Mismatch repair in correction of replication errors and processing of DNA damage.

The primary role of mismatch repair (MMR) is to maintain genomic stability by removing replication errors from DNA. This repair pathway was originally implicated in human cancer through an association between microsatellite instability in colorectal tumors in hereditary nonpolyposis colon cancer (HNPCC) kindreds. Microsatellites are short repetitive sequences which are often copied incorrectly by DNA polymerases because the template and daughter strands in these regions are particularly prone to misalignment. These replication-dependent events create loops of extrahelical bases which would produce frameshift mutations unless reversed by MMR. One consequence of MMR loss is a widespread expansion and contraction of these repeated sequences that affects the whole genome. Defective MMR is therefore associated with a mutator phenotype. Since the same pathway is also responsible for repairing base:base mismatches, defective cells also experience large increases in the frequency of spontaneous transition and transversion mutations. Three different approaches have been used to investigate the function of individual components of the MMR pathway. The first is based on the biochemical characterization of the purified protein complexes using synthetic DNA substrates containing loops or single mismatches. In the second, the biological consequences of MMR loss are inferred from the phenotype of cell lines established from repair-deficient human tumors, from tolerant cells or from mice defective in single MMR genes. In particular, molecular analysis of the mutations in endogenous or reporter genes helped to identify the DNA substrates for MMR. Finally, mice bearing single inactive MMR genes have helped to define the involvement of MMR in cancer prevention.

Spontaneous development of drug resistance: mismatch repair and p53 defects in resistance to cisplatin in human tumor cells.

The contributions of defective mismatch repair and mutated p53 to cisplatin resistance of human tumor cells were analysed. Mismatch repair defects were not associated with a predictable degree of resistance among several tumor cell lines. Repair defective variants of the A2780 ovarian carcinoma cell line which were isolated by selection for a methylation tolerant phenotype and did not express the hMLH1 mismatch repair protein, were highly resistant to cisplatin. Their cisplatin resistance was not a simple consequence of the mismatch repair defect. They were members of a drug-naive subpopulation of A2780 in which a silent hMLH1 gene accompanies a mutated p53. Two complementary approaches indicated that each defect contributes to cisplatin resistance independently and to a different extent. Firstly, separate introduction of a p53 defect into A2780 cells significantly increased their cisplatin resistance; defective hMLH1 provided less extensive protection. Secondly, azadeoxycytidine reactivation of the silent hMLH1 gene or expression of a transfected hMLH1 cDNA sensitized the doubly hMLH1/p53 deficient cells only slightly to cisplatin. Both approaches indicate that defective p53 status is a major determinant of cisplatin resistance and defective mismatch repair is a minor, and independent, contributor. The data have implications for the development of intrinsic cisplatin resistance.

Unmasking a killer: DNA O(6)-methylguanine and the cytotoxicity of methylating agents.

Methylating agents are potent carcinogens that are mutagenic and cytotoxic towards bacteria and mammalian cells. Their effects can be ascribed to an ability to modify DNA covalently. Pioneering studies of the chemical reactivity of methylating agents towards DNA components and their effectiveness as animal carcinogens identified O(6)-methylguanine (O(6)meG) as a potentially important DNA lesion. Subsequent analysis of the effects of methylating carcinogens in bacteria and cultured mammalian cells - including the discovery of the inducible adaptive response to alkylating agents in Escherichia coli - have defined the contributions of O(6)meG and other methylated DNA bases to the biological effects of these chemicals. More recently, the role of O(6)meG in killing mammalian cells has been revealed by the lethal interaction between persistent DNA O(6)meG and the mismatch repair pathway. Here, we briefly review the results which led to the identification of the biological consequences of persistent DNA O(6)meG. We consider the possible consequences for a human cell of chronic exposure to low levels of a methylating agent. Such exposure may increase the probability that the cell's mismatch repair pathway becomes inactive. Loss of mismatch repair predisposes the cell to mutation induction, not only through uncorrected replication errors but also by methylating agents and other mutagens.

Sensitivity to DNA cross-linking chemotherapeutic agents in mismatch repair-defective cells in vitro and in xenografts.

