Isolation and Characterization of Novel xrs2 Mutations in Saccharomyces cerevisiae

Abstract

The Mre11/Rad50/Xrs2 (MRX) complex is involved in DNA damage repair, DNA damage response, telomere control, and meiotic recombination. Here, we constructed and characterized novel mutant alleles of XRS2. The alleles with mutations in the C-terminal conserved domain of Xrs2 were grouped into the same class. Mutant Xrs2 in this class lacked Mre11 interaction ability. The second class, lacking a C-terminal end, showed defects only in telomere control. A previous study showed that this C-terminal end contains a Tel1-association domain. These results indicate that Xrs2 contains two functional domains, Mre11- and Tel1-binding domains. While the Mre11-binding domain is essential for Xrs2 function, the Tel1-binding domain may be essential only for Tel1 function in telomere maintenance. The third class, despite containing a large deletion in the N-terminal region, showed no defects in DNA damage repair. However, some mutants, which showed a reduced level of Xrs2 protein, were partially defective in formation of meiotic DSBs and telomere maintenance. These defects were suppressed by overexpression of the mutant Xrs2 protein. This result suggests that the total amount of Xrs2 protein is a critical determinant for the function of the MRX complex especially with regard to telomere maintenance and meiotic DSB formation.

EUKARYOTES have many different mechanisms for repairing DNA double-stranded breaks (DSBs). In Saccharomyces cerevisiae, genes in the RAD52 epistasis group (RAD50, RAD51, RAD52, RAD54, RAD55, RAD57, RAD59, TID1/RDH54, MRE11, and XRS2) are involved in homologous recombination (Paques and Haber 1999; Symington 2002). Recombination is required for both repair of DSBs and segregation of homologous chromosomes in meiosis.

In addition to mitotic DSB repair and meiotic recombination, the Mre11-Rad50-Xrs2/Nbs1 complex, termed the MRX/N complex, is involved in telomere maintenance and ataxia telangiectasia mutated-related checkpoint response (Haber 1998; D'Amours and Jackson 2001; Usui et al. 2001). Mutations in this complex cause genome instability in yeast (Chen and Kolodner 1999; Williams et al. 2002; Yoshida et al. 2003) and in humans cause some disorders with genome instability and a high risk of cancer as found in Nijmegen breakage syndrome (NBS) and ataxia-telangiectasia-like disorder (A-TLD) (Stewart et al. 1999). The MRX/N complex is well conserved among species. Both Mre11 and Rad50 proteins, encoding a nuclease and a structural maintenance of chromosome-like protein, respectively, are highly conserved in species from yeast to humans. The third component, Xrs2 in budding yeast or Nbs1 in mammal and fission yeast, is relatively less conserved among species. A comparison of amino acid sequences of Xrs2 and Nbs1 from different species revealed that similarity is limited to two regions, namely the N-terminal fork-head-associated (FHA) domain and the C-terminal conserved region (Chahwan et al. 2003). A knockout mouse with a NBS1 null mutation is lethal, indicating that the NBS1 gene is essential for survival in mammalian cells (Zhu et al. 2001). In most NBS patient cells, a 70-kDa Nbs1 protein lacking an N terminus is expressed at low levels (Maser et al. 2001). Therefore, NBS is caused by a hypomorphic allele in the NBS1 gene.

In S. cerevisiae, Mre11, Rad50, and Xrs2 are nonessential for cell growth. Each of the null mutations in MRE11, RAD50, and XRS2 cause similar defects due to dysfunction of the MRX complex. Several non-null mutations in MRE11 and RAD50 have been isolated and shown to be separation-of-function mutant alleles (Alani et al. 1990; Nairz and Klein 1997; Usui et al. 1998). In meiotic recombination, the functions of Mre11 and Rad50 are required for both meiotic DSB formation and processing of the ends (Alani et al. 1990; Johzuka and Ogawa 1995; Ohta et al. 1998). The latter function is suggested by isolation of mutants that are normal in DSB formation but aberrant in processing of the ends: the rad50S, mre11-58, or mre11S mutations (Alani et al. 1990; Nairz and Klein 1997; Tsubouchi and Ogawa 1998). However, it is still not known whether Xrs2 is required for the latter step (processing of the DSB ends) since there have been few reports on systematic isolation and characterization of non-null mutations in XRS2.

