Studies on the genotoxicity of endosulfan in bacterial systems

Endosulfan, an organochlorine pesticide, was subjected to the differential sensitivity assay in repair-deficient and repair-proficient strains of Escherichia coli K12, prophage lambda induction assay in WP2s (lambda) and mutation induction in E. coli K12. The induction of umu gene expression with endosulfan was studied also in Salmonella typhimurium TA1535/pSK1002 cells. The differential sensitivity assay revealed that the recA 13 strain was the most sensitive. Endosulfan induced prophage lambda in E. coli and umu gene expression in S. typhimurium cells; however, the extent of the effects were low. Endosulfan also induced a dose-dependent increase in forward mutations in E. coli K12 cells from ampicillin sensitivity to ampicillin resistance. Our studies indicate the genotoxic potential of endosulfan and the role of the recA gene in the repair of endosulfan-induced DNA damage.     

Microsatellite instability induced by hydrogen peroxide in Escherichia coli

Damage to DNA by reactive oxygen species may be a significant source of endogenous mutagenesis in aerobic organisms. Using a selective assay for microsatellite instability in E. coli, we have asked whether endogenous oxidative mutagenesis can contribute to genetic instability. Instability of repetitive sequences, both in intronic sequences and within coding regions, is a hallmark of genetic instability in human cancers. We demonstrate that exposure of E. coli to low levels of hydrogen peroxide increases the frequency of expansions and deletions within dinucleotide repetitive sequences. Sequencing of the repetitive sequences and flanking non-repetitive regions in mutant clones demonstrated the high specificity for alterations with the repeats. All of the 183 mutants sequenced displayed frameshift alterations within the microsatellite repeats, and no base substitutions or frameshift mutations occurred within the flanking non-repetitive sequences. We hypothesize that endogenous oxidative damage to DNA can increase the frequency of strand slippage intermediates occurring during DNA replication or repair synthesis, and contribute to genomic instability.     

Repair of thymine.guanine and uracil.guanine mismatched base-pairs in bacteriophage M13mp18 DNA heteroduplexes

Repair of thymine.guanine (T.G) and uracil.guanine (U.G) mismatched base-pairs in bacteriophage M13mp18 replicative form (RF) DNA was compared upon transfection into repair-proficient or repair-deficient Escherichia coli strains. Oligonucleotide-directed mutagenesis was used to prepare covalently closed circular heteroduplexes that contained the mismatched base-pair at a restriction recognition site. The heteroduplexes were unmethylated at dam (5'-GATC-3') sites to avoid methylation-directed biasing of repair. In an E. coli host containing uracil-DNA glycosylase (ung+), about 97% of the transfecting U.G-containing heteroduplexes had the U residue excised by the uracil-excision repair system. With the analogous T.G mispair, mismatch repair operated on almost all of the transfecting heteroduplexes and removed the T residue in about 75% of them when the mismatched T was on the minus strand of the RF DNA. Similar preferential excision of the minus-strand's mismatched base was observed whether the heteroduplex RF DNA molecules had only one or both strands unmethylated at dcm (5'-CC(A/T)GG-3') sites and whether the RF DNA was prepared by primer extension in vitro or by reannealing mutant and non-mutant DNA strands. Also, the extent and directionality of repair was the same at a U.G mispair in ung- host cells as at the analogous T.G mispair in ung- or ung+ cells. Only in a mismatch repair-deficient (mutH-) host was the plus strand of the transfecting M13mp18 heteroduplex DNA preferentially repaired. It is suggested that the plus strand nick made by the M13-encoded gene II protein might be employed by a mutH- host to initiate repair on that strand.     

An evaluation of tests using DNA repair-deficient bacteria for predicting genotoxicity and carcinogenicity. A report of the U.S. EPA's Gene-TOX Program