Together with tolerance to killing induced by methylating agents, loss of mismatch repair (MMR) has previously been found to be associated with hypersensitivity to the DNAcross-linking agent 1-(2-chloroethyl)-3-cyclohexyl-nitrosourea(CCNU) in several human tumor cell lines (Aquilina et al., 1998). Here, we have investigated whether MMR might act as an efficient repair pathway and provide protection against the clastogenicity induced by CCNU and whether the hypersensitivity of MMR-defective cells is extended to other cross-linking agents. An increase in cell killing and in the frequency of micronuclei was observed after CCNU exposure in 2 hPMS2-defective clones (clones 6 and 7) compared with the parental HeLa cells. Introduction of a wild-type copy of chromosome 7 in clone 7 led to re-expression of the hPMS2 protein and brought survival and chromosomal damage upon CCNU exposure to parental levels. Our data indicate that MMR protects against the clastogenic damage induced by this drug. The hPMS2-defective HeLa cells were also hypersensitive to killing by mitomycin C. Mitomycin C sensitivity was confirmed in an hMLH1-defective clone derived from Raji cells and in msh2-defective mouse embryo fibroblasts derived from knock-out mice. hPMS2-defective and parental HeLa cells were transplanted into nude mice, and the animals were treated with mitomycin C. While parental cell growth rate was unaffected, the growth of MMR-defective tumor was significantly reduced. Our results indicate that the in vitro hypersensitivity to mitomycin C conferred by loss of MMR is paralleled in vivo and may have implications for the chemotherapy of MMR-defective tumors.

Mismatch repair and p53 independently affect sensitivity to N-(2-chloroethyl)-N'-cyclohexyl-N-nitrosourea.

The contributions of defective mismatch repair (MMR) and the p53-response to cell killing by N-(2-chloroethyl)-N'-cyclohexyl-N-nitrosourea (CCNU) were evaluated. MMR defects were previously shown to be associated with CCNU sensitivity (G. Aquilina et al., Cancer Res., 58: 135-141, 1998). Unexpectedly, eight MMR-deficient variants of the A2780 human ovarian carcinoma cell line were 3-fold more resistant to CCNU than the MMR-proficient parental cells. The variants were members of a preexisting subpopulation of drug-resistant A2780 cells. In addition to deficient expression of the MMR protein hMLH1, an essential component of the hMutL alpha repair complex, the variants exhibited alterations in the expression of other genes that influence drug sensitivity. Although A2780 cells possess a wild-type p53 gene, all of the clones contained a heterozygous G to T tranversion at codon 172. This change resulted in a Val to Phe substitution and was associated with a constitutive production of high levels of p53, which was inactive as a transcriptional activator of bax and p21. The hMLH1/p53 defective variants displayed a less prominent cell cycle arrest and reduced apoptosis after CCNU treatment. In contrast, MMR-defective A2780 variants, which had a similar hMutL alpha defect but retained a wild-type p53, did exhibit the expected CCNU sensitivity. Expression of a dominant-negative p53val135 increased CCNU resistance of both MMR-proficient and MMR-deficient A2780 cells. Thus, defective MMR and p53 influence CCNU sensitivity in opposite directions. Their effects are independent, and sensitization by defective MMR does not require a functional p53 response.

Mismatch repair, G(2)/M cell cycle arrest and lethality after DNA damage.

The role of the mismatch repair pathway in DNA replication is well defined but its involvement in processing DNA damage induced by chemical or physical agents is less clear. DNA repair and cell cycle control are tightly linked and it has been suggested that mismatch repair is necessary to activate the G(2)/M checkpoint in the presence of certain types of DNA damage. We investigated the proposed role for mismatch repair (MMR) in activation of the G(2)/M checkpoint following exposure to DNA-damaging agents. We compared the response of MMR-proficient HeLa and Raji cells with isogenic variants defective in either the hMutLalpha or hMutSalpha complex. Different agents were used: the cross-linker N-(2-chloroethyl)-N'-cyclohexyl-N-nitrosourea (CCNU), gamma-radiation and the monofunctional methylating agent N-methyl-N-nitrosourea (MNU). MMR-defective cells are relatively sensitive to CCNU, while no differences in survival between repair-proficient and -deficient cells were observed after exposure to gamma-radiation. Analysis of cell cycle distribution indicates that G(2) arrest is induced at least as efficiently in MMR-defective cells after exposure to either CCNU or ionizing radiation. As expected, MNU does not induce G(2) accumulation in MMR-defective cells, which are known to be highly tolerant to killing by methylating agents, indicating that MNU-induced cell cycle alterations are strictly dependent on the cytotoxic processing of methylation damage by MMR. Conversely, activation of the G(2)/M checkpoint after DNA damage induced by CCNU and gamma-radiation does not depend on functional MMR. In addition, the absence of a simple correlation between the extent of G(2) arrest and cell killing by these agents suggests that G(2) arrest reflects the processing by MMR of both lethal and non-lethal DNA damage.

