CYS3, a Hotspot of Meiotic Recombination in Saccharomyces cerevisiae: Effects of Heterozygosity and Mismatch Repair Functions on Gene Conversion and Recombination Intermediates

Corresponding author: Alain Nicolas, UMR144, Institut Curie, Section de Recherche, 26 rue d’Ulm, 75248, Paris Cedex 05, France., anicolas{at}curie.fr (E-mail)

Communicating editor: S. JINKS-ROBERTSON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have examined meiotic recombination at the CYS3 locus. Genetic analysis indicates that CYS3 is a hotspot of meiotic gene conversion, with a putative 5'–3' polarity gradient of conversion frequencies. This gradient is relieved in the presence of msh2 and pms1 mutations, indicating an involvement of mismatch repair functions in meiotic recombination. To investigate the role of mismatch repair proteins in meiotic recombination, we performed a physical analysis of meiotic DNA in wild-type and msh2 pms1 strains in the presence or absence of allelic differences at CYS3. Neither the mutations in CYS3 nor the absence of mismatch repair functions affects the frequency and distribution of nearby recombination-initiating DNA double-strand breaks (DSBs). Processing of DSBs is also similar in msh2 pms1 and wild-type strains. We conclude that mismatch repair functions do not control the distribution of meiotic gene conversion events at the initiating steps. In the MSH2 PMS1 background, strains heteroallelic for frameshift mutations in CYS3 exhibit a frequency of gene conversion greater than that observed for either marker alone. Physical analysis revealed no modification in the formation of DSBs, suggesting that this marker effect results from subsequent processing events that are not yet understood.

GENE conversion events, which represent the nonreciprocal transfer of information from one chromatid to another, are manifested during meiosis by the non-Mendelian segregation (NMS) of heterozygous markers. These NMS events are the signature of recombination in that they include the large majority of intragenic recombination events and are associated with the reciprocal exchange of flanking markers about half of the time (FOGEL et al. 1981 ; PETES et al. 1991 ). In Saccharomyces cerevisiae, the average frequency of meiotic gene conversion at an individual site is 2–3% per meiosis, but it can vary from <1% to >50% (FOGEL et al. 1981 ). At specific loci, gradients of conversion frequencies that vary in steepness from one locus to another have been detected: 1–10% across the ARG4 locus (FOGEL et al. 1981 ), 4–15% at DED81 (SCHULTES and SZOSTAK 1990 ), 5–14% at HIS2 (MALONE et al. 1992 ), 1–3% at HIS1, and 17–50% at HIS4 (PETES et al. 1991 ). The molecular mechanism(s) that controls the steepness of these gradients is not yet understood. Two mechanisms have been proposed to explain shallow and steep gradients, respectively. DETLOFF et al. 1992 postulated that conversion of markers exhibiting low postmeiotic segregation (PMS) could reflect a preponderance of restoration-type over conversion-type repair. That is, mismatches located near the site of initiation might be repaired mainly by conversion-type repair. In contrast, if half of the distal mismatches were repaired by restoration-type repair, these events would not be detected and, consequently, would contribute to a twofold gradient, as observed at HIS4. Evidence for restoration-type repair at this locus has been shown recently (KIRKPATRICK et al. 1998 ). Preferential repair resulting in restoration at the low end of the polarity gradient has been observed. Steeper gradients must result from another mechanism. They could be explained if recombination is initiated at fixed sites and if heteroduplex regions extend to varying distances from these sites. Deletion analyses have identified cis-acting regions required for high levels of gene conversion in small chromosomal regions located 5' of the coding sequence of the ARG4 (NICOLAS et al. 1989 ; DE MASSY and NICOLAS 1993 ) and HIS4 (DETLOFF et al. 1992 ) genes. These positions are consistent with the observed gradient of gene conversion frequencies at these loci in that the highest frequency of conversion is at the 5' end. In contrast, the conversion gradient associated with the HIS2 hotspot is of opposite polarity, suggesting that some cis-acting sequences required for the high level of gene conversion at HIS2 are located downstream (MALONE et al. 1992 , MALONE et al. 1994 ; BULLARD et al. 1996 ).

Genetic recombination in yeast is also characterized by the formation of DNA double-strand breaks (DSBs), which occur at recombination hotspots (SUN et al. 1989 ; CAO et al. 1990 ; FAN et al. 1995 ; MAO-DRAAYER et al. 1996 ) and at many locations in the genome, and it is likely that most if not all meiotic-recombination events are initiated by meiosis-specific DSBs (ZENVIRTH et al. 1992 ; WU and LICHTEN 1994 ; KLEIN et al. 1996 ; BAUDAT and NICOLAS 1997 ). The processing of DSBs yields single-strand tails (SUN et al. 1989 ; CAO et al. 1990 ) that invade the homologous chromosome, leading to the formation of Holliday junctions (SCHWACHA and KLECKNER 1995 ) and heteroduplex DNA (LICHTEN et al. 1990 ). In the context of the DSB repair model (SZOSTAK et al. 1983 ), gene conversion gradients could be explained by the distance between the site at which the heteroduplex DNA is initiated and a specific genetic marker undergoing conversion.

Conversion gradients are nearly abolished at both the ARG4 and HIS4 loci in mutants defective in the MSH2 and PMS1 mismatch repair (MMR) genes, which suggests a role for the MMR system in the formation of the conversion gradient (ALANI et al. 1994 ). Msh2p, one of six homologs of the Escherichia coli MutS protein (Msh1p–Msh6p), is a central component of the S. cerevisiae MMR system (REENAN and KOLODNER 1992A , REENAN and KOLODNER 1992B ). It binds mispaired bases and recognizes both single base mispairs and multiple base insertion/deletion mispairs (ALANI 1996 ). Mlh1p and Pms1p are homologs of the bacterial MutL protein (KRAMER et al. 1989 ; REENAN and KOLODNER 1992A ). Genetic and biochemical analyses of MMR in S. cerevisiae have suggested that MMR involves two complexes: two MutS homologs (Msh2p and Msh3p or Msh6p) form a complex that recognizes mispairs, which in turn is bound by a complex of two MutL homologs (Mlh1p and Pms1p, PROLLA et al. 1994 ; MARSISCHKY et al. 1996 ). In addition, purified Msh2 protein binds to synthetic Holliday junctions (ALANI et al. 1997 ). The Msh2p-Holliday junction complex appears to be more stable than the Msh2p-duplex DNA complex, suggesting that Msh2p could coordinate MMR and genetic recombination. The function of these MMR proteins with respect to gene conversion gradients, however, remains to be elucidated. Observations published by ALANI et al. 1994 suggest that the existence of polarity reflects the frequency of heteroduplex formation and/or the processing of this recombination intermediate by MMR-dependent processes.

With the goal of defining the recombination intermediate(s) on which the MMR system acts, we have undertaken a genetical and physical characterization of gene conversion at the CYS3 locus, which is proposed to be a hotspot of gene conversion (CHEREST and SURDIN-KERJAN 1992 ). Our results demonstrate that the CYS3 locus is indeed a gene conversion hotspot, and that the Msh2 and Pms1 proteins act to control the frequency of gene conversion after the formation of initiating DSBs.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Media and sporulation conditions:
Standard medium (YPD) was used for vegetative growth. Presporulation medium was SPS, and sporulation was carried out in 1% potassium acetate, as described previously (DE MASSY and NICOLAS 1993 ). The addition of 0.5 mM reduced glutathione to all media compensated for cysteine auxotrophy. The segregation of cys3 mutations was determined using synthetic minimal medium without glutathione and supplemented for all other auxotrophic markers.

Plasmids:
Fragments of the CYS3 region were isolated from plasmid pStr1.1 (gift of Y. SURDIN-KERJAN) and recloned in pMLC28 (SCHULTES and SZOSTAK 1990 ) or pKS Bluescript. The EcoRI site of pMLC28 was first destroyed to simplify the construction of the cys3-RI strain. The PstI and the two PstI/HindIII fragments (see Figure 1) were inserted into pMLC28 to obtain pMV1, pMV2, and pMV3, respectively. The CYS3 locus was reconstituted from pMV1 and pMV2 (2238 bp PstI-XcaI fragment) to give pMV7.

View larger version (21K): In this window In a new window Download PPT slide   Figure 1. Physical and functional maps of the CYS3 region, plasmid inserts, and probes. The name and the relevant insert of the plasmids used in this report are indicated in the upper part of the figure. Restriction sites are as follows: Bg, BglII; Bst, BstBI; H, HindIII; N, NcoI; P, PstI; RI, EcoRI; St, StyI; Xc, XcaI; and Xh, XhoI. ORFs are mentioned according to OUELLETTE et al. 1993 . Horizontal arrows correspond to the direction of transcription. At the bottom, the black bars correspond to double-strand probes, and the gray bars correspond to the single-strand probes. The positions of the meiotic DSB sites are indicated by vertical arrows.

