Site-specific mutagenesis by O6-alkylguanines located in the chromosomes of mammalian cells: influence of the mammalian O6-alkylguanine-DNA alkyltransferase

Different mutation frequencies and spectra among organs by N-methyl-N-nitrosourea in rpsL (strA) transgenic mice

The frequencies and spectra of N-methyl-N-nitrosourea (MNU)-induced in vivo somatic mutations were determined in rpsL (strA) transgenic mice. The wild-type rpsL gene, which exhibits a streptomycin-sensitive (Sm(S)) phenotype, was used as the rescue marker gene. Studies of mutation spectra among different organs and tissues were simplified using this system because of the short coding sequence (375 bp) of the rpsL gene. MNU administration to transgenic mice significantly elevated the mutation frequencies in various adult organs. Two distinctive patterns of mutation spectrum were observed, depending on the organs tested. Mutations derived from labile organs (spleen and thymus) were predominantly G:C to A:T transitions, as expected for MNU mutagenesis. Stable organs like the liver and brain, however, carried many fewer G:C to A:T transitions but significantly more single base deletions, of which the spectrum was very similar to that of background mutations in the rpsL transgenic mice. This spectrum difference among more and less proliferating organs was confirmed by the predominant occurrence of G:C to A:T transitions in fetal liver cells exposed to transplacental MNU treatment. In addition, most (approximately 90%) of the G:C to A:T transitions induced by MNU were detected in the first nucleotide of some 5'-G-(C or G)-3' sequences, many of which corresponded to the middle guanine residue of 5'-purine-G-(C or G)-3' sequences. It is thus suggested that at particular sites, the neighboring bases in both the 5' side and 3' side seem to influence either the susceptibility to DNA damage or the ability to repair MNU-induced lesions.

Heterozygous DNA mismatch repair gene PMS2-knockout mice are susceptible to intestinal tumor induction with N-methyl-N-nitrosourea

PMS2-deficient (PMS2(-/-)) mice are hypersensitive to N-methyl-N-nitrosourea (MNU)-induced thymic lymphomas based on the failure to initiate mismatch repair (MMR) at O(6)-methylguanine:T mismatches formed after MNU exposure. However, heterozygous PMS2 knockout (PMS2(+/-)) mice do not develop spontaneous tumors, suggesting that they have sufficient MMR function to prevent genomic instability. We hypothesized that in PMS2(+/-) mice, exogenous carcinogens may either mutationally knockout the remaining normal allele leading cells to develop tumors or introduce sufficient DNA adducts and mismatches to overload the lower capacity for MMR, leading in either case to an increased rate of tumor production. In the present study, PMS2(+/-) mice and their littermate PMS2(+/+) mice were monitored for tumor incidence following MNU treatment. Mice were given 50 mg MNU/kg i.p. when 5 weeks old. They demonstrated a similar incidence of thymic lymphomas, suggesting that expression of the single normal PMS2 allele is sufficient to protect the thymus and implying that a single dose of MNU may not efficiently knock out the remaining PMS2 allele in the thymus. Surprisingly, PMS2(+/-) mice were significantly more likely to develop intestinal tumors-both adenomas and adenocarcinomas-after MNU than were PMS2(+/+) mice (2.34 +/- 0.34 tumors per mouse versus 1.34 +/- 0.25 tumors per mouse; P < 0.05). The intestinal tumors were located mainly in the small intestine. However, these tumors in both the PMS2(+/-) mice and PMS2(+/+) mice did not show microsatellite instability characteristic of loss of MMR. These results suggest that a single normal PMS2 allele can protect thymus but not intestine from MNU carcinogenesis. Organ-specific factors might influence MMR- mediated resistance to methylating agents. Heterozygous PMS2 knockout mice may be used as a promising animal model for intestinal tumorigenesis studies involving environmental carcinogens.

Influence of nucleotide excision repair and of dose on the types of vermilion mutations induced by diethyl sulfate in postmeiotic male germ cells of Drosophila

The role of a defect for nucleotide excision repair (NER) in oocytes on the repair of DNA ethyl adducts induced by diethyl sulfate (DES) in male germ cells of Drosophila was analysed. Frequencies of mutations at multiple loci (recessive lethal mutations) and at the vermilion gene induced in NER+ conditions (cross NER+ x NER+) were compared with those fixed in a NER- background (NER- x NER+). The M(NER-)/M(NER+) mutability ratios for two DES concentrations, 10 mM and 15 mM, were 2.21 and 1.49, respectively, indicating that NER repairs part of the DES-induced damage. The majority of 28 fertile vermilion mutations produced by DES in NER- are transitions, both GC-AT (46.4%) and AT-GC (21.4%) transitions are found, the consequences of O6-ethylguanine and O4-ethylthymine, respectively. Transversions (21.5%), one +1 frameshift mutation (3.6%) and two deletions (7.1%) are most likely the result of N-alkylation damage. Furthermore, the DES-induced mutation spectra show interesting differences in relation to the exposure dose. All 10 mutants isolated in this and a previous [L.M. Sierra, A. Pastink, M.J.M. Nivard, E.W. Vogel, DNA base sequence changes induced by DES in postmeiotic male germ cells of Drosophila melanogaster, Mol. Gen. Genet. 237 (1993) 370-374] study from experiments with low DES-effectiveness are exclusively transitions, independent whether the females were of the NER+ or NER-genotype. This indicates that at lower DES effectiveness only O-alkylation damage is relevant, and that N-alkylation damage is repaired. In experiments revealing high DES-effectiveness, vermilion mutations representing N-alkylation damage reached 43% (9/21) with NER- and 26% (7/27) with NER+ females, suggesting (i) that NER becomes involved at high adduct levels because then the base excision repair (BER) may be saturated, and (ii) that this involvement of NER causes the relative decrease from 43% to 26% N-alkylation mediated sequence changes.