Effect of hMSH6 cDNA expression on the phenotype of mismatch repair-deficient colon cancer cell line HCT15.

Mismatch recognition in human cells is mediated primarily by a heterodimer of hMSH2 and hMSH6. Cells mutated in both alleles of the hMSH6 gene are deficient in the correction of base/base mispairs and insertion/deletion loops of one nucleotide and thus exhibit a strong mutator phenotype, evidenced by elevated mutation rates and microsatellite instability, as well as by tolerance to methylating agents. The decrease in replication fidelity associated with a loss of mismatch correction implies that with each division, these cells are likely to acquire new mutations throughout their genomes. Should such secondary mutations occur in genes linked to replication fidelity or involved in the maintenance of genomic stability, they might contribute to the observed mutator phenotype. The human colon tumour line HCT15 represents one such case. Although it carries inactivating mutations in both hMSH6 alleles, it has also been shown to contain a missense mutation in the coding sequence of the proofreading domain of the polymerase-delta gene. In an attempt to find out whether the phenotype of HCT15 cells was indeed brought about solely by the lack of hMSH6, we stably transfected them with a vector carrying the wild-type hMSH6 cDNA. Our results show that although the levels of transgenic hMSH6 were low, expression of the wild-type protein resulted in a substantial restoration of mismatch binding, mismatch repair capacity and the stability of mononucleotide repeats, as well as in the reduction of mutation rates. Although methylation tolerance of the hMSH6-expressing cells was not markedly affected, the G2 cell cycle checkpoint, absent in N-methyl-N'-nitro-N-nitrosoguanidine-treated control cells, was restored.

Mismatch repair and differential sensitivity of mouse and human cells to methylating agents.

The long-patch mismatch repair pathway contributes to the cytotoxic effect of methylating agents and loss of this pathway confers tolerance to DNA methylation damage. Two methylation-tolerant mouse cell lines were identified and were shown to be defective in the MSH2 protein by in vitro mismatch repair assay. A normal copy of the human MSH2 gene, introduced by transfer of human chromosome 2, reversed the methylation tolerance. These mismatch repair defective mouse cells together with a fibroblast cell line derived from an MSH2-/- mouse, were all as resistant to N-methyl-N-nitrosourea as repair-defective human cells. Although long-patch mismatch repair-defective human cells were 50- to 100-fold more resistant to methylating agents than repair-proficient cells, loss of the same pathway from mouse cells conferred only a 3-fold increase. This discrepancy was accounted for by the intrinsic N-methyl-N-nitrosourea resistance of normal or transformed mouse cells compared with human cells. The >20-fold differential resistance between mouse and human cells could not be explained by the levels of either DNA methylation damage or the repair enzyme O6-methylguanine-DNA methyltransferase. The resistance of mouse cells to N-methyl-N-nitrosourea was selective and no cross-resistance to unrelated DNA damaging agents was observed. Pathways of apoptosis were apparently intact and functional after exposure to either N-methyl-N-nitrosourea or ultraviolet light. Extracts of mouse cells were found to perform 2-fold less long-patch mismatch repair. The reduced level of mismatch repair may contribute to their lack of sensitivity to DNA methylation damage.

Increased somatic recombination in methylation tolerant human cells with defective DNA mismatch repair.

