Eukaryotic mismatch repair: an update

The discovery that mutations in mismatch repair genes segregate with hereditary nonpolyposis colon cancer has awakened a great deal of interest in the study of the process of postreplicative mismatch repair. The characterisation of the principal players involved in this important metabolic pathway has been greatly facilitated by the amino acid sequence conservation among functional homologues of bacteria, yeast and mammals. The phenotypes of mismatch repair deficient mutants are also similar in many ways. In humans, mismatch repair malfunction demonstrates itself in the form of a mutator phenotype of the affected cells, an instability of microsatellite sequences and increased levels of somatic recombination. Moreover, mismatch repair deficient cells display also varying levels of tolerance to DNA damaging agents and are thought to be involved in the cell killing mediated by these agents. This article discusses some recent developments in this fast-moving field.     

Genetic requirements and mutational specificity of the Escherichia coli SOS mutator activity

To better understand the mechanisms of SOS mutagenesis in the bacterium Escherichia coli, we have undertaken a genetic analysis of the SOS mutator activity. The SOS mutator activity results from constitutive expression of the SOS system in strains carrying a constitutively activated RecA protein (RecA730). We show that the SOS mutator activity is not enhanced in strains containing deficiencies in the uvrABC nucleotide excision-repair system or the xth and nfo base excision-repair systems. Further, recA730-induced errors are shown to be corrected by the MutHLS-dependent mismatch-repair system as efficiently as the corresponding errors in the rec+ background. These results suggest that the SOS mutator activity does not reflect mutagenesis at so-called cryptic lesions but instead represents an amplification of normally occurring DNA polymerase errors. Analysis of the base-pair-substitution mutations induced by recA730 in a mismatch repair-deficient background shows that both transition and transversion errors are amplified, although the effect is much larger for transversions than for transitions. Analysis of the mutator effect in various dnaE strains, including dnaE antimutators, as well as in proofreading-deficient dnaQ (mutD) strains suggests that in recA730 strains, two types of replication errors occur in parallel: (i) normal replication errors that are subject to both exonucleolytic proofreading and dnaE antimutator effects and (ii) recA730-specific errors that are not susceptible to either proofreading or dnaE antimutator effects. The combined data are consistent with a model suggesting that in recA730 cells error-prone replication complexes are assembled at sites where DNA polymerization is temporarily stalled, most likely when a normal polymerase insertion error has created a poorly extendable terminal mismatch. The modified complex forces extension of the mismatch largely at the exclusion of proofreading and polymerase dissociation pathways. SOS mutagenesis targeted at replication-blocking DNA lesions likely proceeds in the same manner.     

A naturally occurring hPMS2 mutation can confer a dominant negative mutator phenotype

Defects in mismatch repair (MMR) genes result in a mutator phenotype by inducing microsatellite instability (MI), a characteristic of hereditary nonpolyposis colorectal cancers (HNPCC) and a subset of sporadic colon tumors. Present models describing the mechanism by which germ line mutations in MMR genes predispose kindreds to HNPCC suggest a "two-hit" inactivation of both alleles of a particular MMR gene. Here we present experimental evidence that a nonsense mutation at codon 134 of the hPMS2 gene is sufficient to reduce MMR and induce MI in cells containing a wild-type hPMS2 allele. These results have significant implications for understanding the relationship between mutagenesis and carcinogenesis and the ability to generate mammalian cells with mutator phenotypes.     

Correction of hypermutability, N-methyl-N'-nitro-N-nitrosoguanidine resistance, and defective DNA mismatch repair by introducing chromosome 2 into human tumor cells with mutations in MSH2 and MSH6

