamster ovary cells by ring-oxidation under aerobic conditions and by nitroreduction under anaerobic condition, Carcinogenesis 12 (1991) 43–51. [18] M. Maron, B.N. Ames, Revised methods for the Salmonella mutagenicity tests, Mutation Res. 113 (1983) 173–215. [19] OECD guideline for testing of chemicals, Nr. 471, genetic toxicology: bacterial reverse mutation test, adopted 21 July 1997. [20] E.J. Rosenkranz, E.C. McCoy, R. Mermelstein, H.S. Rosenkranz, Evidence for the existence of distinct nitroreductases in S. typhimurium: roles in mutagenesis, Carcinogenesis 3 (1982) 121–123. [21] B. Miller, S. Albertini, F. Locher, V. Thybaud, E. Lorge, Comparative evaluation of the in vitro micronucleus test and the in vitro chromosome aberration test: industrial experience, Mutation Res. 392 (1997) 45–59. [22] J. Cole, D.B. Mc Gregor, M. Fox, R.C. Garner, J. Thacker, Gene mutation assay in cultured mammalian cells, in: D.J. Kirkland (Ed.), Basic mutagenicity tests UKEMS recommended procedures, Cambridge University Press, Cambridge, 1990, pp. 87–114. [23] OECD. Guideline for testing of chemicals, 476, genetic toxicology: in vitro mammalian cell gene mutation tests, adopted 21 July 1997. [24] W.D. Robinson, M.H.L. Green, J. Cole, R.C. Garner, M.J.R. Healy, D. Gatehouse, Statistical evaluation of bacterial/mammalian fluctuation tests, in: D.J. Kirkland (Ed.), Statistical evaluation of mutagenicity test data, Cambridge University Press, Cambridge, 1990, pp. 102–140.
193
[25] F. Romagna, C.D. Staniforth, The automated bone marrow micronucleus test, Mutation Res. 213 (1989) 91–104. [26] OECD guideline for the testing of chemicals, 474, genetic toxicology: mammalian erythrocyte micronucleus test, adopted 21 July 1997. [27] W. Frieauff, F. Romagna, Technical aspects of automatic micronucleus analysis in rodent bone marrow assays, Cell Biol. Toxicol. 10 (1994) 283–289. [28] BigBlueTM transgenic mouse mutagenesis assay system (stratagene, 1 May 1994). [29] J.A. Gossen, A.C. Molijn, G.R. Douglas, J. Vijg, Application of galactose sensitive E. coli strains as selective hosts for lacZ− plasmids, Nucl. Acid Res. 20 (1992) 3254. [30] J.A. Gossen, J. Vijg, A selective system for lacZ− phage using a galactose-sensitive E. coli host, Biotechniques 14 (1993) 326–330. [31] MutaTM mouse: mutation assay instruction manual (Covance Hazleton, version obtained on 15 July 1996). [32] G. Speit, A. Hartmann, The comet assay (single-cell gel test)—A sensitive genotoxicity test for the detection of DNA damage and repair, DNA Repair Protocols 113 (1999) 203– 212. [33] R.R. Tice, E. Agurell, D. Anderson, B. Burlinson, A. Hartmann, H. Kobayashi, Y. Miyamae, E. Rojas, J.-C. Ryu, Y.F. Sasaki, The single cell gel/comet assay: guidelines for in vitro and in vivo genetic toxicology testing, Environ. Mol. Mutagen. 35 (2000) 206–221. [34] W. Frieauff, A. Hartmann, W. Suter, Automatic analysis of slides processed in the comet assay, Mutagenesis 16 (2001) 133–137. [35] P.L. Olive, Cell proliferation as a requirement for development of the contact effect in chinese hamster V79 spheroids, Radiation Res. 117 (1989) 79–92. [36] W.K. Lutz, In vivo covalent binding of organic chemicals to DNA as a quantitative indicator in the process of chemical carcinogenesis, Mutation Res. 65 (1979) 289–356. [37] W.K. Lutz, Quantitative evaluation of DNA binding data for risk estimation and for classification of direct and indirect carcinogens, J. Cancer Res. Clin. Oncol. 