We have studied whether spontaneous intrachromosomal recombination is altered in methylation tolerant human cells with a defect in mismatch repair. Somatic recombination was analysed in HeLaMR cells containing the vector pTPSN, which carries two copies of the gene for hygromycin resistance. The hygromycin genes are both inactivated by an inserted HindIII linker but hygromycin-resistant clones can arise by recombination. The spontaneous rate of recombination in a clone of HeLaMR cells containing a single integrated copy of pTPSN (HeLaG1) was 3.1x10(-6)/cell per generation. Two methylation tolerant variants from HeLaG1 cells (clone 12 and clone 15) were isolated by exposure to MNNG. Clone 12 cells exhibited a 16-fold increase in spontaneous mutation rate at the HPRT gene and extensive microsatellite instability at both mono- and dinucleotide repeats. Microsatellite instability limited to mononucleotide repeats was found in clone 15, whereas the mutation rate at HPRT was not significantly affected. A mismatch binding defect in extracts of clone 15 could be complemented by exogenous GTBP but not by purified hMSH2 protein. These data suggest that clone 15 is defective in GTBP. Extracts of clone 12 were unable to correct a single C:T mispair and complementation by extracts of human colorectal carcinoma cells with known deficiencies in mismatch repair indicated a defect in hMutLalpha. Western blotting with antibodies against different human mismatch repair proteins showed that clone 12 cells did not express hPMS2 protein, but expression of hMLH1, hMSH2 and GTBP appeared normal. The spontaneous recombination rate of clone 12 was 19-fold higher than the parental HeLaG1 cells, whereas no increase was observed in clone 15. Analysis of individual recombinants showed that hygromycin resistance arose exclusively by gene conversion. Our data indicate that mismatch correction regulates somatic recombination in human cells.

N-(2-chloroethyl)-N'-cyclohexyl-N-nitrosourea sensitivity in mismatch repair-defective human cells.

To determine whether loss of mismatch repair (MMR) confers sensitivity to N-(2-chloroethyl)-N'-cyclohexyl-N-nitrosourea (CCNU), the sensitivity of MMR-defective (MMR-) variants was compared to that of their parental cells. Loss of MMR confers between 2- and 5-fold hypersensitivity to CCNU on HeLa, Raji, or Chinese hamster ovary cells. We also examined whether the sensitivity to CCNU is a general feature of MMR-human tumor cells. The majority expressed O6-methylguanine-DNA-methyltransferase (MGMT; Mex+ phenotype) that confers resistance to CCNU independent of their MMR status. The single Mex- MMR- SW48 cells were 4-fold more sensitive to CCNU than the Mex- MMR+ SW620 cells. CCNU sensitivity of the Mex+ cells was analyzed after treatment with the MGMT inhibitor O6-benzylguanine. The MMR- AN3CA, LS174T, LoVo, and DU145 cells were 1.4-4.3-fold more sensitive to CCNU than the MMR+ HeLaS3, HT29, and A2780 cells. Hypersensitivity to CCNU was not seen in the MMR- cell lines DLD1, HEC1A, and HCT116, suggesting that other parameters, besides the MGMT and MMR defects, affect the cell's response to this drug. In contrast, loss of MMR was always associated with tolerance to the methylating agent N-methyl-N-nitrosourea. The sensitivity to CCNU in MMR- cells suggests a possible involvement of this repair pathway in repairing interstrand cross-links and may have implications for clinical treatment of MMR- tumors.

Mismatch repair defects and O6-methylguanine-DNA methyltransferase expression in acquired resistance to methylating agents in human cells.

Fifteen variants with >/=30-fold resistance to N-methyl-N-nitrosourea were isolated from the Burkitt's lymphoma Raji cell line. Eight had received a single treatment with a highly cytotoxic dose. The remainder, including the previously described RajiF12 cell line, arose following multiple exposures to initially moderate but escalating doses. Surprisingly, methylation resistance arose in three clones by reactivation of a previously silent O6-methylguanine-DNA methyltransferase gene. Five clones, including RajiF12, displayed the microsatellite instability and increased spontaneous mutation rates at the hypoxanthine-guanine phosphoribosyltransferase locus, consistent with deficiencies in mismatch repair. Defects in either the hMutSalpha or hMutLalpha mismatch repair complexes were identified in extracts of these resistant clones by in vitro complementation using extracts from colorectal carcinoma cell lines. Defects in hMutLalpha were confirmed by Western blot analysis. Remarkably, five methylation-resistant clones in which mismatch repair defects were demonstrated by biochemical assays did not exhibit significant microsatellite instability.

Reversal of methylation tolerance by transfer of human chromosome 2.

