Mutations in the human DNA mismatch repair gene MSH2 are associated with hereditary nonpolyposis colorectal cancer as well as a significant proportion of sporadic colorectal cancer. The inactivation of MSH2 results in the accumulation of somatic mutations in the genome of tumor cells and resistance to the genotoxic effects of a variety of chemotherapeutic agents. Here we show that the DNA repair and DNA damage-induced apoptosis functions of Msh2 can be uncoupled using mice that carry the G674A missense mutation in the conserved ATPase domain. As a consequence, although Msh2G674A homozygous mutant mice are highly tumor prone, the onset of tumorigenesis is delayed as compared with Msh2-null mice. In addition, tumors that carry the mutant allele remain responsive to treatment with a chemotherapeutic agent. Our results indicate that Msh2-mediated apoptosis is an important component of tumor suppression and that certain MSH2 missense mutations can cause mismatch repair deficiency while retaining the signaling functions that confer sensitivity to chemotherapeutic agents.
The DNA mismatch repair (MMR) system guards against genomic instability, and mutations in the human MMR genes MutS homolog 2 (MSH2) and MutL homolog 1 (MLH1) are the cause of the majority of hereditary nonpolyposis colorectal cancer [HNPCC (1)] . Recent studies indicate that MMR proteins not only protect mammalian genomes by repairing mismatched bases that result from erroneous DNA replication, but also by mediating DNA damage-induced apoptosis as part of the cellular response to endogenous and exogenous agents (2, 3, 4) . These studies showed that cell lines derived from HNPCC and MMR-defective sporadic tumors or MMR-deficient mice displayed increased mutation rates in their genomes and also had increased resistance to the genotoxic effects of a variety of DNA damage-inducing agents, including cisplatin, temozolomide and N-methyl-N′-nitro-N-nitrosoguanidine [MNNG (5, 6, 7, 8, 9, 10)] . In addition, as demonstrated initially in yeast and later in mammalian cells, MMR has been implicated in the removal of endogenous lesions such as mutagenic 8-oxoguanine that is incorporated from the oxidized deoxynucleotide triphosphate pool during DNA replication (11 , 12) . It has been suggested that the failure to clear DNA damage-bearing cells may be responsible in part for the increased mutation frequency in MMR mutant cells and also may confer a selective advantage in tumor cells (13, 14, 15, 16) . This hypothesis is consistent with the observation that MMR deficiency in mouse tissues leads to an elevation in mutation frequency after the mice are exposed to DNA-damaging agents (10 , 17) . These studies were performed with MMR-deficient cell lines that completely lack particular MMR proteins and therefore lack all of the functions of those proteins. However, a significant proportion of HNPCC patients carry missense mutations in MMR genes (1) , and it is unclear how these mutations affect individual MMR protein functions in DNA repair and damage responses.
We therefore decided to generate a mouse line carrying the Msh2G674A missense mutation to assess its impact on MMR and response to DNA damage and examine the consequences with respect to cancer susceptibility. The mutation results in a glycine to alanine change at amino acid residue 674 within the conserved ATPase domain at the COOH-terminal region. This domain is characterized by the Walker “type A” motif GXXXXGKS/T (G denotes the modified G674 amino acid residue) known to coordinate the phosphate groups of ATP in many proteins that hydrolyze ATP (18, 19, 20) . Mutations in this MutS domain in bacteria and yeast result in MMR defects, and overexpression of these mutant proteins was shown to cause dominant mutator phenotypes (21, 22, 23, 24) . The importance of ATP processing for MMR and tumorigenesis is underscored by the significant number of HNPCC missense mutations that are located in the ATP-binding domains of MSH2 (25) .
Here we show that the Msh2G674A mutation has differential effects on the DNA repair and DNA damage response functions. Whereas it caused DNA repair deficiency that resulted in a strong cancer predisposition phenotype in the mice, it did not affect the DNA damage response function of Msh2. As a consequence, tumorigenesis in Msh2G674A/G674A mice was delayed as compared with that in Msh2−/− mice. In addition, unlike Msh2−/− cells, Msh2G674A/G674A mouse embryonic fibroblasts (MEFs) and teratocarcinomas remained sensitive to treatment with genotoxic agents.