Inhibition of O(6)-methylguanine-DNA methyltransferase increases azoxymethane-induced colonic tumors in rats

Azoxymethane (AOM) causes O(6)-methylguanine adduct formation which leads to G-->A transitions. Their repair is carried out by O(6)-methylguanine-DNA methyltransferase (MGMT). To evaluate the importance of this repair event in AOM-induced carcinogenesis, we examined the effect of O(6)-benzylguanine (BG), a potent inhibitor of MGMT, on colonic tumor development. Rats were treated weekly for 2 weeks at 0 and 24 h with BG (60 mg/kg body wt i.p.) or vehicle (40% polyethylene glycol, PEG-400), followed 2 h after the first dose of BG with AOM (15 mg/kg body wt) or vehicle (saline) i.p. Rats were killed 35 weeks later and tumors harvested and DNA extracted. In the AOM-treated groups, BG caused a significant increase in tumor incidence with tumors in 65.9%, versus 30.8% in the AOM/PEG-treated group (P < 0.05). In the BG/AOM group there was also a significant increase in tumor multiplicity, with 2.3 tumors/tumor-bearing rat, versus 1.6 tumors/tumor- bearing rat in the AOM/PEG group (P < 0.05). Since O(6)-methylguanine adducts can cause activating mutations in the K-ras and beta-catenin genes, we examined the effects of BG on these mutations. In the BG group there were seven mutations in codon 12 or 13 of exon 1 of the K-ras gene in 51 tumors examined, compared with no K-ras mutations in 17 tumors analyzed in the AOM/PEG group (P = 0.12). In the BG/AOM group there were 10 mutations in exon 3 of the beta-catenin gene among 48 tumors evaluated, compared with six mutations in 16 tumors analyzed in the PEG/AOM group (P = 0.16). In summary, MGMT inhibition increases AOM-induced colonic tumor incidence and multiplicity in rats.

Mice defective in the DNA mismatch gene PMS2 are hypersensitive to MNU induced thymic lymphoma and are partially protected by transgenic expression of human MGMT

DNA mismatch repair (MMR) stabilizes the cellular genome. Mice defective in the MMR gene PMS2 are susceptible to spontaneous thymic lymphoma and sarcomas. To determine the sensitivity of PMS2 knockout mice to environmental carcinogens and the protective effect of O6-methylguanine DNA methyltransferase (MGMT), heterozygous PMS2 knockout mice and human MGMT (hMGMT) transgenic mice were mated and the PMS2-/- and PMS2+/+ with or without hMGMT offspring were treated at 5 weeks of age with 50 mg/kg N-methyl-N-nitrosourea (MNU). MNU produces carcinogenic O6-methylguanine (O6-meG) adducts, resulting in thymic lymphoma in mice, which can be prevented in normal mice by overexpression of hMGMT. A significantly higher incidence of thymic lymphomas was observed in MNU-treated PMS2-/- mice, compared to wildtype PMS2+/+ mice (100 vs 52%; P < 0.001). The mean latency of lymphomas was also significantly shortened in PMS2-/- mice (81 vs 102 days, P < 0.01). Transgenic expression of hMGMT significantly but incompletely blocked MNU lymphomagenesis in PMS2-/- mice. The incidence of lymphomas in PMS2-/-/hMGMT+ mice was reduced to 80% (P < 0.01) and mean latency increased to 91 days (P < 0.05). Thymic lymphomagenesis was efficiently blocked in PMS2+/+/hMGMT+ mice with rapid repair of O6-meG. Since O6-meG:T mismatches in MMR+ cells may trigger mismatch repair resulting in abortive repair and cell death whereas in the absence of MMR, these mismatches are converted to A:T, we predicted that G to A point mutations in codon 12 of the K-ras gene would occur. In this study, we found G to A point mutations in codon 12 of the K-ras gene in many tumors. Thus, in MMR deficient tissues, methylating agents induce point mutations in cells with a higher rate of cell survival which together are potently carcinogenic in the thymus. These data suggest that PMS2 defective lymphomas may arise by the concerted action of environmental and perhaps endogenous methylation of DNA coupled to genomic instability.

Ethylnitrosourea-induced mutation and molecular analysis of transgenic mice containing the gpt shuttle vector

Novel transgenic mice were developed in order to study the in vivo mutagenesis. The transgenic mice carried pCGK shuttle vector, which contained the Escherichia coli gpt gene as a mutational target, the kanamycin-resistant gene (Kanr) and cos region derived from bacteriophage lambda. The shuttle vector can be recovered from the transgenic mouse genome into the gpt-deficient E. coli by an in vitro packaging method and is selectable as a Kanr phenotype. Mutations induced at the gpt gene can be easily detected with a selective agent, 6-thioguanine (6-TG). In the previous study, the pCGK shuttle vector was incorporated into Chinese hamster CHL/IU cells and the resultant transgenic cell line was shown to be a useful system to study in vitro mutagenesis at the gpt gene. Therefore, an advantage of the shuttle vector is that in vivo mutational data obtained from the transgenic mouse can be compared with those of transgenic cell line in vitro. A transgenic CD-1 mouse line, designated as #128, that carried approximately 50 copies of pCGK shuttle vectors, was selected among 4 transgenic mouse lines. To investigate the sensitivity of the #128 line, the transgenic mice were treated with a single intraperitoneal injection of 250 mg/kg of N-ethyl-N-nitrosourea (ENU) or with 50 mg kg-1 day-1 of ENU for 5 consecutive days, and bone marrow, spleen and liver were dissected to investigate their mutational responses. The background mutant frequency was between 18x10(-6) and 75x10(-6) among all tissues tested. ENU induced significant increases in the mutant frequency above the background level in all three tissues at 14 days after single or 5-day treatment with the chemical. The increases in the mutant frequencies in bone marrow, spleen and liver were 6.4- to 6.8-fold, 3.0- to 5.6-fold and 3.0- to 3.3-fold, respectively. The shuttle vector DNA was recovered from the bone marrow of both spontaneous and ENU-treated mice and the gpt gene was amplified by polymerase chain reaction. The amplified DNA was subject to DNA sequence analysis. Out of 79 spontaneous and 52 ENU-induced mutants, the gpt gene could be amplified from 28 spontaneous and 46 ENU-induced mutants. DNA sequence analysis showed that predominant mutations were identified as A:T to T:A transversions (22 out of 46 sequenced mutants) and G:C to A:T transitions (9/46) in ENU-induced mutants, whereas G:C to T:A transversions (7 out of 28 sequenced mutants) were predominant in spontaneous mutants. These results demonstrate that this transgenic mouse, in combination with the transgenic CHL/IU cell line, is a useful system to study in vivo and in vitro mutational events at the same target gene.