Human cell lines resistant to N-methyl-N-nitrosourea (MNU) were previously assigned to two complementation groups. Members of group I are defective in mismatch correction [S. Ceccotti, G Aquilina, P. Macpherson, M. Yamada, P. Karran, M. Bignami, Processing of O6-methylguanine by mismatch correction in human cell extracts. Current Biol. 6 (1996) 1528-1531]. To identify the mechanism responsible for the less pronounced phenotype of the second complementation group, we characterized the persistence of MNU-induced O6-methylguanine (O6-meGua) and mutation induction at the hypoxanthine guanine phosphoribosyl-transferase (HPRT) locus. Group II clones are unable to repair the premutagenic base O6-meGua and are as mutable by MNU as group I clones and the parental HeLaMR cells. MNU-induced SCE were undetectable in group I clones and drastically reduced in group II in comparison with the parental cells. These observations are consistent with a defective processing of DNA methylation damage by members of both groups. Group II clones exhibit a moderate spontaneous mutator phenotype at the HPRT gene but significant instability at mononucleotide repeat microsatellites. Introduction of a single human chromosome 2 (but not of chromosome 3 or 7) into group II cells partially reverts both MNU resistance and the increased spontaneous mutation rate. The properties of group II variants are consistent with methylation tolerance and a partially defective mismatch repair. We propose that members of group II are defective in the chromosome 2-based mismatch correction gene GTBP/hMSH6.

Processing of O6-methylguanine by mismatch correction in human cell extracts.

Human cell extracts perform an aberrant form of DNA synthesis on methylated plasmids [1], which represents processing of O6-methylguanine (O6-meG). Here, we show that extracts of colorectal carcinoma cells with defects in the mismatch repair proteins that normally correct replication errors do not carry out this synthesis. hMSH2-defective LoVo cell extracts (hMSH for human MutS homologue) performed O6-meG-dependent DNA synthesis only after the addition of the purified hMutS alpha mismatch recognition complex. Processing of O6-meG by mismatch correction requires PCNA and therefore probably DNA polymerase delta and/or epsilon. Mismatch repair-defective cells withstand O6-meG in their DNA [2], making them tolerant to methylating agents. Methylation-tolerant HeLaMR clones, with a mutator phenotype and a defect in either mismatch recognition or correction in vitro, also performed little O6-meG-dependent DNA synthesis. Assays of pairwise combinations of tolerant and colorectal carcinoma cell extracts identified hMLH1 as the missing mismatch repair function in a group of tolerant clones. The absence of processing by extracts of methylation-tolerant cells provides the first biochemical evidence that lethality of DNA O6-meG derives from its interaction with mismatch repair.

Genetic instability and methylation tolerance in colon cancer.

Microsatellite instability was first identified in colon cancer and later shown to be due to mutations in genes responsible for correction of DNA mismatches. Several human mismatch correction genes that are homologous to those of yeast and bacteria have been identified and are mutated in families affected by the hereditary non-polyposis colorectal carcinoma (HNPCC) syndrome. Similar alterations have been also found in some sporadic colorectal cancers. The mismatch repair pathway corrects DNA replication errors and repair-defective colorectal carcinoma cell lines exhibit a generalized mutator phenotype. An additional consequence of mismatch repair defects is cellular resistance, or tolerance, to certain DNA damaging agents.

A mutator phenotype characterizes one of two complementation groups in human cells tolerant to methylation damage.

Sixty % of clones isolated from HeLa cells treated with toxic concentrations of a methylating carcinogen showed increased resistance to the cytotoxicity of N-methyl-N-nitrosourea. D37 values were 6- to 100-fold higher than in the parental cell population. The absence of detectable levels of the repair enzyme O6-methylguanine-DNA methyltransferase indicated that the resistant clones were able to tolerate the presence of O6-methylguanine in their DNA. Analysis of N-methyl-N-nitrosourea survival in the hybrids between tolerant clones and HeLa cells showed that tolerance can be either recessive or codominant. Fusion between tolerant clones indicated two complementation groups. We measured spontaneous mutation rates at microsatellites and at the hypoxanthine-guanine phosphoribosyl transferase (hprt) locus in several tolerant clones. All the clones of Complementation Group I showed unstable microsatellites and 4-8-fold increases in mutation rates at hprt. No significant alterations in spontaneous mutation rates were found in clones of Complementation Group II. The data indicate that tolerance to methylation damage can be conferred by alterations in at least two different gene products and that one of the two groups has the mutator phenotype typical of mismatch correction defective cells.

