Hepatocarcinogen quinoline induces G:C to C:G transversions in the cII gene in the liver of lambda/lacZ transgenic mice (MutaMouse)

DNA Repair and Mutagenesis of ADP-Ribosylated DNA by Pierisin

Pierisin is a DNA-targeting ADP-ribosyltransferase found in cabbage white butterfly (Pieris rapae). Pierisin transfers an ADP-ribosyl moiety to the 2-amino group of the guanine residue in DNA, yielding N2-(ADP-ribos-1-yl)-2'-deoxyguanosine (N2-ADPR-dG). Generally, such chemically modified DNA is recognized as DNA damage and elicits cellular responses, including DNA repair pathways. In Escherichia coli and human cells, it has been experimentally demonstrated that N2-ADPR-dG is a substrate of the nucleotide excision repair system. Although DNA repair machineries can remove most lesions, some unrepaired damages frequently lead to mutagenesis through DNA replication. Replication past the damaged DNA template is called translesion DNA synthesis (TLS). In vitro primer extension experiments have shown that eukaryotic DNA polymerase κ is involved in TLS across N2-ADPR-dG. In many cases, TLS is error-prone and thus a mutagenic process. Indeed, the induction of G:C to T:A and G:C to C:G mutations by N2-ADPR-dG in the hypoxanthine phosphoribosyltransferase gene mutation assay with Chinese hamster cells and supF shuttle vector plasmids assay using human fibroblasts has been reported. This review provides a detailed overview of DNA repair, TLS and mutagenesis of N2-ADPR-dG induced by cabbage butterfly pierisin-1.

Genome-wide direct quantification of in vivo mutagenesis using high-accuracy paired-end and complementary consensus sequencing

Error-corrected next-generation sequencing (ecNGS) is an emerging technology for accurately measuring somatic mutations. Here, we report paired-end and complementary consensus sequencing (PECC-Seq), a high-accuracy ecNGS approach for genome-wide somatic mutation detection. We characterize a novel 2-aminoimidazolone lesion besides 7,8-dihydro-8-oxoguanine and the resulting end-repair artifacts originating from NGS library preparation that obscure the sequencing accuracy of NGS. We modify library preparation protocol for the enzymatic removal of end-repair artifacts and improve the accuracy of our previously developed duplex consensus sequencing method. Optimized PECC-Seq shows an error rate of <5 × 10-8 with consensus bases compressed from approximately 25 Gb of raw sequencing data, enabling the accurate detection of low-abundance somatic mutations. We apply PECC-Seq to the quantification of in vivo mutagenesis. Compared with the classic gpt gene mutation assay using gpt delta transgenic mice, PECC-Seq exhibits high sensitivity in quantitatively measuring dose-dependent mutagenesis induced by Aristolochic acid I (AAI). Moreover, PECC-Seq specifically characterizes the distinct genome-wide mutational signatures of AAI, Benzo[a]pyrene, N-Nitroso-N-ethylurea and N-nitrosodiethylamine and reveals the mutational signature of Quinoline in common mouse models. Overall, our findings demonstrate that high-accuracy PECC-Seq is a promising tool for genome-wide somatic mutagenesis quantification and for in vivo mutagenicity testing.

Quinoline is more genotoxic than 4-methylquinoline in hiHeps cells and rodent liver

Environmental pollutants, such as quinoline (QN) and 4-methylquinoline (4-MeQ), may be genotoxic and carcinogenic. Earlier studies, including in vitro genotoxicity tests, indicated that 4-MeQ is more mutagenic than QN. However, we hypothesized that the methyl group of 4-MeQ favors detoxication over bioactivation, and this factor may be overlooked in in vitro tests that do not incorporate supplementation with cofactors for enzymes that catalyze conjugation reactions. We used human induced hepatocyte cells (hiHeps), which express such enzymes, and compared the genotoxicity of 4-MeQ and QN. We also carried out an in vivo micronucleus (MN) test in rat liver, since 4-MeQ is not genotoxic in rodent bone marrow. In the Ames test and the Tk gene mutation assay, with rat S9 activation, 4-MeQ was more mutagenic than QN. However, QN induced significantly higher MN frequencies in hiHeps and rat liver than did 4-MeQ. Furthermore, QN upregulated genotoxicity marker genes much more than did 4-MeQ. We also investigated the roles of two important detoxication enzymes, UDP-glucuronosyltransferases (UGTs) and cytosolic sulfotransferases (SULTs). When hiHeps were preincubated with hesperetin (UGT inhibitor) and 2,6-dichloro-4-nitrophenol (SULT inhibitor), MN frequencies were elevated approximately 1.5-fold for 4-MeQ, whereas no significant effects were seen for QN. This study shows that QN is more genotoxic than 4-MeQ, when the roles of SULTs and UGTs in detoxication are considered and our results may improve understanding the structure-activity relationships of quinoline derivatives.