112 (1986) 85–91. [38] P. Sagelsdorff, W.K. Lutz, C. Schlatter, The relevance of covalent binding to mouse liver DNA to the carcinogenic action of hexachlorocyclohexane isomers, Carcinogenesis 4 (1983) 1267–1273. [39] P. Sagelsdorff, W.K. Lutz, C. Schlatter, DNA methylation in rat liver by daminozide, 2,2-dimethylhydrazine, and dimethylnitrosamine, Fundamental Appl. Toxicol. 11 (1988) 723–730. [40] J. McCann, E. Choi, E. Yamasaki, B.N. Ames, Detection of carcinogens as mutagens in the Salmonella/microsome test: assay of 300 chemicals, Proc. Natl. Acad. Sci. U.S.A. 72 (1975) 5135–5139. [41] J.M. Gentile, G.J. Gentile, M.J. Plewa, Mutagenicity of selected aniline derivatives to Salmonella following plant activation and mammalian hepatic activation, Mutation Res. 188 (1987) 185–196. [42] M.W. Chou, R.H. Heflich, P.P. Fu, Metabolism of 1-nitrobenzo(a)pyrene by rat liver microsomes to potent mutagenic metabolites, Carcinogenesis 7 (1986) 1837–1844.
194
W. Suter et al. / Mutation Research 518 (2002) 181–194
[43] A.K.D. Hajos, G.W. Winston, Dinitropyrene nitroreductase activity of purified NAD(H)-quinone oxidoreductase: role in rat liver cytosol and induction by aroclor-1254 pretreatment, Carcinogenesis 12 (1991) 697–792. [44] D.B. McGregor, I. Edwards, C.R. Wolf, L.M. Forrester, W.J. Caspary, Endogenous xenobiotic enzyme levels in mammalian cells, Mutation Res. 261 (1991) 29–39. [45] R. Fahrig, M. Engelke, Reinvestigation of in vivo genotoxicity studies in man. Part I: no induction of DNA strand breaks in peripheral lymphocytes after metronidazole therapy, Mutation Res. 395 (1997) 215–221. [46] E. Touati, M. Hofnung, J.M. Thiberge, V. Michel, A. Labigne, P.J. Jenks, Short-term infection with Helicobacter pylori and 1 week exposure to metronidazole does not enhance gastric mutation frequency in transgenic mice, J. Antimicrobic. Chemother. 46 (2000) 987–992. [47] D. Menéndez, E. Rojas, L.A. Herrera, M.C. Lopez, M. Sordo, G. Elizondo, P. Ostrosky-Wegman, DNA breakage due to metronidazole treatment, Mutation Res. 478 (2001) 153–158.
[48] W. Suter, R. Ahiabor, B. Blanco, F. Locher, F. Mantovani, M. Robinson, G. Sreenan, F. Staedtler, T. Swinger, A. Vignutelli, E. Perentes, Evaluation of the in vivo genotoxic potential of three carcinogenic aromatic amines using the BigBlueTM transgenic mouse mutation assay, Environ. Mol. Mutagen. 28 (1996) 354–362. [49] K.B. Declos, C.E. Cerniglia, K.L. Dooley, W.L. Campbell, W. Franklin, R.P. Walker, The role of intestinal microflora in the metabolic activation of 6-nitrochrysene to DNA-binding derivatives in mice, Toxicology 60 (1990) 137–150. [50] Y.F. Sasaki, K. Sekihashi, F. Izumiyama, E. Nishidate, A. Saga, K. Ishida, S. Tsuda, The comet assay with multiple mouse organs: comparison of comet assay results and carcinogenicity with 208 chemicals selected from IARC monographs and U.S. NTP carcinogenicity database, Crit. Rev. Toxicol. 30 (2000) 629–799. [51] D.E. Rickert, Metabolism of Nitroaromatic compounds, Drug Metabol. Rev. 18 (1987) 23–53. [52] R. Rowland, Factors affecting metabolic activity of the intestinal microflora, Drug Metabol. Rev. 19 (1988) 243–261.