Here, we have constructed some truncated mutations at the N or C terminus of Xrs2 and engineered amino acids substitutions in the conserved regions between Xrs2 and human Nbs1. To confirm the role of Xrs2 in the MRX complex, we characterized each mutant for association with Mre11, DNA damage repair, telomere maintenance, meiotic DSB formation, and meiotic recombination. This enabled us to divide the Xrs2 protein into three functional domains. The first domain is a C-terminal conserved region, the Mre11 interaction domain, loss of which induces malfunction of the MRX complex. The second domain is a less conserved region at the C-terminal end, which has a Tel1-interaction domain after the Mre11-binding domain and is required only for telomere maintenance. Finally, the third domain is an N terminus half including a well-conserved FHA domain. Interestingly, in contrast to human Nbs1, this region is not required for any major function of the MRX complex in yeast.

MATERIALS AND METHODS

Strains and plasmids:

All plasmids and yeast strains and their genotypes are shown in Table 1. We used isogenic S. cerevisiae S288C background MSY1179 or W303-1A derivatives for mitotic analysis and isogenic SK1 background NKY1551 derivatives for meiotic analysis. All of the xrs2 mutations were constructed on plasmids as a primary step and then integrated into the original XRS2 locus of the MSY1179 (S288C), W303-1A, or MSY147 (SK1) (Shinohara et al. 1997) strain. The xrs2-84M, -228M, -630, -664, -AA, -F640A, -K641A, -K645A, -GE, and -SH mutant strain were created by two-step gene replacement using XbaI-digested YIplac211-based plasmids (Gietz and Sugino 1988) pMS345, pMS344, pMS346, pMS392, pMS343, pMS471, pMS342, pMS472, pMS347, and pMS393, respectively. The xrs2-314M mutant strain was created by two-step gene replacement using BglII-digested pMS385. The xrs2 null mutant was created by one-step gene replacement using a XhoI-BamHI fragment from pMS294. A parental haploid strain of MSY1758 containing rad50-K18I (rad50S mutation) was constructed by standard genetic cross with a rad50S strain, which was a gift from N. Kleckner. pTAK75 (a gift from T. Usui) was used to make the MSY2175 (tel1) strain. All strains were crossed with each parental strain at least once in the case of S288C and W303-1A and twice in the case of SK1 background and then were used for the assays.

TABLE 1.

Strain and plasmid list

Strain	Genotype	Reference
MSY1179	MATaura3 leu2 lys2 his3	This study
MSY1767	Derivative of MSY1179 with xrs2-84M	This study
MSY1768	Derivative of MSY1179 with xrs2-228M	This study
MSY1770	Derivative of MSY1179 with xrs2-314M	This study
MSY1769	Derivative of MSY1179 with xrs2-630	This study
MSY2045	Derivative of MSY1179 with xrs2-664	This study
MSY1777	Derivative of MSY1179 with xrs2-SH	This study
MSY1772	Derivative of MSY1179 with xrs2-GE	This study
MSY2596	Derivative of MSY1179 with xrs2-F640A	This study
MSY1463	Derivative of MSY1179 with xrs2-K641A	This study
MSY2598	Derivative of MSY1179 with xrs2-K645A	This study
MSY1771	Derivative of MSY1179 with xrs2-AA	This study
MSY1244	Derivative of MSY1179 with xrs2::URA3	This study
MSY1296	Derivative of MSY1179 with mre11::hisG-URA3-hisG	This study
W303-1A	MATaade2-1 can1-100 his3-11,15 leu2-3,112 trp1-1 ura3-1	Thomaset al. (1989)
MSY2140	Derivative of W303-1A with xrs2::URA3	This study
MSY2191	Derivative of W303-1A with xrs2-AA	This study
MSY2674	Derivative of W303-1A with tel1::TRP1, xrs2-AA	This study
MSY2677	Derivative of W303-1A with tel1::TRP1, xrs2::URA3	This study
MSY2175	Derivative of W303-1A with tel1::TRP1	This study
NKY1551	MATa/α ho::LYS2/ho::LYS2 ura3/ura3 leu2::hisG/leu2::hisG lys2/lys2 his4X-LEU2::BamHI-URA3/his4B-LEU2 arg4-Nsp/arg4-Bgl	Storlazzi et al. (1996)
MSY1494	Derivative of NKY1551 with xrs2-84M	This study
MSY1524	Derivative of NKY1551 with xrs2-228M	This study
MSY1992	Derivative of NKY1551 with xrs2-314M	This study
MSY1628	Derivative of NKY1551 with xrs2-630	This study
MSY2015	Derivative of NKY1551 with xrs2-664	This study
MSY1867	Derivative of NKY1551 with xrs2-SH	This study
MSY1509	Derivative of NKY1551 with xrs2-GE	This study
MSY1511	Derivative of NKY1551 with xrs2-AA	This study
MSY1357	Derivative of NKY1551 with xrs2::URA	This study
MSY1758	Derivative of NKY1551 with rad50S-URA3	This study
MSY1762	Derivative of NKY1551 with rad50S-URA3 xrs2-84M	This study
MSY1843	Derivative of NKY1551 with rad5S-URA3 xrs2-228M	This study
MSY2011	Derivative of NKY1551 with rad50S-URA3 xrs2-314M	This study
MSY2034	Derivative of NKY1551 with rad50S-URA3 xrs2-AA	This study
MSY2037	Derivative of NKY1551 with rad50S-URA3 xrs2-630	This study
MSY2317	Derivative of NKY1551 with rad50S-URA3 xrs2::URA3	This study
MSY2223	MATa ho::LYS2 ura3 leu2::hisG trp1::hisG lys2   his4B-LEU2 arg4-Bgl xrs2-84M	This study
MSY2225	MATα ho::LYS2 ura3 leu2::hisG trp1::hisG lys2   his4X-LEU2::BamHI-URA3 arg4-Nsp xrs2-84M	This study
Plasmid	Genotype	Reference
pRS313	ARS-CEN HIS3	Sikorski and Hieter (1989)
pRS423	2μ ori, HIS3	Christianson et al. (1992)
pMS402	pRS313; xrs2-84M	This study
pMS403	pRS423; xrs2-84M	This study
pRS314	ARS-CEN TRP1	Sikorski and Hieter (1989)
pMS387	pRS314; xrs2-84M	This study
pMS388	pRS424; xrs2-84M	This study
pMS398	pRS314; XRS2	This study
pMS399	pRS424; XRS2	This study
pRS316	ARS-CEN URA3	Sikorski and Hieter (1989)
pMS421	pRS426; MRE11	This study
pMS296	pAS2-1; XRS2	This study
pMS278	pAS2-1; xrs2-84M	This study
pMS462	pAS2-1; xrs2-F640A	This study
pMS500	pAS2-1; xrs2-K641A	This study
pMS501	pAS2-1; xrs2-K645A	This study
pMS277	pAS2-1; xrs2-AA	This study
pMS286	pACT2; MRE11	This study