The detection of DNA-damaging agents by repair-deficient bacterial assays is based on the differential inhibition of growth of repair-proficient and repair-deficient bacterial pairs. The various methodologies used are described and recommendations are made for their improved use. In a survey of the literature through April 1979, 91 of 276 papers evaluated contained usable data, resulting in an analysis of 611 compounds that had been assayed in 1 or more of 55 pairs of repair-proficient and repair-deficient strains. The results indicate that (1) a liquid suspension assay is more sensitive than a spot (diffusion) test. In a review of the Escherichia coli polA assay, 45 compounds that gave "No Test" in the spot test were clearly positive or negative in the liquid suspension assay. (2) Of the 21 compounds analyzed by the E. coli polA assay and by other E. coli repair-deficient strains (e.g., rec, uvr, hcr, and exr derivatives of WP2 and AB1157), 10 were in complete agreement in all strains except uvrA strains. This indicates that strains other than polA+/polA- are useful for detecting DNA-damaging agents. However, in selecting strains for use in these assays, care should be taken to consider repair pathway specificity for particular compounds. (3) There was a 78% correspondence between results obtained with E. coli polA and Bacillus subtilis (H17/M45, 17A/45T) rec assay and between E. coli polA and Proteus mirabilis. (4) In a comparison of test results with carcinogenicity data, 44 of 71 (62%) carcinogenic compounds assayed by the polA system were positive, 10 (14%) were negative, and 17 (24%) gave No Test or doubtful results. 7 carcinogens were assayed by other E. coli strains and all were positive. 56 carcinogens were assayed in B. subtilis: 24 (43%) were positive, 9 (16%) were negative, and 23 (41%) gave No Test or doubtful results. Of the 7 carcinogens assayed in P. mirabilis, 6 (86%) were positive and 1 (14%) was negative. (5) The results were analyzed with respect to chemical classes. E. coli polA detected the highest percentage of hydroxylamines and alkyl epoxides. The B. subtilis rec assay detected the highest percentage of nitrosamines and sulfur and nitrogen oxides. It is concluded that some of these test systems are effective tools for the detection of DNA-damaging and potentially carcinogenic compounds, especially if the assay is done in liquid suspension and if more than 1 pair of tester strains is used. Advantages and disadvantages of the assay are discussed and suggestions are made for improvements in the system.     

Effects of oxidative and alkylating damage on microsatellite instability in nontumorigenic human cells

Microsatellite instability is a phenomenon that is well characterized in mismatch repair-deficient tumor cell lines, including the potential etiological role of endogenous DNA damage. However, our understanding of microsatellite mutational mechanisms in repair-proficient, nontumorigenic cells is limited. We determined microsatellite mutation frequencies for human lymphoblastoid cells using an episomal DNA shuttle vector in which a (TTCC/AAGG)(9) microsatellite is inserted in-frame within the herpes simplex virus thymidine kinase (HSV-tk) gene. The responses of plasmid-bearing cells to reactive oxygen species or alkylating agents were compared after treatment with hydrogen peroxide (H(2)O(2)) and N-ethyl-N-nitrosourea (ENU). H(2)O(2) treatment induced a statistically significant increase in overall HSV-tk mutation frequency relative to controls, with catalase reducing the effect. H(2)O(2) treatment increased the mutation frequency within the microsatellite and the HSV-tk coding region to a similar extent (five and six-fold, respectively, relative to the control). Mutational specificity analyses demonstrated that the proportion of mutations within the microsatellite is not statistically different among the H(2)O(2), catalase, and PBS treatment groups. In contrast, treatment of cells bearing the microsatellite vector with ENU altered the mutational spectrum, relative to solvent control. ENU induced the expected base substitutions within the HSV-tk coding region, but did not increase the microsatellite mutation frequency. The low level of microsatellite mutagenesis observed after reactive oxygen species (ROS) insult likely reflects the normal repair processes of these nontumorigenic, repair-competent cells. Our ex vivo experiments demonstrate the manner in which repetitive DNA in normal human cells might respond to endogenous mutagens.     

Low abundance of Escherichia coli microsatellites is associated with an extremely low mutation rate

It is widely assumed that microsatellites are generated by replication slippage, a mutation process specific to repetitive DNA. Consistent with their high mutation rate, microsatellites are highly abundant in most eukaryotic genomes. In Escherichia coli, however, microsatellites are rare and short despite the fact that a high microsatellite mutation rate was described. We show that this high microsatellite instability depends on the presence of the F-plasmid. E. coli cells lacking the F-plasmid have extremely low microsatellite mutation rates. This result provides a possible explanation for the genome-wide low density of microsatellites in E. coli. Furthermore, we show that the F-plasmid induced microsatellite instability is independent of the mismatch repair pathway.     

