Requirement for Msh6, but Not for Swi4 (Msh3), in Msh2-Dependent Repair of Base-Base Mismatches and Mononucleotide Loops in Schizosaccharomyces pombe

Corresponding author: Oliver Fleck, Institute of Cell Biology, University of Bern, Baltzer-Strasse 4, CH-3012 Bern, Switzerland., fleck{at}izb.unibe.ch (E-mail)

Communicating editor: M. LICHTEN

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The msh6 mismatch repair gene of Schizosaccharomyces pombe was cloned, sequenced, and inactivated. Strains bearing all combinations of inactivated msh6, msh2, and swi4 (the S. pombe MSH3 ortholog) alleles were tested for their defects in mitotic and meiotic mismatch repair. Mitotic mutation rates were similarly increased in msh6 and msh2 mutants, both for reversion of a base-base substitution as well as of an insertion of one nucleotide in a mononucleotide run. Tetrad analysis and intragenic two-factor crosses revealed that meiotic mismatch repair was affected in msh6 to the same extent as in msh2 background. In contrast, loss of Swi4 likely did not cause a defect in mismatch repair, but rather resulted in reduced recombination frequency. Consistently, a mutated swi4 caused a two- to threefold reduction of recombinants in intergenic crosses, while msh2 and msh6 mutants were not significantly different from wild type. In summary, our study showed that Msh6 plays the same important role as Msh2 in the major mismatch repair pathway of S. pombe, while Swi4 rather functions in recombination.

IN spite of the high fidelity of DNA polymerase and its 3' exonuclease activity during DNA replication, some base-base mispairs and insertion-deletion loops (IDL) remain (UMAR and KUNKEL 1996 ; JIRICNY 1998 ). In Escherichia coli, such uncorrected replication errors are mainly removed by the MutHLS system. The first step of the repair process consists of mismatch recognition by the MutS homodimeric protein (MODRICH 1991 ; KOLODNER 1996 ; ALLEN et al. 1997 ). In eukaryotes, several MutS and MutL homologues are involved in mismatch repair. Proteins homologous to MutS are generally called Msh (MutS homologue) and proteins homologous to MutL are termed Mlh (MutL homologue) or Pms (postmeiotic segregation). PMS reflects the frequent failure to repair mismatches formed during meiotic recombination and was originally described in Saccharomyces cerevisiae pms1 mutants (WILLIAMSON et al. 1985 ). The binding to mismatches appears to be mediated by heterodimers of two Msh proteins (ACHARYA et al. 1996 ; MARSISCHKY et al. 1996 ; FISHEL and WILSON 1997 ; GRADIA et al. 1997 ; JIRICNY 1998 ; BUERMEYER et al. 1999 ; KOLODNER and MARSISCHKY 1999 ; MARRA and SCHAR 1999 ). From S. cerevisiae to human, up to six Msh proteins (Msh1–6) were discovered, but only three of them (Msh2, Msh3, and Msh6) are involved in nuclear mismatch repair (REENAN and KOLODNER 1992 ; NEW et al. 1993 ; MARSISCHKY et al. 1996 ; FISHEL and WILSON 1997 ). Mutations in several human genes homologous to MutS (MSH2 and MSH6) and to MutL (MLH1, PMS1, and PMS2) were found to be responsible for hereditary nonpolyposis colon cancer (HNPCC; LEACH et al. 1993 ; MIYAKI et al. 1997 ; PROLLA 1998 ; BUERMEYER et al. 1999 ; PLANCK et al. 1999 ).

In S. cerevisiae, msh2 single mutants show a strong mutator phenotype, while the msh3 and msh6 single mutants have much weaker effects on mutation rates. However, the msh3 msh6 double mutant revealed a mutator phenotype that is as strong as that of msh2. These observations led MARSISCHKY et al. 1996 to propose that two types of heterodimers are involved in mismatch recognition in S. cerevisiae: Msh2-Msh3 and Msh2-Msh6. Further studies showed that Msh2-Msh3 heterodimers bind specifically to IDLs, while Msh2-Msh6 heterodimers can bind to small IDLs of one or two unpaired nucleotides and to single-base mispairs (ACHARYA et al. 1996 ; JOHNSON et al. 1996 ; KOLODNER 1996 ; FISHEL and WILSON 1997 ; BUERMEYER et al. 1999 ; KOLODNER and MARSISCHKY 1999 ).

So far, in the fission yeast Schizosaccharomyces pombe, swi4⁺, pms1⁺, and msh2⁺ were cloned, inactivated, and studied (FLECK et al. 1992 ; SCHAR et al. 1997 ; RUDOLPH et al. 1999 ). The msh2⁺ gene product is implicated in mismatch repair and mating-type switching (FLECK et al. 1994 ; RUDOLPH et al. 1999 ). The swi4⁺ gene encodes an amino acid sequence mostly homologous to the proteins of the Msh3 subfamily and is, like msh2⁺, involved in mating-type switching (FLECK et al. 1992 ; NEW et al. 1993 ). Its inactivation produces no or only a weak mutator phenotype, as is the case for msh3 mutants in budding yeast (NEW et al. 1993 ; MARSISCHKY et al. 1996 ; EARLEY and CROUSE 1998 ). In this context it was interesting to isolate the msh6⁺ ortholog of S. pombe, to study its function in mismatch repair, and to compare the defects caused by mutated msh2, msh6, and/or swi4. We determined reversion rates of a base substitution allele and of a thymine insertion in a homonucleotide run of five thymines. In addition, meiotic mismatch repair was tested by intragenic two-factor crosses and tetrad analysis. The results are discussed in comparison with the situation in S. cerevisiae and other species.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains and media:
All strains containing the ade6 mutations 51, 421, 485, 687, M387, and M26 are from the S. pombe collection in Bern. The molecular nature of the mutations was determined by SZANKASI et al. 1988 and SCHAR and KOHLI 1993 . Strains bearing the disrupted msh6 gene were derived from CT-1 (h⁺ msh6::arg3⁺ arg3-D4 his3-D1 ura4-D18, see below). Strains with disrupted msh2 were derived from crosses with strain Ru39 (h^- msh2::his3⁺ his3-D1; RUDOLPH et al. 1999 ). Strains with disrupted swi4 were derived from strain Ru106 (h^- swi4::ura4⁺ ura4-D18), which originated from LH110 (FLECK et al. 1992 ). S. pombe media YEA (yeast extract agar), YEL (yeast extract liquid), MMA (minimal medium agar), and MEA (malt extract agar, sporulation medium) were prepared according to GUTZ et al. 1974 and SCHAR et al. 1997 . Where required, supplements (adenine, histidine, uracil, leucine, arginine, and lysine) were added (0.01% w/v).