We inserted the XhoI-BamHI fragment containing the xrs2-84M gene from pMS278 into pRS314, pRS313 (Sikorski and Hieter 1989), pRS424, and pRS423 (Christianson et al. 1992), resulting in pMS387, pMS402, pMS388, and pMS403, respectively. We inserted the XhoI-BamHI fragment from YCp50-XRS2 (a gift from J. Haber) into pRS314, pMS424, and pRS313, resulting in pMS400, pMS399, and pMS404, respectively. We inserted the PCR-amplified MRE11 fragment (chromosome XIII; nos. 718100–721000) into pRS326, resulting pMS421.

Plasmid and oligonucleotide sequence for xrs2 alleles:

The truncated xrs2 alleles were constructed as follows. First, we made the plasmid pMS256, in which was inserted a fragment from the promoter region (chromosome IV; nos. 1219010–1217570 in the YPD database) and from a downstream region (chromosome IV; nos. 1215002–1214802) of the XRS2 gene created with an NdeI site at the first ATG codon and an NheI site at the stop (TAG) codon in the XhoI-BamHI site of pBluescript KS⁺. Next we inserted each PCR-amplified fragment containing mutated an xrs2 ORF into the NdeI-NheI site of pMS256, resulting in pMS278 (xrs2-84M), pMS279 (xrs2-228M), pMS390 (xrs2-314M), pMS264 (xrs2-630), pMS394 (xrs2-664), pMS459 (xrs2-F640A), pMS273 (xrs2-K641A), pMS458 (xrs2-K645A), pMS274 (xrs2-AA), pMS257 (xrs2-GE), and pMS391 (xrs2-SH). For the xrs2 null (xrs2::URA3) mutation, we inserted each PCR-amplified fragment containing a URA3 ORF from pRS316 (Sikorski and Hieter 1989) into the NdeI-NheI site of pMS256, resulting in pMS294. Then we transferred the XhoI-BamHI fragment from each pMS256-based plasmid into the SalI-BamHI site of YIplac211, resulting in pMS345 (xrs2-84M), pMS344 (xrs2-228M), pMS385 (xrs2-314M), pMS346 (xrs2-630), pMS392 (xrs2-664), pMS471 (xrs2-F640A), pMS342 (xrs2-K641A), pMS472 (xrs2-K645A), pMS343 (xrs2-AA), pMS347 (xrs2-GE), and pMS393 (xrs2-SH).