The cys3-Bg and the cys3-RI strains were made using plasmids bearing a part of the mutated CYS3 gene and the URA3 gene. The BglII and the EcoRI mutations were made by filling in the relevant restriction site overhangs with Klenow DNA polymerase in pMV1 or in pMV7, respectively. An XbaI-SacI fragment containing the URA3 gene (from pHS113, SUN et al. 1989 ) was ligated into the XbaI-SacI-digested cys3-Bg plasmid. The HindIII/XhoI fragment bearing the EcoRI mutation was isolated from pMV7 and ligated to the URA3 plasmid described previously to generate the cys3-RI plasmid. All restriction enzymes were from New England Biolabs (Beverly, MA), with the exception of Bst1107I (Fermentas), which was used instead of XcaI. T4 DNA ligase and Klenow DNA polymerase were from Boehringer Mannheim (Mannheim, Germany).

Strains:
E. coli strains were either DH5 or NPS RKII and were cultured in standard media (LB and 1% thymidine LB, respectively). All S. cerevisiae strains used were derived from the congenic MGD131-102A and MGD131-2C haploids (see Table 1). Genetic markers were introduced into the CYS3 gene by a two-step replacement procedure using URA3 as a selectable marker (SCHERER and DAVIS 1979 ). Both cys3-Bg and cys3-RI plasmids were linearized by XhoI, and each of them was used to transform MGD131-102A and MGD131-2C haploids to uracil prototrophy. Ura^- derivatives were obtained from the primary transformants by selection with 5-fluoroorotate (BOEKE et al. 1984 ). Both constructions were subsequently verified by Southern blot analysis of genomic DNA digested with ClaI (for the cys3-Bg mutation) or AseI (for the cys3-RI mutation).

Strains heteroallelic for CYS3 mutations in the rad50S background were obtained by crossing appropriate segregants from a cross between the rad50S strains ORT305 or ORT311 and a cys3-Bg or cys3-RI strain. A msh2 pms1 homozygous diploid (ORD2048) was constructed by mating RKY1929 and RKY1939 (ALANI et al. 1994 ). The homozygous msh2 pms1 rad 50S diploid ORD2047 was built with segregants from ORD2043 (RKY1935 x ORD311). The msh2 pms1 rad50S haploids ORD2043-9B and ORD2043-13B were used to construct ORD2047, which is homozygous for all three mutations. The correct genotype of these strains was verified by PCR and Southern blot analysis using a radiolabeled 600-bp fragment generated with two oligonucleotides specific to the MSH2 coding region.

Genetic techniques:
Standard procedures for mating and tetrad dissection were used. Cells were grown in presporulation medium (SPS) and sporulated in 1% potassium acetate at 30°. Tetrads were generally dissected after 2–3 days in sporulation medium. Because msh2 and pms1 mutants display a mutator phenotype, msh2 pms1 strains were mated and sporulated after limited growth, as described previously (REENAN and KOLODNER 1992B ; ALANI et al. 1994 ).

For tetrad analysis of diploid strains with two cys3 mutant markers, tester strains ORD2017-1A (his1 cys3-Bg MATa), ORD2022-1A (his1 cys3-Bg MAT), ORD2031-1A (his1 cys3-RI MATa), and ORD2027-29A (his1 cys3-RI MAT) were constructed. cys3-Bg or cys3-RI segregants were mated with the above strains. UV-induced Cys⁺ papillations of resultant diploids allowed for the identification of haploid segregants carrying the cys3-Bg and cys3-RI mutation either singly or together. Meiotic viability of a diploid strain is defined as the percentage of tetrads with four viable spores among the total number of dissected tetrads.

Isolation of DNA and detection of DSBs:
Twenty-five milliliters of meiotic cells were harvested from sporulation medium at each time point (0 to 24 hr). Spheroplasts were produced by incubation with Zymolyase (Chemical Credential, ICN) for 30 min at 30°. DNA was purified as described previously (DE MASSY and NICOLAS 1993 ) using SDS, proteinase K, gentle phenol extractions, and ethanol precipitations. One to two micrograms of DNA were digested by an appropriate restriction enzyme, subjected to electrophoresis on a 0.8% agarose gel, and blotted onto a nylon membrane (Hybond-N+; Amersham, Rainaham, UK). The CYS3 region was examined using the probes shown in Figure 1: (1) the downstream probe a, which is specific to the 3' end of CYS3 and part of the FUN36 ORF, or the a' probe, which is inside the FUN36 ORF, and (2) the upstream probe b, which is the BglII fragment shown in Figure 1, or b', which is internal to the FUN33 ORF. Hybridization with the a or a' probe and the b or b' probe gave the same results, and they will be designated here as probe a or probe b. For studies of breaks located in the YCR47C/48W region of chromosome III (ZENVIRTH et al. 1992 ), the 667-bp XbaI/EcoRV fragment isolated from plasmid PM5239 was used as a probe (gift from L. Grivell). All probes were labeled by the random priming method according to the manufacturer (Pharmacia, Piscataway, NJ) using 50 µCi of 3000 Ci/mM [³²P]dCTP (Amersham). Hybridization conditions were as specified by CHURCH and GILBERT 1984 . The filters were exposed to Amersham Hyperfilms with intensifying screens for 2–7 days or were analyzed with the PhosphorImager system (Molecular Dynamics, Sunnyvale, CA).

For analysis of the processing of DSBs, denatured DNA digests were separated on a 0.8% agarose alkaline gel run in 50 mM NaOH, 1 mM EDTA and blotted to a Genescreen membrane (Dupont-New England Nuclear) using the method of BISHOP et al. 1992 . The DNA was UV cross-linked using the Stratalinker (Stratagene, La Jolla, CA). The exposure was 100 sec at 1200 µJ. Strand-specific ³²P-labeled RNA probes were made from pK Bluescript by in vitro transcription of cloned DNA (Promega, Madison, WI). A standard transcription protocol with T3 or T7 RNA polymerase and 800 Ci/mM [³²P]UTP (Amersham) was used. These probes are indicated in Figure 1 and Figure 5. The conditions of hybridization and washing were similar to those used in nondenaturing Southern analysis.

View larger version (39K): In this window In a new window Download PPT slide   Figure 2. Genotypic composition of exceptional tetrads in two-point crosses in which the cys3-Bg and cys3-RI markers are in the trans-position (upper part) or cis-position (bottom part). In our analysis, the numbers of tetrads with four viable spores were 519 and 295 for the trans- and cis-configurations, respectively. The number of tetrads of each genotype is indicated under each tetrad.

View larger version (37K): In this window In a new window Download PPT slide   Figure 3. Detection, mapping, and quantification of meiotic DSBs at the CYS3 recombination hotspot in rad50S strains. Meiotic cells are harvested from sporulation medium at various times during sporulation (t = 0–24 hr). DNA from the samples is purified; 1–2 µg of DNA is HindIII digested and analyzed by Southern blot. Random-primed-labeled a and b probes are as indicated in Figure 1. The HindIII parental fragment (6710 bp) is at the top of each autoradiogram. In the case of the CYS3 locus, the DSBs are indicated with horizontal arrows of unequal size representing the relative intensity of the two bands. The quantification of DSBs (percentage of DSBs) is calculated as the percentage of radioactivity in both CYS3 DSB fragment groups relative to the total amount of parental plus DSB fragments (including the DEP1 fragment when it was detectable). (A) Detection and mapping of meiotic CYS3 DSBs, probing with the a (left part) or b (right part) fragment. Molecular weight markers are /HindIII (H) and /BstEII (B) digestions. (B) Comparison of DSB formation between homoallelic ORD307 (CYS3/CYS3 on the left part) and heteroallelic ORD2040 (cys3-Bg/cys3-RI on the middle part) at the CYS3 (top) or at YCR47C/48W (bottom) loci. For both loci, genomic DNA was digested by HindIII. In the case of CYS3, the probe was a, as indicated in Figure 1. In the case of the YCR47C/48W locus, an XbaI/EcoRV fragment containing a YCR47C ORF part was labeled as a probe for this region. Other legends are as described in A. The quantification and kinetics of the DSB accumulation at the CYS3 (squares) and the YSC47C/48W (circles) loci in ORD307 (filled squares or circles) or in ORD2040 (open squares or circles) is shown on the right. (C) Extended time course in the cys3-Bg/cys3-RI heteroallelic strain (ORD2040); the quantification and kinetics of the CYS3 DSBs are shown in the right part.

View larger version (45K): In this window In a new window Download PPT slide   Figure 4. Detection of meiotic DSBs at the CYS3 recombination hotspot in the RAD50 strain (ORD149). The DNA was digested with HindIII and blotted as described previously. Each membrane was hybridized with the a or b probe (see Figure 1). A HindIII digestion of DNA from cells harvested at t = 8 hr from the rad50S isogenic strain ORD307 (t = 8 hr) was added as a control in each blot. The vertical open circles indicate the smear corresponding to the transient DSB fragments. A quantitative analysis of labeling in the smears is shown, using the a (diamonds) or the b (squares) probe. The graph indicates the percentage of the labeling in the DSB smear (percent of DSBs) plotted over time in sporulation medium. The percentage is calculated as explained in Figure 3.

