Properties of Natural Double-Strand-Break Sites at a Recombination Hotspot in Saccharomyces cerevisiae

Corresponding author: Robert E. Malone, 204 Biology Bldg. East, Iowa City, IA 52242., robert-malone{at}uiowa.edu (E-mail)

Communicating editor: A. NICOLAS

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

This study addresses three questions about the properties of recombination hotspots in Saccharomyces cerevisiae: How much DNA is required for double-strand-break (DSB) site recognition? Do naturally occurring DSB sites compete with each other in meiotic recombination? What role does the sequence located at the sites of DSBs play? In S. cerevisiae, the HIS2 meiotic recombination hotspot displays a high level of gene conversion, a 3'-to-5' conversion gradient, and two DSB sites located 550 bp apart. Previous studies of hotspots, including HIS2, suggest that global chromosome structure plays a significant role in recombination activity, raising the question of how much DNA is sufficient for hotspot activity. We find that 11.5 kbp of the HIS2 region is sufficient to partially restore gene conversion and both DSBs when moved to another yeast chromosome. Using a variety of different constructs, studies of hotspots have indicated that DSB sites compete with one another for DSB formation. The two naturally occurring DSBs at HIS2 afforded us the opportunity to examine whether or not competition occurs between these native DSB sites. Small deletions of DNA at each DSB site affect only that site; analyses of these deletions show no competition occurring in cis or in trans, indicating that DSB formation at each site at HIS2 is independent. These small deletions significantly affect the frequency of DSB formation at the sites, indicating that the DNA sequence located at a DSB site can play an important role in recombination initiation.

DNA double-strand breaks (DSBs) initiate meiotic recombination in the budding yeast Saccharomyces cerevisiae (SUN et al. 1989 ; reviewed in LICHTEN and GOLDMAN 1995 and in PAQUES and HABER 1999 ). Meiotic recombination does not occur randomly along chromosomes; there are preferred regions in which recombination initiates more frequently than on average (GAME 1992 ; GERTON et al. 2000 ; reviewed in PETES 2001 ). These regions have been termed recombination hotspots. Several natural meiotic recombination hotspots have been studied in detail, including ARG4 (e.g., NICOLAS et al. 1989 ; SUN et al. 1989 , SUN et al. 1991 ; WU and LICHTEN 1995 ), HIS4 (e.g., WHITE et al. 1991 ; DETLOFF et al. 1992 ; FAN et al. 1995 ), CYS3 (e.g., CHEREST and SURDIN-KERJAN 1992 ; DE MASSY et al. 1995 ; VEDEL and NICOLAS 1999 ), and HIS2 (e.g., MALONE et al. 1992 , MALONE et al. 1994 ; BULLARD et al. 1996 ). Each of these hotspots shares the following features: a high level of gene conversion, a conversion polarity gradient, and a high frequency of meiosis-specific DSBs. Another recombination hotspot, the HIS4::LEU2 hotspot, created by the insertion of 77 bp of bacterial DNA during the fusion of HIS4 and LEU2, has also been extensively studied (e.g., CAO et al. 1990 ; SCHWACHA and KLECKNER 1994 ; XU and KLECKNER 1995 ).

That the larger context surrounding a hotspot is important is indicated by many experiments (e.g., WU and LICHTEN 1995 ; BORDE et al. 1999 ), including some examining the HIS2 hotspot (MALONE et al. 1994 ). Insertions and deletions on chromosome VI (the location of HIS2) several kilobases from the DSB sites affect both conversion and DSB formation, and the movement of 5.2 kbp of the HIS2 region to the middle of the right arm of chromosome II is not sufficient to restore any hotspot activity (MALONE et al. 1994 ). These observations led us to propose that large amounts of DNA might be required to allow recognition of recombination initiation sites (MALONE et al. 1994 ). The general consensus from other hotspots also supports the role of long-range chromosome structure. For example, Lichten and colleagues moved a URA3::ARG4::pBR322 construct to several different locations on chromosome III and measured recombination and DSB formation (WU and LICHTEN 1995 ; BORDE et al. 1999 ). They concluded that the normal ARG4 DSB site is not determined solely by either DNA sequence at the DSB site or localized chromatin structure, and they proposed that hotspots can be affected by sequences several kilobases away. In this article we address the question of how much DNA is sufficient to retain hotspot activity by extending the previous 5.2-kbp HIS2 construct on chromosome II.

The URA3::ARG4::pBR322 construct mentioned above displays another interesting property. WU and LICHTEN 1995 discovered that the presence of pBR322 sequences in the construct creates several new sites for meiotic DSB formation; in the construct, meiotic breaks at the nearby ARG4 DSB site are simultaneously reduced. Removal of the pBR322 sequences restores breaks at the insert-borne ARG4 DSB site (WU and LICHTEN 1995 ; OHTA et al. 1999 ). They proposed that competition for DSB formation was occurring and that it may be due to one site's ability to inhibit loading of a recombination initiation complex at the other DSB site (WU and LICHTEN 1995 ; OHTA et al. 1999 ). Competition for DSB formation has been observed at other hotspots. At the HIS4::LEU2 hotspot, DSB site competition both within and between homologous chromosomes has been reported (XU and KLECKNER 1995 ). These authors suggested that there may be steric inhibition (the steric exclusion hypothesis) of the DSB machinery for sites 2 kbp apart (XU and KLECKNER 1995 ). Finally, a 49-bp telomeric repeat sequence inserted into the middle of a chromosome at HIS4 (a few hundred base pairs from the normal DSB site) creates a novel and very active DSB site (FAN et al. 1997 ). Competition between the normal DSB site and the telomeric insertion DSB site was observed in chromosomes containing both (FAN et al. 1997 ).

The HIS2 hotspot contains two sites of meiotic DSB formation located relatively close to one another: One site (DSB-B) is located downstream at +1200 with respect to the start of the coding region, and the other (DSB-C) is located within the coding region at +680 (BULLARD et al. 1996 ). The fact that both DSB sites occur in native yeast DNA makes the HIS2 hotspot an efficacious tool for investigating competition. To examine competition between the two sites, we made small deletions at the DSB sites and examined competition in cis and in trans. If there is competition between break sites at HIS2, one would predict that an increase in DSB formation at one site would inhibit DSB formation at the nearby site and vice versa.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Yeast strains, bacterial strains, and plasmids utilized in this study:
All of the diploid strains used for the experiments described here are derived from RM96-15A, RM182-55C, or RM113-4B, the same strains used in previous studies of the HIS2 hotspot (MALONE et al. 1994 ; BULLARD et al. 1996 ). The genotypes of all yeast strains used are listed in Table 1. Bacterial strains DH5 or DH10B were used for molecular manipulations, with the exception of BMH71-18 mutS (Promega, Madison, WI), which was used for in vitro site-directed mutagenesis. The creation of plasmids used in this study is described in the remainder of the MATERIALS AND METHODS, as well as in Table 2.