Modulation of the toxic and mutagenic effects induced by methyl methanesulfonate in Chinese hamster ovary cells by overexpression of the rat N-alkylpurine-DNA glycosylase

Exposure of mammalian cells to alkylating agents causes transfer of alkyl groups to N- as well as O-atoms of DNA bases. Especially the O-alkylated G and T bases have strong mutagenic properties, since they are capable of mispairing during replication. The mutagenic potential of N-alkylbases is less clear although specific base excision repair (BER) pathways exist which remove those lesions from the DNA. We investigated the relative contribution of N-alkylations to mutation induction at the Hprt gene in cultured Chinese hamster ovary cells (CHO). To this end BER activity in CHO cells was modulated by introduction of an expression vector carrying the rat N-alkylpurine-DNA glycosylase (APDG) gene, which codes for a glycosylase that is able to remove 3-methyladenine and 7-methylguanine from DNA thereby generating apurinic sites. Upon selection of a CHO clone which 10 times overproduced APDG compared to control CHO cells, mutation induction, the mutational spectrum, and cell survival were determined in both cell lines following treatment with methyl methanesulfonate (MMS). The results show that over-expression of APDG renders CHO cells more sensitive for mutation induction as well as cytotoxicity induced by MMS. The involvement of apurinic sites in induction of base pair changes at positions where 3-methyladenine was induced is inferred from the observation that the mutational spectrum of MMS-induced mutations in APDG-CHO cells showed twice as much base pair changes at AT base pairs (33.3%) compared to the spectrum of MMS-induced mutations in CHO-control cells (15.8%).

Tk+/- mouse model for detecting in vivo mutation in an endogenous, autosomal gene

Tk+/- transgenic mice were created using an embryonic stem cell line in which one allele of the endogenous thymidine kinase (Tk) gene was inactivated by targeted homologous recombination. Breeding Tk+/- parents produced viable Tk-/- knockout (KO) mice. Splenic lymphocytes from KO mice were used in reconstruction experiments for determining the conditions necessary for recovering Tk somatic cell mutants from Tk+/- mice. The cloning efficiency of KO lymphocytes was not affected by the toxic thymidine analogues 5-bromo-2'-deoxyuridine (BrdUrd) or trifluorothymidine (TFT), or by BrdUrd in the presence of lymphocytes from Tk+/- animals; however, it was easier to identify clones resistant to BrdUrd than to TFT when Tk+/- cells were present. Tk+/- mice were treated with vehicle or 100 mg/kg of N-ethyl-N-nitrosourea (ENU), and after 4 months, the frequency of Tk mutant lymphocytes was measured by resistance to BrdUrd. The frequency of Tk mutants was 22+/-5.9x10-6 in control animals and 80+/-31x10-6 in treated mice. In comparison, the frequency of Hprt mutant lymphocytes, as measured by resistance to 6-thioguanine, was 2.0+/-1.2x10-6 in control animals and 84+/-28x10-6 in the ENU-treated mice. Analysis of BrdUrd-resistant lymphocyte clones derived from the ENU-treated animals revealed point mutations in the non-targeted Tk allele. These results indicate that the selection of BrdUrd-resistant lymphocytes from Tk+/- mice may be used for assessing in vivo mutation in an endogenous, autosomal gene.

A comparative study of in vivo mutation assays: analysis of hprt, lacI, cII/cI and as mutational targets for N-nitroso-N-methylurea and benzo[a]pyrene in Big Blue mice

We have compared the response of the native hprt gene and the lacI, cII, and cI transgenes in Big Blue B6C3F1 mice following treatment with either N-nitroso-N-methylurea (MNU) or benzo[a]pyrene (BaP). Three weeks after mutagen treatment splenic T cells were isolated from the animals, and samples were either cultured to measure mutation at the native hprt locus or used to extract genomic DNA for transgene mutation analysis. Phage rescued from extracted DNA were plated in the presence of 5-bromo-4-chloro-3-indolyl-beta-d-galactopyranoside (X-gal) to score lacI mutations, or plated on a hflAB lawn to score cII and cI mutants. With MNU hprt mutant frequency increased in a dose-related, sublinear manner up to 78-fold above background at the highest dose tested (20 mg/kg). In comparison, the lacI transgene yielded only a 3.1-fold increase at this dose, and the cII and cI transgenes did not show any increase. With 150 mg/kg BaP a 5.8- and 8.7-fold increase in mutant frequency was observed at hprt and lacI, respectively, while only a 1.3-fold increase was observed at cII. DNA sequencing revealed an increase in GC-->TA transversions among the cII mutants, suggesting that the increase was related to BaP exposure. No significant increase in cI mutant frequency was observed. Therefore, the order of mutation assay sensitivity was hprt>lacI>cII/cI with MNU, and hprt approximately lacI> cII/cI with BaP. While the hflAB selection system offers significant advantages with respect to cost and effort when compared to the lacI assay, additional evaluation of its sensitivity is warranted.