View larger version (57K): In this window In a new window Download PPT slide   Figure 5. Analysis of individual component strands of DSB fragments in the RAD50 strain (ORD149). DNA was digested with HindIII, fractionated on an alkaline gel, and blotted onto a Genescreen membrane. (Top) Southern analysis using single-stranded d or e riboprobes to detect each strand that has its 3' or 5' ends at the DSB site. The HindIII-digested DNA from strain ORD307 (rad50S) at t = 8 hr is shown as a control. Molecular weight markers in the line M were generated by double digestion of genomic DNA with HindIII/XhoI (3344 bp) or HindIII/BstEII (2789 bp). The horizontal arrows indicate the DSB product from the rad50S strain. (Bottom) Partial restriction map of the HindIII parental fragment; d and e strand-specific RNA probes and their orientation are schematized by horizontal arrows. The probe d has the sequence of the top strand, the e probe has that of the bottom strand. Other legends are as described in Figure 1 Figure 2 Figure 3.

Quantification of DSBs:
Quantitative estimation of radioactivity on the blots was done by scanning membranes with the SOFI radioimager (Quartz et Silice) or by analyzing an exposed phosphor screen with a PhosphorImager system (Molecular Dynamics) followed by ImageQuant software analysis (Molecular Dynamics). The quantification of DSBs (percentage) is calculated as the percent of radioactivity in both CYS3 DSB fragment groups relative to the total amount of parental plus DSB fragments (including the DEP1 fragment, when it was detectable). The quantitations were performed by selecting an exposure time that allows accurate determination of the intensity of the parental band (linear response). Sometimes, however, this means that the signal for the DSB fragment is very weak and very difficult to be quantified reliably. This is particularly the case in RAD50 strains. Therefore, the exposures that are presented in the figures, which were chosen to best illustrate weak signals, are not always the same as those used for quantitation.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Genetic analysis of meiotic recombination at the CYS3 locus:
The CYS3 locus, originally called STR1 (CHEREST and SURDIN-KERJAN 1992 ), is located on chromosome I between the FUN54 (also called DEP1) and FUN36 (Figure 1) ORFs (ONO et al. 1992 ; BARTON et al. 1993 ; OUELLETTE et al. 1993 ). In a previous study of the meiotic segregation pattern of the str1-1 allele in a STR1/str1-1 diploid (a one-point cross), it was reported that NMS (3+:1- and 1+:3- conversion events) occurs at a frequency of 15% (18 events out of 111 tetrads analyzed, CHEREST and SURDIN-KERJAN 1992 ).

To confirm that the CYS3 locus is a hotspot of meiotic gene conversion, we introduced two novel markers into this gene and analyzed the frequency of NMS in strains carrying these alleles. These CYS3 mutations were constructed in vitro by filling in natural restriction enzyme sites, and they are located at positions +154 (cys3-Bg) and +808 (cys3-RI) with respect to the first nucleotide (A of the ATG) of the CYS3 coding region (Figure 1). Upon introduction into the wild-type locus by the two-step transplacement procedure (MATERIALS AND METHODS) in haploid strains, the cys3-Bg and cys3-RI mutations, which create +4-bp frameshifts, were found to confer cysteine auxotrophy. CYS3 mutants can grow on media supplemented with either 0.5 mM cysteine or 0.5 mM glutathione. The genotypes of these yeast strains are listed in Table 1.

The meiotic segregation pattern of these mutations was first examined in diploids heterozygous for a single mutation (one-point cross). CYS3/cys3-Bg and CYS3/cys3-RI diploids were sporulated, and unselected tetrads were dissected and analyzed. The results reported in Table 2 show that the cys3-Bg marker exhibits a much higher frequency of NMS (28%) than does the cys3-RI marker (7%). These values bracket the previous report of 15% found for the str1-1 allele, which is mutated at an unknown position (CHEREST and SURDIN-KERJAN 1992 ). Because the average level of gene conversion found at most loci in yeast is 2–3% (FOGEL et al. 1981 ), we conclude that the CYS3 locus is a hotspot of meiotic gene conversion. The vast majority of conversions involving the cys3-Bg and cys3-RI markers are 3+:1- and 1+:3- NMS events without a bias in the direction of conversion (P > 0.05), and they reflect a single conversion event per meiosis. Among 420 tetrads, four 4+:0- and two 0+:4- tetrads were observed. Their occurrence can result from premeiotic events or from two independent events in which each of two sister chromatids is converted in the same meiosis. No PMS (5+3- or 3+5-) events producing CYS3/cys3 sectored colonies were observed, which is in accordance with previous studies showing that small base pair insertion and deletion mutations give only 3+:1- and 1+:3- conversion events in MMR-proficient strains (FOGEL et al. 1981 ; NICOLAS et al. 1989 ; DETLOFF et al. 1992 ).

To further characterize recombination events at the CYS3 locus, we then studied the segregation of the cys3-Bg and cys3-RI markers in two-point crosses, with these markers in the trans (Bg +/+ RI) or cis (Bg RI/+ +) configurations. Unselected tetrad analysis of heteroallelic diploids allows various classes of segregation events to be distinguished: single-site conversions, simultaneous conversion of both sites in one direction (coconversion), reciprocal exchanges, and complex events involving one or both markers on more than two chromatids per meiosis. Our analysis of the results is shown in Table 2, and the genotypic composition of exceptional tetrads is reported in Figure 2. In heteroallelic diploids, spore viability is ~70%, comparable to that observed in one-point crosses. For the trans configuration, the frequency of tetrads exhibiting NMS is very high, as much as 48% (256/529). Among these, NMS events involving the cys3-Bg marker represent 91% (233/256) of the recombinant tetrads, and those involving the cys3-RI marker represent 52% (135/256). Conversions of the cys3-Bg marker alone represent 48% (111/233) of the cases of conversion at this site, and those of the cys3-RI marker alone are 15% (20/135) of the events involving this site. Coconversions are frequent (32%, 82/256), and numerous complex events (17%, 43/256) also occur. We observed that numerous tetrads noticeably exhibited 4+:0- and 0+:4- segregation for either of the two markers, indicating that simultaneous conversion of the sister chromatids at one site is frequent, but that conversion at both sites is less common (3/43 complex tetrads). Reciprocal exchanges are rare (2%, 5/261 recombinant tetrads). The surprising aspect of these data is the higher overall NMS frequency of either marker when both are present, as compared to the NMS frequency of either marker alone. In absolute frequencies, the cys3-Bg marker exhibits 44% NMS (233/529) instead of 28%, as seen in one-point crosses, and the cys3-RI marker exhibits 26% (135/529), compared to 7%.

The results obtained with the two markers in the cis configuration also show that tetrads exhibiting NMS are frequent (34%, see Table 2 and Figure 2). Most events involve NMS of either marker; 39% of recombinant tetrads are single-site conversions. Most involve the cys3-Bg marker (34/39) and, less frequently, the cys3-RI marker (5/39). The other recombinant tetrads correspond to 48% of coconversion, 12% complex events, and 1% reciprocal exchanges. NMS events involving the cys3-Bg marker therefore represent 32% (94/295) of total meioses, a frequency slightly higher than that observed for one-point crosses (P > 0.05). This value is lower (P < 0.01) than that found for the trans-heteroallelic diploids (44%); in this case, NMS events involving the cys3-RI marker represent 21% (62/295) of total meioses compared to 26% for the trans-configuration. Tetrad analysis of gene conversion events in a diploid heterollelic at both ARG4 (arg4-RV +/+ arg4-Bg) and CYS3 (cys3-Bg +/+ cys3-RI) indicates that the increase of NMS at CYS3 is not accompanied by a similar enhancement of NMS at the ARG4 locus, which is located on a different chromosome (data not shown).

To test if cysteine auxotrophy affects the level of NMS events, we examined diploids ORD3703 (+ RI/Bg RI) and ORD3704 (Bg +/Bg RI), which are heterozygous for a single marker but auxotrophic for cysteine (see Table 1). We observed (Table 2) that the cys3-Bg marker exhibits the same frequency of gene conversion (27%) as that measured from a diploid cysteine prototroph in a one-point cross (P = 0.88). In the case of the cys3-RI marker, the frequency of gene conversion is slightly but significantly increased (12% instead of 7%; P = 0.04). These results indicate that the enhanced NMS frequency of the cys3-RI marker observed in two-point crosses could be partially related to cysteine auxotrophy. Further analysis of this unusual marker effect is presented in DISCUSSION.