Toluene Dioxygenase-Catalyzed cis-Dihydroxylation of Quinolines: A Molecular Docking Study and Chemoenzymatic Synthesis of Quinoline Arene Oxides

Molecular docking studies of quinoline and 2-chloroquinoline substrates at the active site of toluene dioxygenase (TDO), were conducted using Autodock Vina, to identify novel edge-to-face interactions and to rationalize the observed stereoselective cis-dihydroxylation of carbocyclic rings and formation of isolable cis-dihydrodiol metabolites. These in silico docking results of quinoline and pyridine substrates, with TDO, also provided support for the postulated cis-dihydroxylation of electron-deficient pyridyl rings, to give transient cis-dihydrodiol intermediates and the derived hydroxyquinolines. 2-Chloroquinoline cis-dihydrodiol metabolites were used as precursors in the chemoenzymatic synthesis of enantiopure arene oxide and arene dioxide derivatives of quinoline, in the context of its possible mammalian metabolism and carcinogenicity.

Hepatocarcinogen 4-methylquinoline induced G:C to C:G transversions in the cII gene in the liver of lambda/lacZ transgenic mice (Muta™Mouse)

We have previously reported that quinoline increased the mutation frequency of the cII gene in the liver of lambda/lacZ transgenic mice (Muta™Mouse), and G:C to C:G transversions were the molecular signature of quinoline-induced mutations. 4-Methylquinoline (4-MeQ) has the highest mutagenicity among quinoline and isomeric methylquinolines according to the Ames test using Salmonella typhimurium TA 100, in the presence of rat liver microsomal enzymes. In this report, we examined the effect of 4-MeQ on mutagenesis in the lambda cII gene in the liver of the Muta™Mouse, and we analyzed the sequences of the mutated genes. The mutation frequency of the liver cII gene was seven times higher in 4-MeQ-treated mice than in control mice. Sequence analysis revealed that 4-MeQ primarily induced G:C to C:G transversions (37 of 45). The specificities of 4-MeQ for target organ and mutation pattern were very consistent with those of quinoline. Thus, we showed that 4-MeQ was also genotoxic in the liver of the Muta™Mouse, and as with quinoline, the G:C to C:G transversion was the molecular signature of the 4-MeQ-induced mutations.

Carcinogenicity of quinoline by drinking-water administration in rats and mice

The carcinogenicity of quinoline was examined by administrating quinoline in the drinking water to groups of 50 F344/DuCrj rats and 50 Crj: BDF1 mice of each sex. In rats, the doses of quinoline were 0, 200, 400, and 800 ppm for males and 0, 150, 300, and 600 ppm for females. In male rats, administration of quinoline was terminated at week 96 due to high mortality caused by tumors. There were significant increases of hepatocellular adenomas, hepatocellular carcinomas, hepatocellular adenomas and/or carcinomas (combined), and liver hemangiomas, hemangiosarcomas, hemangiomas and/or hemangiosarcomas (combined) in both male and female rats, and nasal esthesioneuroepitheliomas and sarcoma NOS (not otherwise specified) in males. In mice, doses of quinoline were 0, 150, 300 and 600 ppm for both males and females. Administration of quinoline was terminated at week 65 in males and at week 50 in females due to high mortality caused by tumors. There were marked increases of hemangiomas, hemangiosarcomas, and hemangiomas and/or hemangiosarcomas (combined) in the retroperitoneum, mesenterium, and liver in males, and in the retroperitoneum, mesenterium, peritoneum, and subcutis in females. Additionally, histiocytic sarcomas were statistically increased in the livers of female mice. Thus the present studies provided clear evidence of carcinogenic activity of quinoline administered in the drinking water in both rats and mice.