Extension of the HIS2 region on chromosome II:
A 2.2-kbp BglII-HindIII fragment from the plasmid pRM94, containing DNA downstream of LYS2 (+5256 to +7475 with respect to the start of the LYS2 coding region), was cloned into YIplac211 (GIETZ and SUGINO 1988 ) digested with BamHI and HindIII to create pGRH1. A 7.3-kbp EcoRI-PmlI fragment from c9146 (ATCC no. 70879), a cosmid containing the entire HIS2 coding region and 40 kbp from downstream DNA, was then cloned into the EcoRI and SmaI sites of pGRH1 to create pGRH2. pGRH2 consists of 1.1 kbp homologous to the downstream end of the 5.2-kbp HIS2 BglII fragment, 6.2 kbp of extended chromosome VI sequences, and 2.2 kbp of chromosome II sequence. Yeast spheroplasts were transformed with pGRH2, and two-step gene replacement (ROTHSTEIN 1991 ) was used to create haploids containing the construct. A 3.3-kbp EcoRI fragment containing HIS2 had been previously deleted from chromosome VI in these strains to eliminate ectopic recombination of HIS2 (MALONE et al. 1994 ). Proper integration and excision were verified by Southern analysis (SOUTHERN 1975 ). Integration of sequences contained on pGRH2 results in the extension of the 5.2-kbp BglII fragment on chromosome II to 11.5 kbp. These haploid strains were subsequently used to create the diploid GRH2-6.

Analysis of double-strand breaks at HIS2:
For the study of DSBs in the region of the HIS2 locus, the rad50-KI81 (rad50S) mutation was utilized (ALANI et al. 1990 ). This mutation was introduced into haploid strains by one-step gene replacement (ROTHSTEIN 1991 ) using the plasmid pNKY349 as described by ALANI et al. 1990 . All transformants were verified genetically by methyl methanesulfonate sensitivity and by Southern analysis. DNA for analysis of DSB formation was isolated as in BULLARD et al. 1996 . All imaging and quantitation was performed using a Molecular Dynamics (Sunnyvale, CA) PhosphorImager (model 445SI) and ImageQuant software.

Deletions at double-strand-break sites B and C:
From previous studies, the breaks at HIS2 were mapped to +1200 ± 50 bp for DSB-B and +680 ± 50 bp for DSB-C (BULLARD et al. 1996 ). A 2.6-kbp HindIII-EcoRI fragment containing HIS2 (-358 to +2284 relative to the start of the coding region) was cloned into plasmid pRS306 (SIKORSKI and HIETER 1989 ) digested with HindIII and EcoRI. This new plasmid, pSJH1, was then subjected to in vitro site-directed mutagenesis to create 50-bp deletions at each DSB site. This was done by creating the following 5' phosphorylated oligonucleotides (Integrated DNA Technologies) to remove sequences at DSB-B and DSB-C, respectively: 244 (5'-TATACTCTTTCATGTTGAACTTCCTAAAGAAAGACAATTTGGGAAC-3') and 243 (5'-GCGGACGTATTGATTTCAATTGCGCCTGGCCATTCACTGATGACGTCCAG-3'). Using these primers and pSJH1 with the Gene Editor in vitro site-directed mutagenesis system (Promega), we created pSJH6 (deletion at DSB-B), pSJH5 (deletion at DSB-C), and pSJH7 (deletions at both sites). These deletions will henceforth be referred to as B, C, and BC, respectively. Proper deletions were confirmed by restriction mapping and by DNA sequencing. These deletion constructs were integrated into the haploid parents of RM193 via transformation and two-step gene replacement. Correct integration of the deletions into the chromosome was confirmed by Southern analysis. All homozygous and heterozygous deletion diploids were created by crossing the appropriate haploid parents.

Allelism testing:
To measure gene conversion at his2-390 or his2-C, allelism testing was performed as in MALONE et al. 1994 .

Creation of a restriction-site polymorphism to determine chromosome-specific DSBs:
To distinguish DSB formation at each site on each chromosome, a new BglII restriction site was created upstream of HIS2 (-1580 relative to the start of the coding region). This was done by site-directed mutagenesis using polymerase chain reaction (PCR) overlap extension (HO et al. 1989 ). First, two PCR fragments with overlapping ends containing the new BglII site were generated using the following primers: 350 (5'-GAATCCAATGAGATCTTTATCCTAGTATAG-3'), 351 (5'-CTATACTAGGATAAAGATCTCATTGGATTC-3'), 349 (5'-CGGTATGATGGGAAAATAGGTG-3'), and 352 (5'-TATCGTATTCGCAGCTAGAGG-3'). One fragment was created using primers 349 and 350, and the other fragment was created using primers 351 and 352, both using pH21 (MALONE et al. 1994 ) as the DNA template. Five nanograms of each purified fragment were used as a DNA template, and primers 349 and 352 were used to "stitch" the two fragments together. This PCR fragment was further amplified using primers 358 (5'-CTGAGATTATACCGTTGTTAATGC-3') and 359 (5'-TAAGGAGAAACATCAATAAGAGAATAG-3'). The resulting PCR fragment was then cut with XhoI and EcoRI and cloned into pRS306 cut with the same restriction enzymes to create pSJH15. The clone was confirmed by restriction digests and the new BglII site was subsequently integrated into the genome via two-step gene replacement. Proper integration was confirmed by Southern analysis. Diploid strains were created by crossing the appropriate haploids.