Genetic studies in msh2 pms1 strains:
At the ARG4 and HIS4 loci, ALANI et al. 1994 found that the 5'–3' conversion gradient is relieved in msh2, pms1, and msh2 pms1 strains because of increased NMS frequencies of markers located at the 3' end of the locus. In one-point crosses, we examined the effect of msh2 pms1 mutations on NMS in strains with the cys3-Bg or cys3-RI alleles (Table 3). Diploids (ORD1455 and ORD1457) were sporulated after minimal growth to avoid the accumulation of lethal mutations (MATERIALS AND METHODS). Spore viability ranged from 50 to 70%. The frequency of NMS involving the cys3-Bg marker in the MMR-deficient background was 18%, and that involving the cys3-RI marker was 21%. In the case of the cys3-Bg marker, the difference between the mutant (18%) and the wild-type strain (28%) is modest but significant (P = 0.034). For the cys3-RI marker, the frequency of NMS events is very significantly (P < 0.005) enhanced in MMR-deficient strains. As expected for MMR mutants, msh2 pms1 mutations lead to the appearance of PMS (5+:3-, 3+:5-, aberrant 5+:3-, and aberrant 4+:4-) at the expense of simple gene conversion events (6+:2- and 2+:6- segregation). These PMS events are indicative of the formation of heteroduplex DNA at CYS3 and of inefficient MMR of these 4-bp insertion mutations. The frequency of PMS among the aberrant events is 60% (18/30) for the cys3-Bg marker and 81% (21/26) for the cys3-RI marker. We note that 40% of the NMS events involving the cys3-Bg marker and 19% at the cys3-RI marker are still simple gene conversions, suggesting either that residual MMR activity persists in the absence of the Msh2 and Pms1 proteins, or that these events result from the repair of a double-strand gap (SZOSTAK et al. 1983 ). In the msh2 pms1 strain, this residual gene conversion frequency of the cys3-Bg marker (7%) is significantly different from that observed in the wild-type strain (27%; P < 0.001). In contrast, conversion of the cys3-RI marker is not significantly modified (7 and 4%, for wild-type and the msh2 pms1 strains, respectively; P = 0.17). These results will be compared with those described for ARG4 and HIS4 (ALANI et al. 1994 ) in DISCUSSION.

Meiotic DSBs in the CYS3 region:
DSBs can be detected either as transient DNA fragments of heterogeneous length in wild-type RAD50 diploid strains (SUN et al. 1989 ) or as discrete bands that accumulate in diploids homozygous for the rad50S mutation (ALANI et al. 1990 ). This mutation has made it possible to map DSB sites and to quantify the extent of breakage at these sites. We examined the formation of meiotic DSBs near CYS3 in rad50S strains by Southern blot analysis. In addition to the parental fragment (6716 bp), the a and b probes hybridize to two bands of lower molecular weight which are seen only in DNA from meiotic cells (Figure 3A). These fragments, as measured in several independent experiments, constitute 8 ± 2 % of the total DNA molecules hybridizing to the probe. The two DSB regions map in the 5' intergenic region of the CYS3 gene. The weaker "proximal" DSB is located around position -160 with respect to the CYS3 ORF, whereas the stronger band ("distal") is located around position -270. Distal DSBs are two- to threefold more frequent than proximal DSBs. A 4.8-kb meiosis-specific DSB fragment can also be seen in Figure 3. This represents a weak DSB site located between FUN34 and DEP1 (Figure 1).

The genetic results presented above show that the frequency of NMS at CYS3 is about twofold greater in a trans-heteroallelic diploid than in a diploid containing only one of the two mutant alleles. To test whether heterozygosities near the DSB sites affect the frequency of DSB formation, we compared DSB levels in rad50S diploids homozygous for the wild-type CYS3 allele (ORD307) or heteroallelic at CYS3 (ORD2040). As a control, we also examined DSB formation at the YCR47C/48W locus on chromosome III (ZENVIRTH et al. 1992 ; GOLDWAY et al. 1993 ; BAUDAT and NICOLAS 1997 ). Figure 3B shows Southern blot analysis of DSB formation at CYS3 (top panels) or at YCR47C/48W (bottom panels) in a wild-type diploid (left panels) or a diploid heteroallelic at CYS3 (right panels). A comparison of the kinetics of formation of DSBs in these strains (Figure 3B) shows that the CYS3 heteroallelic diploid exhibits a 2-hr delay in DSB formation at both loci. It indicates that this delay does not result from heterozygosity at the CYS3 locus, but, very likely, that the CYS3 heteroallelic strain has a different timing for entry or progression into meiotic prophase. At t = 24 hr, the amount of DSBs formed at CYS3 reaches a value of 8.5%, which is similar to that observed in a CYS3/CYS3 strain (8 ± 2%, Figure 3B). The position of the DSBs is the same as that in wild-type strains (Figure 3B). To verify that the level of DSBs in the cys3-Bg/cys3-RI diploid is maximal by 24 hr, we performed an additional experiment that included 32- and 40-hr time points (Figure 3C). The level of DSBs was maximal at 24 hr, whereas it reached a plateau at t = 12 hr in the CYS3/CYS3 diploid. A similar delay was observed in RAD50 strains in which the DSBs are transient (data not shown). This delay was observed at both loci (CYS3 and YCR47C/48W), and it could originate from metabolic defects resulting from the absence of functional Cys3p, which might not be fully compensated for by the addition of glutathione. Taken together, these data allow us to conclude that the presence of heterozygous markers in the CYS3 coding region does not affect the frequency or distribution of DSBs in this region.

The formation of meiotic DSBs at CYS3 in the RAD50 strain background was studied in the diploids ORD149 or ORD2049 (Table 1). With both probes a and b (Figure 4), we observed the appearance of smeared DSB fragments, indicating that both ends of the CYS3 meiosis-specific breaks were resected, as has been found for the ARG4 (SUN et al. 1991 ) and HIS4-LEU2 loci (CAO et al. 1990 ). The smeared products appear as diffuse signals with a bandwidth of 200–300 bp. The bands in the RAD50 sample are of lower molecular weight than the corresponding bands in the rad50S sample. The kinetics and the quantification of band intensities are diagrammed in Figure 4. DSB fragments corresponding to both sides of the break appear with similar kinetics. The smears appear at ~t = 6 hr after transfer into sporulation medium, they reach maximal intensity at ~t = 8 hr, and then disappear between t = 10 and t = 24 hr, at which time ascospores are formed.

To test whether meiotic DSBs at CYS3 are resected as they are at the ARG4 (SUN et al. 1991 ) and HIS4-LEU2 loci (BISHOP et al. 1992 ), DSB fragments were resolved on alkaline denaturing gels and hybridized to single-strand-specific riboprobes (Figure 5). These probes are complementary either to the top strand (e probe) or the bottom strand (d probe), and they are specific to the 5' side of the CYS3 locus (Figure 1). With the e probe, we observed a break-specific band of about the same length as that found in the rad50S background, indicating that this strand is not resected. This strand has its 3' extremity at the DSB site (the 3' end). The d probe, which corresponds to the complementary strand (the 5' end), hybridizes to fragments of heterogeneous length from t = 4 to t = 12 hr. This smear is ~200–300 nucleotides wide, and the fragments at the top of the smear are ~400 bases shorter than the fragment found in the rad50S strain. We address this finding in DISCUSSION. Two other single-strand probes (f and g) were also used to probe the proximal side of the CYS3 DSBs (Figure 1). The 3' end was also observed as an unresected fragment, and the 5' end was observed as a faint smear (data not shown).

DSBs in msh2 pms1 strains:
To identify the recombination intermediate(s) on which the MMR proteins Msh2p and Pms1p act, we studied the processing of DSBs in homozygous msh2 pms1 strains. To rule out the possibility that the absence of Msh2p and Pms1p affects the formation of DSBs, we first examined meiotic DSBs in msh2 pms1 rad50S diploids (Figure 6). We observed that both CYS3-proximal and -distal DSBs are formed, and that no other DSB sites appear either in the CYS3 coding region or further downstream (data not shown). A quantification of CYS3 DSBs in each strain is shown in Figure 6. DSBs accumulate as unresected fragments at the same level in a msh2 pms1 rad50S mutant strain (7 ± 1.5%) as they do in MSH2 PMS1 rad50S strains (8 ± 2%). The ratios of the frequencies of the distal and proximal DSBs remain similar for both strains during the entire time course. We then examined the formation of meiotic DSBs in msh2 pms1 RAD50 cells (ORD2048) with respect to a wild-type strain (ORD2049). The percentage of sporulation (dyads, triads, and tetrads) at 72 hr after the induction of meiosis was similar in both strains (77% in ORD2049, 70% in ORD2048). Figure 7 shows a Southern blot analysis of time course experiments in which both sides of the DSB region were probed (probes a and b, as illustrated in Figure 1) for each strain. In both strains, DSBs are visualized as smeared and transient signals indicative of the formation of resected DSB fragments. The kinetics of signal appearance and disappearance are similar in both strains (Figure 7). At the peak of their accumulation, the frequency of these transient diagnostic fragments is identical (2.0 ± 0.5%) in both msh2 pms1 and wild-type strains. Altogether, this physical analysis shows that the msh2 pms1 mutations do not affect the formation of meiotic DSBs.

