Optimum associations of tester strains for maximum detection of mutagenic compounds in the Ames test

The Ames test is now widely used as a short-term test for the detection of mutagens. Different strains are available with various genetic characteristics, and in the past decade various authors have recommended different associations of strains to give maximum detection potential. However, few studies have been done to compare the sensitivity of individual strains towards a wide range of compounds in a single study. In order to define the best association of strains for screening or regulatory purpose, we have tested 103 direct mutagens (reference genotoxins or in-house compounds) on 7 strains of Salmonella typhimurium: TA1535, TA1537, TA1538, TA97, TA98, TA100 and TA102. 126 different associations of strains have been studied in terms of sensitivity and percentage overlap. Optimum associations of 2, 3, 4 or 5 strains included strains both with and without plasmid pKM101. However, the specificity of detection is greatly diminished by the presence of plasmid pKM101 in the strain, as shown by the high degree of overlap in associations constituted entirely of strains containing the plasmid. The association of strains TA1538 and TA100 detected 86% of the chemicals tested and is therefore recommended for large-scale screening. A rate of detection of 100% was obtained when 6 strains were used. The best associations of 4 and 5 strains, which detected 97 and 99% chemicals respectively, all contained strains TA1537, TA1538 and TA102. Finally, the associations of 4 strains (TA1537, TA1538, TA100, TA102) or 5 strains (TA1535, TA1537, TA1538, TA97, TA102) seemed well adapted to the optimum detection of mutagenic compounds.     

Studies on the effect of ascorbic acid and selenium on the genotoxicity of nitrofurans: nitrofurazone and furazolidone

The genotoxic properties of nitrofurazone and furazolidone were studied using the Ames test and SOS-chromotest. Both compounds were found to act as strong mutagens on the TA97 and TA102 strains of S. typhimurium and to induce the SOS-repair system in the PQ37 strain of E. coli. A good concordance was found between the mutagenic activity and the ability to induce the SOS system. Ascorbic acid and sodium selenite only very slightly lowered the genotoxic effect of the 2 nitrofurans studied both in the Ames test and in the SOS-chromotest.     

Isolation and studies of the mutagenic activity in the Ames test of flavonoids naturally occurring in medical herbs

Quercetin, rhamnetin, isorhamnetin, apigenin and luteolin were isolated from medicinal herbs: Erigeron canadensis L., Anthyllis vulneraria L. and Pyrola chloranta L. The mutagenicity of these naturally occurring flavonoids was tested by the Ames method with S. typhimurium strains TA1535, TA1538, TA97, TA98, TA100 and TA102 in the presence and absence of metabolic activation. Of the above flavonoids only quercetin and rhamnetin revealed mutagenic activity in the Ames test. Quercetin induced point mutations in strains TA97, TA98, TA100 and TA102 of S. typhimurium. The presence of S9 rat liver microsome fraction markedly enhanced the mutagenic activity of quercetin in these strains. Rhamnetin appeared to be a much weaker mutagen in the Ames test. The compound induced mutations in strains TA97, TA98 and TA100 of S. typhimurium but only in the presence of metabolic activation.Comparison of the structure of the studied flavonoids with their mutagenic activity indicates that the mutagenicity of flavonoids is dependent on the presence of hydroxyl groups in the 3′ and 4′ positions of the B ring, and that the presence of a free hydroxy or methoxy group in the 7 position of the A ring also probably contributes to the appearance of mutagenic activity of flavonoids in the Ames test. It also appeared that the presence of methoxy groups, particularly in the B ring of the flavonoid molecule, markedly decreases the mutagenic activity of the compound.     

Mutagenicity studies of methyl-tert-butylether using the Ames tester strain TA102