Medium base pair resolution mapping of double-strand-break sites at HIS2:
Double-strand-break DNA was isolated as described previously (BULLARD et al. 1996 ). Four micrograms of this DNA was then digested with HpaI and run on a 6% (19:1 mono:bis) polyacrylamide gel. The DNA was semidry electrotransferred to Hybond N+ (Amersham Biosciences), and the membrane was UV crosslinked at 60,000 J · m/sec² and baked at 80° for 45 min. PCR was used to generate the DNA fragments necessary to probe for DSB-B and DSB-C. Using pH21 as the DNA template, primers 366 (5'-AACTGTAAAGAGATTACCCATTTCGC-3') and 395 (5'-ATAGTTTATACTCTTTCATGTTGAACTTCC-3') were used to generate a 302-bp fragment [DSB-B upstream probe (BU); +947 to +1247] to probe DSB-B; primers 365 (5'-AACAGATTCTCTGGTGATTGGC-3') and 389 (5'-AATCGGGCAACTGCAACGAAG-3') were used to generate a 359-bp fragment [DSB-C downstream probe (CD); +590 to +947] to probe DSB-C. Probes were labeled with the random primer labeling system (Invitrogen, San Diego) using 200 µCi [³²P]dATP and 200 µCi [³²P]dCTP per reaction, and probes BU and CD were used to measure the breaks at DSB-B and DSB-C, respectively. We estimate that the resolution of medium base pair mapping allows for determination of the positions of DSBs within ±8 bp.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

An 11.5-kbp fragment containing HIS2 is sufficient for significant hotspot activity at HIS2:
To begin to determine the extent of DNA sufficient for hotspot activity at HIS2, we had previously inserted a 5.2-kbp BglII fragment containing HIS2 adjacent to LYS2 on chromosome II. The diploid was also deleted for HIS2 on chromosome VI. This 5.2-kbp BglII fragment contains the complete HIS2 coding region, both DSB sites, 1.9 kbp of upstream sequence, and 2.3 kbp of downstream sequence (MALONE et al. 1994 ). HIS2 gene conversion occurs at low levels (1.2%) in this construct (MALONE et al. 1994 ). To examine meiotic DSBs in this strain, we created an isogenic homozygous rad50S diploid (GRH1-2). There is no detectable DSB formation (data not shown). The failure to detect DSBs, combined with the low frequency of conversion (<1.2%), indicates that the 5.2-kbp BglII fragment present on chromosome II containing HIS2 is not sufficient to cause hotspot activity, even though all local sequences and the DSB sites were present.

Since sequences >1 kbp downstream of HIS2 on chromosome VI are necessary for normal hotspot activity (MALONE et al. 1994 ), we asked whether extending the amount of chromosome VI DNA on chromosome II could restore hotspot activity. We extended the insertion on chromosome II to 11.5 kbp, which resulted in 8.5 kbp of DNA downstream of the HIS2 coding region (Fig 1A). Genetic analysis reveals that hotspot activity at HIS2 is restored to 55% of the normal level on chromosome VI in a diploid (GRH2-6) homozygous for this 11.5-kbp fragment (7.7% gene conversion compared to 14.1%; Table 3). To analyze DSBs in the 11.5-kbp construct, DNA from an isogenic rad50S diploid (GRH4-1) was isolated during meiosis. Breaks were observed at the locations in which they normally occur on chromosome VI (DSB-B at +1200 and DSB-C at +680), although at a lower frequency (the average of four experiments was 1.7%; Fig 1B).

View larger version (42K): In this window In a new window Download PPT slide   Figure 1. Chromosome VI DNA containing the HIS2 region moved to chromosome II. (A) Map of the HIS2 region on chromosome II. The thick line on the chromosome represents chromosome VI DNA. The locations of DSB-B and DSB-C are denoted by vertical arrows, and the expected fragment sizes for DSB-B (B), DSB-C (C), and the parental band (P) are designated below the chromosome. The location of the probe used for this analysis is designated above the chromosome. Bg, BglII; Pm, PmlI. (B) DSB analysis. All DNAs shown were run on the same gel. "II" indicates the diploid GRH4-1 containing the 11.5-kbp HIS2 insertion on chromosome II. Average DSB frequencies for GRH4-1 are 0.8 ± 0.6% and 1.0 ± 0.9% for DSB-B and DSB-C, respectively. VI refers to the diploid SB4-7 with HIS2 at its normal location on chromosome VI containing Aha. Aha increases the frequency of breaks to a total of 12.2%, but does not change the location of DSB-B or DSB-C (BULLARD et al. 1996 ). The time in sporulation (hours) is denoted for each lane above the gel. The Aha makes the parental band (Pa) 382 bp smaller compared to GRH4-1.

Analysis of homozygous deletions at DSB sites shows no evidence for competition:
To determine whether the two breaks at HIS2 competed, we needed small deletions that affected the frequency of DSB formation. We created diploid strains homozygous for 50-bp deletions at both DSB sites (see MATERIALS AND METHODS). To examine the effect on recombination, gene conversion was analyzed using the his2-390 allele (MALONE et al. 1994 ) in these deletion diploids. A diploid homozygous for the deletion (B) at DSB-B showed a reduction in gene conversion from 12.9 to 9.5%, which was significant at the 93% confidence level (G-test; P = 0.07; Table 3). A diploid homozygous for the deletion (C) at DSB-C revealed a significant 2.3-fold increase in conversion from 12.9 to 29.2% (Table 3). When both small deletions (BC) were homozygous, gene conversion was increased 2.1-fold from 12.9 to 27.1% (Table 3).

Since C causes a dramatic increase in gene conversion, there should be a significant increase in DSB formation at HIS2. And since B decreases gene conversion, we predict a decrease in DSB formation at HIS2. The gene conversion data also suggest that events occurring at the two sites are independent. The B deletion decreases gene conversion of the his2-390 allele by 2.8%, regardless of the construct at DSB-C. Likewise, C increases gene conversion of his2-390 by 17%, regardless of the construct at DSB-B. We therefore measured meiotic DSB formation in diploids containing these deletions to test whether competition is occurring.

Examination of meiotic DSBs in the homozygous B diploid revealed no detectable breaks occurring at the region of the DSB-B site (Fig 2B). We conclude that B removes sequences necessary for breaks at DSB-B to occur. No novel breaks were observed, and breaks at the DSB-C site were unaffected. The data show that double-strand breaks at the native DSB-C site occurred at the same location (compare Fig 2B) and at the same frequency (Table 4; compare Fig 2H) as they normally occur when the DSB-B site is present.