PCR gene cloning and physical mapping:
Cloning of the S. pombe msh6⁺ gene was started by PCR using degenerate primers derived from amino acid sequences highly conserved in the MutS family: TGPNM and DELGR. From a set of primers, the use of G1, 5'-CTGGATCCACNGGNCCN(A/C)ATATG-3', and G6, 5'-CTGGATCCCTN(T/G)NCC(T/C)A(A/G)(C/T)T(C/A)GTC-3', allowed successful amplification. We initiated the program by a 4-min denaturation at 94°, 5 cycles of 30 sec at 94°, 30 sec at 45°, 30 sec at 72°, followed by 25 cycles of 20 sec at 94°, 30 sec at 50°, 30 sec at 72°. PCR products were blunt ended by treatment with the large fragment of E. coli DNA polymerase I and cloned into the SmaI site of pBluescript(+). Cloned PCR products were sequenced using a sequencing kit (Amersham, Pharmacia, Dübendorf, Switzerland).

Isolation of the msh6⁺ gene:
A putative 270-bp-long msh6 PCR product was used to screen an ordered S. pombe cosmid library by hybridization (HOHEISEL et al. 1993 ). Fourteen cosmids were isolated that all mapped on the right arm of chromosome III between the adh1 and ade5 loci. Southern blot analysis allowed us to draw a restriction map and to choose one of the cosmids (ICRFc60B039D) for sequencing. Sequencing of this cosmid was performed with PRISM Ready Reaction AmpliTaqFs DyeDeoxyterminator (Perkin Elmer, Norwalk, CT) and sequencing reactions were run on an AbiPRISM sequencer. Primers were first chosen in the 270-bp starting fragment and then 11 forward and 11 reverse primers were designed to complete the 4248-bp sequence containing the entire msh6⁺ gene (accession no. AF207839).

msh6 gene disruption:
A 2.1-kb fragment of the msh6⁺ coding region was amplified by PCR with primers G7fwd (5'-GCATCTCCCACAAAATCGAA-3', nucleotides (nt) 1261–1280) and G2rev (5'-GCTTGAACTGGCTGAAAAAGC-3', nt 3215–3236) and cloned in pUAg (R&D Systems, Abingdon, UK) to give pUAmsh6. The S. pombe arg3⁺ marker (WADDELL and JENKINS 1995 ) was introduced as a blunt-ended fragment into the HpaI site of pUAmsh6, thus interrupting the msh6 gene at nt 1595 (start codon at nt 225). The S. pombe strain h⁺ arg3-D4 his3-D1 ura4-D18 was transformed (ITO et al. 1983 ) with gel-purified 3.9-kb BamHI-XbaI fragment derived from pUAmsh6 and Arg⁺ transformants were selected on MMA supplemented with histidine and uracil. Subsequently, several clones were screened by PCR (Fig 1) with primers 3'arg, 5'- CAACACTCATAAGGAAAAAGC-3' and G1rev2, 5'-ATACAAGGATGTCTCAACTCC-3'. 3'arg derived from the 3' end of the arg3⁺ gene (nt 1791–1811 according to the sequence deposited in the EMBL database, accession no. X63577), and G1rev2 derived from msh6⁺, outside the transformed fragment (nt 3309–3329). Positive clones were tested by Southern hybridization (data not shown) and one was chosen for further studies (CT-1). The structure of the interrupted msh6 gene is shown in Fig 1.

View larger version (26K): In this window In a new window Download PPT slide   Figure 1. Construction of the msh6::arg3+ disruption strain. (A) The arg3+ gene (solid box) was cloned into the HpaI site of plasmid pUAgmsh6, containing a fragment of msh6 (shaded box), previously amplified by PCR with the primers G7fwd and G2rev. (B) The resulting msh6::arg3+ cassette was obtained by digestion with BamHI/XbaI, transformed into S. pombe, and integrated via homologous recombination at the msh6+ locus. (C) Structure of the resulting msh6::arg3+ locus and restriction map. Arg+ transformants were checked for correct disruption by PCR with primers 3'arg and G1rev2 and by Southern blot analysis using restriction enzymes BglII, ClaI, NruI, and ScaI. The probe used for hybridization derived from the 3' region of msh6 (black bar). Restriction sites are as follows: B, BglII; C, ClaI; H, HpaI; N, NruI; S, ScaI.

Genetic tests for mismatch repair:
Fluctuation tests and tetrad analysis were performed according to RUDOLPH et al. 1999 and SCHAR et al. 1997 , respectively. Tetrads were dissected from the cross msh6 ade6-M26 x msh6 ade6⁺. This cross included strains homozygous for the msh6 disruption and heterozygous for the recombination hot spot ade6-M26 (GUTZ 1971 ; SZANKASI et al. 1988 ; SCHUCHERT et al. 1991 ; SCHAR and KOHLI 1994 ). G/A and T/C mismatches are formed when the mutated site is included in heteroduplex DNA. Whole chromatid conversions (WCC) and PMS were counted by examination of the white and red colony color of the spore-derived colonies. ade6 mutants accumulate a red pigment when grown on YEA containing a limited amount of adenine, while ade6⁺ strains form white colonies. Intragenic two-factor crosses with ade6 mutants were performed to study meiotic mismatch repair as described (SCHAR and KOHLI 1993 ; RUDOLPH et al. 1998 ). We included the crosses 421 x 51, 485 x 51, and 485 x M387 (Fig 3). All types of crosses were performed at least three times.

View larger version (146K): In this window In a new window Download PPT slide   Figure 2. Comparison of human (h), S. cerevisiae (Sc), and S. pombe (Sp) homologues belonging to the eukaryotic Msh6 subfamily of bacterial MutS. Sequences were retrieved from GenBank and were aligned using the Pile-Up program (GCG software package, version 10.1, University of Wisconsin) and the Boxshade program (http://www.ch.embnet.org/software/BOX_form.html). Amino acids at a given position are shown in black when at least two residues are identical and in gray when at least two residues are similar. The position of the intron in the S. pombe Msh6 encoding gene between residues K198 and P199 is marked by an arrow. The S. pombe msh6 sequence has been deposited in the EMBL database under accession no. AF207839.