View larger version (31K): In this window In a new window Download PPT slide   Figure 6. Detection and quantification of meiotic DSBs at the CYS3 locus in wild-type and msh2 pms1 strains in the rad50S background. The DNA from the rad50S (ORD307) or the msh2 pms1 rad50S strain (ORD2047) was digested with HindIII and revealed with the a probe. M, mix of HindIII/XhoI plus HindIII/BglII double-digested genomic DNA used as ladders. The kinetic of accumulation of the CYS3 DSBs during the time course is shown on the graph: ORD307 (solid diamonds) and ORD2047 (open diamonds). The percentage of DSBs at CYS3 is calculated as explained in Figure 3.

View larger version (50K): In this window In a new window Download PPT slide   Figure 7. Analysis of CYS3 DSBs in the wild-type (ORD-2049) or msh2 pms1 strain (ORD2048). DNA was digested with HindIII and probed from each side of the CYS3 DSBs (a or b probe). M is a mix of HindIII/XhoI, HindIII/SmaI, HindIII/PstI, and HindIII/BglII. The molecular weights of these fragments are 3587, 3367, 3048, and 2520 bp, revealed with the a probe, and 3344, 3123, 2789, 2612, and 1274 bp, revealed with the b probe. The vertical open circles indicate the smear corresponding to the DSB fragments at the CYS3 or DEP1 locus. The graph shows a quantitative analysis of the diagnostic fragments from DSBs at the CYS3 locus. The percentage of the labeling in the DSB smears on either side of the DSBs with the a (squares) or b (circles) probe is plotted over time in sporulation medium. Solid symbols, wild-type strain; open symbols, msh2 pms1 strain.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

This article reports on the genetic and molecular characterization of recombination events at the CYS3 locus, a hotspot of meiotic gene conversion. In the wild-type strain background, the meiotic segregation analysis of two new frameshift mutations of CYS3 revealed high levels of gene conversion. The cys3-Bg marker located at the 5' end of the gene exhibits a particularly high level of gene conversion (27%). This value is about three- to fourfold higher than that of a similarly positioned mutation in the ARG4 coding region (7% for the arg4-RV marker), and it is even greater than the peak (17%) of gene conversion observed for ARG4 (SCHULTES et al. 1990). The difference in the absolute levels of gene conversion frequencies between CYS3 and ARG4 can be explained by a higher level of DSB formation in the CYS3 promoter, which is 8 ± 2%, i.e., threefold, higher than at ARG4 (2–3%). The frequency of DSB formation at CYS3 is one of the highest found in a wild-type chromosomal segment (BAUDAT and NICOLAS 1997 ). Our cys3-RI marker, located towards the 3' end of the CYS3 coding region, exhibits a lower frequency of gene conversion (7%), which is likely to result from a 5'–3' decreasing gradient of conversion in CYS3, similar to that of the well-characterized ARG4 (FOGEL et al. 1981 ; NICOLAS et al. 1989 ) and HIS4 (DETLOFF et al. 1992 ) hotspots. This conclusion is based on the study of only two markers. However, it is reinforced by our observation that the difference in conversion frequencies of the cys3-Bg and cys3-RI markers is relieved in MMR mutants (msh2 pms1), as observed previously for the ARG4 and HIS4 conversion gradients (ALANI et al. 1994 ). Further discussion of this phenomenon in the context of our study of meiotic DSBs is presented below. We conclude that CYS3 is a very strong hotspot of meiotic gene conversion, with a 5'–3' gradient that resembles that of ARG4 (FOGEL et al. 1981 ; NICOLAS et al. 1989 ), DED81 (SCHULTES and SZOSTAK 1990 ), and HIS4 (DETLOFF et al. 1992 ), but not that of HIS2, where the gradient is of opposite polarity (3'–5', MALONE et al. 1992 ).

Meiotic recombination hotspots in S. cerevisiae are characterized by the formation of transient DNA DSBs during meiotic prophase. Studies in which cis-acting sequences were deleted, or heterologous sequences were inserted, or in which the effects of trans-acting factors were assessed have demonstrated that these DSBs are an early step in the initiation of recombination (reviewed in NICOLAS and PETES 1994 ; LICHTEN and GOLDMAN 1995 ). In the presence of the rad50S mutation, in which DSBs form but are not processed (ALANI et al. 1990 ), two discrete DSB regions can be detected in the 5' region of the CYS3 locus. The two regions are ~150 ± 50 bp apart and undergo breakage to different extents: the distal DSBs are about two- to threefold more frequent than the proximal DSBs. No significant variation in the ratio of DSB formation at these two sites was observed during the course of meiosis. These results are in agreement with the higher-resolution mapping of DSBs at CYS3 (DE MASSY et al. 1995 ). Two DSB regions are also found at the DED82/DED81 divergent promoters (WU and LICHTEN 1994 ), but they are not observed at the ARG4 locus or at HIS4 (XU and PETES 1996 ).

Interestingly, the two regions of DSB formation upstream of CYS3 are also micrococcal nuclease-hypersensitive sites (OHTA et al. 1994 ). The sensitivity of these sites increases two- to fourfold before the appearance of meiotic DSBs and recombination products (4 hr after transferring cells into sporulation medium). Meiotic hypersensitivity to deoxyribonuclease I has also been observed in the YCR47C/48W region and at the ARG4 locus by WU and LICHTEN 1994 . A meiosis-specific change in chromatin structure has also been detected in the wild-type strains at the HIS4 hotspot (FAN and PETES 1996 ). The relationship between these alterations in chromatin structure and the transcriptional state of the adjacent gene remains unclear, and it varies among recombination hotspots (FAN et al. 1995 ). In the case of CYS3, we have observed a threefold increase in the steady-state level of CYS3 mRNA 1 hr after the shift into sporulation medium (M. Vedel, unpublished data), which is before the increase in sensitivity to micrococcal nuclease. RNA polymerase II holoenzyme has been shown to contain SWI/SNF regulators, which are involved in chromatin remodeling (WILSON et al. 1996 ). The increased level of steady-state CYS3 mRNA may reflect an activation process that could act at the level of chromatin to facilitate the accessibility of the region to the Spo11p nuclease, which is responsible for DSB formation (BERGERAT et al. 1997 ; KEENEY et al. 1997 ).

At the ARG4 locus, with few exceptions (e.g., HA-315), there is a good correlation between DSB levels and the frequencies of NMS, suggesting that most recombination events at this locus are initiated by meiotic DSBs, and that these breaks are repaired by copying from the homologous chromosome (DE MASSY and NICOLAS 1993 ). These data are consistent with the observation that double Holliday junctions (a subsequent recombination intermediate) preferentially form between homologs rather than between sister chromatids (SCHWACHA and KLECKNER 1995 ). In the CYS3 region, DSBs represent a 6–10% cleavage per meiotic DNA molecule. Assuming that the two sister chromatids of each homolog are equally prone to DSB formation, this is approximately what is expected if most NMS events (27–28% involving the cys3-Bg marker, which is ~310 or 420 bp from the DSB sites) are initiated by adjacent DSBs. We observe in one-point crosses a high level of conversion at the distant cys3-RI site (7–12%), and in cis- and trans-two-point crosses, there is a high proportion of coconversion events involving both cys3-Bg and cys3-RI (92 and 83% of NMS tetrads, respectively). These results strongly suggest that conversion at the 3' end of the CYS3 locus also depends on the initiating DSBs located in the intergenic promoter region of CYS3. In this respect, recombination at the CYS3 locus is mechanistically similar to that at ARG4 (SUN et al. 1989 ), with prominent DSB-induced conversion events occurring nearby.

