Mutagenicity of azo dyes in the Salmonella/microsome assay using in vitro and in vivo activation

Batch and continuous biodegradation of Amaranth in plain distilled water by P. aeruginosa BCH and toxicological scrutiny using oxidative stress studies

Bacterium Pseudomonas aeruginosa BCH was able to degrade naphthylaminesulfonic azo dye Amaranth in plain distilled water within 6 h at 50 mg l(-1) dye concentration. Studies were carried out to find the optimum physical conditions and which came out to be pH 7 and temperature 30 °C. Amaranth could also be decolorized at concentration 500 mg l(-1). Presence of Zn and Hg ions could strongly slow down the decolorization process, whereas decolorization progressed rapidly in presence of Mn. Decolorization rate was increased with increasing cell mass. Induction in intracellular and extracellular activities of tyrosinase and NADH-DCIP reductase along with intracellular laccase and veratryl alcohol oxidase indicated their co-ordinate action during dye biodegradation. Up-flow bioreactor studies with alginate immobilized cells proved the capability of strain to degrade Amaranth in continuous process at 20 ml h(-1) flow rate. Various analytical studies viz.--HPLC, HPTLC, and FTIR gave the confirmation that decolorization was due to biodegradation. From GC-MS analysis, various metabolites were detected, and possible degradation pathway was predicted. Toxicity studies carried out with Allium cepa L. through the assessment of various antioxidant enzymes viz. sulphur oxide dismutase, guaiacol peroxidase, and catalase along with estimation of lipid peroxidation and protein oxidation levels conclusively demonstrated that oxidative stress was generated by Amaranth.

Genotoxicity Testing of Four Textile Dyes in Two Crosses of Drosophila Using Wing Somatic Mutation and Recombination Test

In this study, four textile dyes, namely Astrazon Yellow, Red, Blue, and Black, were tested for their genotoxic effects in the wing cells of Drosophila melanogaster. Two crosses were used, the standard cross (ST) and the improved high-bioactivation cross (HB), the latter being characterized by increased sensitivity to the genotoxic effects of promutagens and procarcinogens. Three-day-old larvae were exposed to different concentrations of dyes. Commonly known mutagens were applied as positive controls. All concentrations of textile dyes, ethyl methanesulfonate (EMS), and urethane caused a decrease in survival proportional to concentration used. EMS and urethane caused an increase in the number of all types of spots in both standard and high-bioactivation crosses. Compared to ST crosses, the number of induced spots in the HB cross treated with urethane was considerably high. Treatment of the standard and the high-bioactivation crosses with textile dyes gave positive results, apparent from increase in the frequency of the small single spots. Yellow and red dyes also increased the number of large single spots in both crosses, whereas the twin spots were positive only at the highest dose of yellow dye. All these results indicate that D. melanogaster wing spot test can be recommended as a suitable in vivo test for the determination of genotoxicity of textile dyes.

Evaluation of urinary mutagenicity in azo dye manufacture workers

OBJECTIVES

The aim of the study was to evaluate urinary mutagenicity in workers employed in a major chemical plant located near Rouen (France) that produces dichlorobenzidine and azo dyes.

MATERIALS AND METHODS

Samples were obtained from 47 male workers aged 38.9+/-11.3 years (range, 21-58 years), mean duration of employment 9.0+/-8.7 years (range, 1-32 years) for urinary mutagenicity determination with use of the Ames fluctuation test (strains TA 98 and TA 100 with and without metabolic activation) and gas chromatography/mass spectrometry. To assess occupational exposure of workers, urine samples were collected in two series. First, initial just after a one-month holiday (non-exposure). Second, four months later during regular occupational activity. During the same periods, workers completed a questionnaire, which sought information concerning their working conditions, non-occupational factors, and personal habits.

RESULTS

Of the total 47 samples tested, 3 (6%) collected just after a one-month holiday and 6 (12%) samples collected during regular occupational activity were positive in at least one mutagenicity assay. Dichlorobenzidine traces ranging from 1.6 to 8.9 ppb were detected in 4 (8%) samples of the exposed as well as in 4 (8%) samples of non-exposed workers. No significant differences between biological and analytical responses obtained in the non-exposure period and after occupational exposure were observed, however, 5 (11%) workers in this group presented urinary mutagenicity that could be related to occupational exposure.