MATERIALS AND METHODS
Generation of Msh2G674A Mice.
A 3.6-kb HincII fragment containing Msh2 exon 13 was isolated from a 129SvEv bacterial artificial chromosome genomic library and subcloned. A mutation was introduced that changed codon 674 from glycine (GGT) to alanine (GCT) by site-directed mutagenesis (Stratagene Quick Change Kit). A 5.0-kb NotI fragment containing two LoxP sites flanking a neomycin-PGKhygromycin resistance cassette was subcloned into the single SpeI site. The modified HincII fragment was subsequently used to modify the Msh2 genomic locus in bacterial artificial chromosome clone mB183k13 of the RPCI-22 129 mouse genomic library by RecET-mediated recombination (26) . A 24-kb KpnI fragment containing the modified locus was excised from the bacterial artificial chromosome clone and used for gene targeting in WW6 embryonic stem (ES) cells (27) . Three correctly targeted ES cell lines were injected into C57BL/6J blastocysts. Male chimeras from all three lines were mated to C57BL/6J females and transmitted the mutant allele through their germ line. Subsequently, F1 males carrying the mutant allele were mated to Zp3Cre transgenic females (C57BL/6J) to remove the resistance cassette by LoxP-mediated recombination. Male and female mice carrying the modified allele were intercrossed to generate Msh2+/+, Msh2G674A/+, and Msh2G674A/G674A mutant mice.
Reverse Transcription-PCR Analysis.
Total RNA was isolated from Msh2 mutant ES cell lines using Trizol (GibcoBRL). Reverse transcription-PCR was performed with forward primer 5′-CGTAGAGCCAATGCAGACGCT-3′ and reverse primer 5′-GGATGGAAGAAGTCTCCAGC-3′ using the one Tube reverse transcription-PCR reaction kit (Roche) according to the manufacturer’s instructions. The following cycling conditions were used: 30 min at 50°C (1 cycle); 2 min at 94°C, 30 s at 60°C, and 45 s at 68°C (35 cycles); and 7 min at 68°C (1 cycle). The resulting 280-bp fragment was digested with either MnlI to detect the wild-type RNA transcript or AluI to detect the mutant RNA transcript.
Western Blot Analysis.
MEF cell extracts were prepared according to standard procedures, and 50 μg of protein of each cell lysate were separated on a 10% SDS-PAGE gel. Protein was transferred onto a PROTRAN membrane, and the membranes were subsequently incubated with mouse monoclonal antibodies directed against Msh2 (Ab-2; Oncogene), Msh6 (clone 44; BD Biosciences), and β-actin (C-2; Santa Cruz Biotechnology).
Gel Mobility Shift Assays.
Nuclear extracts were prepared as described previously (28) . The invariant sense oligonucleotide 5′-GGGAAGCTGCCAGGCCCCAGTGTCAGCCTCCTATGCTC-3′ was end-labeled with [γ-32P]ATP and annealed in 1× DNA binding buffer [12% glycerol, 20 mm HEPES (pH 7.9), 100 mm KCl, 1 mm DTT, and 5 mm MgCl2] with 3× molar ratios of antisense oligonucleotide 5′-GAGCATAGGAGGCTGACACTGGGGCCTGGCAGCTTCCC-3′ to form a GC homoduplex probe or with 3× molar ratios of antisense oligonucleotide 5′-GAGCATAGGAGGCTGACATTGGGGCCTGGCAGCTTCCC-3′ to form a GT mismatch-containing heteroduplex probe. Twenty μg of nuclear extract were preincubated in 1× DNA binding buffer, 1 μg of poly(dI-dC), and 20 ng of unlabeled homoduplex for 5 min on ice in a total volume of 19 μl. Ten ng of radiolabeled DNA probe were subsequently added, and the binding mixture was incubated on ice for 30 min. For reaction using cold probe competition, cold competitor was included in the preincubation mixture. For adenine nucleotide exchange experiments, ATP or ATP-γ-S was added 15 min after the addition of the DNA probe. The reaction mixture was then subjected to electrophoresis in a 6% polyacrylamide gel in 1× Tris-borate EDTA buffer. The gels were dried, and the percentage of relative binding of Msh2-Msh6 and Msh2G674A-Msh6 complexes to G/T oligonucleotide probe in the presence of increasing amounts of cold competitor or increasing ATP concentration was quantified using a STORM PhosphorImager with ImageQuant software (Molecular Dynamics) and calculated as [(Msh2-Msh6-probe complex/Msh2-Msh6-probe complex + free probe) × 100].