Genotoxicity risk assessment of diversely substituted quinolines using the SOS chromotest

Quinolines are aromatic nitrogen compounds with wide therapeutic potential to treat parasitic and microbial diseases. In this study, the genotoxicity of quinoline, 4-methylquinoline, 4-nitroquinoline-1-oxide (4-NQO), and diversely functionalized quinoline derivatives and the influence of the substituents (functional groups and/or atoms) on their genotoxicity were tested using the SOS chromotest. Quinoline derivatives that induce genotoxicity by the formation of an enamine epoxide structure did not induce the SOS response in Escherichia coli PQ37 cells, with the exception of 4-methylquinoline that was weakly genotoxic. The chemical nature of the substitution (C-5 to C-8: hydroxyl, nitro, methyl, isopropyl, chlorine, fluorine, and iodine atoms; C-2: phenyl and 3,4-methylenedioxyphenyl rings) of quinoline skeleton did not significantly modify compound genotoxicities; however, C-2 substitution with α-, β-, or γ-pyridinyl groups removed 4-methylquinoline genotoxicity. On the other hand, 4-NQO derivatives whose genotoxic mechanism involves reduction of the C-4 nitro group were strong inducers of the SOS response. Methyl and nitrophenyl substituents at C-2 of 4-NQO core affected the genotoxic potency of this molecule. The relevance of these results is discussed in relation to the potential use of the substituted quinolines. The work showed the sensitivity of SOS chromotest for studying structure-genotoxicity relationships and bioassay-guided quinoline synthesis.

Quantifying in vivo somatic mutations using transgenic mouse model systems

This chapter describes the use of the bacteriophage cII positive selection somatic mutational assay with the Muta™Mouse transgenic model system. The assay is similar to others involving a transgenic target, including the cII and lacI assays in the Big Blue(®) Mouse, lacZ in the MutaMouse, and the gpt delta assay. Briefly, high-molecular-weight DNA is purified from the tissue of interest and used as substrate during in vitro packaging reactions, where the λ transgenes are excised from the genome and assembled into viable phage. Phage containing the mutational targets is then adsorbed into an appropriate bacterial host, and mutations sustained in vivo are detected and quantified by either standard recombinant screening or selection assays. Mutant frequencies are reported as the ratio of mutant phage to total phage units analyzed. The λ-based transgenic mouse assays are used to study and characterize in vivo mutagenesis as well as for mutagenicity assessment of chemicals and other agents. These models permit the enumeration of mutations sustained in virtually any tissue of the mouse and are both sensitive and robust. Application of the assays is simple, not requiring resources beyond those commonly found in most academic laboratories.

iMARS--mutation analysis reporting software: an analysis of spontaneous cII mutation spectra

The sensitivity of any mutational assay is determined by the level at which spontaneous mutations occur in the corresponding untreated controls. Establishing the type and frequency at which mutations occur naturally within a test system is essential if one is to draw scientifically sound conclusions regarding chemically induced mutations. Currently, mutation-spectra analysis is laborious and time-consuming. Thus, we have developed iMARS, a comprehensive mutation-spectrum analysis package that utilises routinely used methodologies and visualisation tools. To demonstrate the use and capabilities of iMARS, we have analysed the distribution, types and sequence context of spontaneous base substitutions derived from the cII gene mutation assay in transgenic animals. Analysis of spontaneous mutation spectra revealed variation both within and between the transgenic rodent test systems Big Blue Mouse, MutaMouse and Big Blue Rat. The most common spontaneous base substitutions were G:C-->A:T transitions and G:C-->T:A transversions. All Big Blue Mouse spectra were significantly different from each other by distribution and nearly all by mutation type, whereas the converse was true for the other test systems. Twenty-eight mutation hotspots were observed across all spectra generally occurring in CG, GA/TC, GG and GC dinucleotides. A mutation hotspot at nucleotide 212 occurred at a higher frequency in MutaMouse and Big Blue Rat. In addition, CG dinucleotides were the most mutable in all spectra except two Big Blue Mouse spectra. Thus, spontaneous base-substitution spectra showed more variation in distribution, type and sequence context in Big Blue Mouse relative to spectra derived from MutaMouse and Big Blue Rat. The results of our analysis provide a baseline reference for mutation studies utilising the cII gene in transgenic rodent models. The potential differences in spontaneous base-substitution spectra should be considered when making comparisons between these test systems. The ease at which iMARS has allowed us to carry out an exhaustive investigation to assess mutation distribution, mutation type, strand bias, target sequences and motifs, as well as predict mutation hotspots provides us with a valuable tool in helping to distinguish true chemically induced hotspots from background mutations and gives a true reflection of mutation frequency.