Temperature-dependent antimutagenic activity of acrolein in Escherichia coli

The effect of temperature on the antimutagenic activity of acrolein was investigated using UV-irradiated E. coli B. When incubated at lower temperatures (30°C or 37°C), acrolein greatly reduced the mutation frequency in WP2 (wild-type strain), but no such effect was observed with WP2s and ZA159 (excision repair-deficient strains). The antimutagenic activity of acrolein increased when cells were incubated at higher temperatures (40°C or 42°C). Particularly in excision repair-deficient strains, the antimutagenic activity was observed only at higher temperatures. In heat shock response-deficient background, however, the antimutagenic activity was observed at 30°C even in the excision repair-deficient strains.     

Antimutagenic effect of garlic (Allium sativum L.) on 4NQO-induced mutagenesis in Escherichia coli WP2

Aqueous extract prepared from garlic bulbs markedly suppressed the mutagenesis in both E. coli WP2 trp- and E. coli WP2 trp- uvrA- induced by 4-nitroquinoline 1-oxide (4NQO), but not that induced by UV. Cellular toxicity, inhibition of the expression of the Trp+ phenotype and delay of the first cell division after 4NQO treatment were not observed in the presence of the extract. Since the extract showed identical antimutagenic effects against 4NQO in both test strains but no effect on the mutagenesis of UV, it seems that the extract might act by inactivating the electrophilic group(s) of 4NQO or inhibiting its metabolic activation.     

Mutagen sensitivity of Drosophila melanogaster. VI. Evidence from the excision-defective mei-9AT1 mutant for the timing of DNA-repair activity during spermatogenesis

The sex-linked recessive lethal test has been used to compare mutation induction by ethyl methanesulfonate and methyl methanesulfonate in spermatogenic stages of the DNA repair-deficient mei-9AT1 mutant and a repair-proficient control strain. For both agents, the data demonstrate that induced mutation rates are similar in both strains for the meiotic and post-meiotic broods. Conversely, for spermatogonial broods, the data indicate that the excision-deficient strain exhibits a 4-8 fold increase in induced mutation rate in comparison to the excision-proficient control strain. These experiments suggest that the low mutability of gonial cells normally observed for these agents is due to effective excision-repair processes which function until the commencement of meiosis. From alkylation mutagenesis experiments with repair-deficient E. coli strains, we note that the mei-9 strain exhibits pleiotropic mutant phenotypes very similar to those displayed by the uvr D mutant. By analogy with these studies, we speculate that mei-9, like uvr D, is deficient in a DNA unwinding protein.     

Mutational hotspot variability in an ultraviolet-treated shuttle vector plasmid propagated in xeroderma pigmentosum and normal human lymphoblasts and fibroblasts

The mutagenesis shuttle vector, pZ189, was treated with ultraviolet (u.v.) radiation in vitro and passed through a DNA repair-deficient lymphoblastoid cell line derived from a patient with xeroderma pigmentosum complementation group A (XP-A) (XP12BE(EBV)) and a DNA repair-proficient lymphoblastoid cell line (GM606(EBV)). After u.v. treatment, plasmid survival was lower and mutation frequency higher with the XP-A cells mirroring the survival and mutagenesis of the host cells. The nature of the mutations in the suppressor tRNA marker gene was determined by direct sequence analysis. The G.C to A.T transition was the dominant (85%) base substitution mutation with the XP lymphoblasts and was the major (56%) base substitution mutation with the repair-proficient lymphoblasts. We found a G.C to A.T transition mutational hotspot with the XP lymphoblasts not seen in our previous experiments with fibroblasts from the same patient. Comparison of the data presented here with our results with DNA repair-deficient and DNA repair-proficient fibroblasts suggests that hotspot variability is not due to genetic polymorphism or repair capacity of the cells. Instead it appears that cellular factors can influence the probability of mutagenesis of modified DNA at particular sites.     

Induction of repairable DNA damage in Escherichia coli and interaction with DNA in vitro by the radical cation of chlorpromazine

Studies were performed to determine the DNA interactions of and the induction of cytotoxic effects by the radical cation (CPZ+.) formed enzymatically from chlorpromazine (CPZ): in the presence of native DNA the lifetime of CPZ+. is markedly increased. The decreased reactivity of CPZ+. in the presence of native DNA and the concomitant increased viscosity of CPZ+.-DNA complexes strongly support the assumption that CPZ+. does form intercalation complexes with DNA. The relative strong bacteriotoxicity of CPZ+. hindered the accurate determination of mutagenesis in various Salmonella indicator strains, but a test for repairable DNA damage in Escherichia coli using various repair-deficient strains indicated that the cytotoxic action of CPZ+. is in part due to DNA alterations which can be excised in wild-type DNA repair-proficient strains. After activation of CPZ with long wavelength UV light, genetic effects are observed in S. typhimurium strain TA98, as well as in the E. coli tester strains. The possible role of CPZ+. in the photosensitization of CPZ is discussed.     