View larger version (68K): In this window In a new window Download PPT slide   Figure 2. DSB analysis of strains containing deletions at the HIS2 DSB sites. Diploids examined were (A) SJH5-0, (B) SJH5-16, (C) SJH5-1, (D) SJH5-9, (E) SJH5-6, and (F) SJH5-10. A–D were run on the same gel. E–F were run on the same gel. The relevant genotype of each diploid is shown above each gel. Numbers above each lane refer to hours in sporulation. Sites of DSB formation are denoted with horizontal arrows. B and C represent the normal sites of DSB formation, and C' represents breaks forming at a location indistinguishable from DSB-C in C/C strains. C# indicates that both DSB-C and DSB-C' contribute to the observed band in diploids heterozygous for C. B# refers to breaks at DSB-B in strains heterozygous for this site (B/+). Lane M represents size markers consisting of HIS2 restriction fragments extending different distances from the upstream BglII site (5.2, 3.8, 3.2, 2.6, and 2.1 kbp from top to bottom). (G–L) Bar graphs showing the percentage of DSBs in all experiments done for each of the deletion diploids (see Table 4). (M) Map of HIS2 region on chromosome VI. Locations of DSBs, relevant fragments that arise due to the formation of DSBs (B and C), the parental BglII fragment (P), and the location of the probe used in the analyses are shown. Bg, BglII; H, HindIII; Ea, EagI; WT, a strain with no deletions at either DSB-B or DSB-C.

Examination of DSB formation in the homozygous C diploid showed that breaks occurred at a position indistinguishable from the normal DSB-C site (Fig 2C). We refer to this break site as DSB-C'. In addition, there was a dramatic increase (8.4-fold) in the frequency of DSB formation at this site from 2.0 to 16.7% (Table 4). Neither the amount of breaks at DSB-B (2.4% in the wild-type diploid vs. 2.5% in the homozygous C diploid) nor the location of the break at the DSB-B site was affected by C (compare Fig 2C and Fig 2I).

In the BC diploid, there was no detectable DSB formation at the DSB-B site (Fig 2D), consistent with analysis of the B diploid (Fig 2B). Breaks at the DSB-C' site occurred at a frequency (15.7%) not significantly different (t-test; P = 0.78) from that observed in the C diploid (16.7%; Table 4; compare Fig 2J). Thus, both gene conversion frequency and DSB formation in the BC diploid indicate that the properties conferred by each deletion are additive and independent. To further test this hypothesis, we examined diploids heterozygous for the deletions.

Analysis of heterozygous C diploids shows no evidence for competition:
Examination of recombination occurring in diploids heterozygous for other hotspots has resulted in the consensus view that recombination initiates more frequently on the "hotter" chromosome; this generates a disparity of gene conversion and an unequal distribution of DSBs on the two homologous chromosomes (e.g., LICHTEN and GOLDMAN 1995 ; XU and KLECKNER 1995 ). However, previous data from HIS2 indicated that not all alterations affecting hotspots act in this way. A 382-bp deletion (Aha) in the DNA 150 bp downstream of DSB-B at HIS2 stimulated conversion and DSBs when homozygous (BULLARD et al. 1996 ). However, when Aha was heterozygous, no disparity of conversion was observed, and DSBs were stimulated on both chromosomes equally (BULLARD et al. 1996 ). Bullard et al. interpreted this to mean that the Aha deletion increased the frequency of initiation after chromosomes were already associated without affecting the sites where DSBs occurred. Given this precedent, it was important to examine strains heterozygous for C or BC for a bias in the initiation of recombination. All experiments were done with the deletions in coupling with or in repulsion to the his2-390 marker.

In two crosses in which C was heterozygous, the his2-390 allele converted at an average frequency of 11.9% (Table 3). This value was significantly reduced (G-test; P < 0.01) from the homozygous C diploid frequency of 29.2% but was not significantly different (G-test; P > 0.39) from the wild-type (no deletions) value of 12.9%. Two crosses in which BC were both heterozygous had an average gene conversion frequency of 11.8% (Table 3). This value was significantly reduced (G-test; P < 0.01) from the 27.1% frequency observed in the BC homozygote but did not significantly differ (G-test; P > 0.36) from the wild-type value. If recombination initiates more frequently on one chromosome than on the other, then the allele on the chromosome with more frequent initiation should be preferentially lost during gene conversion with a resultant loss of parity (LICHTEN and GOLDMAN 1995 ; PETES 2001 ). In all strains in which C was coupling with his2-390, conversion was biased in the 3+:1- direction, where the "-" refers to the his2-390 allele (Table 3). When C was located in repulsion to his2-390, conversion was biased in the 1+:3- direction (Table 3). If the disparity observed above were due to a preferential initiation of recombination on the chromosome containing C, then the conversion of the his2-C allele should also show a bias in favor of 3+:1C. Table 3 shows that every cross containing a heterozygous C exhibits the predicted disparity. The conversion data are consistent with preferential initiation on the chromosome containing C.

To locate the breaks and to confirm that C stimulates DSB formation even when heterozygous, we measured meiotic DSBs in rad50S diploids. On the basis of the analysis of deletion homozygotes, one would expect 2.4% DSBs at DSB-B and 9.2% DSBs at DSB-C/C' in the C heterozygotes, if each chromosome retains its intrinsic break frequency (i.e., if each break site acted independently). In diploids heterozygous for C, breaks at the DSB-B site averaged 1.5% and breaks at the DSB-C/C' site averaged 8.1% (Table 4). All breaks occurred at the normal locations and no new sites were observed (Fig 2E). If each chromosome in the BC heterozygotes retains its intrinsic break frequency, we would expect 1.2% DSBs at DSB-B and 9.2% DSBs at DSB-C/C'. In the BC heterozygotes, the expected 50% reduction was observed at the DSB-B site, and breaks in the DSB-C/C' region were similar to that observed in C heterozygotes (Table 4; Fig 2F). These data are consistent with the hypothesis that the breaks at DSB-B and DSB-C/C' occur independently and that the DSB-C' site initiates preferentially over DSB-C. To further examine this hypothesis, we created strains to distinguish DSB formation on each chromosome within the same experiment.

Individual chromosome analysis of DSBs shows no evidence for competition in cis or in trans:
To test the independence of the DSB sites on each chromosome, a single base substitution creating a BglII restriction site (G) was made 1.6 kbp upstream of the HIS2 coding region to differentiate between the two homologous chromosomes. Control experiments indicate that the G mutation (homozygous or heterozygous) has no effect on gene conversion (Table 3) or on DSB frequency (Table 4; Fig 3).