CONCLUSIONS

The study suggests that some industrial hygiene problems, revealed in the analysis of questionnaire responses and confirmed by our evaluation, could be undoubtedly and easily solved to improve working conditions of the employees.

o-Aminoazotoluene does induce the enzymes of its own mutagenic activation in mouse liver

The objective of this study was to investigate the CYP1A1 and CYP1A2 mRNAs and enzyme activities in mouse liver during induction with o-aminoazotoluene (OAT) as well as the capability of the hepatic S9-fraction from OAT-treated mice to induce its own activation to mutagens in the Ames test using S. typhymurium strain TA98. The data obtained indicate that when used at appropriate doses, OAT is a PAH-type inducer of mouse hepatic microsomal monooxygenases, which activity is not less than that of the known inducer 3,4-benzo[alpha]pyrene. In the absence of S9-fraction enzymes no OAT-mediated mutagenicity was observed in the Ames test. In the presence of the S9-fraction from OAT-pretreated mice, OAT induced as high revertant numbers, as it did in the presence of the S9 fraction from the liver of Aroclor 1254-treated mice. Thus, OAT does induce the enzymes of its own mutagenic activation in mouse liver.

Mutagenicity of different textile dye products in Salmonella typhimurium and mouse lymphoma cells

Textile dyes used within the European Union (EU) were examined for available published and unpublished mutagenicity data. Fifty-three dye products that had so far not been investigated for mutagenicity were tested in the bacterial reverse mutation assay with Salmonella typhimurium (Ames test) according to a modification of the OECD 471 guidelines (instead of five strains, only TA98 and/or TA100 were used with and without metabolic activation (S9-mix)). About 28% (15 out of 53) of the dye samples were positive in the Ames test. Fifteen samples showed positive results with TA98, two were positive in TA100. The mutagenicity of nine Ames-positive textile dye products was further investigated in the mouse lymphoma assay (MLA) (OECD 476). Sixty-seven percent (6 out of 9) induced genotoxic effects in the MLA. The induction rates (IR) were between 2.1 and 132 in the bacterial reverse mutation assay and between 2.1 and 15.2 in the MLA. The results confirm previous findings that dye products are marketed that are not sufficiently tested and that show mutagenic effects in in vitro tests.

Investigating the sources of the mutagenic activity found in a river using the Salmonella assay and different water extraction procedures

In the routine São Paulo state (Brazil) surface water quality-monitoring program, which includes the Salmonella microsome mutagenicity assay as one of its parameters, a river where water is taken and treated for drinking water purposes has repeatedly shown mutagenic activity. A textile dyeing facility employing azo-type dyes was the only identifiable source of mutagenic compounds. We extracted the river and drinking water samples with XAD4 at neutral and acidic pH and with blue rayon, which selectively adsorbs polycyclic compounds. We tested the industrial effluent, raw, and treated water and sediment samples with YG1041 and YG1042 and compared the results with the TA98 and TA100 strains. The elevated mutagenicity detected with YG-strains suggested that nitroaromatics and/or aromatic amines were causing the mutagenicity detected in the samples analyzed. Positive responses for the blue rayon extracts indicated that mutagenic polycyclic compounds were present in the water samples analyzed. The mutagen or mixture of mutagens present in the effluent and water samples cause mainly frameshift mutations and are positive with and without metabolic activation. The Salmonella assay combined with different extraction procedures proved to be very useful in the identification of the origin of the pollution and in the identification of the classes of chemical compounds causing the mutagenic activity in the river analyzed.

Comparison of the mutagenic activity of XAD4 and blue rayon extracts of surface water and related drinking water samples