Cell-Free Extracts and MMR Assay.
The MMR proficiency of MES cell line cytosolic extracts was measured using M13mp2 DNA substrates and subsequent transfection of bacterial cells as described previously (29) . Repair efficiency is expressed (in percentage) as 100 × (1 − the ratio of the percentages of mixed bursts obtained from extract-treated and untreated samples). The substrates used are described in the Fig. 3 ⇓ legend.
Microsatellite Instability (MSI) Analysis.
Mutations in microsatellite sequences were assayed by PCR of single target molecules. Equal amounts of tail DNA isolated from 10 mice each of Msh2+/+, Msh2−/−, and Msh2G674A/G674A mouse strain were pooled and diluted to 0.5–1.5 genome equivalents. Cycling reactions for the three markers analyzed, U12235, D7Mit91, and D17Mit123, were performed as described previously (30) .
MEF Survival Analysis.
MEF cells (2 × 104) of each Msh2 genotype were seeded onto a single well of a 24-well plate in 10% FCS/DMEM. On the following day, the cells were exposed to cisplatin, 6-thioguanine, or MNNG at different drug concentrations for 24 h or for different time periods. After drug exposure, the cells were washed once with PBS, washed once with PBS:methanol (1:1), fixed in 0.5 ml of 100% methanol, and air dried. The cells were subsequently stained with 0.1% crystal violet and washed extensively with PBS, and the dye was extracted in 10% acetic acid. The dye concentration was determined by measuring absorption at A600 nm, and the percentage of cell survival was calculated as (treated cells/untreated cells × 100). The experiments were performed for three different MEF strains for each Msh2 genotype and repeated at least three times for each strain. Cisplatin (Bedford Laboratories), 6-thioguanine (Sigma), and MNNG (Sigma) dilutions in culture medium were prepared fresh each time before use. For exposure to MNNG, 20 μm O6-benzylguanine (Sigma) was added to the medium.
Msh2+/+, Msh2−/−, and Msh2G674A/G674A ES cells (2 × 106) were implanted into BALB/c nu/nu mice by s.c. injection into the flank regions. A single dose of cisplatin (10 μg·g−1 body weight) was administered i.p. 48 h after tumor implantation in half of the mice. The remainder of the mice received a single dose of saline as untreated control. Tumor growth at the inoculation sites was monitored daily by measuring the tumor size using a Vernier caliper, and tumor volume was calculated [volume = length × (width)2/2]. The values for each time point were calculated as the mean of 10 replicates.
Generation of Msh2G674A Mutant Mice.
The mutant mouse line was created by a knockin gene targeting strategy (Fig. 1A) ⇓ . Analysis of 22 litters of the F2 offspring showed that the Msh2G674A allele was transmitted in a normal Mendelian ratio with 40 Msh2+/+, 89 Msh2G674A/+, and 37 Msh2G674A/G674A. None of the heterozygous or homozygous animals displayed any developmental abnormalities. Molecular analysis showed that the Msh2G674A mutation allowed normal gene expression and did not interfere with the stability of the mutant protein (Fig. 1, C and D) ⇓ . In eukaryotes, MSH2 forms complexes with either MSH6 or MSH3 to initiate the repair of single base mutations (MSH2-MSH6) or larger insertion/deletion mutations (MSH2-MSH3) (31, 32, 33, 34, 35, 36, 37) . The formation of these complexes is important for the stability of the MutS proteins, and immunohistochemical analysis in tumor cells showed that the loss of MSH2 is frequently associated with the loss of MSH6 (38) . Western blot analysis of cell extracts derived from Msh2G674A/G674A mice showed that the mutation did not alter the stability of either mutant Msh2G674A or Msh6 protein in the cells (Fig. 1D) ⇓ . In addition, immunohistochemical analysis indicated that the subcellular distribution of the mutant Msh2G674A or the Msh6 protein was not affected (data not shown).