Different ras gene mutational frequencies in thyroid papillary carcinomas in Japan and Thailand

The incidence and pattern of ras oncogene mutations in human malignancies demonstrate geographic and racial differences. For example, specificity of alterations is found in cholangiocellular carcinomas in Thai patients with a different etiology from those in Japanese patients. In the present study, a comparison of ras gene mutations in thyroid papillary carcinomas from Japanese and Thai patients was performed using single-strand conformation polymorphism and direct sequencing analyses. The incidence of ras mutation differed markedly in Japanese (two of 24 carcinomas, 8.3%) and Thai (five of 10 carcinomas, 50%) patients. In addition, all but one ras mutation occurred at codon 12 of the K-ras gene in the Thai cases. These results suggest that thyroid cancers in Thailand may be due to specific genetic and/or environmental factors.

Mutation spectra resulting from carcinogenic exposure: from model systems to cancer-related genes

The events leading to cancer are complex and interactive. Alteration of cancer genes, such as the tumor suppressor gene p53, plays a central role in this process. Analysis of the frequency, type and site of mutations in important cancer-related genes may provide clues to the identification of etiological factors and sources of exposure. In this chapter we have selected a few examples of environmental human carcinogens and have attempted to use the knowledge of their mechanisms of mutagenesis, as derived from in vitro cell systems, as a key to understanding the complexity of p53 mutation spectra in tumors arising at the putative target organ. The analysis will focus on environmental exposure to UV radiation. The examples of tobacco smoke, dietary aflatoxin and vinyl chloride will be also briefly discussed.

A comparison of the genotoxicity of ethylnitrosourea and ethyl methanesulfonate in lacZ transgenic mice (Muta Mouse)

We compared the induction of gene mutations and chromosomal aberrations by ethylating agents in lacZ transgenic mice (Muta Mouse). Chromosomal aberrations were detected by the peripheral blood micronucleus assay. Gene mutations were detected in the lacZ transgene. A small amount of blood was sampled from a tail vessel during the expression time for fixation of gene mutations in vivo; this enabled us to detect and compare clastogenicity and gene mutations in the identical mouse. Single intraperitoneal injections of ENU (50-200 mg/kg) and EMS (100-400 mg/kg) strongly induced micronucleated reticulocytes (MN) detectable in peripheral blood 48 h after treatment. The maximum MN frequencies induced were 6.6% and 3.3% for ENU (100 mg/kg) and EMS (400 mg/kg), respectively (the control value was 0.3%). lacZ mutant frequency (MF) was analyzed in bone marrow and liver 7 days after treatment. Spontaneous MFs were 2.0-4.6 x 10(-6). MF in bone marrow was increased by ENU to 3.4 x 10(-5) at 200 mg/kg and induced by EMS to 1.8 x 10(-5) at 400 mg/kg. In liver, however, both chemicals at their highest doses induced only slight increases in MF. The induction of both micronuclei and lacZ mutations in bone marrow by both ENU and EMS correlated better with O6-ethylguanine adducts than with N7-ethylguanine adducts. The mutants (19 for ENU and 12 for EMS) were subjected to DNA sequence analysis. Among EMS-induced mutations, 75% were GC to AT transitions, which were probably caused by O6-ethylguanine. Among ENU-induced mutations, in contrast, 40% occurred as AT base pair substitutions (6 AT to TA transversions and 2 AT to GC transitions) (no such mutations were induced by EMS). These results, together with the known reactivity of ENU to thymine suggest that thymine adducts play a significant role in the ENU mutagenesis.

Structure of the hydrogen bonding complex of O6-methylguanine with cytosine and thymine during DNA replication

During DNA replication, mutations occur when an incorrect dNTP is incorporated opposite a carcinogen-modified nucleotide. We have probed the structures of the interaction between O 6-methylguanine ( O 6mG) and cytosine and thymine during replication by kinetic means in order to examine the structure during the rate determining step. The kinetics of incorporation of dCTP and dTTP opposite O 6mG and three analogs, S 6-methyl-6-thioguanine, O 6-methyl-1-deazaguanine and O 6-methylhypoxanthine, have been measured with four polymerases, the Klenow fragment of DNA polymerase I, the Klenow fragment with the proof-reading exonuclease inactivated, Taq and Tth polymerases. In the insertion of dTTP opposite O 6mG, a large decrease in V max/ K m was observed only upon modification of the N1 position. This result is consistent with a Watson-Crick type configuration. For the incorporation of dCTP, the V max/ K m was significantly decreased only with removal of the exocyclic amino group at the 2 position. The pH dependence of the ratio of incorporation of dCTP and dTTP was independent of pH at physiological pH. This result suggests that dCTP is incorporated via an uncharged complex such as the wobble configuration.

Spectrum of Bmp5 mutations from germline mutagenesis experiments in mice

Over 40 years of mutagenesis experiments using the mouse specific-locus test have produced a large number of induced germline mutations at seven loci, among them the short ear locus. We have previously shown that the short ear locus encodes bone morphogenetic protein 5 (BMP5), a member of a large family of secreted signaling molecules that play key roles in axis formation, tissue differentiation, mesenchymalepithelial interactions, and skeletal development. Here we examine 24 chemical- and radiation-induced mutations at the short ear locus. Sequence changes in the Bmp5 open reading frame confirm the importance of cysteine residues in the function of TGF beta superfamily members. The spectrum of N-ethyl-N-nitrosourea-induced mutations also provides new information about the basepair, sequence context, and strand specificity of germline mutations in mammals.