The combination of mutagenicity tests and selective extraction methodologies can be useful to indicate the possible classes of genotoxic organic contaminants in water samples. Treated and source water samples from two sites were analyzed: a river under the influence of an azo dye-processing plant discharge and a reservoir not directly impacted with industrial discharges, but contaminated with untreated domestic sewage. Organic extraction was performed in columns packed with XAD4 resin, that adsorbs a broad class of mutagenic compounds like polycyclic aromatic hydrocarbons (PAHs), arylamines, nitrocompounds, quinolines, antraquinones, etc., including the halogenated disinfection by-products; and with blue rayon that selectively adsorbs polycyclic planar structures. The organic extracts were tested for mutagenicity with the Salmonella assay using TA98 and TA100 strains and the potencies were compared. A protocol for cleaning the blue rayon fibers was developed and the efficiency of the reused fibers was analyzed with spiked samples. For the river water samples under the influence of the azo-type dye-processing plant, the mutagenicity was much higher for both blue rayon and XAD4 extracts when compared to the water from the reservoir not directly impacted with industrial discharges. For the drinking water samples, although both sites showed mutagenic responses with XAD4, only samples from the site under the influence of the industrial discharge showed mutagenic activity with the blue rayon extraction, suggesting the presence of polycyclic compounds in those samples. As expected, negative results were found with the blue rayon extracts of the drinking water collected from the reservoir not contaminated with industrial discharges. In this case, it appears that using the blue rayon to extract drinking water samples and comparing the results with the XAD resin extracts we were able to distinguish the mutagenicity caused by industrial contaminants from the halogenated disinfection by-products generated during water treatment.

Genotoxicity evaluation of polluted ground water in human peripheral blood lymphocytes using the comet assay

This paper presents the genotoxicity experiments with the ground water collected from an area under the influence of textile dyeing and bleaching industries in Tirupur, Tamilnadu, India. The alkaline single cell gel electrophoresis (SCGE) assay was performed in vitro with human peripheral blood lymphocytes. The cells were exposed to two doses of non-volatile organic agents extracted from ground water samples. Ground water samples were collected from 12 locations distributed in and around Tirupur and extracts were taken at different pHs (without pH adjustment and acidic pH 2.0). The persistence of the DNA damage after exposure to the organic extracts was also studied. All the samples were found to contain substances capable of inducing DNA damage in human lymphocytes. Extracts from acidified waters (pH=2.0) were found to induce more DNA damage than extracts from without pH adjustment (natural pH). The DNA damage was not fully repaired after incubation for 2h at 37 degrees C. The chemical characterization of the sub-fractions revealed the existence of aromatic amines in the extracts, which may be responsible for the DNA damaging activity of the water samples. The results of this investigation demonstrate the application of the comet assay in environmental monitoring studies.

Using base-specific Salmonella tester strains to characterize the types of mutation induced by benzidine and benzidine congeners after reductive metabolism

Although benzidine (Bz), 4-aminobiphenyl (ABP), 3,3'-dichlorobenzidine HCl (DCBz), 3,3'-dimethylbenzidine (DMBz), 3,3'-dimethoxybenzidine (DMOBz) and the benzidine congener-based dye trypan blue (TB) produce primarily frameshift mutations in Salmonella typhimurium, the base-substitution strain TA100 also responds to these compounds when S9 is present. Performing DNA sequence analysis, other investigators have shown that ABP induces frameshift, base-pair and complex mutations. Also, it was found that an uninduced hamster liver S9 preparation with glucose-6-phosphate dehydrogenase, FMN, NADH and four times glucose 6-phosphate gave a stronger mutagenic response than the conventional plate incorporation with rat S9 activation mixture for all the compounds tested. Using the base-specific tester strains of S. typhimurium (TA7001-TA7006) with the above reductive metabolic activation system, we surveyed these compounds for the ability to produce specific base-pair substitutions after reductive metabolism. Bz was weakly mutagenic in TA7005 (0.04 revertants/microg). ABP was mutagenic in TA7002 (1.4 revertants/microg), TA7004 (0.6 revertants/microg), TA7005 (2.98 revertants/microg) and TA7006 (0.4 revertants/microg). DCBz was weakly mutagenic in TA7004 (0.01 revertants/microg). It was concluded that benzidine induced some CG->AT transversions in addition to frameshift mutations. ABP induced TA->AT, CG->AT, and CG->GC transversions as well as GC->AT transitions. DCBz induced only GC->AT transitions. Because DMBz, DMOBz and TB were not mutagenic in this base-substitution mutagen detection system, their mutagenic activity was attributed strictly to frameshift mechanisms.