Detailed review of transgenic rodent mutation assays

Induced chromosomal and gene mutations play a role in carcinogenesis and may be involved in the production of birth defects and other disease conditions. While it is widely accepted that in vivo mutation assays are more relevant to the human condition than are in vitro assays, our ability to evaluate mutagenesis in vivo in a broad range of tissues has historically been quite limited. The development of transgenic rodent (TGR) mutation models has given us the ability to detect, quantify, and sequence mutations in a range of somatic and germ cells. This document provides a comprehensive review of the TGR mutation assay literature and assesses the potential use of these assays in a regulatory context. The information is arranged as follows. (1) TGR mutagenicity models and their use for the analysis of gene and chromosomal mutation are fully described. (2) The principles underlying current OECD tests for the assessment of genotoxicity in vitro and in vivo, and also nontransgenic assays available for assessment of gene mutation, are described. (3) All available information pertaining to the conduct of TGR assays and important parameters of assay performance have been tabulated and analyzed. (4) The performance of TGR assays, both in isolation and as part of a battery of in vitro and in vivo short-term genotoxicity tests, in predicting carcinogenicity is described. (5) Recommendations are made regarding the experimental parameters for TGR assays, and the use of TGR assays in a regulatory context.

Nitrogen-substitution effect on in vivo mutagenicity of chrysene

We have previously reported the in vivo mutagenicity of aza-polycyclic aromatic hydrocarbons (azaPAHs), such as quinoline, benzo[f]quinoline, benzo[h]quinoline, 1,7-phenanthroline and 10-azabenzo[a]pyrene. The 1,10-diazachrysene (1,10-DAC) and 4,10-DAC, nitrogen-substituted analogs of chrysene, were shown to exhibit mutagenicity in Salmonella typhimurium TA100 in the presence of rat liver S9 and human liver microsomes in our previous report, although DACs could not be converted to a bay-region diol epoxide, the ultimate active form of chrysene, because of their nitrogen atoms. In the present study, we tested in vivo mutagenicity of DACs compared with chrysene using the lacZ transgenic mouse (Mutatrade markMouse) to evaluate the effect of the nitrogen substitution. DACs- and chrysene-induced mutation in all of the six organs examined (liver, spleen, lung, kidney, bone marrow and colon). The mutant frequencies obtained with chrysene showed only small differences between the organs examined and ranged from 1.5 to 3 times the spontaneous frequency. The 4,10-DAC was more mutagenic than chrysene in all the organs tested. The highest lacZ mutation frequency was observed in the lung of 4,10-DAC-treated mice and it was 19 and 6 times the spontaneous frequency and the frequency induced by chrysene, respectively. The 1,10-DAC induced lacZ mutation in the lung with a frequency 4.3- and 1.5-fold higher than in the control and chrysene-treated mice, respectively, although the mutant frequencies in the other organs of 1,10-DAC-treated mice were almost equivalent to those of chrysene-treated mice. Not only chrysene but also DACs depressed the G:C to A:T transition and increased the G:C to T:A transversion in the liver and lung. These results suggest that the two types of nitrogen substitutions in the chrysene structure may enhance mutagenicity in the mouse lung, although they showed no difference in the target-organ specificity and the mutation spectrum.

In vivo mutagenicity of benzo[f]quinoline, benzo[h]quinoline, and 1,7-phenanthroline using the lacZ transgenic mice