In several instances, "marker effects" have been reported in which a heterozygous state decreases the frequencies of conversion of adjacent markers by affecting the initiation of recombination or the processing of heteroduplex intermediates (reviewed in NICOLAS and ROSSIGNOL 1989 ; KEENEY and KLECKNER 1996 ; ROCCO and NICOLAS 1996 ). As another example, in an inverted repeat assay system, a single mismatch has been found to modify the rate of mitotic recombination between otherwise identical sequences in MMR-competent cells (DATTA et al. 1997 ). In this case, the incidence of crossing over was reduced approximately fourfold. Marker effects increasing the gene conversion frequencies of adjacent markers were also found for the buff-YS17 mutation in Sordaria brevicollis (WHITEHOUSE 1982 ) and in the ade6-M26 mutation in Schizosaccharomyces pombe (ZAHN-ZABAL and KOHLI 1996 and references within). In both cases, these mutations are likely to stimulate the initiation of recombination as they undergo a higher level of NMS than do nearby markers, with respect to what is measured in one-point crosses. The result of our genetic analysis at CYS3 showing that gene conversion frequencies in two-point crosses are higher than in one-point crosses, especially in the case of the trans-heteroallelic (RI +/+ Bg) diploid (P < 0.001), was therefore unexpected. It was not observed in the extensive one-point and multipoint analyses of NMS frequencies at the ARG4 locus (FOGEL et al. 1971 ). Three nonexclusive hypotheses may explain the significant increase in conversion frequencies involving the cys3-Bg and cys3-RI markers when both are present. First, the cysteine auxotrophy of diploids seems to slightly enhance the NMS frequency of only the cys3-RI marker, which cannot explain the observed marker effect. The second possibility is that DSB formation is stimulated. To test this hypothesis, we examined the formation of meiotic DSBs in the absence of heterozygosities (CYS3/CYS3) and in the heteroallelic diploid (cys3-Bg/cys3-RI). In the rad50S background, both strains display a similar DSB frequency in the CYS3 promoter region. Additional DSBs could have resulted from an interaction between homologs before DSB formation (KEENEY and KLECKNER 1996 ; ROCCO and NICOLAS 1996 ). Our physical examination of the pattern of DNA fragments in the CYS3 region or in adjacent regions during meiosis did not reveal any evidence of such an event. We conclude that the presence of both markers in the CYS3 coding region probably does not affect the formation of initiating DSBs. The third and more probable hypothesis is that this marker effect results from changes in the processing of the recombination intermediates at a stage after DSB formation. The classes of tetrads obtained from the trans- and cis-heteroallelic diploids, presented in Figure 2, define two prominent categories of events. First, we measure a high level of coconversion, which contributes to the overall increase in NMS. The cys3-Bg and the cys3-RI markers are ~400 and 1000 bp distant from the DSBs, respectively. In S. cerevisiae, the average meiotic gene conversion tract length has been estimated at 1.5–3.7 kb, depending on the locus (DETLOFF and PETES 1992 ). Our results indicate that the majority of the aberrant events involving the 3' marker could be the result of a heteroduplex that extends through the 5' end of the CYS3 gene. They also suggest that the presence of heterozygosity may influence the extent of heteroduplex formation so that both markers are more likely to be involved without a change in the direction of the mismatch repair parameters. Second, we detect numerous complex events involving more than two chromatids per meiosis. The existence of these exceptional tetrads suggests that the presence of two heterozygous markers on the same tract of heteroduplex DNA might initiate a second round of recombination events promoted by MMR, as proposed by HASTINGS 1984 . Secondary DNA lesions, such as DSBs and gaps created by converging excision repair processes operating on opposite strands of the same heteroduplex intermediate, could then be repaired by interaction of the affected chromatid with its intact sister or nonsister chromatids to yield three or four recombinant products per meiosis. Another "marker effect" (i.e., an increase in nonreciprocal recombination events) has been observed upon introduction of nine heterozygosities located between an artificially created nontandem duplication of the MAT locus in S. cerevisiae (BORTS and HABER 1987 ).

The high frequency of DSB formation at the CYS3 locus allowed us to study the processing of the broken ends in an isogenic RAD50 strain. As described for the ARG4 locus (SUN et al. 1991 ) and the HIS4-LEU2 construct (CAO et al. 1990 ), we observed smeared DSB fragments corresponding to processing intermediates. We used alkaline gels and strand-specific probes to follow the processing of the CYS3 DSBs and observed the production of 3' overhanging tails on either side of the DSBs. In our experimental conditions, these tails are 400–700 nt long at CYS3, with a constant heterogeneity of ~200–300 nt. In the case of the ARG4 locus, SUN et al. 1989 observed an 800-nt 3' overhanging tail with a gradient of resection. The analysis by SUN et al. 1989 was performed with an SK1 strain in which the ARG4 region was carried on a single copy plasmid; they used a nondenaturing Southern blot procedure to examine the DSB fragments. Our results with CYS3 are more similar to those reported by BISHOP et al. 1992 for the HIS4-LEU2 hotspot. Using alkaline gels, they determined that in the SK1 background, the 5' strands are always at least ~600 nt shorter than the 3' strands. At CYS3, we observed that the 5' strand is at least 400 nt shorter than the 3' strand. Consequently, the cys3-Bg marker, located ~400 bp from the DSBs, is included in the region of single-strand resection, which could explain its high frequency of gene conversion. The mechanism and the protein(s) that create the single-strand tail and control its length remain to be discovered (KEENEY et al. 1997 ).

Finally, an important result of our characterization of the CYS3 hotspot of meiotic gene conversion is a confirmation that the MMR genes are involved in determining the frequency of meiotic gene conversion, as reported by ALANI et al. 1994 . At the CYS3 locus, we observed in the msh2 pms1 mutant strains an important increase of NMS frequency at the cys3-RI marker (in more than 25% of the meioses), while the NMS frequency at the cys3-Bg marker is significantly diminished (18% instead of 27%; P = 0.034). Both modifications in NMS abolish the gradient of NMS observed in the wild-type strain. We don't know if a modification of the ratio of conversion-type to restoration-type repair, as observed for the HIS4 locus, contributes to the observed polarity gradient in the wild-type strain at CYS3. If this is the case, a steep gradient may reflect an excess of restoration-type over conversion-type repair at the 3' end of CYS3. Another interpretation of this effect, as proposed by ALANI et al. 1994 , is that MMR regulates both the length of heteroduplex DNA and the formation of symmetric heteroduplex DNA. In our msh2 pms1 strains, the frequency of aberrant 4:4 tetrads at CYS3 is similar (<1%) to that at ARG4 and much lower than that at HIS4 (ALANI et al. 1994 ). This result indicates that aberrant 4:4 more likely results from multiple independent events rather than from the processing of DNA intermediates with regions of symmetric heteroduplex.

Our physical analysis of meiotic DSB formation in MMR mutants allowed us to test whether MMR proteins are directly or indirectly involved in the formation of DSBs or in the processing of gene conversion intermediates, apart from their function in the recognition and correction of heteroduplexes containing mispaired bases. Our results indicate that the frequency and the positioning of DSBs in a msh2 pms1 rad50S mutant strain is similar (7 ± 1.5%) to that found in the rad50S strain. We conclude that MMR proteins are not involved in DSB formation. MSH2 and PMS1 could also be implicated indirectly in the resection of DSBs, which would affect both recombination and DSB repair. Using neutral agarose gels, we find that both the intensity of the transient signal and the kinetics are similar in wild-type and msh2 pms1 strains. We have also analyzed the processing of the 5' and 3' strands on alkaline agarose gels with strand-specific riboprobes. In six strains differing with respect to their allelic composition at CYS3 (CYS3/CYS3, cys3-Bg/CYS3, and cys3-RI/CYS3), and either wild-type or msh2 pms1, the 3' strand exhibits the same length as in the rad50S strains (data not shown). We have also examined the 5' strand. Under our experimental conditions, we did not detect convincing differences in the length of the resected strand among these strains.

The Msh2 protein has been recently shown to bind to synthetic Holliday junctions (ALANI et al. 1997 ), which suggests that another stage at which the MMR system could influence recombination is in the formation and resolution of Holliday junctions (for a review see STAHL 1996 ).

	ACKNOWLEDGMENTS

We thank Y. Surdin-Kerjean and L. Grivell or providing pStr1.1 and PM5239 plasmids, respectively, all members of our laboratory for helpful discussion, F. Fabre and G. Simchen for comments about the manuscript, and K. Smith for English grammar corrections. We also thank S. Le Bilcot for her technical assistance. This work was successively performed in the Centre National de la Recherche Scientifique URA1354, URA1292, and UMR144 and has been supported by grants from the Groupement de Recherches et d'Etudes sur les Génomes, the Association de Recherche Contre le Cancer, the Ligue Nationale contre le Cancer, the A.C.C. SV8 program from the Ministère de l'Education Nationale, de l'Enseignement Supérieur, et de la Recherche, the Human Frontier Science Programs (RG493/95), and the European Community (Human Capital Mobility Meiosis Network and Biotech program).

Manuscript received September 10, 1997; Accepted for publication September 23, 1998.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALANI, A., 1996  The Saccharomyces cerevisiae Msh2 and Msh6 proteins form a complex that binds specifically to duplex oligonucleotides containing mismatched DNA base pairs. Mol. Cell. Biol. 16:5604-5615[Abstract].

ALANI, E., R. PADMORE, and N. KLECKNER, 1990  Analysis of wild type and rad50 mutants of yeast suggests an intimate relationship between meiotic chromosome synapsis and recombination. Cell 61:419-436[Medline].

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

ALANI, E., S. LEE, M. F. KANE, J. GRIFFITH, and R. D. KOLODNER, 1997  Saccharomyces cerevisiae MSH2, a mispaired base recognition protein, also recognizes Holliday junctions in DNA. J. Mol. Biol. 265:289-301[Medline].