Genotoxicity of some sulfur dyes on tadpoles (Rana hexadactyla) measured using the comet assay

This report presents the results of a genotoxicity study to evaluate the DNA damage caused by four sulfur dyes used in the textile and tannery industries. Alkaline single-cell gel electrophoresis assay (SCGE) was performed on erythrocytes from Rana hexadactyla tadpoles following whole-body exposure to increasing concentrations of the dyes. The dyes, along with their active ingredients, were Sandopel Basic Black BHLN, Negrosine, Dermapel Black FNI, and Turquoise Blue. The dye-treated tadpoles showed significant DNA damage, measured as mean DNA length:width ratio, when compared with unexposed control animals. Among the four tested dyes Sandopel Basic Black BHLN appears to be highly genotoxic, Dermapel Black FNI was least genotoxic, and Negrosine and Turquoise Blue were moderately toxic to R. hexadactyla tadpoles. The tadpoles showed a significant reduction in DNA damage when placed in dechlorinated tap water after exposure for a 24-hr period to the dye solutions.

Genotoxicity of o-aminoazotoluene (AAT) determined by the Ames test, the in vitro chromosomal aberration test, and the transgenic mouse gene mutation assay

o-Aminoazotoluene (AAT) has been evaluated as a possible human carcinogen (Class 2B) by the International Agency for Research on Cancer (IARC). The Ames test found it to be mutagenic in the presence of a metabolic activation system, whereas it has little clastogenicity either in vitro or in vivo in the chromosomal aberration assay. AAT is also carcinogenic in the lung or liver of mice and rats given long-term administrations. Therefore, metabolites generated in the liver etc. may have gene mutation activity, and carcinogenesis would occur. We examined the mutagenicity of AAT in a gene mutation assay, using lacZ transgenic mice (MutaMice) and a positive selection method. AAT showed positive results for organs with metabolic functions, such as liver and colon and other organs. Positive results were also seen in an Ames test in the presence of metabolic activation and negative results seen in a chromosomal aberration test. Therefore, AAT had the potential to cause gene mutation in the presence of metabolic activation systems in vitro and the same reaction was confirmed in vivo with organs with metabolic function, such as liver and colon, but little clastogenicity in vitro or in vivo. Thus, metabolites with gene mutation activity may be responsible for the carcinogenicity of AAT. The transgenic mouse mutation assay proved to be useful for concurrent assessment of in vivo mutagenicity in multiple organs and to supplement the standard in vivo genotoxicity tests, such as the micronucleus assay which is limited to bone marrow as the only target organ.

Genotoxicity studies on the azo dye Direct Red 2 using the in vivo mouse bone marrow micronucleus test

The clastogenicity of the azo dye Direct Red 2 (DR2) has been investigated using the murine bone marrow micronucleus assay. A potent dose-dependent response was observed following oral gavage of DR2 up to 4 mg/kg, after which significant toxicity to the erythroid compartment was observed. The route of administration had a significant effect on the frequency of micronucleus formation: intraperitoneal injection was approximately two-fold less clastogenic than the equivalent dose delivered orally (p<0.05). The requirement for activation of DR2 by intestinal microflora was indicated by the fact that mice given acid-treated water prior to administration of DR2 showed a significant reduction (40%; p<0.001) in micronucleated polychromatic erythrocyte formation. The implications of these findings for the health and safety of occupationally exposed workers are discussed.

The SOS chromotest: a review

The SOS chromotest is reviewed through over 100 publications corresponding to the testing of 751 chemicals. 404 (54%) of these chemicals present a genotoxic activity detectable in the SOS chromotest. Their SOS inducing potencies span more than 8 orders of magnitude. For 452 compounds, the results obtained in the SOS chromotest could be compared to those obtained in the Ames test. It was found that 373 (82%) of these compounds give similar responses in both tests (236 positive and 137 negative responses). Thus the discrepancies between both tests concern 79 compounds (18%). A case by case analysis shows that many of these compounds are at the same time very weak SOS inducers and very weak mutagens. Thus we think that, most of the time, the discrepancies between the two tests may be accounted for by differences in the interpretation of the results rather than by the experimental results themselves. However, there are some compounds which are clearly SOS inducers but devoid of mutagenic activity in the Ames test (such as quinoline-1-oxide) and to a larger extent, clearly mutagenic compounds which do not induce the SOS response in the SOS chromotest (such as benzidine, cyclophosphamide, acridines, ethidium bromide). We also analyzed the correlation between SOS induction, mutagenesis and carcinogenesis according to the classification of Lewis. For 65 confirmed carcinogens (class 1), the sensitivity, i.e., the capacity to identify carcinogens, was 62% with the SOS chromotest and 77% with the Ames test. For 44 suspected carcinogens (class 2), the sensitivity was 66% with the SOS chromotest and 68% with the Ames test. Thus, we confirmed previous observations made on 83 compounds that there is a close correlation between the results given by both bacterial tests. The capacity of the Ames test to identify carcinogens is higher than that of the SOS chromotest. However, because the number of false positive compounds was lower in the SOS chromotest, the specificity, i.e., the capacity to discriminate between carcinogens and non-carcinogens of the SOS chromotest, appeared higher than that of the Ames test. Thus, the results of the SOS chromotest and of the Ames test can complement each other. The SOS chromotest is one of the most rapid and simple short-term test for genotoxins and is easily adaptable to various conditions, so that it could be used as an early--perhaps the earliest--test in a battery.