The human DNA mismatch repair genes hMSH2 and hMSH6 encode the proteins that, together, bind to mismatches to initiate repair of replication errors. Human tumor cells containing mutations in these genes have strongly elevated mutation rates in selectable genes and at microsatellite loci, although mutations in these genes cause somewhat different mutator phenotypes. These cells are also resistant to killing by certain drugs and are defective in mismatch repair. Because the elevated mutation rates in these cells may lead to mutations in additional genes that are causally related to the other defects, here we attempt to establish a cause-effect relationship between the hMSH2 and hMSH6 gene mutations and the observed phenotypes. The endometrial tumor cell line HEC59 contains mutations in both alleles of hMSH2. The colon tumor cell line HCT15 contains mutations in hMSH6 and also has a sequence change in a conserved region of the coding sequence for DNA polymerase delta, a replicative DNA polymerase. We introduced human chromosome 2 containing the wild-type hMSH2 and hMSH6 genes into HEC59 and HCT15 cells. Introduction of chromosome 2 to HEC59 cells restored microsatellite stability, sensitivity to N-methyl-N'-nitro-N-nitrosoguanidine treatment, and mismatch repair activity. Transfer of chromosome 2 to HCT15 cells also reduced the mutation rate at the HPRT locus and restored sensitivity to N-methyl-N'-nitro-N-nitrosoguanidine treatment and mismatch repair activity. The results demonstrate that the observed defects are causally related to mutations in genes on chromosome 2, probably hMSH2 or hMSH6, but are not related to sequence changes in other genes, including the gene encoding DNA polymerase delta.     

Mismatch repair co-opted by hypermutation

Mice homozygous for a disrupted allele of the mismatch repair gene Pms2 have a mutator phenotype. When this allele is crossed into quasi-monoclonal (QM) mice, which have a very limited B cell repertoire, homozygotes have fewer somatic mutations at the immunoglobulin heavy chain and lambda chain loci than do heterozygotes or wild-type QM mice. That is, mismatch repair seems to contribute to somatic hypermutation rather than stifling it. It is suggested that at immunoglobulin loci in hypermutable B cells, mismatched base pairs are "corrected" according to the newly synthesized DNA strand, thereby fixing incipient mutations instead of eliminating them.     

The role of 3'-5' exonucleolytic proofreading and mismatch repair in yeast mitochondrial DNA error avoidance

In the D171G/D230A mutant generated at conserved aspartate residues in the Exo1 and Exo2 sites of the 3'-5' exonuclease domain of the yeast mitochondrial DNA (mtDNA) polymerase (pol-gamma), the mitochondrial genome is unstable and the frequency of mtDNA point mutations is 1500 times higher than in the wild-type strain and 10 times higher than in single substitution mutants. The 10(4)-fold decrease in the 3'-5' exonuclease activity of the purified mtDNA polymerase is associated with mismatch extension and high rates of base misincorporation. Processivity of the purified polymerase on primed single-stranded DNA is decreased and the Km for dNTP is increased. The sequencing of mtDNA point mutations in the wild-type strain and in proofreading and mismatch-repair deficient mutants shows that mismatch repair contributes to elimination of the transitions while exonucleolytic proofreading preferentially repairs transversions, and more specifically A to T (or T to A) transversions. However, even in the wild-type strain, A to T (or T to A) transversions are the most frequent substitutions, suggesting that they are imperfectly repaired. The combination of both mismatch repair and proofreading deficiencies elicits a mitochondrial error catastrophe. These data show that the faithful replication of yeast mtDNA requires both exonucleolytic proofreading and mismatch repair.     

Human MLH1 deficiency predisposes to hematological malignancy and neurofibromatosis type 1

Heterozygous germ-line mutations in the DNA mismatch repair genes lead to hereditary nonpolyposis colorectal cancer. The disease susceptibility of individuals who constitutionally lack both wild-type alleles is unknown. We have identified three offspring in a hereditary nonpolyposis colorectal cancer family who developed hematological malignancy at a very early age, and at least two of them displayed signs of neurofibromatosis type 1 (NF1). DNA sequence analysis and allele-specific amplification in two siblings revealed a homozygous MLH1 mutation (C676T-->Arg226Stop). Thus, a homozygous germ-line MLH1 mutation and consequent mismatch repair deficiency results in a mutator phenotype characterized by leukemia and/or lymphoma associated with neurofibromatosis type 1.     

Detection of a mutator phenotype in cancer cells by inter-Alu polymerase chain reaction

A mutator phenotype due to a DNA mismatch repair deficiency is usually detected by typing a number of microsatellite markets. Here, eight hereditary nonpolyposis colon cancer patients with microsatellite instability were investigated by inter-Alu PCR, known to amplify DNA segments that may represent preferential targets of replication errors. Among 40-60 bands revealed in a single PCR experiment, more than 20% were found altered in tumoral DNA samples compared to matched normal samples from the same patient. Shifts and changes in signal intensity accounted for most of the alterations, whereas gains or losses of bands were rare. Certain bands were affected only in a single patient, whereas the instabilities in others were common. These results suggest that some genomic regions are more susceptible than others to the expression of a mutator phenotype. Four such bands altered in at least five patients were characterized further and shown to be unstable because of contractions of the Alu poly(A) tails. Interestingly, none of the bands representing loci shown previously to be polymorphic in the population displayed instability in the tumoral samples. Inter-Alu PCR appears to be a robust, cost-effective, and sensitive technique for revealing the mutator phenotype in cancer cells.     