View larger version (21K): In this window In a new window Download PPT slide   Figure 3. Formation and repair of mismatches in intragenic two-factor crosses. The ade6 mutants that were crossed are shown on the left. When both mutated sites are included in heteroduplex DNA during recombination, two mismatches are formed in the same heteroduplex. Two mismatch repair systems can correct these mismatches. The MMR system recognizes all mismatches except C/C and has long excision tracts, leading to frequent cocorrection of both mismatches (SCHAR and KOHLI 1993 ; RUDOLPH et al. 1998 ). The minor system, requiring nucleotide-excision repair factors, corrects C/C mismatches and has short excision tracts (FLECK et al. 1999 ). MMR-mediated repair is indicated by long double arrows. Repair by the minor system, which is accompanied by short excision tracts, is indicated by short double arrows (only processing of C/C mismatches is shown). Prototrophic recombinants (right side) can arise when the underlined bases are replaced. (A) Cross 421 x 51. Only non-C/C mismatches can be formed (distance of 90 bp). These are frequently corepaired by the MMR system. (B) Cross 485 x 51. Heteroduplexes with either a C/C and a T/G, or a G/G and a C/A can be produced (distance of 22 bp). The C/C can be independently repaired by the minor system or is corepaired with the T/G mismatch by the MMR pathway. (C) Cross 485 x M387. Heteroduplexes can contain either two C/C or two G/G mismatches (distance of 26 bp). The C/C mismatches are not a substrate of MMR but can be often independently repaired by the minor system.

Intergenic two-factor crosses:
Recombination rates were measured between the intervals leu2-120 x lys7-2 (chromosome I) and ade1-40 x lys4-95 (chromosome II). Parental strains were mixed on MEA and incubated for 2–3 days at 25°. Vegetative cells were killed by treatment with a 1:500 (v/v) dilution of snail enzyme (Helix pomatia juice, Biosepra, Villeneuve-la-Garenne, France) and spore suspensions were plated on nonselective YEA medium. After 4–5 days of growth at 30°, colonies were checked for their genotypes by replica plating on different types of MMA with respect to the included supplements.

Iodine staining:
Colonies grown on MEA at 25° were treated with iodine vapor, which stains spores but not vegetative cells. Colonies from homothallic strains turned homogeneously brown, while heterothallic (nonsporulating) strains are iodine negative. Homothallic strains with a defect in mating-type switching have reduced switching efficiency and form mottled colonies (EGEL et al. 1984 ).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identification of the S. pombe msh6⁺ gene:
The aim of this work was to study the function of the S. pombe msh6⁺ gene in mismatch repair. We were particularly interested to learn about the roles of S. pombe msh6⁺ and swi4⁺ (the S. pombe MSH3 ortholog) in mismatch repair and whether their functions are similar or different to those devoted to MSH6 and MSH3 of other organisms. swi4 is known to be involved in mating-type switching, a special mitotic recombination event (FLECK et al. 1992 ), but an involvement in mismatch repair could not be demonstrated so far. Conservation of the MutLS system among species and the discovery of two members of the MSH family in S. pombe, namely Msh2 and Swi4, suggested that a Msh6 ortholog exists in S. pombe too. We cloned the msh6⁺ gene as described in MATERIALS AND METHODS. It is contained in cosmid ICRFc60B039D, which maps between adh1 and ade5 on chromosome III (HOHEISEL et al. 1993 ). Sequencing revealed a short open reading frame (ORF) starting at nt 225 and ending at nt 818 and a long ORF starting at nt 935 and ending at nt 4101 (the 4248-bp sequence containing msh6⁺ has been deposited in the EMBL database under accession no. AF207839). The following features suggested the existence of a 117-nt-long intron between the two ORFs (nt 819 to 934). First, 5' and 3' splice sites (GTATGG and TAG, respectively) at the ends of the putative intron and a branch site (CTAAC) close to the 3' splice site are present, which correspond to respective consensus sequences (PRABHALA et al. 1992 ). Second, the peptide sequences deduced from the nucleotide sequences 5' and 3' to the intron show homology to both bacterial MutS and eukaryotic MSH proteins. The deduced amino acid sequence of S. pombe msh6⁺ is 35–42%, identical with MSH6 orthologs of other organisms (highest homology to S. cerevisiae Msh6). It is 26% identical with MSH4 and MSH5 proteins, 28% identical with MSH2 proteins, and 27–32% identical with Msh3 proteins. The fission yeast Msh6 amino acid sequence contains the general MSH motifs (Fig 2). It has an extended N-terminal region, which is common with Msh6 proteins of other organisms. In addition, Msh6 contains several motifs, which are apparently specific to the MSH6 subgroup. Computer analysis of the msh6⁺ sequence revealed a putative GAL1-TATA site starting at nt 121 (ATG at nt 225) and a CTCATGTT sequence at nt 145, which might serve as a transcription initiation site (Sigscan program, GCG software package, University of Wisconsin).

A major role of the S. pombe msh6⁺ gene in mitotic mismatch repair:
To study the phenotypes caused by mutated msh6 in S. pombe, we inactivated this gene by gene targeting using the arg3⁺ gene as selection marker (Fig 1). A mutator phenotype of msh6 mutants was noted by mere visual inspection. ade6 mutants form red colonies on YEA medium (GUTZ et al. 1974 ; RUDOLPH et al. 1999 ). In mismatch repair-proficient ade6 strains the red color is quite homogeneous. In contrast, msh6::arg3⁺ ade6 colonies developed white sectors within 1–2 weeks of growth, indicative of the occurrence of mutations in ade genes acting upstream of ade6 in the purine biosynthesis pathway. A similar observation had been made with msh2 mutants but not with swi4 mutants (FLECK et al. 1994 ; RUDOLPH et al. 1999 ).

To obtain quantitative values of mutation rates we measured reversions to Ade⁺ of two defined ade6 mutations: ade6-51 (a C to T transition at nt 2141) and ade6-687 (a T insertion in a run of 5 thymines at nt 1216). These mutations were representatively used to study repair of two main types of replicative errors processed by the mismatch repair (MMR) system: base-base mispairs mainly caused by misincorporation and insertion-deletion mismatches produced by strand slippage in repetitive DNA (STRAND et al. 1993 ; UMAR and KUNKEL 1996 ; MARRA and SCHAR 1999 ). In all strain backgrounds, reversion rates were 10 times lower for the base substitution ade6-51 (5 x 10^-10 in wild type) than for the insertion ade6-687 (5 x 10^-9 in wild type). In msh6 strains, reversion rates of both types of ade6 mutations were 15 times increased as compared to wild type and were in the same range as those found in msh2 strains (Table 1). The msh2 msh6 double mutant showed reversion rates similar to msh6 and msh2 single mutants. In contrast the swi4 mutation either alone or in combination with msh2 and/or msh6 did not significantly alter mutation rates.