The primer sets for the N-terminal truncated xrs2 mutation series were 5′-GGAAAATTTATCGCTAGCCTTTTCTTCTTTTG-3′ for a common primer of the STOP end and 5′-CCCATATGAAAGTTGGCGAAAC-3′ (xrs2-84M), 5′-AGCATATGAGGCTGAATAATATC-3′ (xrs2-228M), and 5′-CCATATGAAGGGTGCATCTTCAAG-3′ (xrs2-314M) for the primers of each first ATG end of the truncate alleles. The primer sets for the C-terminal truncated xrs2 mutation series were 5′-GATAACTATAAACATATGTGGGTAGTAC-3′ for a common primer of the first ATG end and 5′-CGGCTAGCTTAGGCTATTTTCCCATTTTTC-3′ (xrs2-630) and 5′-GGCTAGCTTTGAGTGTTATTTTTACCCTC-3′ (xrs2-664) for the primers of each STOP end of the truncate alleles. The xrs2-F640A, xrs2-K641A, xrs2-K645A, xrs2-AA (K641, 645A), xrs2-GE (G31E), and xrs2-SH (S47A, H50A) mutations were produced by site-directed mutagenesis using PCR reaction that was performed with the following primers: 5′-CCGTACCTTGACAAAAGTCTTGGCATTCTTGCG-3′ and 5′-CCGAAATCAAAGGCGCACAAAG-3′ (xrs2-F640A); 5′-GAAATTCTTGCCTTTATGCC-3′ and 5′-GCGACTTTTGTCAAGGTACGTC-3′ (xrs2-K641A); 5′-CCGTACCGCGACAAAAGTCTTGAAATTCTTGCG-3′ and 5′-CCGAAATCAAAGGCGCACAAAG-3′ (xrs2-K645A); 5′-GAAATTCTTGCCTTTATGCC-3′ and 5′-GCGACTTTTGTCGCGGTACGTCCAAAATC-3′ (xrs2-AA); 5′-GGGATCCTTCCCAATTTTCCAATG-3′ and 5′-CAGGCCTTCAAAACTTATAGTATAGAAAGATCAAG-3′ (xrs2-GE); and 5′-GCGCGCCAGGCTATCACATTCAAATGG-3′ and 5′-AATACTTTTATCATTTTTAATTATCAGTGG-3′ (xrs2-SH). The F640A and K645A mutations were marked with a created EagI site. The K641A and K641, -645A mutations were chased with a created NruI site, G31E mutation was marked with a NdeI site, and the S47A and H50A mutations were marked with a BssHII site. All of the PCR-amplified fragments were checked for their DNA sequences before use.

Immunoprecipitation and Western blotting:

The immunoprecipitation experiment was carried out as described (Usui et al. 2001). Immunoprecipitates were separated by SDS-polyacrylamide gel electrophoresis and then transferred onto a polyvinylidene difluoride membrane (Immobilon-P, Millipore, Bedford, MA) and analyzed by Western blotting. Bands were visualized with Alexa Fluor 680-labeled secondary antibodies (Molecular Probes, Eugene, OR) or IR dye 800-labeled secondary antibodies (Rockland, Gilbertsville, PA) using an Odyssey infrared imaging system (LI-COR Biosciences). Antibodies used in this assay were UWM45 (anti-Xrs2) and no. 59567 (anti-Mre11) for Western blotting, and guinea pig anti-Xrs2 and no. 59567 for immunoprecipitate (IP) assay (a gift from J. Petrini).

Two-hybrid analysis:

Each subfragment of the coding region of full-length XRS2, xrs2-84M, xrs2-F640A, xrs2-K641A, xrs2-K645A, xrs2-AA, and xrs2-630 was cloned into the NdeI-BamHI site of plasmid pAS2-1 (Clontech Laboratories), resulting in pMS296, pMS278, pMS462, pMS500, pMS501, and pMS277, respectively. The PCR-amplified MRE11 fragment was cloned into the NcoI-BamHI site of plasmid pACT2 (Clontech Laboratories), resulting in pMS286. Yeast strain AH109 (Clontech Laboratories) was cotransformed with a pair of plasmids, which are pAS2-1 based and pACT2 based. The transformants were selected on SD-Trp, Leu plates. For β-galactosidase liquid assay, 12 independent cotransformants carrying both plasmids were grown in SD-Trp, Leu liquid media until log phase. β-Galactocidase activity was calculated as 1 unit = 1000 × A₄₀₅/(t × V × OD₆₀₀), where t is time of incubation in minutes and V is volume of cells in milliliters.