Enhanced intestinal adenomatous polyp formation in Pms2-/-;Min mice

Analysis of two human familial cancer syndromes, hereditary nonpolyposis colorectal cancer and familial adenomatous polyposis, indicates that mutations in either one of four DNA mismatch repair gene homologues or the adenomatous polyposis coli (APC) gene, respectively, are important for the development of colorectal cancer. To further investigate the role of DNA mismatch repair in intestinal tumorigenesis, we generated mice with mutations in both Apc and the DNA mismatch repair gene, Pms2. Whereas Pms2-deficient mice do not develop intestinal tumors, mice deficient in Pms2 and heterozygous for Min, an allele of Apc, develop approximately three times the number of small intestinal adenomas and four times the number of colon adenomas relative to Min and Pms2+/-;Min mice. Although Pms2 deficiency clearly increases adenoma formation in the Min background, histological analysis indicated no clear evidence for progression to carcinoma.     

Removal of polymerase-produced mutant sequences from PCR products

Heteroduplex DNA lacking d(GATC) methylation is subject to mismatch-provoked double-strand cleavage at d(GATC) sites in a reaction dependent on MutH, MutL, MutS, and ATP. We have exploited this reaction to develop a method for removal of polymerase-produced mutant sequences that arise during sequence amplification by PCR. After denaturation and reannealing, the PCR product pool is subjected to MutH, MutL, and MutS mismatch repair proteins under double-strand cleavage conditions, followed by isolation of uncleaved product by size selection. Use of an Escherichia coli lac forward mutation assay has shown that this procedure reduces the incidence of polymerase-induced mutant sequences by an order of magnitude. Twenty mutants that originated from three independent PCR amplification reactions and survived MutHLS treatment all were found to contain an infrequently occurring A.T --> T.A transversion mutation at a unique position within the product. By contrast, the majority of mutations in untreated PCR products were transitions occurring throughout the amplified region, although frameshifts and transversions also were observed. The MutHLS method thus can be used to effectively remove the majority of mutant sequences produced by polymerase errors during PCR amplification.     

Drug-induced sleep disorders]

Mismatch repair genes are involved in increasing the fidelity of replication by specific repair of DNA polymerase incorporation errors. In Escherichia coli, the best studied mismatch repair (MMR) pathway is the methyl-directed long patch repair system which is mediated by three gene products; MutS, MutL and MutH. These are conserved in higher eukaryotes. Mutations in human homologues of these proteins have been shown to be implicated in hereditary non-polyposis colorectal cancer (HNPCC). Alterations in the coding regions of MMR genes result in a mutator phenotype with marked instability of microsatellite sequences, indicative of a deficiency in DNA repair.     

Escherichia coli MutY protein has a guanine-DNA glycosylase that acts on 7,8-dihydro-8-oxoguanine:guanine mispair to prevent spontaneous G:C-->C:G transversions

Low rates of spontaneous G:C-->C:G transversions would be achieved not only by the correction of base mismatches during DNA replication but also by the prevention and removal of oxidative base damage in DNA. Escherichia coli must have several pathways to repair such mismatches and DNA modifications. In this study, we attempted to identify mutator loci leading to G:C-->C:G transversions in E.coli. The strain CC103 carrying a specific mutation in lacZ was mutagenized by random miniTn 10 insertion mutagenesis. In this strain, only the G:C-->C:G change can revert the glutamic acid at codon 461, which is essential for sufficient beta-galactosidase activity to allow growth on lactose. Mutator strains were detected as colonies with significantly increased rates of papillae formation on glucose minimal plates containing P-Gal and X-Gal. We screened approximately 40 000 colonies and selected several mutator strains. The strain GC39 showed the highest mutation rate to Lac+. The gene responsible for the mutator phenotypes, mut39 , was mapped at around 67 min on the E.coli chromosome. The sequencing of the miniTn 10 -flanking DNA region revealed that the mut39 was identical to the mutY gene of E.coli. The plasmid carrying the mutY + gene reduced spontaneous G:C-->T:A and G:C-->C:G mutations in both mutY and mut39 strains. Purified MutY protein bound to the oligonucleotides containing 7,8-dihydro-8-oxo-guanine (8-oxoG):G and 8-oxoG:A. Furthermore, we found that the MutY protein had a DNA glycosylase activity which removes unmodified guanine from the 8-oxoG:G mispair. These results demonstrate that the MutY protein prevents the generation of G:C-->C:G transversions by removing guanine from the 8-oxoG:G mispair in E.coli.     