Replication of O6-methylguanine-containing DNA by repair and replicative DNA polymerases

The biological consequences of O6-methylguanine (m6G) in DNA are well recognized. When template m6G is encountered by DNA polymerases, replication is hindered and trans-lesion replication results in the preferential incorporation of dTMP opposite template m6G. Thus, unrepaired m6G in DNA is both cytotoxic and mutagenic. Yet, cell lines tolerant to m6G in DNA have been isolated, which indicates that some cellular DNA polymerases may replicate m6G-containing DNA with reasonable efficiency. Previous reports suggested that mammalian pol beta could not replicate m6G-containing DNA, but we find that pol beta can catalyze trans-lesion replication; however, the lesion must reside in the optimal context for pol beta activity, single- or short nucleotide gapped substrates. Primed single-stranded DNA templates, with or without template m6G, were poor substrates for pol beta as reported in earlier studies. In contrast, trans-lesion replication by bacteriophage T4 DNA polymerase was observed for primed single-stranded DNA templates. Replication of m6G-containing DNA by T4 DNA polymerase required the gp45 accessory protein that clamps the polymerase to the DNA template. The rate-limiting step in replicating m6G-containing DNAs by both DNA polymerases tested was incorporation of dTMP across from the lesion.

Lack of the DNA repair protein O6-methylguanine-DNA methyltransferase in histologically normal brain adjacent to primary human brain tumors

Exposure to exogenous alkylating agents, particularly N-nitroso compounds, has been associated with increased incidence of primary human brain tumors, while intrinsic risk factors are currently unknown. The DNA repair protein O6-methylguanine-DNA methyltransferase (MGMT) is a major defense against the carcinogenicity of N-nitroso compounds and other alkylators. We report here that in 55% (64/117) of cases, histologically normal brain tissue adjacent to primary human brain tumors lacked detectable MGMT activity [methyl excision repair-defective (Mer-) status]. The incidence of Mer- status in normal brain tissue from brain tumor patients was age-dependent, increasing from 21% in children 0.25-19 years of age to 75% in adults over 50. In contrast, Mer- status was found in 12% (5/43) of normal brain specimens from patients operated for conditions other than primary brain tumors and was not age-dependent. The 4.6-fold elevation in incidence of Mer- status in brain tumor patients is highly significant (chi2 = 24; p < or = 0.001). MGMT activity was independent of age in the lymphocytes of brain tumor patients and was present in lymphocytes from six of nine tumor patients whose normal brain specimen was Mer-. DNA polymerase beta, apurinic/apyrimidinic endonuclease, and lactate dehydrogenase activities were present in all specimens tested, including Mer- specimens from brain tumor patients. Our data are consistent with a model of carcinogenesis in human brain in which epigenetically regulated lack of MGMT is a predisposing factor and alkylation-related mutagenesis is a driving force.

Database and software for the analysis of mutations at the lacI gene in both transgenic rodents and bacteria

The use of transgenic rodents for the study of genetic toxicology has increased dramatically in the past several years. A great deal of the recent work has employed the lacI locus in transgenic mice. In addition to the transgenic data, a substantial amount of information exists regarding mutation of the lacI gene in bacteria. In an effort to centralize the information regarding mutations in the lacI gene in both rodents and bacteria, we have created a computerized database that contains information about DNA sequence alterations on about 500 mutations in transgenic rodents and 8,000 mutations in bacteria. We have also produced a software package for the analysis of the lacI database. Routines have been developed for the analysis of single base substitutions, including programs to (i) determine if two mutational spectra are different; (ii) determine if mutations show a DNA strand bias; (iii) determine the frequency of transitions and transversions; (iv) display the number and kind of mutations observed at each base in the coding region; (v) perform nearest neighbor analysis; and (vi) display mutable amino acids in the lacI protein. The software runs only on IBM-compatible machines running Microsoft Windows. The software and lacI database are freely available via the internet (http:/(/)sunsite.unc.edu/dnam/mainpage.++ +html).

Analysis of in vivo mutation induced by N-ethyl-N-nitrosourea in the hprt gene of rat lymphocytes

The rat lymphocyte hprt assay measures in vivo mutagenicity by quantifying the frequency of 6-thioguanine-resistant (TGr) spleen lymphocytes cultured in vitro. In this study we have examined the types of mutations induced in the hprt gene of TGr lymphocyte clones from female Fischer 344 rats exposed to 100 mg/kg N-ethyl-N-nitrosourea (ENU). Hprt exons 3 and 8 were amplified from DNA extracted from each of 249 clones, and the resulting products were screened for mutant:wild-type heteroduplex formation by denaturing gradient gel electrophoresis. The analysis revealed 59 clones with mutations in exon 3, and 20 clones with mutations in exon 8. DNA sequence analysis of the heteroduplexes identified 84 mutations: all of the mutations were base pair substitutions, and 88% were mutations of A:T base pairs. At least 82% were induced independently. These results suggest that the mutations found in TGr rat lymphocytes from ENU-treated rats were due mainly to ethylthymidine adducts. In addition, a comparison of these results with previously reported in vivo ENU mutational profiles indicates that the types of mutation detected by heteroduplex screening of rat hprt exons 3 and 8 are representative of mutation in the entire protein coding sequence.