Determination of γ-ray sensitivity:

Each strain was pre-grown in liquid YPAD overnight. Each cell culture was resuspended in fresh YPAD to be OD₆₀₀ = 0.2, then the cells were grown for 3 hr and irradiated with γ-rays using a Shimadzu Isostron RTGS-21 (Shimadzu, Tokyo). After serial dilution in PBS, cells were plated onto SC plates and colonies were counted after 4 days of incubation. Survival at each dose was determined from the ratio of irradiated to unirradiated colony numbers. The assay was repeated more than three times, and the average value was presented numerically.

Determination of MMS sensitivity:

Methyl methanesulfonate (MMS) sensitivity of each strain was determined in liquid complete media (YPAD) culture as described (Usui et al. 2001). MMS-treated cells were mixed with sodium thiosulfate to neutralize MMS. After serial dilution in PBS, cells were plated onto YPAD plates and colonies were counted after 3 days of incubation. Survival at each dose was determined from the ratio of treated to untreated colony numbers. The assay was repeated more than three times, and the average value was presented numerically.

Analysis of telomere length:

Southern blot analysis was carried out for the examination of telomere structure. Genomic DNA digested with XhoI was separated on 0.7% agarose in TBE buffer. DNA was transferred and UV crosslinked onto a nylon membrane (Hybond-N, Amersham, Buckinghamshire, UK) and then probed with a ³²P-labeled EcoRI-fragment from pNH3 (a gift from F. Ishikawa). Blots were visualized using a phosphorimager, BAS1500 (Fuji).

Return-to-growth assay and physical analysis in meiosis:

Meiotic time-course experiments were carried out as described (Cao et al. 1990; Shinohara et al. 2003). Presporulation culture was carried out in presporulation medium for 17 hr. Return-to-growth assay (Sherman and Roman 1963) was carried out as described in a previous work (Shinohara et al. 1992). The cells were collected, washed twice with water, and then resuspended in sporulation medium (SPM) to initiate meiosis. Aliquots of cells were withdrawn and genomic DNAs were prepared as described (Shinohara et al. 1997). For DSB analysis, genomic DNAs were digested with PstI and subject to electrophoresis in a 0.7% agarose gel for 18 hr at 15 V/cm in TAE buffer. For crossover analysis, genomic DNAs were digested with XhoI and analyzed on a 0.6% agarose gel for 42 hr. DNAs were transferred onto nylon membrane (Hybond-N, Amersham) by capillary transfer and analyzed by Southern hybridization after UV crosslinking. The probe used for DSB detection and crossover assay was the 1.6-kb EcoRI-PstI fragment from pNKY291 (a gift from Nancy Kleckner). Blots were visualized using a phosphorimager, BAS1500 (Fuji Film), and quantified using Image Gauge software (Fuji Film).

RESULTS

Construction of the novel xrs2 mutations:

When yeast Xrs2 and human Nbs1 proteins are compared, two highly conserved regions are found, namely an N-terminal FHA domain and a C-terminal domain. The FHA domain is known as a phosphoprotein recognition domain (Durocher et al. 1999). Xrs2 seems to lack an apparent BRCA1 C-terminus motif found in Nbs1. To analyze the function of the domains of Xrs2 encoding a 854-amino-acid protein, we constructed five truncated xrs2 mutant genes lacking either a FHA domain (xrs2-84M, -228M, -314M) or a C-terminal region (xrs2-630, -664) and three point mutations, each possessing amino acid changes in the FHA domain (xrs2-GE, -SH) or in the C-terminal conserved domain (xrs2-AA) (Figure 1). The xrs2-84M, -228M, and -314M are N-terminal deletion mutant alleles with translation starts from the 84th, 228th, and 314th methionine, encoding Xrs2 proteins with 771-, 627-, 541-amino-acid residues from the N terminus, respectively. The xrs2-630 and -664 mutants have deletions of 224 and 190 amino acid residues from the C terminus. The xrs2-AA (K641A and K645A) mutant has amino acid substitutions in the highly conserved lysine residues in the C-terminal conserved domain. The xrs2-SH (S47A and H50A) and xrs2-GE (G31E) mutants have amino acid substitution(s) in the FHA domain (Figure 1). The same amino acid substitutions in the FHA domain cause loss of function in Cds1 (fission yeast), Dun1, and Rad53 (budding yeast) proteins (Sun et al. 1998; Boddy et al. 2000; Bashkirov et al. 2003). Each mutant allele was used to replace the wild-type XRS2 gene locus on the yeast chromosome. As a control, the xrs2 null mutation, in which an entire XRS2 ORF was replaced with the URA3 ORF, was used. We constructed each mutant allele in the isogenic strain background of S288C or W303-1A and SK1 for mitotic and meiotic analyses, respectively.