Bio-antimutagenic effects of tannic acid on UV and chemically induced mutagenesis in Escherichia coli B/r

Tannic acid suppressed the mutagenesis in E. coli B/r WP2 trp- induced by UV or 4-nitroquinoline 1-oxide (4NQO), but not that induced by gamma-rays or N-methyl-N'-nitro-N-nitrosoguanidine (MNNG). The depression of mutations induced by UV was most remarkable in the DNA-repair-proficient strain (WP2). Tannic acid, however, showed no bio-antimutagenic effect in the excision repair-deficient strain (WP2s uvrA- or ZA159 uvrB-) under the test conditions where no cellular toxicity was observed. The effect ceased within 30 min after UV irradiation. The inhibition of the expression of Trp+ phenotype and the delay of the first cell division after UV irradiation were not observed in the presence of tannic acid. From these results we conclude that tannic acid may enhance the excision-repair system probably by activating the repair enzymes or by interacting with DNA.     

DNA mismatch repair and the significance of a sebaceous skin tumor for visceral cancer prevention

DNA mismatch repair is a postreplicative DNA repair cascade ensuring genomic integrity. Inactivating germline mutations in DNA mismatch repair genes are responsible for hereditary non-polyposis colorectal carcinoma syndrome (HNPCC), which predisposes to various types of visceral cancer. Most associated tumors exhibit high-grade microsatellite instability. Some patients develop skin tumors of the sebaceous glands. This combined occurrence is known as Muir-Torre syndrome, which has a high probability of an underlying DNA mismatch repair defect. This is also true for individuals selected solely on the basis of sebaceous neoplasias, tumors with the highest frequency of high-grade microsatellite instability. This article focuses on the recent advances in molecular diagnostics for the detection of DNA mismatch repair defects in patients with sebaceous neoplasias, and the potential significance for the secondary prevention of visceral cancer in these patients.     

Rat liver S9 preparations contain comutagenic activity

Rat liver S9 preparations contain material which causes enhancement of UV mutagenesis in Escherichia coli WP2. This comutagenic activity is present in S9 preparations from both uninduced and Aroclor-induced rats. Strains of E. coli which are defective in the uvr-dependent excision repair pathway fail to show comutagenic action by S9. The comutagenic material is heat-labile and non-dialyzable, suggesting that it might be protein. This differs from the small amount of mutagenic material present in rat liver S9, as the latter is dialyzable and can be demonstrated in the repair-deficient strain E. coli WP2s (uvrA).     

Antimutagenic Effects of Autoxidized Linoleic and Oleic Acids on UV-Induced Mutagenesis in Escherichia coli

The antimutagenic effects of autoxidized linoleic and oleic acids on mutagenesis by UV irradiation were investigated in Escherichia coli B/r WP2 and WP2s uvrA. When added to an agar medium, these autoxidized acids greatly reduced the number of Trp⁺ revertants without significant effects on survival in WP2, but no such effect was observed with WP2s uvrA. The presence of autoxidized linoleic acid decreased the survival of WP2s uvrA greatly and CM571 recA somewhat. It thus appears that the autoxidized unsaturated fatty acid has antimutagenic effects on the wild type strain and lethal effects on the genetic repair-deficient strains.     