Elevated PMS frequency caused by a mutated msh6:
Tetrad analysis is a direct approach to obtaining information on mismatch repair during meiotic recombination. Aberrant tetrads in one-factor crosses are mainly of the types 6⁺:2^- and 2⁺:6^- (WCC) or 5⁺:3^- and 3⁺:5^- (PMS). Repair of a mismatch in the heteroduplex leads to either 6⁺:2^-, 2⁺:6^-, or 4⁺:4^- segregation. The latter type represents restoration events, which are not distinguishable from normal 4⁺:4^- tetrads. The failure to repair a mismatch in heteroduplex DNA results in PMS events. We dissected tetrads from the cross msh6 ade6-M26 x msh6 ade6⁺. In the wild-type cross, 5.1% WCC and no PMS were found among 1018 tetrads (GUTZ 1971 ). In the msh6 mutant cross, the WCC frequency decreased to 1.1%, while PMS increased to 3.4% among 796 tetrads (Table 2). Among the aberrant events, the PMS frequency increased from <1.9% in wild type to 75% in msh6 background. In the same type of cross, 85% PMS events were found for msh2 (RUDOLPH et al. 1999 ).

Tetrad analysis also allows determining viability of spores derived from four-spored asci. In the msh6 mutant we observed a spore viability of 90%. A spore viability of 86% was found for msh2 (RUDOLPH et al. 1999 ) and >92% for wild type (SCHAR et al. 1997 ). Thus, like mutated msh2, inactivation of msh6 has only a weak effect on spore viability.

A mutated swi4, but not msh6 or msh2, caused reduced recombination frequencies:
To measure a possible effect of MMR mutations on recombination frequencies we performed the intergenic crosses leu2 x lys7 and ade1 x lys4 (Table 3). Compared to wild type, we found an approximately two- to threefold reduction when swi4 was mutated, while msh2 or msh6 did not significantly change the recombination frequencies.

Intragenic two-factor crosses:
Crosses involving very close markers were performed to study repair of mismatches produced during meiotic recombination (Table 4 and Fig 3). During meiosis, recombination between homologous but nonidentical sequences creates mismatch-containing heteroduplex DNA. We studied the effect of the various inactivated msh genes on intragenic recombination within the ade6 gene using defined and closely linked mutations. Two independent repair systems are known to operate in S. pombe: (1) the long-patch MMR system (MutLS pathway), which efficiently repairs all mismatches except C/C, and (2) a minor short-patch repair system, which repairs C/C mismatches and in the absence of the MMR system also other mismatches (SCHAR and KOHLI 1993 ; RUDOLPH et al. 1998 ; FLECK et al. 1999 ). Recently some factors of the nucleotide-excision repair pathway were identified as components of the minor system (FLECK et al. 1999 ). It was estimated that the MMR system has excision-resynthesis tracts of 100 nucleotides unidirectional from a mismatch, while the minor pathway leads to excision-resynthesis tracts of 10 nucleotides unidirectionally (SCHAR and KOHLI 1993 ).

We measured prototroph frequencies of the crosses 421 x 51, 485 x 51, and 485 x M387 (Table 4). In all three crosses, mismatches that arise in heteroduplex DNA are separated by <100 nucleotides (Fig 3). Thus, they are in most cases corepaired on the same strand by the long-patch MMR system, which prevents formation of prototrophic recombinants. One exception is the cross 485 x M387 where two C/C mismatches can be produced in the same heteroduplex. They are not a substrate of MMR, but can be independently repaired by the short-patch repair system, which frequently results in prototrophs. Prototrophic recombinants will be generated when the two mismatches in the same heteroduplex are repaired independently toward wild-type information, i.e., when the bases of the opposite strands are replaced. Prototrophs can also be produced when one of the mismatches is repaired toward wild-type information and the other remains unrepaired. In this case, one of the daughter cells that are formed after replication conserves the mutation, while the other will have inherited the wild-type information and thus is a prototroph. The distance between the two point mutations is so short that recombinants arising by a simple crossing over are considered to be rare.

In the cross 421 x 51 only non-C/C mismatches are generated in heteroduplex DNA at a distance of 90 bp apart. We found a prototroph frequency of 15 x 10^-6 for the wild-type cross (Table 4). The frequency is rather low because repair of one mismatch by the MMR system is in most cases accompanied by corepair of the second mismatch on the same strand. When either msh6 or msh2 was mutated, the prototroph frequency increased 6- to 7-fold, and a 13-fold increase was found with the msh2 msh6 double mutant. The increase is likely caused by more frequent independent repair of the mismatches by the short-patch repair system due to the absence of functional MMR.

In the 485 x 51 cross, one heteroduplex contains a C/C and a T/G mismatch, and the other heteroduplex contains a G/G and a C/A mismatch (Fig 3). Prototrophic recombinants can be formed in the same way as in the cross 421 x 51. However, one difference is that the short-patch repair system can repair the C/C mismatch, which is not a substrate of MMR. When the MMR system acts before the minor system, the long-patch excision-resynthesis will frequently corepair the C/C mismatch with the G/T mismatch on the same strand, thus preventing formation of prototrophs. In contrast, when the first event is processing of the C/C mismatch by the short-patch repair system, corepair of the G/T mismatch will be rather rare. Thus, although the mismatches are separated by only 22 bp, a relatively high frequency of prototrophs is expected in wild type. In fact, we found a frequency of 15 x 10^-6, which is the same as in the cross 421 x 51, where the distance of the mismatches is 90 bp. The prototroph frequencies of msh2, msh6, and msh2 msh6 mutant crosses were 17-, 12-, and 25-fold increased, respectively. Thus, as in the cross 421 x 51, independent repair of the mismatches occurred more frequently when the MMR system was defective.

In the 485 x M387 cross, one heteroduplex contains two C/C and the other heteroduplex two G/G mismatches. The MMR system frequently corepairs the two G/G, preventing formation of prototrophs. The minor system can frequently repair the two C/C mismatches independently and without competition by MMR. This explains the high number of prototrophs (223 x 10^-6) generated in wild type. In contrast to the other crosses, msh6, msh2, and msh6 msh2 mutants did not significantly alter the frequency of prototroph formation. These data confirm that msh2 is not involved in the short-patch C/C correcting pathway (RUDOLPH et al. 1998 ; FLECK et al. 1999 ) and show that the same is true for msh6.