Figure 1.—

A diagram of the comparison of human Nbs1 and yeast Xrs2 proteins and the construction of the truncated Xrs2 proteins. (A) Yeast Xrs2 is a homolog of human Nbs1, but Xrs2 and Nbs1 are relatively less conserved except in two domains: the FHA domain in the N terminus and the ∼35-amino-acid region in the C terminus. Solid boxes indicate these two conserved domains. (B) Comparison of the amino acid sequences of Xrs2 and Nbs1 in the conserved domains. The identical and homologous amino acids are shaded, and the mutation sites and substituted amino acid residues are shown.

Expression of the Xrs2 mutant proteins and Mre11-Xrs2 complex formation:

First, we performed Western blot analysis using an anti-Xrs2 antibody to check the amount of mutant Xrs2 proteins (Figure 2A). Whereas the expected molecular weight of the native XRS2 gene product is 96 kDa, the mutant gene products of the xrs2-84M, -228M, -314M, -630, and -664 are 87, 70, 60, 71, and 75 kDa, respectively. Western blot analysis of log-phase cell extracts showed that both the wild-type Xrs2 and the mutant proteins possessed the correct expected molecular weight. In addition, total amounts of Xrs2-84M and Xrs2-228M protein were approximately fivefold lower than in wild type (Figure 2A). All mutant proteins, except the xrs2-84M product, exhibited heterogeneous mobility on the gel, suggesting that they are phosphorylated as wild-type proteins (D'Amours and Jackson 2001).

Figure 2.—

Expression of mutant Xrs2 proteins and interaction between Mre11 and mutant Xrs2 proteins. (A) Western blotting analysis against Xrs2 proteins. The same amounts of protein in extracts from wild type (MSY1179), xrs2-GE (MSY1772), -SH (MSY1777), -84M (MSY1767), -228M (MSY1768), -314M (MSY1770), -AA (MSY1771), -630 (MSY1769), and -664 (MSY2045) were applied on each lane or five times the total amounts of protein were loaded on the three lanes from the right. (B) Western blotting of whole-cell extract (W) or Xrs2 IP from indicated strains in log phase with (+) or without (−) γ-ray irradiation. (C) Western blotting of Mre11 immunoprecipitates from extracts of the indicated strains in log phase. (D) Interaction between Xrs2 proteins and Mre11 proteins were analyzed by a two-hybrid system. β-Galactosidase activity was measured by liquid assay as described in materials and methods. BD shows pAS2-1-based plasmids containing the following subfragments: Vec, pAS2-1 alone; Xrs2, full-length XRS2 ORF; 84M, xrs2-84M ORF; 630, xrs2-630 ORF; F640A, xrs2-F640A; K641A, xrs2-K641A; K645A, xrs2-K645A. ACT shows pACT2-based plasmids containing the following subfragments: Vec, pACT2 alone; Mre11, full-length MRE11 ORF.

Since Rad50 and Xrs2 do not interact directly but form a complex through the Mre11 protein (Usui et al. 1998), we examined the ability of the mutant Xrs2 proteins to bind to the Mre11 protein by co-immunoprecipitation using an anti-Xrs2 antibody (Figure 2B). Cell lysates were prepared from log-phase cells with or without γ-ray irradiation. As shown previously (Usui et al. 1998), Western blotting of the immunoprecipitates of wild-type cells indicates that Mre11 is co-immunoprecipitated with Xrs2. Mre11 was not recovered from an extract of the xrs2 null mutant or the mre11 null mutant, which is consistent with previous reports. Among various xrs2 mutant cells, Mre11 was precipitated from the xrs-84M, -228M, -314M, -664, -SH, and -GE extracts, but not from the xrs2-630 or xrs2-AA mutants. These results indicate that the C-terminal region from 630 to 664 of Xrs2, particularly lysine 641 and 645, is critical for Mre11 binding to Xrs2. The relative ratio of Mre11 to Xrs2 in the immunoprecipitates was significantly lower in the xrs2-84M and xrs2-664 mutants and slightly lower in the xrs2-228M mutant compared to wild type. Furthermore, in the xrs2-84M and xrs2-228M mutants, the total amounts of Mre11 in the immunoprecipitates increased slightly after irradiation. However, no significant differences were found in the mobility of Xrs2 in cells irrespective of whether they were irradiated.