DNA adduct formation by 12 chemicals with populations potentially suitable for molecular epidemiological studies

DNA adduct formation, route of absorption, metabolism and chemistry of 12 hazardous chemicals are reviewed. Methods for adduct detection are also reviewed and approaches to sensitivity and specificity are identified. The selection of these 12 chemicals from the Environmental Protection Agency list of genotoxic chemicals was based on the availability of information and on the availability of populations potentially suitable for molecular epidemiological study. The 12 chemicals include ethylene oxide, styrene, vinyl chloride, epichlorohydrin, propylene oxide, 4,4'-methylenebis-2-chloroaniline, benzidine, benzidine dyes (Direct Blue 6, Direct Black 38 and Direct Brown 95), acrylonitrile and benzyl chloride. While some of these chemicals (styrene and benzyl chloride, possibly Direct Blue 6) give rise to unique DNA adducts, others do not. Potentially confounding factors include mixed exposures in the work place, as well the formation of common DNA adducts. Additional research needs are identified.

Metabolism of azo dyes: implication for detoxication and activation

Azo dyes are consumed and otherwise utilized in varying quantities in many parts of the world. Such widely used chemicals are of great concern with regard to their potential toxicity and carcinogenic properties. Their metabolism has been studied extensively and is significant for detoxication and metabolic activation. Both oxidative and reductive pathways are involved in these processes. The majority of azo dyes undergo reduction catalyzed by enzymes of the intestinal microorganisms and/or hepatic enzymes including microsomal and soluble enzymes. The selectivity of substrate and enzyme may to a large extent be determined by the oxygen sensitivity of reduction since a normal liver is mainly aerobic in all areas, whereas the microorganisms of the lower bowel exist in an anaerobic environment. However, it should be pointed out that the pO2 of centrilobular cells within the liver is only a fraction that of air, where pO2 = 150 torr. Therefore, an azo dye reduction experiment performed aerobically may not be an accurate predictor of reductive metabolism in all areas of the liver. Many of the azo dyes in common use today have highly charged substituents such as sulfonate. These resist enzymic attack and for the most part are poorly absorbed from the intestinal tract, providing poor access to the liver, the major site of the mixed-function oxidase system. Lipophilic dyes, such as DAB, which are often carcinogenic, readily access oxidative enzymes and are activated by both mixed-function oxidase and conjugating systems. Reduction of the carcinogenic dyes usually leads to loss of carcinogenic activity. By contrast, most of the highly charged water-soluble dyes become mutagenic only after reduction. Even then, most of the fully reduced amines required oxidative metabolic activation. An outstanding example is the potent human bladder carcinogen benzidine, which derives from the reduction of several azo dyes. Many problems regarding mutagenic and carcinogenic activation remain to be solved. At the present time, it is apparent that both oxidative and reductive pathways yield toxic products. Toxicologic assessment of azo dyes must consider all pathways and particularly the oxygen sensitivity of azoreduction. This is critical in the treatment of waste from chemical plants where there is a great need for soil bacteria which catalyze reduction aerobically. Consideration of secondary pathways are also of great concern. For example, azoreduction of carcinogenic dyes such as DAB removes carcinogenic activity although oxidative metabolism of the primary amines yield mutagenic products. Such apparent dilemmas must be dealt with when considering metabolism/toxicity relationships for azo dyes.