Mismatch Binding Activity in Msh2G674A/G674A Cell Extracts.
We next studied the mismatch binding activities of nuclear extracts isolated from Msh2+/+, Msh2−/−, and Msh2G674A/G674A ES cells. Using gel mobility shift assays, we did not detect any significant differences between Msh2+/+ and Msh2G674A/G674A extracts in their DNA binding affinity using an oligonucleotide substrate containing a G/T mismatch. In addition, proteins in both extracts bound with similar, albeit lower, affinity to a homoduplex oligonucleotide substrate, whereas Msh2−/− extracts did not show any binding activity (Fig. 2A ⇓ ; data not shown). In contrast to the Msh2+/+ extracts, the mutant Msh2G674A/G674A extracts were partially resistant to ATP-dependent mismatch release, even at concentrations that exceed normal physiological conditions (Fig. 2B) ⇓ . The addition of the poorly hydrolyzable ATP-γ-S analog also resulted in mismatch release in Msh2+/+ extracts but not in the Msh2G674A/G674A extracts (data not shown). These results are consistent with previous studies in yeast and suggest that the resistance of Msh2G674A/G674A cell extracts to ATP-dependent release from mismatched DNA is caused by defective or altered ATP binding resulting from the substitution of alanine for glycine in the P-loop (24) .
MMR Deficiency in Msh2G674A/G674A Cells.
To test the impact of the Msh2G674A mutation on DNA MMR, we measured the repair activity in ES cell extracts using substrates containing G·G mismatches, single-base insertion/deletion mismatches, or 2-base insertion/deletion mismatches with a nick either 3′ or 5′ to the mismatched base (Fig. 3) ⇓ . Whereas both Msh2+/+ and heterozygous Msh2G674A/+ extracts repaired all of these substrates, extracts prepared from homozygous Msh2G674A/G674A cells did not. The repair defect in the Msh2G674A/G674A extracts was comparable with the defect that was observed in Msh2−/− extracts.
MSI in Msh2G674A/G674A Mice.
We assessed the in vivo mutator phenotype in the Msh2G674A/G674A mice by analyzing MSI in tail genomic DNA. We found that at the dinucleotide marker D7Mit91, 28% (38 of 134) of alleles tested were unstable in Msh2G674A/G674A mice; in contrast, 9% (11 of 118) of alleles in Msh2+/+ genomes were unstable. Similarly, 20% (24 of 123) of the alleles at the mononucleotide marker U12235 in Msh2G674A/G674A mice were unstable, compared with 3% (7 of 244) unstable alleles that were found previously in Msh2+/+ animals (39) . This analysis indicated that the genomes of Msh2G674A/G674A mice displayed a highly significant increase in mutation frequency at these two markers (P < 0.0001 for D7Mit91 and U12235, Msh2G674A/G674A compared with Msh2+/+). Furthermore, the MSI at these loci in the Msh2G674A/G674A animals was comparable with the MSI observed in Msh2−/− mice [D7Mit91, 27% (36 of 132); U12235, 21% (19 of 92)]. These results indicate that the Msh2G674A mutation impairs the repair function of the protein and significantly increases the mutator phenotype in the genomes of the mutant mice.
Survival and Cancer Susceptibility in Msh2G674A Mutant Mice.