As a reciprocal experiment, we next performed coimmunoprecipitation using an anti-Mre11 antibody (Figure 2C). Mutant Xrs2 proteins from xrs2-84M, -228M, and -664 in co-immunoprecipitates of Mre11 were recovered in lower quantities compared to wild type. These results suggest that xrs2-84M, -228M, and xrs2-664 mutations result in an unstable interaction between Xrs2 and Mre11.

In the xrs2-AA and xrs2-630 mutants, the Xrs2 protein showed no co-immunoprecipitation with Mre11, even when the anti-Mre11 antibody was used. Our results indicate that the conserved lysine residues at 641 and 645 are very important for interactions between Mre11 and Xrs2 (see below).

Interaction between Xrs2 and Mre11 is weakened by the xrs2-84M mutation:

The co-immunoprecipitation assay revealed that the total amounts of Mre11 were significantly reduced in the xrs2-84M immunoprecipitates. Also the total amount of Xrs2 protein in the xrs2-84M mutant cells was reduced relative to wild-type cells (Figure 2A). From this it is difficult to evaluate whether the reduced amount of MRX complex in xrs2-84M is due to either a weak interaction between Mre11 and mutant Xrs2 proteins or a reduced amount of free Xrs2 protein. To distinguish between these possibilities, we next performed a two-hybrid analysis to quantify the interaction between the mutant Xrs2 and Mre11 proteins. Expression of Xrs2 or Mre11 proteins using the ADH promoter ensured the expression of a high level of each protein (data not shown). We could detect an interaction between full-length Xrs2 and Mre11 as previously reported (Ogawa et al. 1995). A significant interaction was observed between the mutant Xrs2-84M protein and Mre11, but β-galactosidase activity for this combination was 2.3 times lower than that for the wild-type pair. This result indicates that the N-terminal region of Xrs2 is important for interactions between Mre11 and Xrs2 proteins.

Effect of xrs2 mutations on DNA damage repair:

We examined each of the xrs2 mutant strains for sensitivity to ionizing radiation (Figure 3, A-i and A-ii), treatment with MMS (Figure 3, B-i and B-ii), camptothecin (CPT), and hydroxyurea (HU) (Figure 3C). The xrs2 null mutant was ∼10⁵-fold more sensitive than the wild type to γ-ray irradiation at a dose of 625 Gy (Figure 3A-ii). The xrs2-630 mutant was as sensitive to γ-rays as the xrs2 null mutant. On the other hand, the xrs2-664 mutant was as resistant to γ-rays as wild type, and the xrs2-AA mutant showed an intermediate sensitivity: 3-fold more sensitive than wild type to γ-rays after 625 Gy irradiation (Figure 3A-ii). Two mutations in the FHA domain, xrs2-GE and -SH, showed little sensitivity to γ-ray irradiation. Unexpectedly, the xrs2-84M, -228M, and also xrs2-314M mutants, which possess large amino acid deletions in the N-terminal region, showed no significant difference in sensitivity to γ-rays from the wild type (Figure 3A-i). A similar sensitivity for the xrs2 mutants to MMS (Figure 3, B-i and B-ii), CPT, or HU (Figure 3C) was observed. The same results were obtained in the pure SK1 or W303 genetic background (data not shown). The resistance of these xrs2 mutant strains, which lack an N-terminal or a C-terminal region, to DNA damage indicates that the N-terminal region of 1–313 and the C-terminal region of 664–854 are dispensable for DNA repair function of the Xrs2 protein.

Figure 3.—

γ-ray, MMS, camptothecin, and hydroxyurea sensitivity of each xrs2 mutant allele. γ-ray sensitivity (A-i and A-ii) and MMS sensitivity (B-i and B-ii) were assessed in the following strains: wild type (MSY1179), xrs2-GE (MSY1772), -SH (MSY1777), -84M (MSY1767), -228M (MSY1768), -314M (MSY1770), -AA (MSY1771), -F640A (MSY2596), -K641A (MSY1463), -K645A (MSY2598), -630 (MSY1769), -664 (MSY2045), and mre11 (MSY1296) as described in materials and methods. Two graphs (i and ii) are shown with a different range of y-axis. To facilitate comparison, the curves of wild-type and xrs2 null mutant strains are shown in each graph. (C and D) Wild-type and xrs2 mutant strains were grown to log phase and cell cultures were serially diluted, spotted onto the YPAD plates with or without 20 μm CPT, plates with or without 50 mm HU, and plates with or without 2 mg/ml PLM.