View larger version (100K): In this window In a new window Download PPT slide   Figure 3. Chromosome-specific DSB formation in heterozygous C diploids. Diploids examined were: (A) SJH7-11, (B) SJH7-12, and (C) SJH7-13. All lanes shown were run on the same gel. The relevant genotype of each diploid is indicated above each gel. Numbers above each lane represent hours in sporulation, and the size markers are as described in Fig 2. DSBs are denoted with horizontal arrows. BG and CG refer to DSB-B and DSB-C on the chromosome marked with the new BglII site (G), and B and C refer to DSB-B and DSB-C on the original chromosome. C' refers to DSB-C' and C'G refers to DSB-C' on the chromosome with G. (D–F) Bar graphs showing the percentage of DSBs in all experiments done for each of the deletion diploids. NA, not applicable. CD, cannot distinguish the individual contribution of DSB-C' and DSB-BG to DSB formation due to their close proximity. (G) Map of the locations of DSBs. The location of the probe used in the analysis is shown. Restriction sites are denoted as in Fig 2, except G refers to the new BglII site.

Since the G mutation has no effect on recombination, diploids containing a heterozygous G mutation were used to examine DSBs occurring on each homologous chromosome. On the basis of the analysis of DSB formation in previous experiments (Table 4), one would expect 1.0% DSBs at DSB-C and 8.1% at DSB-C' in the C heterozygote (SJH7-13) if each break site acted independently. The values observed were 1.2 and 8.2% for DSB-C and DSB-C', respectively, indicating that the presence of C on one chromosome does not affect the normal DSB-C site on the homolog and vice versa (Table 4; Fig 3C). If each chromosome in the C heterozygote retains its intrinsic break frequency, we would expect 1.2% breaks at each DSB-B site. The values observed were 1.2 and 1.3% (Table 4; Fig 3). All break sites appear to act independently rather than competitively.

Medium base pair resolution mapping of DSBs in normal and deletion strains:
To determine at higher resolution where the B and C deletions are located relative to DSB-B and DSB-C, as well as to determine the locations of any newly formed breaks in our studies, we employed the method of medium base pair resolution mapping (MBM), as used in previous studies of hotspots (e.g., LIU et al. 1995 ; DIAZ et al. 2002 ). Analysis of DSB-B by MBM in a rad50S diploid (SJH5-0) reveals a pattern of 19 breaks (Fig 4A). These 19 breaks fall within 200 bp (from position +1145 to +1345 bp, relative to the start of the HIS2 coding region). Of the 19 breaks, 6 fall within the 50-bp B deletion, including 2 of the 4 stronger breaks. Although 13 of the 19 sites where breaks normally occur are not removed by B, no detectable breaks were observed at DSB-B in homozygous B diploids (Fig 2B and Fig D), indicating that the sequence removed by B is important for DSB formation. Analysis of DSB-C in SJH5-0 reveals 20 breaks scattered within 290 bp (from +530 to +840), with the strongest breaks located within 80 bp (from +615 to +695; Fig 4B). Three of the break sites are removed by C, which clearly does not prevent breaks from forming.

View larger version (61K): In this window In a new window Download PPT slide   Figure 4. Medium base pair resolution mapping of DSB sites. (A) Locations of breaks at DSB-B in SJH5-0. (B) Location of breaks at DSB-C in SJH5-0. (Inset) A different exposure of a subset of breaks located between +737 and +947 found at DSB-C. For A and B, the positions of deletions B and C are shown on the left side of the gel. The map on the left side of the gel represents the HIS2 region, and the numbers to the left indicate the positions of the size markers. Size markers contain 0.25 pg of DNA per band and were run in the presence of 4 µg of mitotic genomic DNA (deleted for HIS2) cut with HpaI to ensure proper comparison among all lanes. (C and D) Locations of breaks at DSB-C' in strains with DSB-B intact or with B, respectively. DSBs for A–D are denoted by horizontal arrows. WT, see Fig 2.

Double-strand breaks were mapped at DSB-C' in C homozygotes. This site has 20 breaks located within 250 bp (Fig 4C); the majority of breaks are located within 100 bp. Eight of the breaks located downstream of DSB-C' in the C homozygote map at positions indistinguishable from the original breaks at DSB-C (e.g., compare inset of Fig 4B with 4C and 4D). The alignment of breaks upstream of DSB-C' with the normal positions in DSB-C cannot be made directly because of the 50-bp C deletion. Subject to the resolution of the technique, the positions appear to be very similar. We also note that the range of breaks at DSB-C' is 50 bp shorter than the range of breaks at DSB-C (compare Fig 4B TO 4C and Fig 4D). This suggests that the breaks are limited to the same region in C as they are in the wild-type strain. The MBM shows that the pattern of DSB formation at DSB-C' is the same whether or not DSB-B is present (Fig 4C and Fig D); this was also observed for DSB-C whether or not B is present (data not shown). These results are consistent with the break sites acting independently.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

How much DNA is sufficient to restore hotspot activity?
As is true for all hotspots studied (reviewed in PETES 2001 ), our previous work on the HIS2 hotspot (MALONE et al. 1994 ) indicated that considerable amounts of DNA are necessary for hotspot activity at HIS2 on chromosome VI. Whereas WU and LICHTEN 1995 and BORDE et al. 1999 have addressed the question of how chromosomal location (and presumably chromosome structure) affect a hotspot, relatively few experiments have addressed how much DNA is sufficient for hotspot activity. That is, how much DNA is needed to recreate a hotspot in a different chromosomal position? One such experiment involved moving a 12.5-kbp fragment containing the ARG4 locus onto a yeast artificial chromosome (YAC; ROSS et al. 1992 ). This construct increases the amount of crossing over on the YAC 20-fold in the interval containing the ARG4 DNA fragment (ROSS et al. 1992 ). Conversion frequencies in the YAC construct of two mutations in the 12.5-kb insert-borne ARG4 were 4.5% (12/269) for the arg4-RV allele [normally 8.0% (283/3517); data taken from NICOLAS et al. 1989 and SCHULTES and SZOSTAK 1990 ] and 0.6% (2/353) for the arg4-Bg allele [normally 1.3% (15/1192); data taken from NICOLAS et al. 1989 and SCHULTES and SZOSTAK 1990 ]. Thus, 56% of the normal level of conversion was observed at the RV allele and 46% of the normal level of gene conversion was observed at the Bg allele. When ROSS et al. 2000 examined DSBs on the ARG4 YAC construct, they reported that DSBs were located at positions indistinguishable from the normal positions on chromosome VIII; however, these authors did not report the frequency of DSBs on the YAC construct. The conversion and DSB data taken together indicate that 12.5 kbp is capable of partially restoring the ARG4 recombination hotspot. We note that the YAC contains both phage and bacterial DNA, and available data suggest that procaryotic DNA may have unusual meiotic recombination properties when present in yeast. ROSS et al. 2000 discuss this and mention that phage DNA appears to be quite cold for recombination. We therefore felt it important to examine the question of sufficiency in a normal yeast chromosome.