When cohorts of Msh2G674A/G674A, Msh2G674A/+, and Msh2+/+ mice were followed for a period of 12 months, the overall survival and cancer susceptibility of the Msh2G674A/G674A mice were clearly affected. None of the Msh2G674A/G674A mice died in the first 3 months of life, and by 6 months, >90% of Msh2G674A/G674A mice were still alive (Fig. 4) ⇓ . By 9 months of age, >60% of Msh2G674A/G674A mice were alive; however, the number of surviving animals declined rapidly in the next 3 months, and all of the remaining mice died by 12 months of age. Only one Msh2G674A/+ and none of the Msh2+/+ mice died during the same period of time. The reduced survival in the Msh2G674A/G674A mutant mice was caused by an increase in cancer predisposition. Most of the animals that died and were available for analysis had developed non-Hodgkin’s lymphomas (10 of 16 animals, 63%) between the ages of 9 and 12 months (Supplementary Table 1). A smaller number of mice between the ages of 7 and 10 months (3 of 16 animals, 19%) developed gastrointestinal adenocarcinomas. One animal at 9 months of age developed a squamous basal cell carcinoma of the skin (1 of 16 animals, 6%). In 2 animals that were 10 months of age, no obvious tumors could be detected (2 of 16 animals, 12%). Although the tumor spectrum in the Msh2G674A/G674A mice resembled that seen in previously studied Msh2−/− mouse strains (40, 41, 42) , we noted a striking difference in survival. Whereas the 50% survival of Msh2−/− mice on various mixed genetic backgrounds was reported at approximately 6 months of age, it took between 9 and 10 months for 50% of the Msh2G674A/G674A mice to die. To directly compare the survival between the Msh2G674A/G674A and Msh2−/− lines on a similar genetic background, we generated a cohort of Msh2−/− and Msh2G674A/G674A mice that were backcrossed several times onto the C57BL/6 background. This comparison confirmed that the difference in survival between the Msh2−/− and Msh2G674A/G674A mice was highly significant (Fig. 4 ⇓ ; P = 0.001, log-rank test).
DNA Damage Response in Msh2G674A/G674A MEF Cells.
The difference in survival between the Msh2−/− and Msh2G674A/G674A mice suggested that the mutant protein retained some function important for tumor suppression. In recent years, several studies demonstrated that MSH2-deficient human colorectal cancer cell lines as well as mouse embryonic fibroblast lines have an increased resistance to treatment with a variety of DNA-damaging agents including cisplatin, MNNG, and 6-thioguanine. It was proposed that the resistance to DNA damage-induced apoptosis in MMR-deficient cancer cells might provide a selective advantage in the initial stages of tumorigenesis (16) . We therefore analyzed the genotoxic response to treatment with DNA-damaging agents in Msh2+/+, Msh2−/−, and Msh2G674A/G674A MEF lines. Consistent with previous results, Msh2−/− MEF cells were largely resistant to treatment with cisplatin at the drug levels tested (Fig. 5, A and B) ⇓ . In contrast, both Msh2+/+ and Msh2G674A/G674A MEF lines were sensitive to cisplatin exposure. The differences in sensitivity between the Msh2+/+ or Msh2G674A/G674A cells and Msh2−/− cells were highly significant (P ≤ 0.0003). The Msh2−/− cells also displayed increased resistance to treatment with 6-thioguanine and, to a lesser extent, to treatment with MNNG, whereas Msh2+/+ and Msh2G674A/G674A cells showed higher sensitivity at the same drug concentrations (Supplementary Fig. 1 ⇓ ). The cisplatin sensitivity in Msh2+/+ and Msh2G674A/G674A cells was associated with significant increases in apoptosis as assessed by terminal deoxynucleotidyl transferase-mediated nick end labeling assay (Fig. 5C ⇓ ; P < 0.0001 for both Msh2+/+ and Msh2G674A/G674A compared with untreated cells). In contrast, no significant increase in the number of apoptotic cells was seen in Msh2−/− cells when compared with untreated cells. Interestingly, there was also a significant increase in the number of terminal deoxynucleotidyl transferase-mediated nick end labeling-positive cells in the untreated Msh2G674A/G674A MEF cultures compared with the untreated Msh2+/+ or Msh2−/− cell cultures (Ps < 0.001).