Methyl-tert-butylether (MTBE) is an oxygenate widely used in the United States as a motor vehicle fuel additive to reduce emissions and as an octane booster [National Research Council, Toxicological and Performance Aspects of Oxygenated Motor Vehicle Fules, National Academy Press, Washington, DC, 1996]. But it is the potential for MTBE to enter drinking water supplies that has become an area of public concern. MTBE has been shown to induce liver and kidney tumors in rodents but the biochemical process leading to carcinogenesis is unknown. MTBE was previously shown to be non-mutagenic in the standard Ames plate incorporation test with tester strains that detect frame shift (TA98) and point mutations (TA100) and in a suspension assay using TA104, a strain that detects oxidative damage, suggesting a non-genotoxic mechanism accounts for its carcinogenic potential. These strains are deficient in excision repair due to deletion of the uvrB gene. We hypothesized that the carcinogenic activity of MTBE may be dependent upon a functional excision repair system that attempts to remove alkyl adducts and/or oxidative base damage caused by direct interaction of MTBE with DNA or by its metabolites, formaldehyde and tert-butyl alcohol (TBA), established carcinogens that are mutagenic in some Ames strains. To test our hypothesis, the genotoxicity of MTBE-induced DNA alterations was assayed using the standard Ames test with TA102, a strain similar to TA104 in the damage it detects but uvrB + and, therefore, excision repair proficient. The assay was performed (1) with and without Aroclor-induced rat S-9, (2) with and without the addition of formaldehyde dehydrogenase (FDH), and (3) with human S-9 homogenate. MTBE was weakly mutagenic when tested directly and moderately mutagenic with S-9 activation producing between 80 and 200 TA102 revertants/mg of compound. Mutagenicity was inhibited 25%-30% by FDH. TA102 revertants were also induced by TBA and by MTBE when human S-9 was substituted for rat S-9. We conclude that MTBE and its metabolites induce a mutagenic pathway involving oxidation of DNA bases and an intact repair system. These data are significant in view of the controversy surrounding public safety and the environmental release of MTBE and similar fuel additives.     

Mutagenic potential of toluidine blue evaluated in the Ames test.

Toluidine blue is a vital, metachromatic thiazine dye which is used as an adjunct in clinical examination for the early detection of asymptomatic recurrent or secondary primary carcinoma in individuals who are at high risk for developing oral cancer. Because available data on the mutagenicity of toluidine blue was limited and contradictory, this study was conducted to evaluate the mutagenic potential of toluidine blue in the in vitro Ames Salmonella test. Tester strains TA97a, TA98, TA100 and TA102 were used. Toluidine blue was tested at concentrations of 0.1, 1.0, 10, 50, 100, 250 and 500 micrograms/plate, with and without S9 microsomal activation, and positive and negative controls were included. Results from tests without S9 showed a significant increase (p less than 0.05) in number of revertants in TA102 and in TA97a with 50 and 100 micrograms toluidine blue/plate, respectively. In tests with S9 activation, doses of toluidine blue ranging from 10 to 250 micrograms/plate induced dose-related increases in the number of revertants in all 4 strains. The results of this study indicate that toluidine blue has a mutagenic effect in the Ames test.     

Revised methods for the Salmonella mutagenicity test

The methods for detecting carcinogens and mutagens with the Salmonella mutagenicity test were described previously (Ames et al., 1975b). The present paper is a revision of the methods. Two new tester strains, a frameshift strain (TA97) and a strain carrying an ochre mutation on a multicopy plasmid (TA102), are added to the standard tester set. TA97 replaces TA1537. TA1535 and TA1538 are removed from the recommended set but can be retained at the option of the investigator. TA98 and TA100 are retained. We discuss other special purpose strains and present some minor changes in procedure, principally in the growth, storage, and preservation of the tester strains. Two substitutions are made in diagnostic mutagens to eliminate MNNG and 9-aminoacridine. Some test modifications are discussed.     

Mutagenicity, comutagenicity, and antimutagenicity of erythrosine (FD and C red 3), a food dye, in the Ames/Salmonella assay

Erythrosine (diNa, tetraiodofluorescein) was nonmutagenic to the Ames/Salmonella typhimurium strains TA97a, TA98, TA100, TA102, and TA104, to a concentration of 2 mg/plate. No mutative intermediates were detected on metabolism by rat caecal cell-free extracts or rat liver S9 mixture; or on incubation with the comutagens, harman and norharman (+/- S9). Instead, an unexpected dose-dependent suppression in spontaneous reversion frequencies was observed (maximum approximately equal to 35% decrease). Erythrosine was antimutagenic to benzo[a]pyrene, but it did not decrease the mutagenicity of the other adduct-forming mutagen, 4-nitroquinoline N-oxide. The food dye was strongly antimutagenic to the bifunctional alkylating agent, mitomycin C, though it did not exhibit a similar effect on the mutagenicity of the corresponding monofunctional agent, methyl methanesulphonate. It partially depressed the mutagenic potentials of sodium azide. The antimutagenic effect of erythrosine on an intercalating agent, ethidium bromide, was discernible only at the highest dose (2 mg/plate). These results have been interpreted in terms of a genointeractive role of erythrosine. Erythrosine produced differential toxic effects in repair-deficient (TA97a, TA98, TA100) and repair-proficient (TA102, TA104) Salmonella tester strains; survival of the repair-deficient strains was found to be decreased. Photoinduced potentiation of erythrosine toxicity was observed, although light irradiation in the presence of erythrosine did not modify the reversion frequencies of the tester strains. The evidence strongly suggests that erythrosine, which exhibits nonmutagenicity in the Ames/Salmonella test, can interact with DNA repair enzymes and/or with DNA.     