DNA replication arrest and tolerance to DNA methylation damage.

Inhibition of DNA replication by different DNA damaging agents has been investigated in HeLaMR cells and a methylation damage-tolerant variant HeLa5A1. In synchronous HeLaMR and HeLa5A1 cells exposed to N-ethyl-N-nitrosourea or ionizing radiation in mid-G1 phase, DNA synthesis was inhibited in the following S phase. N-methyl-N-nitrosourea-induced replication inhibition in HeLaMR cells was delayed until the second S phase after treatment. In contrast, N-methyl-N-nitrosourea treatment of HeLa5A1 cells affected neither the timing nor the extent of the first or second S phases. Both radiation and chemical treatment inhibited replication of an episomal plasmid and of genomic DNA in unison. Inhibition was observed at levels of DNA damage that did not directly damage the plasmid molecules. Thus, DNA replication inhibition occurs immediately after ionizing radiation or ethylation damage, but methylation damage requires processing through one cell cycle to generate an inhibitory signal. The inhibitory signal appears to act in trans on undamaged DNA. Although methylation-tolerant cells are responsive to inhibition after gamma-irradiation, methylation damage does not produce inhibitory signals to which they respond.

Spontaneous mutations at aprt locus in a mammalian cell line defective in mismatch recognition.

Clone B is a CHO cell line that shows a moderate mutator phenotype as a consequence of a defect in mismatch recognition. To identify the classes of mutation that accumulate spontaneously in a functional gene, we isolated and sequenced 54 clone B spontaneous mutants at the adenine phosphoribosyltransferase gene. This spectrum was compared to 42 mutants collected in the parental cells. Rates of AT-->TA transversions and frameshifts were strikingly increased in clone B (almost eight- and sixfold, respectively). Minor increases were also observed for GC-->TA transversions and GC-->AT transition rates. Frameshifts occurred in repeated sequences, and a large proportion were losses of 2 bases occurring in dinucleotide runs of a type similar to microsatellite sequences. AT-->TA transversions clustered in regions of secondary structure and their formation might be explained by slippage-mediated mechanisms. These data indicate that an important function of mismatch recognition is in repair of extrahelical bases generated by misalignment during DNA replication.

A mismatch recognition defect in colon carcinoma confers DNA microsatellite instability and a mutator phenotype.

We have analyzed spontaneous mutations in the adenine phosphoribosyltransferase gene of Chinese hamster clone B cells that exhibit a mutator phenotype because of defective mismatch binding. The mutator phenotype conferred increases in a limited number of mutational classes. The rates of transitions and most transversions were not significantly increased. The rates of A to T transversions and -2 frameshifts were strikingly elevated. These mutations were in repeated elements and 5 of 9 of the frameshifts were dinucleotide deletions in DNA sequences resembling microsatellites. The mismatch binding protein that is defective in the mutator line is a G-T mismatch recognition factor. Band-shift analysis indicated that the preferred substrate for the mismatch recognition protein is duplex DNA containing an extrahelical mono- or dinucleotide within repeated sequences. In agreement with a role in preventing minus frameshifts, a defective binding protein conferred an instability in clone B microsatellite DNA. A mismatch binding defect was also detected in Lo Vo, a human colorectal carcinoma cell line. Extracts of clone B or a second mismatch binding-deficient line, Raji-F12, did not complement Lo Vo extracts, indicating that these lines share a common defect. Our data provide a mechanistic explanation for the relation between defective mismatch recognition and the microsatellite instability of human colon cancer.

Endogenous DNA damage and spontaneous mutagenesis in cultured mammalian cells.

This review describes the main molecular mechanisms of endogenous mutagenesis, with reference to the possible sources of the different DNA lesions, the types of mutation originated and the specific defense strategies employed by the cell. The analysis of mutations induced by mutagenic agents that mimic the effect of endogenous sources of damage is also considered, since it provides information on the possible effect of intracellular mutagens. Furthermore spontaneous mutational spectra of different target genes of mammalian cells are described and compared and the contribution of the different mutational sources is discussed.