Substantial amounts of DNA are required for the normal hotspot activity of HIS2 at its normal location on chromosome VI. Two insertions, located 2.5 kbp and 4.2 kbp from the 3' end of the HIS2 gene, reduced gene conversion about threefold, demonstrating that distant sequences can affect gene conversion at HIS2 on chromosome VI (R. E. MALONE and S. LINDQUIST, unpublished data). We conclude that an extensive region around HIS2 is required for hotspot activity, perhaps to create the necessary higher-order chromosome structure. While a 5.2-kbp fragment containing HIS2 inserted on chromosome II has no hotspot activity (MALONE et al. 1994 ), this article demonstrates that 11.5 kbp of HIS2 DNA is sufficient to allow some recognition of sequences where breaks normally occur. We note that the size (11.5 kbp) and the percentage of restoration (50%) is similar to that observed in the ARG4 YAC discussed earlier. Taken together, this suggests that 10–15 kb of DNA surrounding a hotspot locus may be adequate to maintain much of the structure allowing recombination activity.

The two HIS2 DSB sites do not compete:
Understanding how DSB sites interact is important in defining how recombination initiates. Previous studies have reported that two DSB sites located from 400 bp to 17 kbp apart can compete with one another for DSB formation (e.g., FAN et al. 1997 ; OHTA et al. 1999 ). However, we note that all studies in which competition was examined utilized one DSB site created by the presence of unusual DNA and a second normal yeast DSB site (see below). We define unusual DNA in yeast recombination as DNA from procaryotes (e.g., Escherichia coli, pBR322, phage ) or telomeric repeat sequences normally located at the ends of chromosomes. The small deletions we examined at HIS2 altered the frequency of conversion and DSB formation, giving us the opportunity to examine competition without having to insert unusual DNA. The data obtained support the hypothesis that the DSB sites at HIS2 do not compete with each other.

First, gene conversion frequencies in the wild type and in all three deletion homozygotes (B, C, and BC) suggest that each DSB site contributes independently to recombination. In homozygous deletion diploids, we estimate that a single DSB-B site contributes 1.4% conversion, a single DSB-C site contributes 4.7% conversion, and a single DSB-C' site contributes 13.5% conversion of his2-390. The conversion frequencies in homozygotes appear to be additive and independent. For example, a homozygous B deletion reduces conversion by 2.8% in diploids whether DSB-C or DSB-C' is present.

Second, DSB frequencies in the wild type and in deletion homozygotes indicate that each DSB site acts independently. The amount of DSBs contributed by the different sites in homozygous diploids [DSB-B = 2.5%, DSB-C = 2.2%, DSB-C' (C) = 16.2%, and B = 0%] was unchanged by the presence or absence of the other site. Since breaks at DSB-C' are increased over eightfold, a competition model argues that breaks should decrease at DSB-B in a C homozygote. Breaks at DSB-B did not decrease. Likewise, the B deletion did not cause an increase in breaks at DSB-C (or -C'). None of the analyses on the homozygous deletions provides any support for competition.

The third observation arguing that the break sites do not compete is the analysis of the DSBs occurring in diploids heterozygous for the deletions. All of the data obtained from heterozygotes (e.g., C/+) indicate that no competition exists between different sites. In experiments where breaks on the individual chromosomes could be distinguished, no competition was observed between DSB-B and DSB-C (or -C') in cis or in trans. Nor was there competition between sites at the same position on homologs (e.g., DSB-C vs. DSB-C'). This is contrary to the competition observed for HIS4::LEU2, the other analysis where breaks have been analyzed on each chromosome (XU and KLECKNER 1995 ). In that study, altering breaks at one site affected breaks on the homologous chromosome at the same position (in trans) as well as on the same chromosome at the other break site (in cis; XU and KLECKNER 1995 ).

We note that our HIS2 data are consistent with results published by DE MASSY and NICOLAS 1993 , who observed that deletions eliminating DSB formation immediately upstream of ARG4 had no detectable effect on DSB formation at the neighboring DED81 hotspot (2.1 kbp from the ARG4 break site). DE MASSY and NICOLAS 1993 concluded that the DED81 hotspot is not affected by the presence or absence of the ARG4 hotspot. Like the experiments done at HIS2, these ARG4 experiments examined DSB sites that occur without unusual DNA. Another observation consistent with the lack of competition observed between the breaks at HIS2 is the classic genetic observation that gene conversion events do not show interference (MORTIMER and FOGEL 1974 ). If DSB sites competed, gene conversion events should not occur randomly, and the occurrence of one conversion should decrease the probability of a second occurring nearby.

Possible reasons for the difference between HIS2 and regions showing competition:
There are several possible explanations for why our results differ from previous work showing competition between DSB sites. First, the recombination hotspot at HIS2 may be unusual. This is formally possible, if only because one break occurs in the coding region, which is apparently rare in yeast (BAUDAT and NICOLAS 1997 ). We think that this is unlikely, because the properties of recombination (e.g., conversion gradient, DSBs, parity, association with crossing over) observed at HIS2 mimic those observed at other hotspots. In addition, both HIS2 breaks are eliminated in a recombination initiation mutant (BULLARD et al. 1996 ), just like other hotspots (e.g., CAO et al. 1990 ).