Aspects of chloroquine mutagenicity

Using the Ames plate reversion and fluctuation tests, the mutagenic activity of chloroquine was tested in the new tester strains of Salmonella typhimurium, TA97, TA102, and Escherichia coli strains WP2, WP2hcr, WP6 and WP67. The E. coli transconjugants obtained from the mating transfer of R-plasmid(s) in strains TA97 and TA102 respectively to E. coli WP2, i.e. EE97 and EE102, were also tested. Chloroquine reverted strain TA97 from histidine dependence to independence and also reverted E. coli strains EE97 and EE102 from tryptophan dependence to independence. The E. coli strains WP2, WP2hcr; WP6 and WP67 and S. typhimurium TA102 were not affected. S. typhimurium TA97 could be reverted with 250 ng/ml of chloroquine (therapeutic blood level of chloroquine is 300 ng/ml). Reversion generally occurred optimally at the relatively lower concentrations of chloroquine i.e. 25, 50 micrograms/ml than at higher concentrations. From the properties of the reverted tester strains, the results indicated that chloroquine per se mediated frameshift reversion.     

The involvement of dimethyl sulfoxide in a bacteriotoxic response of the Ames assay tester strains TA98 and TA100

Dimethyl sulfoxide is a widely accepted and recommended solvent in which to dissolve compounds to be tested for mutagenicity via the Ames Salmonella/mammalian microsome assay. Using tester strains TA98 and TA100, we observed a bacteriotoxic response with various fractions isolated from beer when dissolved in DMSO but not when dissolved in water. Further characterization of the role of solvent in simple model systems consisting of butanol, DMSO and bacteria strongly suggests a chemical reaction occurs between dimethyl sulfoxide and specific chemical constituents of the test substance, nutrient broth, or the Ames bacterial strains. The result of such an interaction could be misinterpreted as a toxic response to the test substance when, in fact, the bacteriotoxicity could be due to another compound, chemically distinct from the test substance.     

Genotoxicity of industrial wastewaters obtained from two different pollution sources in northern India: a comparison of three bioassays

The genotoxicity of industrial wastewater samples from Aligarh and Ghaziabad cities was compared using the Ames plate incorporation test, the Ames fluctuation test and the Allium cepa test. While TA102 and TA104 strains exhibited the highest sensitivity against the Aligarh sample (AWW) in terms of the slope (m) of the dose-response curve in the plate incorporation assay, TA98 and TA97a were the most sensitive strains based on the induction factor, Mi(p). TA98 once again, was the most sensitive strain against the test sample from Ghaziabad (GWW) in terms of 'Mi(p)' while TA102 was the most sensitive strain on the basis of the slope (m). TA100 displayed the highest susceptibility towards the samples from Aligarh in the fluctuation test. However, TA102 and TA98 responded maximally to GWW in this bioassay. The mutagenicity of the test samples seemed to be partly mediated by reactive oxygen species (ROS) as evidenced by the use of free radical scavengers. Mannitol brought about the maximum decline in the number of revertants of TA102 by the Aligarh sample, whereas such a reduction in case of Ghaziabad sample was exhibited with superoxide dismutase. Both the test water samples induced various anaphase aberrations in the root cells of Allium cepa. Fragmentation of the chromosome was the predominant effect of the Aligarh water sample while the Ghaziabad sample induced chromosome stickiness. The crucial roles of heavy metals and pesticides in the genotoxicity of AWW and GWW, respectively, have been suggested. In view of the problem associated with the interpretations of data, we recommend that all the test bioassays should be carried out in the presence of ROS scavengers for the fool proof evaluation of the genotoxicity of water samples.     