BARTON, A. B., D. B. KABACK, M. W. CLARK, T. KENG, and B. F. F. OUELLETTE et al., 1993  Physical localization of yeast CYS3, a gene whose product resembles the rat -cystathionase and Escherichia coli cystathionine -synthase enzymes. Yeast 9:363-369[Medline].

BAUDAT, F. and A. NICOLAS, 1997  Clustering of meiotic double-strand breaks on yeast chromosome III. Proc. Natl. Acad. Sci. USA 94:5213-5218[Abstract/Free Full Text].

BERGERAT, A., B. DE MASSY, D. GADELLE, P.-C. VAROUTAS, and A. NICOLAS et al., 1997  An atypical topoisomerase II from archaea with implications for meiotic recombination. Nature 386:414-417[Medline].

BISHOP, D. K., D. PARK, L. XU, and N. KLECKNER, 1992  DMC1: a meiosis-specific yeast homolog of E. coli recA required for recombination, synaptonemal complex formation, and cell cycle progression. Cell 69:439-456[Medline].

BOEKE, J. D., F. LACROUTE, and G. R. FINK, 1984  A positive selection for mutants lacking orotidine-5'-phosphate decarboxylase activity in yeast: 5-fluoroacetic acid resistance. Mol. Gen. Genet. 197:345-346[Medline].

BORTS, R. H. and J. E. HABER, 1987  Meiotic recombination in yeast: alteration by multiple heterozygosities. Science 237:1459-1465[Abstract/Free Full Text].

BULLARD, S. A., S. KIM, A. M. GALBRAITH, and R. E. MALONE, 1996  Double strand breaks at the HIS2 recombination hotspot in Saccharomyces cerevisiae.. Proc. Natl. Acad. Sci. USA 93:13054-13059[Abstract/Free Full Text].

CAO, L., E. ALANI, and N. KLECKNER, 1990  A pathway for generation and processing of double-strand breaks during meiotic recombination in S. cerevisiae.. Cell 61:1089-1101[Medline].

CHEREST, H. and Y. SURDIN-KERJAN, 1992  Genetic analysis of a new mutation conferring cysteine auxotrophy in Saccharomyces cerevisiae: updating of the sulfur metabolism pathway. Genetics 130:51-58[Abstract].

CHURCH, G. M. and W. GILBERT, 1984  Genomic sequencing. Proc. Natl. Acad. Sci. USA 81:1991-1995[Abstract/Free Full Text].

DATTA, A., M. HENDRIX, M. LIPSITCH, and S. JINKS-ROBERTSON, 1997  Dual roles for DNA sequence identity and the mismatch repair system in the regulation of mitotic crossing-over in yeast. Proc. Natl. Acad. Sci. USA 94:9757-9762[Abstract/Free Full Text].

DE MASSY, B. and A. NICOLAS, 1993  The control in cis of the position and the amount of the ARG4 meiotic double-strand break of Saccharomyces cerevisiae.. EMBO J. 12:1459-1466[Medline].

DE MASSY, B., V. ROCCO, and A. NICOLAS, 1995  The nucleotide mapping of DNA double-strand breaks at the CYS3 initiation site of meiotic recombination in Saccharomyces cerevisiae.. EMBO J. 14:4589-4598[Medline].

DETLOFF, P. and T. D. PETES, 1992  Measurements of excision repair tracts formed during meiotic recombination in Saccharomyces cerevisiae. Mol. Cell. Biol. 12:1805-1814[Abstract/Free Full Text].

DETLOFF, P., M. A. WHITE, and T. D. PETES, 1992  Analysis of a gene conversion gradient at the HIS4 locus in Saccharomyces cerevisiae.. Genetics 132:113-123[Abstract].

FAN, Q.-Q. and T. D. PETES, 1996  Relationship between nuclease-hypersensitive sites and meiotic recombination hotspot activity at the HIS4 locus of Saccharomyces cerevisiae.. Mol. Cell. Biol. 16:2037-2043[Abstract].

FAN, Q.-Q., F. XU, and T. D. PETES, 1995  Meiosis-specific double-strand DNA breaks at the HIS4 recombination hotspot in the yeast Saccharomyces cerevisiae: control in cis and trans. Mol. Cell. Biol. 15:1679-1688[Abstract].

FOGEL, S., D. D. HURST and R. K. MORTIMER, 1971 Gene conversion in unselected tetrads from multipoint crosses. The Second Stadler Symposia pp. 89–110.

FOGEL, S., R. K. MORTIMER and K. LUSNAK, 1981 Mechanism of meiotic gene conversion or "wandering on a foreign strand," pp. 289–339 in The Molecular Biology of Yeast Saccharomyces, edited by J. N. STRATHERN, J. E. JONES and J. BROACH. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

GOLDWAY, M., A. SHERMAN, D. ZENVIRTH, T. ARBEL, and G. SIMCHEN, 1993  A short chromosomal region with major roles in yeast chromosome III meiotic disjunction, recombination and double strand breaks. Genetics 133:159-169[Abstract].

HASTINGS, P. J., 1984  Measurement of restoration and conversion: its meaning for the mismatch repair hypothesis of conversion. Cold Spring Harbor Symp. Quant. Biol. 49:49-53[Abstract/Free Full Text].

KEENEY, S. and N. KLECKNER, 1996  Communication between homologous chromosomes: genetic alterations at a nuclease-hypersensitive site can alter mitotic chromatin structure at that site both in cis and in trans.. Genes Cells 1:475-489[Abstract].

KEENEY, S., C. GIROUX, and N. KLECKNER, 1997  Meiosis-specific DNA double-strand breaks are catalyzed by Spo11, a member of a widely conserved protein family. Cell 88:375-384[Medline].

KIRKPATRICK, D. T., M. DOMINSKA, and T. D. PETES, 1998  Conversion-type and restoration-type repair of DNA mismatches formed during meiotic recombination in Saccharomyces cerevisiae.. Genetics 149:1693-1705[Abstract/Free Full Text].

KLEIN, S., D. ZENWIRTH, V. DROR, A. B. BARTON, and D. B. KABACK et al., 1996  Patterns of meiotic double-strand breakage on native and artificial yeast chromosomes. Chromosoma 105:205-284.

KRAMER, W., B. KRAMER, M. WILLIAMSON, and S. FOGEL, 1989  Cloning and nucleotide sequence of DNA mismatch repair gene PMS1 from Saccharomyces cerevisiae: homology of PMS1 to procaryotic MutL and HexB. J. Bacteriol. 171:5339-5346[Abstract/Free Full Text].

LICHTEN, M. and A. S. H. GOLDMAN, 1995  Meiotic recombination hotspots. Annu. Rev. Genet. 29:423-444[Medline].

LICHTEN, M., C. GOYON, N. P. SCHULTES, D. TRECO, and J. W. SZOSTAK et al., 1990  Detection of heteroduplex DNA molecules among the products of Saccharomyces cerevisiae meiosis. Proc. Natl. Acad. Sci. USA 87:7653-7657[Abstract/Free Full Text].

MALONE, R. E., S. BULLARD, S. LUNDQUIST, S. KIM, and T. TARKOWSKI, 1992  A meiotic gene conversion gradient opposite to the direction of transcription. Nature 359:154-155[Medline].

MALONE, R. E., S. KIM, S. A. BULLARD, S. LUNDSQUIST, and L. HUTCHINGS-CROW et al., 1994  Analysis of a recombination hotspot for gene conversion occurring at the HIS2 gene of Saccharomyces cerevisiae.. Genetics 137:5-18[Abstract].

MAO-DRAAYER, Y., A. M. GALBRAITH, D. L. PITTMAN, M. COOL, and R. E. MALONE, 1996  Analysis of meiotic recombination pathways in Saccharomyces cerevisiae.. Genetics 144:71-86[Abstract].

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].

NAG, D. K. and T. D. PETES, 1993  Physical detection of heteroduplexes during meiotic recombination in the yeast Saccharomyces cerevisiae.. Mol. Cell. Biol. 13:2324-2331[Abstract/Free Full Text].

NICOLAS, A. and T. D. PETES, 1994  Polarity of meiotic gene conversion in fungi: contrasting views. Experientia 50:242-252[Medline].

NICOLAS, A. and J. L. ROSSIGNOL, 1989  Intermediates in homologous recombination revealed by marker effects in Ascobolus.. Genome 31:528-535.

NICOLAS, A., D. TRECO, N. P. SCHULTES, and J. W. SZOSTAK, 1989  An initiation site for meiotic gene conversion in the yeast Saccharomyces cerevisiae.. Nature 338:35-39[Medline].

OHTA, K., T. SHIBATA, and A. NICOLAS, 1994  Changes in chromatin structure at recombination initiation sites during yeast meiosis. EMBO J. 13:5754-5763[Medline].