A differential DNA-repair test using mixtures of E. coli K12 strains in liquid suspension and animal-mediated assays

The feasibility of performing tests for repairable DNA damage in animal assay procedures was investigated by using repair-proficient and repair-deficient derivatives of E. coli K12 strain 343/113, including mutations in the uvrB, recA, polA and dam genes. To avoid variations in the relative recovery of viable cells from different samples, the strains were further marked with auxotrophic growth requirements, so that mixtures could be treated and the survival of each strain determined individually on media containing the corresponding growth factors. Spot tests were performed with the various strains to re-assess the necessity of using a combination of repair deficiencies, when genotoxic agents of differing mode of action are to be detected. Liquid suspension tests on mixtures of the different strains, furthermore, confirmed that the survival of the individual strains can be determined separately on selective media after treatment with methyl methanesulfonate (MMS) and methyl nitrosourea (MNU). These tests were also used to demonstrate that dimethyl nitrosamine (DMNA) is activated by Aroclor-1254-induced rat-liver S9 fractions to genotoxic products, as measured by the low survival of a recA derivative compared with the repair-proficient wild-type strain. Intrasanguineous host-mediated assays using the present derivatives of E. coli K12/343/113 showed that the various strains, injected simultaneously into mice, could be recovered in amounts sufficient for the individual determination of the relative survival in liver, spleen, lungs, kidneys, pancreas and the blood stream of the host animals. Using a mixture of the repair-proficient parent and the recA derivative inoculated into mice that were subsequently treated with MMS, NMU or DMNA, we found that these chemicals induce a larger decrease in survival in the recA strain as compared with the wild-type in cells recovered from the liver and the spleen. The order of genotoxic potency so determined was DMNA greater than MNU greater than MMS; this is similar to the ranking of the carcinogenicity of these compounds in rodents and probably also reflects the various degrees of DNA alkylation in cells of the livers of the treated animals. The general usefulness of the host-mediated differential DNA repair assay for detecting genotoxic factors in various organs of animals remains to be assessed by using chemical mutagens of different modes of action.     

Antimutagenicity of cinnamaldehyde and vanillin in human cells: Global gene expression and possible role of DNA damage and repair

Vanillin (VAN) and cinnamaldehyde (CIN) are dietary flavorings that exhibit antimutagenic activity against mutagen-induced and spontaneous mutations in bacteria. Although these compounds were antimutagenic against chromosomal mutations in mammalian cells, they have not been studied for antimutagenesis against spontaneous gene mutations in mammalian cells. Thus, we initiated studies with VAN and CIN in human mismatch repair-deficient (hMLH1(-)) HCT116 colon cancer cells, which exhibit high spontaneous mutation rates (mutations/cell/generation) at the HPRT locus, permitting analysis of antimutagenic effects of agents against spontaneous mutation. Long-term (1-3 weeks) treatment of HCT116 cells with VAN at minimally toxic concentrations (0.5-2.5mM) reduced the spontaneous HPRT mutant fraction (MF, mutants/10(6) survivors) in a concentration-related manner by 19-73%. A similar treatment with CIN at 2.5-7.5microM yielded a 13-56% reduction of the spontaneous MF. Short-term (4-h) treatments also reduced the spontaneous MF by 64% (VAN) and 31% (CIN). To investigate the mechanisms of antimutagenesis, we evaluated the ability of VAN and CIN to induce DNA damage (comet assay) and to alter global gene expression (Affymetrix GeneChip) after 4-h treatments. Both VAN and CIN induced DNA damage in both mismatch repair-proficient (HCT116+chr3) and deficient (HCT116) cells at concentrations that were antimutagenic in HCT116 cells. There were 64 genes whose expression was changed similarly by both VAN and CIN; these included genes related to DNA damage, stress responses, oxidative damage, apoptosis, and cell growth. RT-PCR results paralleled the Affymetrix results for four selected genes (HMOX1, DDIT4, GCLM, and CLK4). Our results show for the first time that VAN and CIN are antimutagenic against spontaneous mutations in mammalian (human) cells. These and other data lead us to propose that VAN and CIN may induce DNA damage that elicits recombinational DNA repair, which reduces spontaneous mutations.     