Bacterial mutagenesis and hepatocyte unscheduled DNA synthesis induced by chrysoidine azo-dye components

5 azo dye components of Gurr chrysoidine 'Y' have been separated, synthesised and identified. Dyes with a methyl substitution (particularly between the two amino groups) were more mutagenic in Salmonella typhimurium strain TA100 with control rat liver S9 than the non-methylated counterpart (range 66-1992 revertants at 50 micrograms/plate). Mutagenicity was also catalysed by human-liver S9 and pre-treatment of rats with either phenobarbitone or beta-naphthoflavone enhanced the activation ability of S9 by greater than 4-fold. Using the most potent promutagenic component (2,4-diamino-3-methylazobenzene), the use of inhibitors of cytochrome P450 (metyrapone: 1.0 mM; alpha-naphthoflavone: 0.075 mM; DPEA: 0.125 mM) and of the flavin monooxygenase (methimazole: 0.75 mM) suggested a major role for cytochrome P448 in the activation of chrysoidine to mutagens. The ability of chrysoidine components to induce unscheduled DNA synthesis in rat hepatocytes in vitro was demonstrated and ranged between 11.92 and 23.5 net nuclear grains at a dose level of 2.5 micrograms/incubation. Since each dye was equi-potent, methyl substitution had little influence on genetic toxicity in hepatocytes.

The effect of alkoxy substituents on the mutagenicity of some phenylenediamine-based disazo dyes

16 phenylenediamine-based disazo dyes were examined in the Salmonella/mammalian microsome assay with strains TA98, TA100 and TA1538. All of the dyes contain an alkoxy group ortho to one of the azo linkages. Increasing the size of this alkoxy substituent from 1 to 4 carbons led to a decrease in mutagenic activity in certain instances while no change was noted in other cases. Comparison of the mutagenicity of the disazo dyes with their potential reductive-cleavage products suggests that (1) the reductive-cleavage products are not solely responsible for the mutagenicity of the disazo dyes, and (2) significant reductive-cleavage of the disazo dyes is not taking place in the standard Salmonella assay.

Evaluation of azo food dyes for mutagenicity and inhibition of mutagenicity by methods using Salmonella typhimurium

The mutagenicity of 4 azo dyes (FD&C Yellow No. 5, FD&C Yellow No. 6, FD&C Red No. 40 and amaranth) that are widely used to color food has been evaluated. 4 different methods were used: (1) the standard Ames plate-incorporation assay performed directly on the dyes in the absence of S9 and in the presence of rat- or hamster-liver S9; (2) application of the standard plate assay to ether extracts of aqueous solutions of the dyes; (3) a variant of the standard assay, using hamster liver S9, preincubation, flavin mononucleotide (FMN) and other modifications designed to facilitate azo reduction; and (4) reduction of the dyes with sodium dithionite, followed by ether extraction and the standard plate assay. Assays that include chemical reduction (methods 3 and 4) were included because azo compounds ingested orally are reduced in the intestine with the release of free aromatic amines. No mutagenic activity was seen for any of the azo dyes tested by using the standard Ames plate assay (method 1). Ether extracts of some samples of FD&C Yellow No. 6, FD&C Red No. 40 and amaranth were active (method 2), but only at high doses, generally 250 mg-equivalents or more per plate. These results indicate the presence of low levels of ether-extractable mutagenic impurities. The FMN preincubation assay (method 3) gave negative results for all dye samples tested. Most batches of FD&C Red No. 40 tested had mutagenic activity that was detectable when the ether extract of less than 1 mg of dithionite-reduced dye was plated in the presence of S9 (method 4). This finding implies that an impurity in these samples of FD&C Red No. 40 can be reduced to yield an ether-extractable mutagen. Dithionite-reduced samples of FD&C Yellow No. 6 and amaranth showed ether-extractable mutagenic activity only at much higher doses than those at which activity was seen with most dithionite-reduced samples of FD&C Red No. 40 (method 4). FD&C Yellow No. 5 showed no mutagenic activity with this method. Mutagenic activity was not detected when FD&C Red No. 40 was tested by using the azo reduction preincubation assay with FMN (method 3).(ABSTRACT TRUNCATED AT 400 WORDS)

Evaluation of the SOS chromotest

In the present investigation, the SOS chromotest with E. coli PQ37 was evaluated. The potential to identify different kinds of bacterial mutagens was examined. 124 chemicals of different chemical classes were tested. Their responses in the SOS chromotests were compared to reported test results obtained with the Ames test.