Cisplatin Sensitivity in Msh2G674A/G674A Teratocarcinomas.
In the next set of experiments, we studied the in vivo cisplatin sensitivity of Msh2 mutant teratocarcinomas in athymic nude mice. ES cells have the capacity to develop into teratocarcinomas when injected into immunodeficient athymic nude mice and were shown to provide a suitable model system to study drug sensitivity. In addition, the use of Msh2−/− ES cells as xenografts in athymic nude mice demonstrated a major impairment in the cisplatin responsiveness of the tumor in vivo (43 , 44) . Because the results in the Msh2G674A/G674A MEF lines indicated that the Msh2G674A mutation would behave differently in this model system, we studied tumor growth in nude mice injected with Msh2+/+, Msh2−/−, or Msh2G674A/G674A ES cells and measured their responsiveness to cisplatin treatment. We found that without cisplatin treatment, ES cells of all three Msh2 genotypes rapidly developed into teratocarcinomas after implantation (Fig. 6, A–C ⇓ ). However, treatment with cisplatin had different effects on tumor growth in the three cell lines. Whereas Msh2−/− cells did not respond to cisplatin treatment, and tumor growth occurred at similar rates in cisplatin-treated and untreated mice (Fig. 6B) ⇓ , the growth of both Msh2+/+ and Msh2G674A/G674A tumors was significantly suppressed after cisplatin treatment (Fig. 6, A and C) ⇓ . These results are consistent with the observations made in the MEF cells and demonstrate that although the Msh2G674A mutation has a severe impact on DNA mismatch repair, it still confers a wild-type-specific response to cisplatin-induced DNA damage. In this regard, Msh2G674A is a separation of function mutation.
We generated a mouse line that carries the Msh2G674A missense mutation and studied the consequences on individual MMR functions and cancer susceptibility. The analysis of Msh2G674A mice showed that this mutation results in MMR deficiency and increased cancer susceptibility in homozygous mutant mice. In contrast, DNA mismatch repair was not significantly impaired in heterozygous mutant mice, indicating that the mutation does not act in a dominant manner. Our analysis of this mouse line demonstrates that the roles of MMR in the prevention of DNA replication errors and DNA damage-induced apoptosis can be separated by Msh2 missense mutations, and both functions are important for tumor suppression. Although, the Msh2G674A/G674A mutant mice show a strong cancer predisposition phenotype with a tumor spectrum similar to that of Msh2−/− mice, the extended survival in these animals in the first 9 months of life is consistent with the idea that the loss of the DNA damage-induced apoptotic response could provide a selective advantage for tumor cells in the initial stages of tumorigenesis (45 , 46) . We also observed a rapid decline in survival of the Msh2G674A/G674A animals between 9 and 12 months of age, which might be explained by the eventual accumulation of genomic mutations in cells that are not cleared by apoptosis. These cells could then accelerate tumorigenesis in the older Msh2G674A/G674A mice, once the initial barrier to tumorigenesis is overcome. Our results indicate that the increased mutation rates caused by MMR deficiency are sufficient to drive tumorigenesis and that it is the combination of increase in mutation rates and defective apoptosis that cooperates to result in tumorigenesis. Our results also demonstrate that the DNA damage-induced apoptosis function of Msh2 can delay but not prevent tumorigenesis.