Response of repair-competent and repair-deficient Escherichia coli to three O6-substituted guanines and involvement of methyl-directed mismatch repair in the processing of O6-methylguanine residues

Plasmids containing a site-specifically incorporated O6-methyl- (m6G), O6-ethyl- (e6G), or O6-benzylguanine (b6G) within the ATG initiation codon of the lacZ' gene were used to transform Escherichia coli that were repair proficient or deficient in one or both of the E. coli O6-alkylguanine-DNA alkyltransferases, the uvr(ABC) excision repair system, the recA-mediated recombination system, or the methylation-directed mismatch repair system. Colonies were scored phenotypically for adduct-induced mutations. With plasmids containing either e6G or b6G, the frequency of adduct-induced mutation was low and independent of the repair proficiency of the strain transformed. Plasmids containing an m6G residue elicited similar responses in all but the mismatch repair-deficient strain. The generally low mutagenicity of all the O6-substituted guanines was interpreted as reflecting an adduct-induced arrest of replication of the modified strand while the unmodified complementary strand was replicated normally. Studies of the involvement of mismatch repair in m6G mutagenesis showed that m6G:T base pairs were more readily processed than m6G:C base pairs, indicating that mismatch repair involving m6G residues occurs after replication. These data support a model in which the E. coli methylation-directed mismatch repair system diverts plasmids containing promutagenic m6G:T base pairs into replication-arrested complexes providing another line of defense against O6-methylguanine mutagenicity in addition to O6-alkylguanine-DNA alkyltransferase repair and excision repair mechanisms.

Crystal and solution structures of d(CGC[e6G]AATTCGCG)-drug complexes reveal conformational polymorphism of O6-ethyl-guanine:cytosine base pair

O6-ethyl-guanine (e6G) is a relatively persistent alkylation lesion caused by the exposure of DNA to carcinogen N-ethyl-N-nitrosourea. We have studied the structural consequences of the e6G incorporation in DNA by X-ray crystallography and NMR. We have obtained crystals of the modified DNA dodecamer d(CGC[e6G]AATTCGCG) complexed to several minor groove binding drugs including Hoechst 33258, Hoechst 33342, netropsin, and SN6999. The space group of the crystals from those complexes is P2(1)2(1)2(1). However the crystal structure of the SN6999 complex is not isomorphous to that from the other three complexes. In all four refined crystal structures the drugs bind in the narrow minor groove at or close to the central AATT region of the dodecamer B-DNA duplex. The DNA conformation is influenced by the binding of drugs. The eight independent e6G:C base pairs have a conformation ranging from one with three-centered hydrogen bonds between the bases to a wobble conformation with two hydrogen bonds. The ethyl group of the eight e6G bases is mostly in the proximal orientation to N7. Our 1D and 2D-NMR studies of the same (free) dodecamer reveal that the e6G:C base pairs in the duplex are likely to adopt a wobble conformation in solution. Those results suggest that the e6G:C base pair has a dynamic equilibrium among various conformations, which may present an ambiguous signal to cells. In contrast, the e6G:T base pair adopts a Watson-Crick-like conformation. This may be a plausible explanation of why thymine is found preferentially incorporated across the e6G during replication.

Spectrum of spontaneous missense mutations causing cyclic AMP-resistance phenotypes in cultured S49 mouse lymphoma cells differs markedly from those of mutations induced by alkylating mutagens

Mutants of S49 mouse lymphoma cells resistant to cytolysis by analogs of cyclic AMP (cAMP) generally have missense mutations in the gene encoding the regulatory subunit of cAMP-dependent protein kinase. We have compared the mutations in 95 spontaneous isolates with those in 60 mutagen-induced isolates by sequence analysis of amplified cDNAs. Twenty-nine single basepair substitutions in 19 codons produced selectable phenotypes. The spontaneous mutant spectrum was dominated by a CpG transition hotspot in the codon for Arg334. This and other nearby CpG sites were found to be methylated in genomic S49 cell DNA by restriction enzyme analyses. Most of the remaining spontaneous mutants had either G-C-->C-G or T-A-->G-C transversions, which have been associated with damage caused by oxygen radicals. In contrast, the majority of mutants induced with the alkylating mutagens ethyl methanesulfonate (EMS) and N-methyl-N'-nitro-N-nitrosoguanidine had G-C-->A-T mutations at non-CpG sites; in addition, EMS induced several A-T-->G-C, A-T-->T-A, and G-C-->T-A substitutions. A single ICR191-induced mutant analyzed had a unique A-T-->G-C lesion. A number of spontaneous and mutagen-induced isolates had closely linked double or triple substitutions, and two isolates had tandem triple substitutions.

Inducible transcription of the dnaA gene from Streptomyces lividans 66

The dnaA gene of Streptomyces lividans was cloned using the Escherichia coli medium-copy-number vector pSU18 and E. coli strain TC1963, which can by-pass the requirement for the DnaA protein. Its regulatory region was subcloned in the Streptomyces probe vector pIJ4083. Primer extension and S1 mapping studies allowed the identification of a class I Streptomyces promotor (P2). An additional, previously unknown promoter type (P1) was found by S1 mapping. The presence of two DnaA box motifs between P1 and P2 suggests that the transcriptions of the S. lividans dnaA gene is autoregulated by its gene product. It was shown that the transcription of the dnaA gene is significantly induced by mitomycin C, an agent known to inhibit DNA replication. The data suggest that, as in E. coli, one of the regulatory mechanisms governing the transcription of the dnaA gene in S. lividans is probably related to the SOS response network.