The amyloid peptide induces early genotoxic damage in human preneuron NT2

The antimutagenic activities of benzalacetone (4-phenyl-3-buten-2-one) and its structurally-related compounds were evaluated through their use as post-treatments for the UV-induced mutagenesis in Escherichia coli WP2s (uvrA) and the gamma-induced mutagenesis in Salmonella typhimurium TA2638, the latter of which is sensitive to oxidants. Structure-activity relationships were studied between IC(50) activity values, i.e. the dose (micromol/ml) at which the mutation frequency is reduced to 50% of the control, and electronic and hydrophobicity properties of the studied molecules. Benzalacetone and benzalacetone analogs, cinnamaldehyde and trans-1,1,1-trifluoro-4-phenyl-3-buten-2-one (TF), inhibited both forms of mutagenesis, but methyl cinnamate, cinnamic acid and cinnamamide did not. The IC(50) values of TF, for UV-induced mutagenesis and gamma-induced mutagenesis, were 0.028 and 0.045 micromol/ml, respectively, and one order of magnitude lower than those of cinnamaldehyde and benzalacetone. The three antimutagenic analogs listed in order of decreasing activity are: TF>cinnamaldehyde>benzalacetone. This order is proportional to the electron-withdrawing property of the terminal group attached to an alpha,beta-unsaturated carbonyl moiety in the side chain that is known to play an important role in the antimutagenicities of benzalacetone and related compounds. In UV-induced mutagenesis in E. coli WP2s, mono-substituted benzalacetones - the ring-substituents of which have electron-withdrawing properties - showed antimutagenic activity that correlated with their electronic property. In gamma-induced mutagenesis in S. typhimurium TA2638, the antimutagenic activities of mono-substituted benzalacetones were proportional to the substituent hydrophobicities (pi). The different effects on both the mutation-induced systems is suggested to be related to the relative permeability of the cell membranes and the different sensitivities to mutagens between E. coli WP2s and S. typhimurium TA2638. In addition, the antimutagenic activity against gamma-induced mutagenesis could be due to the ability of parent compounds or their derivatives to scavenge long-lived organic radicals; the radicals have been described to be generated as a result of the X-irradiation of cells by Koyama et al. [Mutat. Res. 421 (1998) 45].     

Biological effects of dyes on bacteria. VI. Mutation induction by acridine orange and methylene blue in the dark with special reference to Escherichia coli WP6 (polA1)

Acridine orange (AO) and methylene blue (MB) in the dark were shown to be weak to moderate mutagens (induction of resistance to T5 phage) in repair-deficient strains of Escherichia coli B/r. However, strain WP2 (wild-type) was not mutated by AO in the dark, in confirmation of earlier data. The presence of 2 microM AO reduced by 41% the spontaneous mutation rate in strain WP2, from 4.1 to 2.4 mutants/10(8) cells/generation. In the polymerase I-deficient strain WP6 (polA1), 2 microM AO increased the mutation rate in the dark 14-fold. We propose that both spontaneous and AO-induced mutagenesis in the absence of light occur at the site of semiconservative DNA replication. If the intercalation mechanism for the effects in the absence of light is valid, the wild-type strain (WP2) may be resistant to frameshift mutagenesis induced by intercalated compounds, while the polymerase I-deficient strain (WP6) may be highly suceptible to the presence of an intercalated dye such as AO at the DNA-replication fork. MB and AO likely act through different mechanisms since MB is only a moderate mutagen in strain WP6 and the other repair-deficient strains tested.     

Evaluation of drug candidates in a battery of short-term genetic toxicology assays: overview

2-Hydroxy-3-methoxybenzaldehyde (omicron-vanillin), the antimutagenic effect of which has been reported on mutagenesis induced by 4-nitroquinoline 1-oxide (4NQO) in Escherichia coli WP2s, enhanced N-methyl-N'-nitro-N-nitrosoguanidine (MNNG)-induced mutagenesis in the same strain. A remarkable enhancement of mutagenesis provoked by N-methyl-N-nitrosourea (MNU) was also observed by the addition of omicron-vanillin. No enhancing effect was observed on mutagenesis induced by other mutagens such as methyl methanesulfonate (MMS), dimethylsulfate, N-ethyl-N'-nitro-N-nitrosoguanidine (ENNG), N-ethyl-N-nitrosourea (ENU), ethyl methanesulfonate, diethylsulfate, 4NQO and furylfuramide (AF-2). On the contrary, omicron-vanillin greatly suppressed AF-2- and 4NQO-induced mutagenesis and showed a slight suppressing effect against mutagenesis induced by MMS, ENNG and ENU. One possible explanation for the enhancing effect of omicron-vanillin on the mutagenesis induced by MNNG or MNU in E. coli WP2s may be inhibition of an inducible adaptive response. Among 7 derivatives of omicron-vanillin, 2-hydroxy-3-ethoxy-benzaldehyde, omicron-hydroxybenzaldehyde and m-methoxybenzaldehyde showed an enhancing effect on MNNG-induced mutagenesis.