The DNA repair defect in the Msh2G674A/G674A mutant mice is consistent with previous studies in bacteria and yeast and indicates that the ATPase domain is essential for the activation of the repair processes that facilitate the removal of mismatched bases (21, 22, 23, 24) . However, in contrast to the Msh2-null allele, the Msh2G674A mutation did not significantly affect the cellular response to DNA damage-inducing agents, indicating that normal ATP processing with subsequent repair is not essential for the apoptosis signaling function of Msh2. Different models have been developed in the past to explain the resistance of MMR-deficient cell lines to DNA-damaging agents. One model suggested that DNA repair-competent cells engage in futile repair cycles after treatment with alkylating agents because MMR is a strand-specific mechanism and is always directed to the newly synthesized strand (3 , 5) . Because DNA adducts in the template strand cannot be removed, the MMR reaction is continuously initiated upon repair synthesis, leading ultimately to the formation of double strand breaks that provide a signal for apoptosis. Alternatively, it was proposed that the binding of MSH2-MSH6 and also MLH1-PMS2 complexes to damaged bases at the replication fork could block DNA replication or other processes such as transcription and repair, leading to cell cycle arrest and cell death (47) . The molecular analysis of Msh2G674A/G674A cells shows that although the mutant Msh2G674A protein is capable of mismatch binding, it does not allow normal MMR to proceed and therefore supports the latter notion. Our results are consistent with the idea that MMR components can function as sensors for genetic damage (16 , 48) and are also in agreement with a recent model by Brown et al. (49) , which proposes that MSH2-MLH1 complexes act as molecular scaffolds that physically link downstream effectors involved in DNA damage response pathways such as the ATM (ataxia telangietasia mutated) gene product and checkpoint kinase 2 (CHK2). In this model, MSH2-bound CHK2 and MLH1-bound ATM complexes interact at the sites of DNA damage, resulting in the phosphorylation of CHK2 by ATM and the subsequent activation of the S-phase checkpoint and apoptotic pathways. Furthermore, MMR-mediated apoptosis appears to be activated through p53-dependent and p53-independent pathways and also involves the activation of p73 (9 , 50 , 51) . In contrast to MMR-deficient cell lines, which display variable defects in the induction of p53 and p73, Msh2G674A/G674A MEF cells have normal induction of both proteins, indicating that normal Msh2 ATPase activity is not required for this response (data not shown). The presence of mutant Msh2G674A protein in Msh2G674A/G674A ES cells might allow the formation of mutant Msh-Mlh complexes that are capable of signaling cell cycle arrest and apoptosis and provide an explanation for the increased number of apoptotic cells that we observed in the untreated Msh2G674A/G674A MEF cultures. In these cells, the DNA repair defect caused by the Msh2G674A mutation prevents the removal of misincorporated bases or oxidized bases such as 8-oxoguanine; however, it does not interfere with the recognition and binding of such lesions. The persistence of these DNA lesions in the genome and their continuous recognition by the mutant MMR complexes may in turn result in checkpoint activation and increased apoptosis. This hypothesis is supported by preliminary studies that revealed decreased cell proliferation and alterations in the cell cycle in Msh2G674A/G674A MEF cells. 5 The availability of Msh2G674A mutant mice will aid in future studies to investigate the role of Msh2 in these processes.
Our results also suggest that a subset of tumors that carry MSH2 missense mutations will remain responsive to treatment with chemotherapeutic agents, a finding that may have important implications for the treatment of HNPCC patients. Different MSH2 missense mutations will likely have varied effects on DNA repair and apoptosis. Therefore, determining the genotype/phenotype correlations of MMR point mutations in HNPCC patients may provide valuable information for treatment and prognosis.
We thank Drs. Lisa Edelmann and Matthew Scharff for critical reading of the manuscript.
Grant support: NIH Grants CA76329 and CA93484 (to W. E.), CA84301 and ES11040 (to R. K.), CN05117 (to M. L.) and Center Grant CA13330 (to the Albert Einstein College of Medicine); a Deutsche Krebshilfe fellowship (to S. J. S.); and an Irma T. Hirschl Career Scientist Award (to W. E.).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Notes: D. P. Lin, Y. Wang, and S. J. Scherer contributed equally to this work. Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org).
Requests for reprints: Winfried Edelmann, Department of Cell Biology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461. Phone: (718) 430-2030; Fax: (718) 430-8574; E-mail:
↵5 S. J. Scherer and W. Edelmann, unpublished observations.
- Received September 18, 2003.
- Revision received October 20, 2003.
- Accepted November 5, 2003.
- ©2004 American Association for Cancer Research.