Site-specific mutagenesis induced by single O6-alkylguanines (O6-n-propyl, O6-n-butyl, O6-n-octyl) in vivo

The mutagenic activity of a series of longer chain O6-n-alkylguanine residues (O6-n-propyl, O6-n-butyl, O6-n-octyl) has been analyzed using a plasmid molecule (pUC 9) in which single O6-alkylguanines were positioned in the unique Pstl recognition site by shot gun ligation (Nucleic Acids Res. 13, 3305-3316 (1985)) of overlapping synthetic oligonucleotides. After transfection of these vectors into E. coli cells having normal DNA repair systems, progeny plasmids were produced, of which 2.6%, 2.8% and 4.3% were mutated in their Pstl site when containing O6-n-propylguanine, O6-n-butylguanine, O6-n-octylguanine, respectively. DNA sequence analysis of mutant plasmid genomes revealed that O6-n-propylguanine and O6-n-butylguanine induced exclusively G-->A transitions located specifically at the preselected site. O6-n-octylguanine induced apart from G-->A transitions (70%) also targeted G-->T transversions (30%). These results indicate that the mutation frequency of longer chain O6-alkylguanines can be substantial in cells with normal repair systems and that the mutation pattern depends on the nature of the alkyl group.

Excision of 3-methylguanine from alkylated DNA by 3-methyladenine DNA glycosylase I of Escherichia coli

Escherichia coli has two DNA glycosylases for repair of DNA damage caused by simple alkylating agents. The inducible AlkA DNA glycosylase (3-methyladenine [m3A] DNA glycosylase II) removes several different alkylated bases including m3A and 3-methylguanine (m3G) from DNA, whereas the constitutively expressed Tag enzyme (m3A DNA glycosylase I) has appeared to be specific for excision of m3A. In this communication we have reexamined the substrate specificity of Tag by using synthetic DNA rich in GC base pairs to facilitate detection of any possible methyl-G removal. In such DNA alkylated with [3H]dimethyl sulphate, we found that m3G was excised from double-stranded DNA by both glycosylases, although more efficiently by AlkA than by Tag. This was further confirmed using both N-[3H]methyl-N-nitrosourea- and [3H]dimethyl sulphate-treated native DNA, from which Tag excised m3G with an efficiency that was about 70 times lower than for AlkA. These results can explain the previous observation that high levels of Tag expression will suppress the alkylation sensitivity of alkA mutant cells, further implying that m3G is formed in quantity sufficient to represent an important cytotoxic lesion if left unrepaired in cells exposed to alkylating agents.

The prevention of thymic lymphomas in transgenic mice by human O6-alkylguanine-DNA alkyltransferase

Nitrosoureas form O6-alkylguanine-DNA adducts that are converted to G to A transitions, the mutation found in the activated ras oncogenes of nitrosourea-induced mouse lymphomas and rat mammary tumors. These adducts are removed by the DNA repair protein O6-alkylguanine-DNA alkyltransferase. Transgenic mice that express the human homolog of this protein in the thymus were found to be protected from developing thymic lymphomas after exposure to N-methyl-N-nitrosourea. Thus, transgenic expression of a single human DNA repair gene is sufficient to block chemical carcinogenesis. The transduction of DNA repair genes in vivo may unravel mechanisms of carcinogenesis and provide therapeutic protection from known carcinogens.

Quantitative relationship between ethylated DNA bases and gene mutation at two loci in CHO cells

In a previous study we showed that the formation of O6-ethylguanine (O6-EtGua) in the DNA of CHO cells in culture correlated with mutations induced by ethylnitrosourea (ENU) and diethylsulfate (DES) at the hypoxanthine-guanine-phosphoribosyltransferase (hprt) locus but not at the Na, K-ATPase locus. This study was extended to another ethylating agent, ethyl methanesulfonate (EMS). DNA adduct formation and induction of mutation at the two gene loci were determined simultaneously in CHO cells after EMS exposure. The extent of ethylation at the N7 and O6 positions of guanine and at the N3 site of adenine were measured and the possible correlations with 6-thioguanine resistance (6-TGr) and ouabain resistance (ouar) mutations were investigated. A good correlation between the levels of ethylation at O6 guanine and mutation frequency at hprt gene by all three ethylating agents was observed. In the case of the ouar locus, the frequency of O6-EtGua adducts correlated with mutation induction by EMS and ENU but not by DES. Although both EMS and DES have similar reaction mechanisms, these results highlight differences in their mutational specificity. The comparison of this type of analysis with mutational spectra revealed that correlation studies may be inadequate to analyse multicomponent phenomena like mutation formation.

Mutagenesis by O6 meG residues within codon 12 of the human Ha-ras proto-oncogene in monkey cells

The first or/and the second guanines of the human Ha-ras codon 12 (normally GGC) were substituted by O6 meG residues and the modified sequence was subsequently introduced into an SV40-based shuttle vector able to replicate in both simian cells and bacteria. After replication in simian COS7 cells (proficient in O6-alkyl-guanine transferase), plasmid DNA was extracted and mutations were screened in E. coli DH5 alpha cells. The vast majority of the mutations induced by O6 meG were G----A transitions. The mutation frequency observed at the second guanine of codon 12 (12G2 position: 3.75% +/- 0.4) was higher than the one observed at the first guanine (12G1 position: 1.09% +/- 0.6). This difference was confirmed by the results obtained when two adjacent O6 meG residues were positioned within codon 12. The higher mutation frequency observed for the 12G2 position could be attributed to differential repair or/and variation in polymerase fidelity. These results are in agreement with animal experiments where alkylating agents gave rise to mutation on G2 position of codon 12.