A second explanation for differences between HIS2 and published reports arguing for competition is that almost all of the experiments demonstrating competition between DSB sites were done at or near the HIS4 locus on chromosome III (XU and KLECKNER 1995 ; FAN et al. 1997 ; OHTA et al. 1999 ). The lack of competition observed between neighboring natural DSB sites at both HIS2 and ARG4 might be due to their location in regions of chromosomal architecture different from those that exist at HIS4, HIS4::LEU2, or the his4::ARG4::pBR322 construct. Perhaps something unique in the HIS4 region is responsible for the observed competition between DSB sites. However, not all hotspots at HIS4 on chromosome III show competition. NAG and KURST 1997 reported an instance where a 140-bp synthetic palindromic insertion created a break site at HIS4 that exhibited 3.4% DSBs. This new site did not affect the location or the frequency of breaks occurring at the normal site (1.3% in wild type vs. 1.5% in the palindromic insertion diploid). These latter data suggest that the region around the HIS4 locus probably does not have unique properties.

A third possibility is that the presence of unusual DNA in the constructs used to examine recombination affected competition. It is perhaps not entirely surprising that procaryotic sequences could affect recombination, since they did not evolve in an environment containing chromatin or meiotic recombination. Both pBR322 DNA and a small 77-bp insertion of bacterial DNA appear to be quite recombinogenic (WU and LICHTEN 1995 ; XU and KLECKNER 1995 ; BORDE et al. 1999 ). These sequences create breaks, which also clearly compete effectively with nearby yeast DSB sites. In contrast, phage DNA seems to be rather cold (ROSS et al. 1992 , ROSS et al. 2000 ); artificial chromosomes containing DNA have low frequencies of recombination. A 49-bp telomere repeat sequence insert creates a high frequency of DSBs at its location and competes with the nearby normal HIS4 hotspot (FAN et al. 1997 ). Although clearly yeast DNA, the telomeric repeat is now located in the middle of a chromosome arm when inserted at HIS4.

Competition has been observed in one situation where only normal yeast sequences are present. PECINA et al. 2002 created a SPO11-GAL4DB (DNA-binding domain) fusion construct that fully complemented a spo11 null mutation. Examination of breaks in a strain containing the construct showed that several regions containing UAS_GAL sites display novel meiotic DSBs. The UAS_GAL break at YCR048W reduced (i.e., competed with) the DSBs at nearby sites. No unusual DNA was present in these experiments. We explain this competition by suggesting that tethering the Spo11 protein to a Gal4 DNA-binding domain gives UAS_GAL sites an intrinsic advantage over normal sites, which have to bind Spo11 protein (and the initiation complex) in a normal fashion. Here we presume that the competition is not due to the intrinsic features of the DNA sequence, but rather to the localized recombination protein.

Models for competition:
Competition has been reported to occur if an alteration in one site changing the frequency of DSB formation has an inverse effect on the frequency of DSB formation at another nearby site. It has been explained by models such as limited availability of recombination factors (BORDE et al. 1999 ), steric exclusion (XU and KLECKNER 1995 ), and relief of chromosomal stress by Spo11 protein complexes (KEENEY 2001 ).

The three proposed models for competition cannot easily explain both the results at HIS2 and loci showing competition. If breaks occur randomly at natural DSB sites among meiotic cells, their low (1–10%) frequency indicates that breaks occurring at different natural DSB sites would typically occur in different cells. This would seem to preclude most obvious forms of competition. In the sites showing competition, the hotter DSB site is always the one containing the unusual DNA. Competition could be explained if the majority of hot DSB sites containing unusual DNA are "primed" for recombination initiation. In contrast, only a fraction of natural sites in any given cell would be in a state capable of attracting the recombination factors. Thus, unusual hotspots would always be able to compete with natural hotspots within the same individual cells. What might the priming be? The priming could be due to the absence of protein ("open" DNA) or due to the presence of a protein bound at the unusual DNA (e.g., Rap1 protein binding to the telomeric insert; FAN et al. 1997 ) leading to preferential recruitment of the recombination complex. Consistent with this idea is the observation that Gal4DB-Spo11p gives UAS_GAL sites a competitive advantage over normal nearby DSB sites. As in the situation postulated above, the majority of UAS_GAL sites would contain the Gal4DB-Spo11 protein, whereas natural sites could attract only the initiation complex with normal low probability. Although all competitive hotspots known contain unusual DNA, not all unusual DNA creates DSB sites that are competitive (e.g., the 140-bp palindromic insertion at HIS4; NAG and KURST 1997 ). These data would predict that the 140-bp palindromic sequence would not be primed in every cell and therefore is acting similarly to a natural noncompetitive hotspot.

The role of sequences located at the sites of DSB formation:
One view of recombination hotspots (reviewed in PETES 2001 ) suggests that the actual sequences where DSBs occur may not play a major role in the initiation of recombination at hotspots (e.g., DE MASSY and NICOLAS 1993 ). Consistent with this view, comparison of the DNA sequence at the actual locations of the DSBs at HIS4, ARG4, CYS3, and HIS2 show no obvious sequence homology. Despite this, some previous data have indicated that sequences at the site of a DSB might be important; for example, a 142-bp deletion removing the DSB site in the ARG4 gene reduced gene conversion from 8 to 2% and eliminated detectable DSBs (NICOLAS et al. 1989 ; SUN et al. 1989 ).

The deletion of DSB-B decreases conversion of his2-390 by 2.8% and eliminates detectable breaks, confirming that the sequence located at the DSB-B site is necessary for recombination initiation at that site. The deletion of DSB-C has the opposite effect: conversion increases 2.3-fold, and DSBs increase 8.4-fold. In fact, the DSB frequency (16.7%) found at the DSB-C' site in the C strain is the highest reported in yeast without the insertion of unusual DNA. Since deleting DSB-C has such a strong effect on recombination, one can formally argue that it is important; that is, the loss of that DNA greatly stimulates recombination. We realize that it remains a possibility that instead of eliminating a negative element, the C deletion might create a "better" DSB site. Because of the strong DSBs observed in the C strain, we searched the yeast genome using the fusion sequence created by C to determine if such a sequence is located near any known DSB sites. No such DNA sequences were found near any known DSB sites. Taking the data from the B and C experiments together, we argue that the DNA sequence where breaks occur plays a significant role.