All three types of crosses including swi4 mutants showed a reduction of prototroph frequencies (Table 4). In addition, the crosses with double and triple mutants additionally mutated in msh2 and/or msh6 gave fewer prototrophs than respective crosses with an intact swi4⁺ wild-type gene. The swi4-dependent reduction was in about the same range as that measured with the intergenic crosses (Table 3). One explanation for the decrease of prototroph frequencies is that mutated swi4 causes a general reduction in meiotic recombination frequencies.

Msh6 is not involved in mating-type switching:
Swi4 and Msh2 (originally identified as Swi8) are both involved in the termination step of mating-type switching (EGEL et al. 1984 ; FLECK et al. 1992 , FLECK et al. 1994 ; RUDOLPH et al. 1999 ). Homothallic strains defective in either swi4 or msh2 form mottled colonies on sporulation medium when stained with iodine vapor, while colonies of switching-proficient strains are homogeneously brown. The mottled phenotype is caused by a reduced frequency of correct switching. In addition, heterothallic colonies frequently segregate as iodine negative. The segregants show arrangements in the mating-type region that are caused by incorrect termination of the switching process (EGEL et al. 1984 ; FLECK et al. 1990 , FLECK et al. 1992 , FLECK et al. 1994 ). We tested whether Msh6 is also implicated in mating-type switching. Cells from a homothallic msh6 strain were grown to colonies on sporulation medium and subsequently treated with iodine vapors. All colonies showed a homogeneously stained brown color, like switching-proficient strains. Thus, mating-type switching does not require msh6.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The MSH family is characterized by a conserved region located between two almost invariable motifs (TGPNM and DELGR) in the C-terminal part of the amino acid sequences. We performed PCR with degenerate primers directed to these motifs and identified the S. pombe msh6⁺ gene. To study the function of Msh6 in vivo, the msh6 gene was disrupted and analyzed as a single mutation and in combination with msh2 and swi4 mutations. Mitotic mutation rates caused by msh6 were equivalent to those seen in msh2, not only in repair of base-base mispairs but also of mononucleotide loops. In contrast, the mutation rates of swi4 strains were found to be as low as that of wild type in both types of reversion assays. Thus, repair of base-base mispairs mediated by the MMR system seems to be similar in S. pombe and S. cerevisiae, while they likely differ in repair of mononucleotide loops. msh6 mutants of S. cerevisiae are severely affected in repair of base-base mispairs but only slightly in repair of insertion-deletion mismatches (MARSISCHKY et al. 1996 ; GREENE and JINKS-ROBERTSON 1997 ; SIA et al. 1997 ; EARLEY and CROUSE 1998 ). The latter was explained by a partial compensation by the functional Msh3 protein. Consistently, in S. cerevisiae, repair of small loops is similarly affected in a msh3 msh6 double mutant as in a msh2 mutant. Thus, S. cerevisiae has two MMR pathways depending on either Msh2-Msh3 or Msh2-Msh6, while S. pombe seems to have only one. The hypothesis of a single Msh2-Msh6 pathway for both types of mutations in S. pombe is strengthened by the finding that all double mutants as well as the triple mutant exhibited mutation rates similar to those of msh6 and msh2 single mutants. In addition, we found that inactivation of msh6 caused a strong increase in GT repeat instability similar to that of msh2 and pms1, while swi4 had nearly no effect (MANSOUR et al. 2001 , accompanying article). Thus, repair of both base-base mispairs and small loops with one or two unpaired nucleotides requires Msh2-Msh6, while Swi4 plays no or only a minor role.

Tetrad analysis and determination of PMS frequencies allows us to estimate repair efficiencies of mismatches formed during meiotic recombination. A repair event is represented by WCC, while unrepaired mismatches cause PMS. We dissected tetrads from the cross msh6::arg3⁺ ade6-M26 x msh6::arg3⁺ ade6⁺, where G/A and T/C can be produced, and compared the data with those of a repair-proficient wild-type cross (GUTZ 1971 ). We found that msh6 inactivation increased the frequency of PMS events and concomitantly decreased the WCC frequency (Table 2). A similar effect was observed for msh2 and pms1 mutants in S. cerevisiae and S. pombe (ALANI et al. 1994 ; SCHAR et al. 1997 ; FLECK et al. 1999 ; RUDOLPH et al. 1999 ). The frequency of aberrant events (PMS + WCC) in the cross ade6⁺ x ade6-M26 is not significantly different among wild type (5.1 ± 0.7%), msh2 (6.1 ± 1.1%; RUDOLPH et al. 1999 ), pms1 (4.4 ± 1.1%; SCHAR et al. 1997 ), and msh6 (4.5 ± 0.7%). These results imply that the MMR system preferentially repairs mismatches at ade6-M26 toward gene conversion. Repair toward restoration would result in undetectable 4⁺:4^- events, while the failure of restoration-type repair would cause additional PMS events and thus increase the frequency of aberrant events in MMR mutants.

To extend the analysis on meiotic mismatch repair, intragenic two-factor crosses were also performed. Such crosses with closely situated mutations in the ade6 gene originally led to the discovery of two pathways acting on mismatches during meiosis (SCHAR and KOHLI 1993 ). The long-patch repair pathway, efficiently correcting most types of mismatches except C/C, was found to be equivalent to the MMR systems of other organisms. The short-patch system repairs C/C mismatches and, with low efficiency, other types of mismatches and requires components of nucleotide-excision repair (SCHAR and KOHLI 1993 ; SCHAR et al. 1997 ; RUDOLPH et al. 1998 , RUDOLPH et al. 1999 ; FLECK et al. 1999 ). With the intragenic two-factor crosses prototrophic recombinants can be determined. The types and distances of the mismatches that can be produced influence the frequency of prototroph formation, which directly depends on substrate specificity and repair efficiency of the two pathways. However, as formation of mismatches in heteroduplex DNA requires recombination, also the crossover frequency and the extension of heteroduplex tracts can modulate prototroph frequencies. We performed the crosses 421 x 51, 485 x 51, and 485 x M387. In all three crosses, the MMR system often corepairs the two mismatches in the same heteroduplex, with the exception of the two C/C produced in the cross 485 x M387 (Fig 3). Thus when MMR is defective, a strong increase of prototroph frequencies is expected for the crosses 421 x 51 and 485 x 51, but not for 485 x M387. Consistently, and in agreement with previous work, inactivation of msh2 caused increased prototroph frequencies in the crosses 421 x 51 and 485 x 51, involving no or only one C/C mismatch, but not in the cross 485 x M387. We found that prototroph frequencies in msh6 mutants were in the same range as those of the msh2 crosses (Table 4). Thus, loss of Msh6 uncovers meiotic short-patch repair of C/C and other types of base-base mismatches to an extent similar to that of inactivation of Msh2. In addition, Msh6, like Msh2, has no function in the short-patch repair system.