Complex genetic predisposition to cancer in an extended HNPCC family with an ancestral hMLH1 mutation

Hereditary non-polyposis colorectal cancer (HNPCC) is characterised by a genetic predisposition to develop colorectal cancer at an early age and, to a lesser degree, cancer of the endometrium, ovaries, urinary tract, and organs of the gastrointestinal tract other than the colon. In the majority of families the disease is linked to mutations in one of the two mismatch repair genes, hMSH2 or hMLH1. We have found a novel hMLH1 nonsense mutation in a Swiss family with Lynch syndrome, which has been transmitted through at least nine generations. A different tumour spectrum of neoplasms of the skin, soft palate, breast, duodenum, and pancreas was observed in three branches of this family, where there was a virtual absence of colonic tumours. The hMLH1 mutation could not be detected in members of these branches suggesting that at least a second genetic defect predisposing to cancer is segregating in part of the kindred.     

Detection of K-ras mutations in lung carcinomas: relationship to prognosis

The K-ras mutation is one of the most common genetic alterations found in human lung cancer. To evaluate the prognostic value of ras gene alterations in lung cancer in a U.S. population, we have screened 173 human lung tumors, which included 127 adenocarcinomas, 37 squamous carcinomas, and 9 adenosquamous carcinomas, for mutations in the K-ras gene using the combination of the PCR and denaturing gradient gel electrophoresis. Forty-three tumors contained K-ras mutations. Of these, 41 were identified among the adenocarcinomas (32%), 1 among the squamous carcinomas (2.7%), and 1 among the adenosquamous carcinomas (11%). Forty of these mutations were found in codon 12 and consisted of 24 G to T transversions, 12 G to A transitions, 2 G to C transversions, and 1 double GG to TT mutation. Two other G to T transversions were found in codon 13, and 1 A to C transversion was found in codon 61. The data showed that gender did not seem to affect the incidence and the types of the K-ras mutations or amino acid changes. Examination of the mutations in adenocarcinomas in relation to overall survival showed no difference in adenocarcinomas with K-ras mutations compared with K-ras-negative adenocarcinomas. However, the substitution of the wild-type GGT (glycine) at codon 12 with a GTT (valine) or a CGT (arginine) showed a strong trend (P = 0.07) toward a poorer prognosis compared with wild-type or other amino acid substitutions. Substitution of the wild-type glycine for aspartate (GAT) showed a strong trend (P = 0.06) for a better outcome than the valine or arginine substitution. Although these trends will require larger patient populations for verification, these data suggest that the prognostic significance of K-ras mutations may depend on the amino acid substitution in the p21(ras) protein.     

Identification of proteins that interact with a protein of interest: applications of the yeast two-hybrid system

Colorectal cancer remains a major health problem. Few therapies are effective apart from surgery, and survival has increased little in recent years. This is despite the fact that screening by colonoscopy can potentially remove nearly all colorectal tumours before they become malignant. Molecular genetics has identified some inherited mutations (such as at APC and the mismatch repair loci) that predispose to colon cancer and some somatic mutations (such as at APC and p53) that cause sporadic colon tumours. We review the likely role of these and other genes in colorectal tumorigenesis. We also highlight areas of relative ignorance in colon cancer and emphasise that many important genes, especially those that cause invasion and metastasis, remain to be identified. Colorectal cancer is, however, a well characterised tumour, as regards both its natural history and its histopathology; there are consequently good prospects for advances in colon cancer genetics, with probable benefits for its treatment. We anticipate: (a) that new genes predisposing to colon tumours, including those conferring relatively minor risks, will be characterised; (b) genes and proteins important in invasion and metastasis will be identified; (c) the network of protein interactions in which molecules such as APC are involved will be elucidated; (d) large-scale studies of somatic mutations in tumours will provide accurate predictions of prognosis and suggest optimal therapeutic regimens; and (e) new potential targets for therapy will be identified. Whilst molecular genetics is by no means sufficient for progress in preventing and treating colon cancer, it is a necessary and central part of such advances.     