Genotoxic activity and potency of 135 compounds in the Ames reversion test and in a bacterial DNA-repair test

Compounds of various chemical classes were comparatively assayed in the Ames reversion test with his- S. typhimurium strains TA1535, TA157 , TA1538, TA98, TA100, and, in part, TA97 , and in a DNA-repair test with trp- E. coli strains WP2 (repair-proficient), WP67 (uvrA- polA-) and CM871 (uvrA- recA- lexA-). A liquid micromethod procedure for the assessment of the minimal inhibitory concentration (MIC) of test compounds, using the same reagents as the Ames test, was set up and calibrated in its technical details. Other techniques (spot test and treat-and-plate method) were applied to a number of compounds in order to obtain more complete information on their DNA-damaging activity in E. coli. From a qualitative standpoint, the results obtained in the reversion test and in the DNA-repair test (liquid micromethod) were overlapping for 96 (59 positive and 37 negative) out of 135 compounds (71.1%). There was disagreement for 39 compounds (28.9%), 9 of which were positive only in the reversion test (8 requiring metabolic activation and 5 genotoxic in the treat-and-plate method). 30 compounds were positive only in the lethality test, showing a direct DNA-damaging activity, which in half of the cases was completely eliminated by S9 mix. Although the experimental protocol intentionally included several compounds already reported as nonmutagenic carcinogens or as noncarcinogenic mutagens, the overall accuracy was 64.5% for the reversion test and 72.4% for the DNA-repair test, as evaluated for 75 compounds classified according to their carcinogenic activity. Quantitation of results was obtained in the Ames test by relating the net number of revertants to nmoles of compound and in the DNA-repair test by means of a formula relating the difference and ratio of MICs in repair-proficient and -deficient bacteria to nmoles of compound. Following these criteria, the genotoxic potency varied over a 4.5 X 10(7)-fold range among compounds positive in the reversion test and over a 6 X 10(9)-fold range among compounds damaging E. coli DNA. The genotoxic potencies in the two bacterial systems were correlated within the majority of the chemical classes under scrutiny.     

Mutagenic and genotoxic effects of Anilofos with micronucleus,chromosome aberrations,sister chromatid exchanges and Ames test

We aimed to evaluate the mutagenic effect of Anilofos, organophosphate pesticide, by using Ames/Salmonella/microsome test. Its cytotoxic and genotoxic effects were also determined by chromosome aberration (CA), sister chromatid exchange (SCE) and micronucleus (MN) test in human peripheral blood lymphocytes. In the Ames test, five different concentrations of Anilofos were examined on TA97, TA98, TA100 and TA102 strains in the absence and presence of S9 fraction. According to the results all concentrations of this pesticide have not shown any mutagenic activity on TA97, TA100 and TA102 strains in the absence and presence of S9 fraction. But, 10, 100 and 1000 µg/plate concentrations of Anilofos were determined to be mutagenic on TA98 strain without S9 fraction. Lymphocytes were treated with various concentrations (25, 50, 100 and 200 µg/ml) of Anilofos for 24 and 48 h. The results of the assays showed that Anilofos did not induce SCE frequency, replication index and MN formation at all concentrations for both treatment periods. Anilofos significantly increased CA frequency at 100 and 200 µg/ml concentrations at 24 h treatment periods and at 50, 100 and 200 µg/ml concentrations in 48 h treatment periods. Additionally, it was determined that this pesticide decreased mitotic index and nuclear division index significantly. It was concluded that Anilofos has genotoxic and cytotoxic effects in human peripheral lymphocytes.     

Role of bacterial metabolism in transformation of non-mutagenic compounds into mutagens. II. Participation of bacteria producing urease in degradation of pesticides--urea derivatives]

Participation of bacteria producing urease: Proteus mirabilis and Staphylococcus epidermidis in degradation of pesticides--urea derivatives, was investigated. Four new compounds were studies: N-/(3-trifluoromethyl)phenyl/N'-(2-cyanoethyl)-urea (IPO 6584), N-(4-chlorophenyl)-N'-(2-hydroxyethyl)-N'-ethylurea (PO 6236), N-(4-chlorophenyl)-N'-(2-hydroxyethyl), N'-propyl-urea (IPO 6237), N-(2-hydroxyethyl), N-methyl-N'-(3,4 dichlorophenyl)-urea (IPO 3102), pesticide reference standard--thiram fungicide--tetramethyl-bis-thiocarbamyl disulfide, and rodenticide-alpha-naphthylthiourea (ANTU). Investigated compounds were incubated with cells of P. mirabilis 4508 ans S. epidermidis CCM 2448 and commercial preparation of urease from beans. Mutagenicity of resulting metabolites was then studies by the Ames test. All compounds were decomposed by bacteria used in this study, as well with beans urease with different activity. Reaction products did not exhibit mutagenic activity for test strains S. typhimurium his- TA97a, TA98, TA100 and TA102.     