Concerning swi4, we found that crosses with strains bearing a swi4 defect resulted in a decrease of prototroph frequencies. The decrease was similar in all types of crosses and thus independent from the distances and the types of mismatches that can be produced. Fewer prototrophs were also found with double and triple mutants additionally defective for msh2 and/or msh6 when compared to respective crosses with the swi4⁺ wild-type gene (Table 4). Thus, the reduction in prototroph frequencies is likely not due to a defect in either the long-patch or short-patch repair pathway. As mentioned above, the frequency of prototrophs derived from intragenic two-factor crosses also depends on recombination. We performed intergenic two-factor crosses and found that swi4, but not msh2 or msh6, caused a decrease in recombinants (Table 3). Similar to the intragenic crosses, a two- to threefold reduction to wild-type crosses was observed. We propose that Swi4 is involved in meiotic recombination but not in repair of base-base mismatches arising during either meiosis or vegetative growth.

Swi4 and Msh2 are both involved in the termination step of mating-type switching in conjunction with Swi10 and Rad16, which form a heterodimeric 5' endonuclease, also implicated in nucleotide-excision repair (EGEL et al. 1984 ; FLECK et al. 1992 , FLECK et al. 1994 ; RODEL et al. 1992 ; CARR et al. 1994 ). It was proposed that correct termination requires binding of Msh2-Msh3 to a stem-loop structure either by recognizing DNA loops or by branched DNA, which is then processed by Rad16-Swi10 (RUDOLPH et al. 1999 ). In this study, we found that Msh6 likely has no function in mating-type switching. In budding yeast, Msh2-Msh3 and Rad1-Rad10, the homologues of Rad16-Swi10, but not Msh6, act in the same pathway of mitotic recombination (SAPARBAEV et al. 1996 ; SUGAWARA et al. 1997 ). It was suggested that branched structures, which can be formed during recombination by strand invasion or single-strand annealing, are bound by Msh2-Msh3 and subsequently processed by Rad1-Rad10 (SUGAWARA et al. 1997 ). In this respect, the mechanisms of mating-type switching are similar in S. cerevisiae and S. pombe. However, Swi4 in S. pombe seems to have also a Msh2-independent role in meiotic recombination, as crossover frequencies appeared to be reduced in intergenic two-factor crosses (Table 3). In S. cerevisiae, such a function was described for the meiosis-specific MutS homologues Msh4 and Msh5 (ROSS-MACDONALD and ROEDER 1994 ; HOLLINGSWORTH et al. 1995 ). Interestingly, homologues of Msh4 or Msh5 were not identified in S. pombe so far, although 95% of the genome was already sequenced (http://www.sanger.ac.uk/Projects/S_pombe/).

Our studies revealed that msh6 mutants and msh2 mutants were similarly affected, both in repair of base-base mismatches and of loops with one or two unpaired nucleotides (this work; MANSOUR et al. 2001 , accompanying article). In contrast, inactivated swi4 caused no or only slight effects. As already discussed, the relative contribution of Msh3 and Msh6 in Msh2-dependent repair clearly differs from the S. cerevisiae MMR system. Also in mammals, specific roles in MMR were assigned to the various MSH proteins. Mice deficient for MSH2 or MSH6, but not for MSH3, present a predisposition to cancer (DE WIND et al. 1995 , DE WIND et al. 1999 ; REITMAIR et al. 1995 ; EDELMANN et al. 1997 ). Inactivation of both MSH3 and MSH6 does not significantly increase the development of cancer compared to MSH6^-/- mice, although such mice developed more intestinal tumors, which were considered to be HNPCC-like (DE WIND et al. 1999 ). HNPCC in humans is frequently correlated with mutated MSH2, less frequently with mutated MSH6, and so far not found to be associated with a mutated MSH3 gene (FISHEL and WILSON 1997 ; MIYAKI et al. 1997 ; PROLLA 1998 ). However, some sporadic types of tumors were also found in patients with MSH3 mutations. Repetitive DNA of one or several nucleotides per repeat unit is drastically destabilized in MSH2 deficient cell lines but not when MSH3 is inactivated (INOKUCHI et al. 1995 ). Loss of MSH6 causes instability of mononucleotide repeats but not of other repeats (PAPADOPOULOS et al. 1995 ). A recent study revealed that MSH2-MSH6 mediates repair not only of base-base mismatches but also of loops with up to eight unpaired nucleotides (GENSCHEL et al. 1998 ). The studies on mice and humans implicate MSH2-MSH6 as the major recognition complex, while MSH2-MSH3 rather has a secondary function in mismatch repair. In this respect, the S. pombe MMR system seems to be closer to mammalian MMR than to that of S. cerevisiae.

	ACKNOWLEDGMENTS

We thank Marc Crouzet and Jean-Paul Javerzat for helpful technical advice and stimulating discussions, and Nadja Pohl for kindly providing cosmids. This work was supported by the French Ligue contre le Cancer and the Swiss National Science Foundation.

Manuscript received May 22, 2000; Accepted for publication January 19, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ACHARYA, S., T. WILSON, S. GRADIA, M. F. KANE, and S. GUERRETTE et al., 1996  hMSH2 forms specific mispair-binding complexes with hMSH3 and hMSH6. Proc. Natl. Acad. Sci. USA 93:13629-13634[Abstract/Free Full Text].

ALANI, E., R. A. REENAN, and R. D. KOLODNER, 1994  Interaction between mismatch repair and genetic recombination in Saccharomyces cerevisiae.. Genetics 137:19-39[Abstract].

ALLEN, D. J., A. MAKHOV, M. GRILLEY, J. TAYLOR, and R. THRESHER et al., 1997  MutS mediates heteroduplex loop formation by a translocation mechanism. EMBO J. 16:4467-4476[Medline].

BUERMEYER, A. B., S. M. DESCHENES, S. M. BAKER, and R. M. LISKAY, 1999  Mammalian DNA mismatch repair. Annu. Rev. Genet. 33:533-564[Medline].

CARR, A. M., H. SCHMIDT, S. KIRCHHOFF, W. J. MURIEL, and K. S. SHELDRICK et al., 1994  The rad16 gene of Schizosaccharomyces pombe: a homolog of the RAD1 gene of Saccharomyces cerevisiae.. Mol. Cell. Biol. 14:2029-2040[Abstract/Free Full Text].