ONO, B. I., K. TANAKA, K. NAITO, C. HEIKE, and S. SHINODA et al., 1992  Cloning and characterization of the CYS3 (CYI1) gene of Saccharomyces cerevisiae.. J. Bacteriol. 174:3339-3347[Abstract/Free Full Text].

OUELLETTE, B. F., M. W. CLARK, T. KENG, R. G. STORMS, and W. ZHONG et al., 1993  Sequencing of chromosome I from Saccharomyces cerevisiae: analysis of a 32 kb region between the LTE1 and SPO7 genes. Genome 36:32-42[Medline].

PETES, T. D., R. E. MALONE and L. S. SYMINGTON, 1991 Recombination in yeast, pp. 407–521 in The Molecular and Cellular Biology of the Yeast Saccharomyces, edited by R. BROACH, E. W. JONES and J. R. PRINGLE. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

PROLLA, T. A., Q. PANG, E. ALANI, R. D. KOLODNER, and R. M. LISKAY, 1994  MLH1, PMS1, and MSH2 interactions during the initiation of DNA mismatch repair in yeast. Science 265:1091-1093[Abstract/Free Full Text].

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

REENAN, R. and R. KOLODNER, 1992b  Characterization of insertion mutation in the Saccharomyces cerevisiae MSH1 and MSH2 genes: evidence for separate mitochondrial and nuclear functions. Genetics 132:975-985[Abstract].

ROCCO, V. and A. NICOLAS, 1996  Sensing of nonhomology lowers the initiation of meiotic recombination in yeast. Genes Cells 1:645-661[Abstract].

SCHERER, S. and R. W. DAVIS, 1979  Replacement of chromosome segments with altered DNA sequences constructed in vitro.. Proc. Natl. Acad. Sci. USA 76:4951-4955[Abstract/Free Full Text].

SCHULTES, N. P. and J. W. SZOSTAK, 1990  Decreasing gradients of gene conversion on both sides of the initiation site for meiotic recombination at the ARG4 locus in yeast. Genetics 126:813-822[Abstract].

SCHWACHA, A. and N. KLECKNER, 1995  Identification of double Holliday junctions as intermediates in meiotic recombination. Cell 83:783-791[Medline].

STAHL, F., 1996  Meiotic recombination in yeast: coronation of the double-strand break repair model. Cell 87:965-968[Medline].

SUN, H., D. TRECO, N. P. SCHULTES, and J. W. SZOSTAK, 1989  Double-strand breaks at an initiation site for meiotic gene conversion. Nature 338:87-90[Medline].

SUN, H., D. TRECO, and J. W. SZOSTAK, 1991  Extensive 3'-overhanging, single-stranded DNA associated with the meiosis-specific double-strand breaks at the ARG4 recombination initiation site. Cell 64:1155-1161[Medline].

SZOSTAK, J. W., T. L. ORR-WEAVER, R. J. ROTHSTEIN, and F. W. STAHL, 1983  The double-strand break repair model for conversion and crossing-over. Cell 33:25-35[Medline].

WHITEHOUSE, H. L. K., 1982 Genetic Recombination, Understanding the Mechanisms. John Wiley & Sons, Chichester.

WILSON, C., J. D. M. CHAO, A. N. IMBALZANO, G. R. SCHNITZLER, and R. E. KINGSTON et al., 1996  RNA polymerase II holoenzyme contains SWI/SNF regulators involved in chromatin remodeling. Cell 88:235-244.

WU, T.-C. and M. LICHTEN, 1994  Meiosis-induced double-strand break sites determined by yeast chromatin structure. Science 263:515-518[Abstract/Free Full Text].

XU, F. and T. D. PETES, 1996  Fine-structure mapping of meiosis-specific double-strand DNA breaks at a recombination hotspot associated with an insertion of telomeric sequence upstream of the HIS4 locus in yeast. Genetics 143:1115-1125[Abstract].

ZAHN-ZABAL, M. and J. KOHLI, 1996  The distance dependence of the fission yeast ade6-M26 marker effect in two-factor crosses. Curr. Genet. 29:530-536[Medline].

ZENVIRTH, D., T. ARBEL, A. SHERMAN, M. GOLWAY, and S. KLEIN et al., 1992  Multiple sites for double-strand breaks in whole meiotic chromosomes of Saccharomyces cerevisiae.. EMBO J. 11:3441-3447[Medline].

This article has been cited by other articles:

				H. Murakami and A. Nicolas Locally, Meiotic Double-Strand Breaks Targeted by Gal4BD-Spo11 Occur at Discrete Sites with a Sequence Preference Mol. Cell. Biol., July 1, 2009; 29(13): 3500 - 3516. [Abstract] [Full Text] [PDF]

				K. F. Cooper, M. J. Mallory, V. Guacci, K. Lowe, and R. Strich Pds1p Is Required for Meiotic Recombination and Prophase I Progression in Saccharomyces cerevisiae Genetics, January 1, 2009; 181(1): 65 - 79. [Abstract] [Full Text] [PDF]

				H. Sasanuma, K. Hirota, T. Fukuda, N. Kakusho, K. Kugou, Y. Kawasaki, T. Shibata, H. Masai, and K. Ohta Cdc7-dependent phosphorylation of Mer2 facilitates initiation of yeast meiotic recombination Genes & Dev., February 1, 2008; 22(3): 398 - 410. [Abstract] [Full Text] [PDF]

				L. Kauppi, M. Jasin, and S. Keeney Meiotic crossover hotspots contained in haplotype block boundaries of the mouse genome PNAS, August 14, 2007; 104(33): 13396 - 13401. [Abstract] [Full Text] [PDF]

				H. Sasanuma, H. Murakami, T. Fukuda, T. Shibata, A. Nicolas, and K. Ohta Meiotic association between Spo11 regulated by Rec102, Rec104 and Rec114 Nucleic Acids Res., February 28, 2007; 35(4): 1119 - 1133. [Abstract] [Full Text] [PDF]

				J. E. Stone and T. D. Petes Analysis of the Proteins Involved in the in Vivo Repair of Base-Base Mismatches and Four-Base Loops Formed During Meiotic Recombination in the Yeast Saccharomyces cerevisiae Genetics, July 1, 2006; 173(3): 1223 - 1239. [Abstract] [Full Text] [PDF]

				A. Antonell, O. de Luis, X. Domingo-Roura, and L. A. Perez-Jurado Evolutionary mechanisms shaping the genomic structure of the Williams-Beuren syndrome chromosomal region at human 7q11.23 Genome Res., September 1, 2005; 15(9): 1179 - 1188. [Abstract] [Full Text] [PDF]

				L. Maloisel, J. Bhargava, and G. S. Roeder A Role for DNA Polymerase {delta} in Gene Conversion and Crossing Over During Meiosis in Saccharomyces cerevisiae Genetics, July 1, 2004; 167(3): 1133 - 1142. [Abstract] [Full Text] [PDF]

				S. J. Haring, G. R. Halley, A. J. Jones, and R. E. Malone Properties of Natural Double-Strand-Break Sites at a Recombination Hotspot in Saccharomyces cerevisiae Genetics, September 1, 2003; 165(1): 101 - 114. [Abstract] [Full Text] [PDF]

				J. L. Argueso, A. W. Kijas, S. Sarin, J. Heck, M. Waase, and E. Alani Systematic Mutagenesis of the Saccharomyces cerevisiae MLH1 Gene Reveals Distinct Roles for Mlh1p in Meiotic Crossing Over and in Vegetative and Meiotic Mismatch Repair Mol. Cell. Biol., February 1, 2003; 23(3): 873 - 886. [Abstract] [Full Text]

				C. Soustelle, M. Vedel, R. Kolodner, and A. Nicolas Replication Protein A Is Required for Meiotic Recombination in Saccharomyces cerevisiae Genetics, June 1, 2002; 161(2): 535 - 547. [Abstract] [Full Text] [PDF]

				J. L. Argueso, D. Smith, J. Yi, M. Waase, S. Sarin, and E. Alani Analysis of Conditional Mutations in the Saccharomyces cerevisiae MLH1 Gene in Mismatch Repair and in Meiotic Crossing Over Genetics, March 1, 2002; 160(3): 909 - 921. [Abstract] [Full Text] [PDF]

				J. A. Clikeman, S. L. Wheeler, and J. A. Nickoloff Efficient Incorporation of Large (>2 kb) Heterologies Into Heteroduplex DNA: Pms1/Msh2-Dependent and -Independent Large Loop Mismatch Repair in Saccharomyces cerevisiae Genetics, April 1, 2001; 157(4): 1481 - 1491. [Abstract] [Full Text]

				T.-F. Wang, N. Kleckner, and N. Hunter Functional specificity of MutL homologs in yeast: Evidence for three Mlh1-based heterocomplexes with distinct roles during meiosis in recombination and mismatch correction PNAS, November 23, 1999; 96(24): 13914 - 13919. [Abstract] [Full Text] [PDF]