Mutagenicity of 4-nitroquinoline N-oxide after its complexation with copper(II) 2-chlorophenoxyacetate

Using the Ames test withSalmonella typhimurium strains TA97, TA100 and TA102 it was found that the mutagenicity of 4-nitroquinoline N-oxide was much lower than that of the tetrakis[μ-2-chlorophenoxyacetato(1−)-O,O′]bis(4-nitroquinoline N-oxide-O)dicopper(II) complex. This work was supported by theMinistry of Education of the Slovak Republic in Grant no. 95/5195/416.     

Measurement of drug-metabolizing systems in Salmonella typhimurium strains G46, TA1535, TA100, TA1538 and TA98

Salmonella typhimurium strains which are commonly used in the Ames test for screening potential carcinogens were examined for a number of drug-metabolizing systems. Neither cytochrome P-450 itself nor two activities catalyzed by the cytochrome P-450 system in mammalian cells, i.e., benzpyrene monooxygenase and ethoxycoumarin O-deethylation, could be detected. Nor do these bacterial strains demonstrate any ability to detoxify epoxides by hydrating them or to conjugate p-nitrophenol with glucuronic acid.On the other hand, S. typhimurium strains G46, TA1535, TA100, TA1538 and TA98 contain considerable amounts of acid-soluble thiols, approx. 5–10% of which is glutathione. These bacteria can also enzymatically conjugate glutathione with 1-chloro-2,4-dinitrobenzene (CDNB) and can reduce oxidized glutathione using NADPH as cofactor.Thus, enzymatic and non-enzymatic reaction of immediate carcinogens with thiol groups in S. typhimurium may have a significant effect on the outcome of the Ames test in certain cases.     

A comparison of aflatoxin B1-induced cytotoxicity,mutagenicity and prophage induction in Salmonella typhimurium mutagen tester strains TA1535, TA1538, TA98 and TA100

Treatment of Ames mutagen tester strains with aflatoxin B1 (AFB1) and S9 mix results not only in the production of a poten mutagen, but induces a pathway that leads to the induction of prophages present in all Ames tester strains.Characterization of the prophage induction and mutagenic response following AFB1 treatment showed that plasmid pKM101 dramatically enhances mutagenesis, but suppressed prophage induction. Spontaneous release of phage by TA98 and TA100 was also lower than in TA1535 and TA1538.In addition to mutagenesis and prophage induction, survival of all 4 tester strains was quantitated after AFB1 treatment. The data show that the frameshift tester strains (TA1538 and TA98) are more sensitive to the bactericidal action of AFB1 than the base-pair tester strains (TA1535 and TA100), survival being significantly affected above 100 ng. One of several hypotheses examined was the difference in the number and types of prophages present in base-pair tester strains that are not detectable in the frame-shift tester strains.These data suggest that prophage induction can detect DNA damage that is non-mutagenic; and that it is important to characterize the lysogenic nature of the Ames strains since it may influence the observed histidine revertant rate and the survival of the tester strain.     

Structure-mutagenicity relationships in a series of indolo[3,2-c]quinoline-1,4-diones that have shown cytotoxic properties on leukemia cells.

A series of seven 6-methylindolo[3,2-c]quinoline-1,4-diones substituted either in the 2 position or in 3 position by various groups were examined for their ability to induce mutation in the Ames test at several concentrations in four strains of Salmonella typhimurium (TA97, TA98, TA100, and TA102). First, relationships were established between their mutagenic activities and either the nature or the position of the substituent on the quinonic nucleus. Compounds substituted in the 2 position were less mutagenic than the 3 isomers. In the second study, the mutagenic properties were compared to the in vitro antitumor activity. Interestingly, some very cytotoxic quinones were only weak mutagens. So where the cytotoxicity is similar, the less mutagenic compounds may be suitable for clinical use as antitumor drugs, in order to avoid important side effects; the Ames test can then be used guide the selection of molecules for further in vivo antitumor screening. It can also be very helpful in selecting the best candidate molecules to be synthesized.     