Neurofibromatosis and early onset of cancers in hMLH1-deficient children

Hereditary nonpolyposis colon cancer is a common hereditary disorder caused by the germ-line mutations of DNA mismatch repair (MMR) genes, especially hMLH1 and hMSH2. We report here the first identification of human compounds with a homozygous inactivation of a MMR gene. In a typical hereditary nonpolyposis colon cancer family, MMR-deficient children conceived from matings between heterozygotes for a hMLH1 deleterious mutation exhibited clinical features of de novo neurofibromatosis type I and early onset of extracolonic cancers. This observation demonstrates that MMR deficiency is compatible with human development but may lead to mutations during embryogenesis. On the basis of clinical symptoms observed in MMR-deficient children, we speculate that the neurofibromatosis type 1 gene is a preferential target for such alterations.     

Mutation in the DNA mismatch repair gene homologue hMLH1 is associated with hereditary non-polyposis colon cancer

The human DNA mismatch repair gene homologue hMSH2, on chromosome 2p is involved in hereditary non-polyposis colon cancer (HNPCC). On the basis of linkage data, a second HNPCC locus was assigned to chromosome 3p21-23 (ref. 3). Here we report that a human gene encoding a protein, hMLH1 (human MutL homologue), homologous to the bacterial DNA mismatch repair protein MutL, is located on human chromosome 3p21.3-23. We propose that hMLH1 is the HNPCC gene located on 3p because of the similarity of the hMLH1 gene product to the yeast DNA mismatch repair protein, MLH1, the coincident location of the hMLH1 gene and the HNPCC locus on chromosome 3, and hMLH1 missense mutations in affected individuals from a chromosome 3-linked HNPCC family.     

Mutation of hMSH3 and hMSH6 mismatch repair genes in genetically unstable human colorectal and gastric carcinomas

Mutations within microsatellite sequences, consisting of additions or deletions of repeat units, are known as the replication/repair error positive (RER+) phenotype or micorsatellite instability (MI). Microsatellite instability has been demonstrated in hereditary and sporadic colorectal carcinomas and is usually observed in noncoding regions of genomic DNA. However, relatively few coding region targets of MI have been identified thus far. Using PCR, we amplified regions encompassing (A)8 and (C)8 microsatellite tracts within hMSH3 and hMSH6 from 31 RER+ sporadic colorectal tumors, 8 hereditary colon cancers, 23 RER+ gastric carcinomas, and 32 RER- gastric tumors. Mutations were found in 11 (36%) of 31 sporadic colon carcinomas, 4 (50%) of 8 hereditary colorectal cancers, and 5 (22%) of 23 RER+ gastric carcinomas, but in only 2 (6%) of 32 RER- gastric carcinomas. These frameshift mutations cause premature stop codons downstream that are predicted to abolish normal protein function. Our results and those of others suggest that DNA mismatch repair genes, such as hMSH3 and hMSH6, are targets for the mutagenic activity of upstream mismatch repair gene mutations and that this enhanced genomic instability may accelerate the accumulation of mutations in RER+ tumors.     

Mutation in the mismatch repair gene Msh6 causes cancer susceptibility

Mice carrying a null mutation in the mismatch repair gene Msh6 were generated by gene targeting. Cells that were homozygous for the mutation did not produce any detectable MSH6 protein, and extracts prepared from these cells were defective for repair of single nucleotide mismatches. Repair of 1, 2, and 4 nucleotide insertion/deletion mismatches was unaffected. Mice that were homozygous for the mutation had a reduced life span. The mice developed a spectrum of tumors, the most predominant of which were gastrointestinal tumors and B- as well as T-cell lymphomas. The tumors did not show any microsatellite instability. We conclude that MSH6 mutations, like those in some other members of the family of mismatch repair genes, lead to cancer susceptibility, and germline mutations in this gene may be associated with a cancer predisposition syndrome that does not show microsatellite instability.