Mutational spectrum at the Hprt locus in splenic T cells of B6C3F1 mice exposed to N-ethyl-N-nitrosourea

We have determined the mutational spectrum of N-ethyl-N-nitrosourea (ENU) in exon 3 of the hypoxanthine (guanine) phosphoribosyltransferase gene (Hprt) in splenic T cells following in vivo exposure of male B6C3F1 mice (5-7 weeks old) to ENU. Hprt- mutants were isolated by culturing splenic T cells in microtiter dishes containing medium supplemented with interleukin 2, concanavalin A, and 6-thioguanine. DNA was extracted from 6-thioguanine-resistant colonies and amplified by the polymerase chain reaction (PCR) using primers flanking Hprt exon 3. Identification of mutant sequences and purification of mutant DNA from contaminating wild-type Hprt DNA was accomplished by denaturing-gradient gel electrophoresis. Purified mutant DNA was then sequenced. Treatment of mice with ENU at 40 mg/kg of body weight produced a Hprt- mutant frequency of 7.3 x 10(-5) in splenic T cells, approximately 35-fold above background levels. Sixty-nine of the 521 Hprt- mutants analyzed contained mutations in exon 3 (13%). Transversions and transitions at A.T base pairs dominated the spectrum; 62 of the 69 exon 3 mutations were at A.T base pairs (14 different sites). Thirteen of 14 thymine bases undergoing mutation (61 of 62 mutations at A.T bases) were located on the nontranscribed strand of exon 3. The majority of the remaining mutations (6 of 69) were transitions at a single G.C base pair. These results suggest the importance of thymidine alkylation in ENU-induced mutagenesis in vivo. The mouse Hprt- T-cell cloning/sequencing assay described here may represent a useful system for studying the molecular mechanism of chemically induced mutation occurring in vivo in an endogenous gene.

Repair of O6-ethylguanine in DNA protects rat 208F cells from tumorigenic conversion by N-ethyl-N-nitrosourea

O6-Ethylguanine (O6-EtGua) is one of about a dozen different alkylation products formed in the DNA of cells exposed to the alkylating N-nitroso carcinogen N-ethyl-N-nitrosourea (EtNU). We have evaluated selectively the relative capacity of cells for the specific enzymatic repair of O6-EtGua as a determinant for the probability of malignant conversion. Eleven O6-EtGua-repair-proficient (R+) variant subclones were isolated from the O6-EtGua-repair-deficient (R-) clonal rat fibroblast line 208F by selection for resistance to 1,3-bis-(2-chloroethyl)-1-nitrosourea (frequency, approximately equal to 10(-5). Contrary to the 208F wild-type cells, all variants expressed O6-methylguanine-DNA methyltransferase activity, while both kinds of cells were deficient for repair of the DNA ethylation products O2- and O4-ethylthymine. After exposure to EtNU (less than or equal to 500 micrograms/ml; 20 min), cells were analyzed for the formation of piled-up foci in monolayer culture and of anchorage-independent colonies in semisolid agar medium. Depending on the EtNU concentration, the frequencies of piled-up foci and agar colonies, respectively, in the R+ variants were as low as 1/28th and 1/56th of those in the R- wild type. Contrasting with the cells from R+ variant-derived agar colonies, cells from 208F (R-) agar colonies gave rise to highly malignant tumors when implanted subcutaneously into syngeneic rats. No significant differences in the frequencies of piled-up foci were found between wild-type and variant cells after exposure to the major reactive metabolite of benzo[a]pyrene, (+)-7 beta, 8 alpha-dihydroxy-9,10 alpha-epoxy-7,8,9,10 alpha-tetrahydrobenzo[a] pyrene, for which stable binding to guanine O6 in cellular DNA has not been observed. The relative capacity of cells for repair of O6-alkylguanine is, therefore, a critical determinant for their risk of malignant conversion by N-nitroso carcinogens.

Use of shuttle vectors to study the molecular processing of defined carcinogen-induced DNA damage: mutagenicity of single O4-ethylthymine adducts in HeLa cells

We developed a simian virus 40 based shuttle vector system to study the molecular consequences of distinct carcinogen-induced DNA lesions in human cells. To establish the mutagenicity of O4-ethylthymine adducts, oligonucleotides carrying a single O4-ethylthymine adduct at a unique position were ligated into the vector molecules. Following replication in HeLa cells on average 23% of the progeny molecules carried a mutation in the region of modification. The vast majority of these mutations represented single T----C transitions at the position of the modified base, most probably as a consequence of mispairing of the O4-ethylthymine residues during replication. To a minor extent the O4-ethylthymine adduct may also induce T----A transversions or double point mutations. The in vivo mutation frequency of the adduct was found to be comparable to that of a C-A mismatch at the same position, but was lower than that expected from in vitro experiments with adducted DNA templates and purified DNA polymerases.

O6-methylguanine in the SV40 origin of replication inhibits binding but increases unwinding by viral large T antigen

To study the effect of the potentially cytotoxic base O6-methylguanine (O6-meG) on the initiation of DNA replication, double-stranded oligonucleotides corresponding to the SV40 origin of replication were constructed in which O6-meG replaced guanine in one strand. Out of 14 methylated residues, 10 were present in the Binding sites for T antigen (3 in Binding Site 1 and 7 in Binding Site 2). Binding of purified T antigen to the substituted oligonucleotide was considerably reduced in comparison to the unsubstituted one, as measured by nitrocellulose filter binding. Both the ATP-dependent and ATP-independent binding of T antigen were affected by the presence of the methylated base. Band shift analysis revealed an altered pattern of delayed-migrating complexes of T antigen with the O6-meG-containing oligonucleotide. Competition experiments, in which unmodified oligonucleotides containing Binding Site 1 or 2 were included in the binding assays, indicated that the affinity of T antigen for the O6-meG modified sites was reduced. When partially duplex oligonucleotides containing either Binding Site 1 or Site 2 of the origin of replication were used as substrates for the helicase activity of T antigen, the presence of O6-meG increased the extent of T antigen catalysed displacement of single-stranded DNA fragments.