Phenanthrene, a simplest angular polycyclic aromatic hydrocarbon with a bay-region in its molecule, is reported to be non-mutagenic, although most angular (non-linear) polycyclic aromatic hydrocarbons, such as benzo[a]pyrene and chrysene, are known to show genotoxicity after metabolic transformation into a bay-region diol epoxide. On the other hand, benzo[f]quinoline (BfQ), benzo[h]quinoline (BhQ), and 1,7-phenanthroline (1,7-Phe), which are all aza-analogs of phenanthrene, are mutagenic in the Ames test using Salmonella typhimurium TA100 in the presence of a rat liver S9 fraction. In this report, we undertook to investigate the in vivo mutagenicity of BfQ, BhQ and 1,7-Phe by an in vivo mutation assay system using the lacZ transgenic mouse (Muta Mouse). BfQ and BhQ only slightly induced mutation in the liver and lung, respectively. BfQ- and BhQ-induced cII mutant spectra showed no characteristics compared with that of the control. These results suggest that the in vivo mutagenicities of BfQ and BhQ were equivocal. On the other hand, 1,7-Phe induced a potent mutation in the liver and a weak mutation in the lung. Furthermore 1,7-Phe depressed the G:C to A:T transition and increased the G:C to C:G transversion in the liver like quinoline, a hepatomutagen possessing the partial structure of 1,7-Phe, compared with the spontaneous mutation spectrum. These results suggest that the in vivo mutagenicity of 1,7-Phe might be caused by the same mechanism as that of quinoline, which induced the same mutational spectrum change (G:C to C:G transversion).

DNA adducts and mutagenic specificity of the ubiquitous environmental pollutant 3-nitrobenzanthrone in Muta Mouse

3-nitrobenzanthrone (3-NBA) is an extremely potent mutagen in the Salmonella reversion assay and a suspected human carcinogen identified in diesel exhaust and in ambient airborne particulate matter. To evaluate the in vivo mutagenicity of 3-NBA, we analyzed the mutant frequency (MF) in the cII gene of various organs (lung, liver, kidney, bladder, colon, spleen, and testis) in lambda/lacZ transgenic mice (Muta Mouse) after intraperitoneal treatment with 3-NBA (25 mg/kg body weight injected once a week for 4 weeks). Increases in MF were found in colon, liver, and bladder, with 7.0-, 4.8-, and 4.1-fold increases above the control value, respectively, whereas no increase in MF was found in lung, kidney, spleen, and testis. Simultaneously, induction of micronuclei in peripheral blood reticulocytes was observed. The sequence alterations in the cII gene recovered from 41 liver mutants from 3-NBA-treated mice were compared with 32 spontaneous mutants from untreated mice. Base substitution mutations predominated for both the 3-NBA-treated (80%) and the untreated (81%) groups. However, the proportion of G:C-->T:A transversions in the mutants from 3-NBA-treated mice was higher (49% vs. 6%) and the proportion of G:C-->A:T transitions was lower than those from untreated mice (10% vs. 66%). The increase in MF in the liver was associated with strong DNA binding by 3-NBA, whereas in lung, in which there was no increase in MF, a low level of DNA binding was observed (268.0-282.7 vs. 8.8-15.9 adducts per 10(8) nucleotides). DNA adduct patterns with multiple adduct spots, qualitatively similar to those formed in vitro after activation of 3-NBA with nitroreductases and in vivo in rats, were observed in all tissues examined. Using high-pressure liquid cochromatographic analysis, we confirmed that all major 3-NBA-DNA adducts produced in vivo in mice are derived from reductive metabolites bound to purine bases (70-80% with deoxyguanosine and 20-30% with deoxyadenosine in liver). These results suggest that G:C-->T:A transversions induced by 3-NBA are caused by misreplication of adducted guanine residues through incorporation of adenine opposite the adduct (A-rule).

Regional mutagenicity of heterocyclic amines in the intestine: mutation analysis of the cII gene in lambda/lacZ transgenic mice

Transgenic mouse assays have revealed that the mouse intestine, despite its resistance to carcinogenesis, is sensitive to the mutagenicity of some heterocyclic amines (HCAs). Little is known, however, about the level and localization of that sensitivity. We assessed the mutagenicity of four orally administered (20 mg/kg per day for 5 days) HCAs-2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) hydrochloride, 2-amino-3-methylimidazo[4,5-f]quinoline (IQ), 2-amino-3,4-dimethylimidazo[4,5-f]quinoline (MeIQ), and 3-amino-1-methyl-5H-pyrido[4,3-b]indole (Trp-P-2) acetate-in the intestine of male MutaMice. Two weeks after the last administration, we isolated epithelium from the small intestine, cecum, and colon and analyzed lacZ and cII transgene mutations. PhIP increased the lacZ mutant frequency (MF) in all the samples, and in the small intestine, cII and lacZ MFs were comparable. In the cII gene, G:C to T:A and G:C to C:G transversions were characteristic PhIP-induced mutations (which has also been reported for the rat colon, where PhIP is carcinogenic). In the small intestine, PhIP increased the cII MF to four-fold that of the control, but IQ, MeIQ, and Trp-P-2 did not have a significant mutagenic effect. In the cecum, cII MFs induced by IQ and MeIQ were 1.9 and 2.7 times those in the control, respectively. The MF induced by MeIQ in the colon was 3.1 times the control value. Mutagenic potency was in the order PhIP>MeIQ>IQ; Trp-P-2 did not significantly increase the MF in any tissue. The cecum was the most susceptible organ to HCA mutagenicity.