DE WIND, N., M. DEKKER, A. BERNS, M. RADMAN, and H. TE RIELE, 1995  Inactivation of the mouse MSH2 gene results in mismatch repair deficiency, methylation tolerance, hyperrecombination, and predisposition to cancer. Cell 82:321-330[Medline].

DE WIND, N., M. DEKKER, N. CLAIJ, L. JANSEN, and Y. VAN KLINK et al., 1999  HNPCC-like cancer predisposition in mice through simultaneous loss of MSH3 and MSH6 mismatch-repair protein functions. Nat. Genet. 23:359-362[Medline].

EARLEY, M. C. and G. F. CROUSE, 1998  The role of mismatch repair in the prevention of base pair mutations in Saccharomyces cerevisiae.. Proc. Natl. Acad. Sci. USA 95:15487-15491[Abstract/Free Full Text].

EDELMANN, W., K. YANG, A. UMAR, J. HEYER, and K. LAU et al., 1997  Mutation in the mismatch repair gene MSH6 causes cancer susceptibility. Cell 91:467-477[Medline].

EGEL, R., D. H. BEACH, and A. J. KLAR, 1984  Genes required for initiation and resolution steps of mating-type switching in fission yeast. Proc. Natl. Acad. Sci. USA 81:3481-3485[Abstract/Free Full Text].

FISHEL, R. and T. WILSON, 1997  MutS homologs in mammalian cells. Curr. Opin. Genet. Dev. 7:105-113[Medline].

FLECK, O., L. HEIM, and H. GUTZ, 1990  A mutated swi4 gene causes duplications in the mating-type region of Schizosaccharomyces pombe.. Curr. Genet. 18:501-509[Medline].

FLECK, O., H. MICHAEL, and L. HEIM, 1992  The swi4⁺ gene of Schizosaccharomyces pombe encodes a homologue of mismatch repair enzymes. Nucleic Acids Res. 20:2271-2278[Abstract/Free Full Text].

FLECK, O., C. RUDOLPH, A. ALBRECHT, A. LORENTZ, and P. SCHÄR et al., 1994  The mutator gene swi8 effects specific mutations in the mating-type region of Schizosaccharomyces pombe.. Genetics 138:621-632[Abstract].

FLECK, O., E. LEHMANN, P. SCHÄR, and J. KOHLI, 1999  Involvement of nucleotide-excision repair in msh2 pms1-independent mismatch repair. Nat. Genet. 21:314-317. [see comments][Medline].

GENSCHEL, J., S. J. LITTMAN, J. T. DRUMMOND, and P. MODRICH, 1998  Isolation of MutSbeta from the human cells and comparison of the mismatch repair specificities of MutSbeta and MutSalpha. J. Biol. Chem. 273:19895-19901[Abstract/Free Full Text].

GRADIA, S., S. ACHARYA, and R. FISHEL, 1997  The human mismatch recognition complex hMSH2-hMSH6 functions as a novel molecular switch. Cell 91:995-1005[Medline].

GREENE, C. N. and S. JINKS-ROBERTSON, 1997  Frameshift intermediates in homopolymer runs are removed efficiently by yeast mismatch repair proteins. Mol. Cell. Biol. 17:2844-2850[Abstract].

GUTZ, H., 1971  Site specific induction of gene conversion in Schizosaccharomyces pombe.. Genetics 69:317-337[Free Full Text].

GUTZ, H., H. HESLOT, U. LEUPOLD and N. LOPRIENO, 1974 Schizosaccharomyces pombe, pp. 395–446 in Handbook of Genetics, Vol. 1, edited by R. C. KING. Plenum Press, New York.

HOHEISEL, J. D., E. MAIER, R. MOTT, L. MCCARTHY, and A. V. GRIGORIEV et al., 1993  High resolution cosmid and P1 maps spanning the 14 Mb genome of the fission yeast S. pombe.. Cell 73:109-120[Medline].

HOLLINGSWORTH, N. M., L. PONTE, and C. HALSEY, 1995  MSH5, a novel MutS homolog, facilitates meiotic reciprocal recombination between homologs in Saccharomyces cerevisiae but not mismatch repair. Genes Dev. 9:1728-1739[Abstract/Free Full Text].

INOKUCHI, K., M. IKEJIMA, A. WATANABE, E. NAKAJIMA, and H. ORIMO et al., 1995  Loss of expression of the human MSH3 gene in hematological malignancies. Biochem. Biophys. Res. Commun. 214:171-179[Medline].

ITO, H., Y. FUKUDA, K. MURATA, and A. KIMURA, 1983  Transformation of intact yeast cells treated with alkali cations. J. Bacteriol. 153:163-168[Abstract/Free Full Text].

JIRICNY, J., 1998  Replication errors: cha(lle)nging the genome. EMBO J. 17:6427-6436[Medline].

JOHNSON, R. E., G. K. KOVVALI, L. PRAKASH, and S. PRAKASH, 1996  Requirement of the yeast MSH3 and MSH6 genes for MSH2 dependent genomic stability. J. Biol. Chem. 271:7285-7288[Abstract/Free Full Text].

KOLODNER, R., 1996  Biochemistry and genetics of eukaryotic mismatch repair. Genes Dev. 10:1433-1442[Free Full Text].

KOLODNER, R. and G. T. MARSISCHKY, 1999  Eukaryotic DNA mismatch repair. Curr. Opin. Genet. Dev. 9:89-96[Medline].

LEACH, F. S., N. C. NICHOLAIDES, N. PAPADOPOULOS, B. LIU, and J. JEN et al., 1993  Mutations of a MutS homolog in hereditary nonpolyposis colorectal cancer. Cell 75:1215-1225[Medline].

MANSOUR, A. A., C. TORNIER, E. LEHMANN, M. DARMON, and O. FLECK, 2001  Control of GT repeat stability in Schizosaccharomyces pombe by mismatch repair factors. Genetics 158:77-85[Abstract/Free Full Text].

MARRA, G. and P. SCHÄR, 1999  Recognition of DNA alterations by the mismatch repair system. Biochem. J. 338:1-13.

MARSISCHKY, G. T., N. FILOSI, M. F. KANE, and R. KOLODNER, 1996  Redundancy of Saccharomyces cerevisiae MSH3 and MSH6 in MSH2-dependent mismatch repair. Genes Dev. 10:407-420[Abstract/Free Full Text].

MIYAKI, M., M. KONISHI, K. TANAKA, Y. R. KIKUCHI, and M. MURAOKA et al., 1997  Germline mutation of MSH6 as the cause of hereditary nonpolyposis colorectal cancer. Nat. Genet. 17:271-272[Medline].