Another observation supporting the importance of the sequence at DSB sites is the contrast in effects observed between deletions at DSB sites and deletions in surrounding DNA. The data presented here clearly show that B and C affect only the site at which they are located. In contrast, alterations in surrounding regions can affect both DSB sites. For example, BULLARD et al. 1996 showed that a deletion (Aha) downstream of the two DSB sites at HIS2 increases DSBs equally at both sites. Even when Aha is heterozygous, both DSB sites are equally affected (BULLARD et al. 1996 ). This suggests that there are two kinds of sequences at hotspots: those sequences that affect the overall probability of initiation and those sequences, located at the DSB site, that affect only the probability of initiating at that specific site.

Although genetic recombination, unlike replication or transcription, can initiate anywhere, some regions of the genome do so more efficiently. Consistent with the idea proposed by many investigators (reviewed in PETES 2001 ) that chromosome structure plays an important role in hotspots, we have found that at least 11.5 kbp is needed to partially retain hotspot activity at HIS2. In addition to long-range determinants, the sequences at break sites play an important and independent role in DSB formation. Further analyses of specific natural hotspots in yeast will determine whether the properties observed at HIS2 are general. While global studies of the genome investigating DSBs are critical, the work presented here reinforces the importance of studying specific hotspots to obtain an understanding of the initiation of meiotic recombination.

	ACKNOWLEDGMENTS

We thank Kelley Foreman, Luke Lautner, April Kopps, and Timothy Walker for comments on this manuscript and Jan Fassler and Greg Gingerich for experimental assistance. Alex J. Jones was supported by an undergraduate fellowship from the Howard Hughes Foundation during part of this work. This research was funded by a National Science Foundation grant MCB97-28557 to R.E.M.

Manuscript received March 19, 2003; Accepted for publication May 27, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

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]

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]

BOLIVAR, F., R. L. RODRIGUEZ, P. J. GREENE, M. C. BETLACH, and H. L. HEYNEKER et al., 1977  Construction and characterization of new cloning vehicles. II. A multipurpose cloning system. Gene 2:95-113.[Medline]

BORDE, V., T. C. WU, and M. LICHTEN, 1999  Use of a recombination reporter insert to define meiotic recombination domains on chromosome III of Saccharomyces cerevisiae.. Mol. Cell. Biol. 19:4832-4842.[Abstract/Free Full Text]

BULLARD, S. A., S. KIM, A. M. GALBRAITH, and R. E. MALONE, 1996  Double strand breaks at the HIS2 recombination hot spot 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]

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., 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]

DIAZ, R. L., A. D. ALCID, J. M. BERGER, and S. KEENEY, 2002  Identification of residues in yeast Spo11p critical for meiotic DNA double-strand break formation. Mol. Cell. Biol. 22:1106-1115.[Abstract/Free Full Text]

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

FAN, Q. Q., F. XU, M. A. WHITE, and T. D. PETES, 1997  Competition between adjacent meiotic recombination hotspots in the yeast Saccharomyces cerevisiae.. Genetics 145:661-670.[Abstract]

GAME, J. C., 1992  Pulsed-field gel analysis of the pattern of DNA double-strand breaks in the Saccharomyces genome during meiosis. Dev. Genet. 13:485-497.[Medline]

GERTON, J. L., J. DERISI, R. SHROFF, M. LICHTEN, and P. O. BROWN et al., 2000  Inaugural article: global mapping of meiotic recombination hotspots and coldspots in the yeast Saccharomyces cerevisiae.. Proc. Natl. Acad. Sci. USA 97:11383-11390.[Abstract/Free Full Text]

GIETZ, R. D. and A. SUGINO, 1988  New yeast-Escherichia coli shuttle vectors constructed with in vitro mutagenized yeast genes lacking six-base pair restriction sites. Gene 74:527-534.[Medline]

HO, S. N., H. D. HUNT, R. M. HORTON, J. K. PULLEN, and L. R. PEASE, 1989  Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene 77:51-59.[Medline]

KEENEY, S., 2001  Mechanism and control of meiotic recombination initiation. Curr. Top. Dev. Biol. 52:1-53.[Medline]

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

LIU, J., T. C. WU, and M. LICHTEN, 1995  The location and structure of double-strand DNA breaks induced during yeast meiosis: evidence for a covalently linked DNA-protein intermediate. EMBO J. 14:4599-4608.[Medline]

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. LUNDQUIST, 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]

MORTIMER, R. K., and S. FOGEL, 1974 Genetical interference and gene conversion, pp. 263–275 in Mechanism in Recombination, edited by R. F. GRELL. Plenum, New York.

NAG, D. K. and A. KURST, 1997  A 140-bp-long palindromic sequence induces double-strand breaks during meiosis in the yeast Saccharomyces cerevisiae.. Genetics 146:835-847.[Abstract]

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. C. WU, M. LICHTEN, and T. SHIBATA, 1999  Competitive inactivation of a double-strand DNA break site involves parallel suppression of meiosis-induced changes in chromatin configuration. Nucleic Acids Res. 27:2175-2180.[Abstract/Free Full Text]

PAQUES, F. and J. E. HABER, 1999  Multiple pathways of recombination induced by double-strand breaks in Saccharomyces cerevisiae.. Microbiol. Mol. Biol. Rev. 63:349-404.[Abstract/Free Full Text]

PECINA, A., K. N. SMITH, C. MEZARD, H. MURAKAMI, and K. OHTA et al., 2002  Targeted stimulation of meiotic recombination. Cell 111:173-184.[Medline]

PETES, T. D., 2001  Meiotic recombination hot spots and cold spots. Nat. Rev. Genet. 2:360-369.[Medline]

ROSS, L. O., D. TRECO, A. NICOLAS, J. W. SZOSTAK, and D. DAWSON, 1992  Meiotic recombination on artificial chromosomes in yeast. Genetics 131:541-550.[Abstract]

ROSS, L. O., D. ZENVIRTH, A. R. JARDIM, and D. DAWSON, 2000  Double-strand breaks on artificial chromosomes in yeast. Chromosoma 109:226-234.[Medline]

ROTHSTEIN, R., 1991  Targeting, disruption, replacement, and allele rescue: integrative DNA transformation in yeast. Methods Enzymol. 194:281-301.[Medline]

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, 1994  Identification of joint molecules that form frequently between homologs but rarely between sister chromatids during yeast meiosis. Cell 76:51-63.[Medline]

SIKORSKI, R. S. and P. HIETER, 1989  A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae.. Genetics 122:19-27.[Abstract/Free Full Text]