Analysis of HPRT and supF mutations caused by pierisin-1, a guanine specific ADP-ribosylating toxin derived from the cabbage butterfly

Pierisin-1, an ADP-ribosylating toxin derived from the cabbage butterfly, Pieris rapae, induces apoptosis in various mammalian cell lines. We recently reported that the target for ADP ribosylation by pierisin-1 is the 2'-deoxyguanosine residue in DNA. To examine whether pierisin-1 would induce mutations in mammalian cell genes, we conducted a mutational analysis for the hypoxanthine-guanine phosphoribosyltransferase (HPRT) locus in pierisin-1-treated Chinese hamster lung (CHL) cells. N(2)-(ADP-ribos-1-yl)-2'-deoxyguanosine was detected by the (32)P-postlabeling method in CHL cells after treatment with pierisin-1 at doses of 2-32 ng/mL; adduct levels were 1.1-12.0 per 10(6) nucleotides. Pierisin-1 induced mutations in the HPRT gene dose-dependently, and the frequency was 38 times higher than the control, at a dose of 32 ng/mL. To confirm that mono(ADP-ribosyl)ated dG itself leads to mutations, the pierisin-1-treated DNA of plasmid pMY189 bearing the supF gene was used for mutational analysis. The mutation frequency of the supF gene treated with 2-8 micro g/mL of pierisin-1 was 17-40-fold the control value. Mutation spectrum analysis showed that single base substitutions dominated in both HPRT and supF genes. Among these, transversions were predominant, and more than 70% of the base substitutions occurred at G:C base pairs in both genes. The most frequent mutations were G:C to C:G, followed by G:C to T:A in HPRT gene, whereas G:C to T:A transversions dominated in the supF gene. Our results indicate that pierisin-1 produced N(2)-(ADP-ribos-1-yl)-2'-deoxyguanosine and this guanine-adduct could lead to mutations in the HPRT and supF genes. These findings could provide very useful information for understanding the biological significance of pierisin-1.

Effect of 10-aza-substitution on benzo[a]pyrene mutagenicity in vivo and in vitro

Benzo[a]pyrene (BaP), an environmental carcinogen, shows genotoxicity after metabolic transformation into the bay-region diol epoxide, BaP-7,8-diol 9,10-epoxide. 10-Azabenzo[a]pyrene (10-azaBaP), in which a ring nitrogen is located in the bay-region, is also a carcinogen and shows mutagenicity in the Ames test in the presence of the rat liver microsomal enzymes. In order to evaluate the effect of aza-substitution on in vivo genotoxicity, BaP and 10-azaBaP were assayed for their in vivo mutagenicity using the lacZ-transgenic mouse (MutaMouse). BaP was potently mutagenic in all of the organs examined (liver, lung, kidney, spleen, forestomach, stomach, colon, and bone marrow), as described in our previous report, whereas, 10-azaBaP was slightly mutagenic only in the liver and colon. The in vitro mutagenicities of BaP and 10-azaBaP were evaluated by the Ames test using liver homogenates prepared from several sources, i.e. CYP1A-inducer-treated rats, CYP1A-inducer-treated and non-treated mice, and humans. BaP showed greater mutagenicities than 10-azaBaP in the presence of a liver homogenate prepared from CYP1A-inducer-treated rodents. However, 10-azaBaP showed mutagenicities similar to or more potent than BaP in the presence of a liver homogenate or S9 from non-treated mice and humans. These results indicate that 10-aza-substitution markedly modifies the nature of mutagenicity of benzo[a]pyrene in both in vivo and in vitro mutagenesis assays.