Genotoxic evaluation of extracts from Aplysina fulva, a Brazilian marine sponge

A range of biologically active secondary metabolites with pharmacological application has been reported to occur in marine sponges. The present study was undertaken to provide a set of data on the safety of a hydro-alcoholic extract (ALE) and an aqueous fraction (AQE) from Aplysina fulva Pallas, 1766 (Aplysinidae, Verongida, Porifera). Salmonella typhimurium strains TA97, TA98, TA100 and TA102, Escherichia coli strains PQ65, OG40, OG100, PQ35 and PQ37 and Balb/c 3T3 mouse fibroblasts were used to detect induction of DNA lesions by ALE and AQE. Assays used for these analyses were a bacterial (reverse) mutation assay (Ames test), the SOS-chromotest and the comet assay. Both extracts presented identical infrared 2-oxazolidone spectra. ALE treatment induced a higher frequency of type-4 comets, indicative of increasing DNA migration, in the alkaline comet assay. ALE also induced a weak genotoxic effect, as expressed by the induction factor (IF) values in the test with E. coli strain PQ35 (IF=1.5) and by cytotoxic effects in strains PQ35, PQ65 and PQ37. Positive SOS induction (IF=1.7) was detected in strain PQ37 treated with diluted AQE. No genotoxic effects were observed in strains PQ35, PQ65, OG40 and OG 100 after treatment with AQE dilutions. Using the bacterial (reverse) mutation test and survival assays with or without S9 mix, after 60min of pre-incubation, we observed for strain TA97 treated with ALE a weak mutagenic response (MI=2.2), while cytotoxic effects were seen for strains TA98, TA100 and TA102. AQE did not show mutagenic activity in any of the strains tested, but a weak cytotoxic effect was noted in strain TA102. Our data suggest that both ALE and AQE from A. fulva induce DNA breaks leading to cytotoxicity and mutagenicity under the conditions used.     

The Salmonella mutagenicity of industrial, surface and ground water samples of Aligarh region of India

The genotoxicity of three water bodies, viz. industrial waste water of Aligarh city, ground water pumped out from the industrial area of Aligarh, and river water of Yamuna, downstream of Agra, was carried out by means of Ames plate incorporation test and the Ames fluctuation test. All the test samples were significantly mutagenic in both the testing systems. The ground water and river water samples were subjected to XAD concentration prior to the mutagenicity/genotoxicity testing, while the industrial waste water was used directly. Whereas TA98, TA102 and TA104 strains have been found to be maximally sensitive in the Ames plate incorporation assay conducted for various water samples, TA98 and TA100 strains were the most responsive strains in the Ames fluctuation test. The apparent disparity in the sensitivity patterns of various Ames strains by plate incorporation and fluctuation assays could be attributed to a large extent to the different conventional ways of interpretation of the data in these systems.     

Further examination of the effects of nitrosylation on Alternaria alternata mycotoxin mutagenicity in vitro

Previously, Alternaria extract and metabolite mutagenicities+/-nitrosylation were characterized using Ames Salmonella strains TA98 and TA100, which are both reverted at GC sites. To examine other targets for mutation, the metabolites Altertoxin I (ATX I), Altenuene (ALT), Alternariol (AOH), Alternariol monomethyl ether (AME), Tentoxin (TENT), Tenuazonic acid (TA) and Radicinin (RAD) were reexamined+/-nitrosylation, using Ames Salmonella strain TA97, sensitive to frameshift mutations at a run of C's, as well as strains TA102 and TA104, reverted by base pair mutations at AT sites and more sensitive to oxidative damage. ATX I was also assessed for mammalian mutagenicity at the Hprt gene locus in Chinese hamster V79 lung fibroblasts and rat hepatoma H4IIE cells. When tested from 1 to 100 microg/plate without nitrosylation, ATX I was mutagenic in TA102+/-rat liver S9 for activation and weakly mutagenic in TA104+/-S9, demonstrating direct-acting AT base pair mutagenicity. AOH was also directly mutagenic at AT sites in TA102+/-S9 while AME was weakly mutagenic in TA102+/-S9 and TA104+S9. Nitrosylation of ATX I enhanced mutagenicity at AT sites in TA104+/-S9 but produced little change in TA102+/-S9 compared to native ATX I. However, nitrosylated ATX I generated a potent direct-acting frameshift mutagen at C sites in TA97+/-S9. While ATX I was not mutagenic in either V79 cells or H4IIE cells, 5 and 10 microg/ml nitrosylated ATX I produced a doubling of 6-thioguanine resistant V79 colonies and 0.5 and 1 microg/ml were mutagenic to H4IIE cells, becoming toxic at higher concentrations. These results suggest ATX I, AME and AOH induce mutations at AT sites, possibly through oxidative damage, with nitrosylation enhancing ATX I frameshift mutagenicity at runs of C's. Nitrosylated ATX I was also directly mutagenic in mammalian test systems.