MODRICH, P., 1991  Mechanisms and biological effects of mismatch repair. Annu. Rev. Genet. 25:229-253[Medline].

NEW, L., K. LIU, and G. F. CROUSE, 1993  The yeast gene MSH3 defines a new class of eukaryotic MutS homologues. Mol. Gen. Genet. 239:97-108[Medline].

PAPADOPOULOS, N., N. C. NICOLAIDES, B. LIU, R. PARSONS, and C. LENGAUER et al., 1995  Mutations of GTBP in genetically unstable cells. Science 268:1915-1917[Abstract/Free Full Text].

PLANCK, M., A. KOUL, E. FERNEBRO, A. BORG, and U. KRISTOFFERSSON et al., 1999  hMLH1, hMSH2 and hMSH6 mutations in hereditary non-polyposis colorectal cancer families from southern Sweden. Int. J. Cancer 83:197-202[Medline].

PRABHALA, G., G. J. ROSENBERG, and N. F. KÄUFER, 1992  Architectural features of pre-mRNA introns in the fission yeast Schizosaccharomyces pombe.. Yeast 8:171-182[Medline].

PROLLA, T. A., 1998  DNA mismatch repair and cancer. Curr. Opin. Cell Biol. 10:311-316[Medline].

REENAN, R. A. and R. D. KOLODNER, 1992  Isolation and characterization of two Saccharomyces cerevisiae genes encoding homologs of the bacterial HexA and MutS mismatch repair proteins. Genetics 132:963-973[Abstract].

REITMAIR, A. H., R. SCHMITS, A. EWEL, B. BAPAT, and M. REDSTON et al., 1995  MSH2 deficient mice are viable and susceptible to lymphoid tumours. Nat. Genet. 11:64-70[Medline].

RÖDEL, C., S. KIRCHHOFF, and H. SCHMIDT, 1992  The protein sequence and some intron positions are conserved between the switching gene swi10 of Schizosaccharomyces pombe and the human excision repair gene ERCC1.. Nucleic Acids Res. 20:6347-6353[Abstract/Free Full Text].

ROSS-MACDONALD, P. and G. S. ROEDER, 1994  Mutation of a meiosis-specific MutS homolog decreases crossing over but not mismatch correction. Cell 79:1069-1080[Medline].

RUDOLPH, C., O. FLECK, and J. KOHLI, 1998  Schizosaccharomyces pombe exo1 is involved in the same mismatch repair pathway as msh2 and pms1.. Curr. Genet. 34:343-350[Medline].

RUDOLPH, C., C. KUNZ, S. PARISI, E. LEHMANN, and E. HARTSUIKER et al., 1999  The msh2 gene of Schizosaccharomyces pombe is involved in mismatch repair, mating-type switching, and meiotic chromosome organization. Mol. Cell. Biol. 19:241-250[Abstract/Free Full Text].

SAPARBAEV, M., L. PRAKASH, and S. PRAKASH, 1996  Requirement of mismatch repair genes MSH2 and MSH3 in the RAD1-RAD10 pathway of mitotic recombination in Saccharomyces cerevisiae.. Genetics 142:727-736[Abstract].

SCHÄR, P. and J. KOHLI, 1993  Marker effects of G to C transversions on intragenic recombination and mismatch repair in Schizosaccharomyces pombe.. Genetics 133:825-835[Abstract].

SCHÄR, P. and J. KOHLI, 1994  Preferential strand transfer and hybrid DNA formation at the recombination hotspot ade6-M26 of Schizosaccharomyces pombe.. EMBO J. 13:5212-5219[Medline].

SCHÄR, P., M. BAUR, C. SCHNEIDER, and J. KOHLI, 1997  Mismatch repair in Schizosaccharomyces pombe requires the MutL homologous gene pms1: molecular cloning and functional analysis. Genetics 146:1275-1286[Abstract].

SCHUCHERT, P., M. LANGSFORD, E. KÄSLIN, and J. KOHLI, 1991  A specific DNA sequence is required for high frequency of recombination in the ade6 gene of fission yeast. EMBO J. 10:2157-2163[Medline].

SIA, E. A., R. J. KOKOSKA, M. DOMINSKA, P. GREENWELL, and T. P. PETES, 1997  Microsatellite instability in yeast: dependence on repeat unit size and DNA mismatch repair genes. Mol. Cell. Biol. 17:2851-2858[Abstract].

STRAND, M., T. A. PROLLA, R. M. LISKAY, and T. D. PETES, 1993  Destabilization of simple repetitive DNA in yeast by mutations affecting DNA mismatch repair. Nature 365:274-276[Medline].

SUGAWARA, N., F. PÂQUES, M. COLAIÁCOVO, and J. E. HABER, 1997  Role of Saccharomyces cerevisiae MSH2 and MSH3 repair proteins in double-strand break-induced recombination. Proc. Natl. Acad. Sci. USA 94:9214-9219[Abstract/Free Full Text].

SZANKASI, P., W. D. HEYER, and J. KOHLI, 1988  DNA sequence analysis of the ade6 gene of Schizosaccharomyces pombe. Wild-type and mutant alleles including the recombination hot spot allele ade6-M26. J. Mol. Biol. 204:917-925[Medline].

UMAR, A. and T. KUNKEL, 1996  DNA-replication fidelity, mismatch repair and genome instability in cancer cells. Eur. J. Biochem. 238:297-307[Medline].

WADDELL, S. and J. R. JENKINS, 1995  arg3⁺, a new selection marker system for Schizosaccharomyces pombe: application of ura4⁺ as a removable integration marker. Nucleic Acids Res. 23:1836-1837[Free Full Text].

WILLIAMSON, M. S., J. C. GAME, and S. FOGEL, 1985  Meiotic gene conversion mutants in Saccharomyces cerevisiae. I. Isolation and characterization of pms1-1 and pms1-2. Genetics 110:609-646[Abstract/Free Full Text].

This article has been cited by other articles:

				T. Yamada-Inagawa, A. J. S. Klar, and J. Z. Dalgaard Schizosaccharomyces pombe Switches Mating Type by the Synthesis-Dependent Strand-Annealing Mechanism Genetics, September 1, 2007; 177(1): 255 - 265. [Abstract] [Full Text] [PDF]

D. R. Denver, S. Feinberg, S. Estes, W. K. Thomas, and M. Lynch Mutation Rates, Spectra and Hotspots in Mismatch Repair-Deficient Caenorhabditis elegans Genetics, May 1, 2005; 170(1): 107 - 113. [Abstract] [Full Text]