Telomere length control in xrs2 mutants:

The MRX complex is important in telomere length maintenance (Chamankhah and Xiao 1999; Tsukamoto et al. 2001). To determine which domain in the Xrs2 protein is needed for telomere function of the protein, we analyzed the steady-state length of the telomere in various xrs2 mutant alleles (Figure 4, A and B). As reported previously, the xrs2 null mutation causes a shortening of telomere lengths as in mre11 (Figure 4A, lane 3; Figure 4B, lanes 4 and 5) and the telomeres in the xrs2-630, which is defective in DNA damage repair, was as short as in the xrs2 null mutant. And the xrs2-AA mutant cells, which are slightly defective in DNA damage repair, also were indistinguishable from the xrs2 null mutant (Figure 4A, lane 9; Figure 4B, lanes 6 and 7). The xrs2-664 and xrs2-84M mutant cells also exhibited telomere shortening, although they were proficient in DNA damage repair. The xrs2-228M mutant cells showed slightly shorter telomeres than those in wild type, but were longer than those in the xrs2-84M mutant. Finally, the xrs2-314, -GE, and -SH mutant cells maintained a wild-type length of telomeres or slightly longer telomeres than those in wild type (xrs2-SH; Figure 4A, lane 5). These results indicate that the function of the Xrs2 protein in telomere homeostasis is functionally separable from that of DNA repair.

Figure 4.—

Effects of each xrs2 mutation in telomere length control. (A) Telomere structures in wild type (MSY1179), xrs2-GE (MSY1772), -SH (MSY1777), -84M (MSY1767), -228M (MSY1768), -314M (MSY1770), -AA (MSY1771), -630 (MSY1769), and -664 (MSY2045) were obtained as described in materials and methods. (B) Telomere structures in mutants of domain B and the double mutants with tel1. Strains used were wild type (W303-1A), xrs2 null (MSY2140), xrs2-AA (MSY2191), -K641A (MSY1463), -K645A (MSY2598), -F640A (MSY2596), tel1 (MSY2175), tel1 xrs2-AA (MSY2674), and tel1 xrs2 null (MSY2677).

Next, we examined the xrs2 mutants with the tel1 mutations in telomere function (Figure 4B). The tel1 single mutant showed short telomeres as previously reported (Figure 4B, lanes 14 and 15) (Chan et al. 2001). Telomeres in the tel1 xrs2-AA as well as the tel1 xrs2 double-mutant cells were as short as those in the tel1 or xrs2 single mutant (Figure 4B), indicating that the xrs2-AA mutation as well as the xrs2 null mutation is epistatic to tel1 mutation in telomere function.

Genetic analyses on meiotic phenotypes of xrs2 mutants:

Mre11 and Rad50, which are components of the MRX complex, are required for both formation of meiotic DSBs and processing of DSB ends. However, meiosis in the absence of XRS2 function has not yet been examined in detail. We analyzed meiotic phenotypes of various xrs2 mutants. All analyses for meiosis were carried out using a background of the SK1 strain, which enters meiosis in a very rapid and synchronous manner (Kane and Roth 1974). While the wild type showed 98% spore viability, the xrs2 null mutant generated very few (0.58%) viable spores (Table 2). The spore viability of xrs2-SH, -GE, and -664 cells was 95.4, 95.6, and 94.6%, respectively. While the xrs2-664 mutant showed high spore viability, the xrs2-630 mutant generated few viable spores. This result indicates that the function of the region from 630 to 664 of Xrs2 (domain B) is essential for viable spore formation. The xrs2-AA mutation greatly reduced viable spore formation, but produced significantly higher spore viability than the xrs2 null or xrs2-630 mutants did. Interestingly, three N-terminal truncated mutants, namely xrs2-84M, -228M, and -314M, which showed no mitotic defects, had a reduced spore viability of 46.7, 89.2, and 89.0%, respectively (Table 2). And viability of spores cultured in liquid SPM was reduced in the xrs2-84M mutant (Table 2) compared to that produced on SPM plates. However, the difference of spore viabilities under different culture conditions was not observed for wild type or for xrs2-228M (data not shown). Curiously, all three N-terminal truncated mutants showed a high number of four, two, and zero viable spores/tetrad but showed no increase in one or three viable spores. We could not find nonmater spores, indicative of nondisjunction of chromosome III, among two-viable-spore tetrads. This may be due to the small number of tetrads that we were able to recover. Alternatively, this may be caused by the other meiotic defect rather than by that in meiosis I. These results suggest that the function of the N-terminal region of Xrs2 is necessary for meiosis.

TABLE 2.

Spore viability

		Viable spores per ascusc
Straina	Spore viability (%)b	4	3	2	1	0
Wild type	98.1	121	8	1	0	0
xrs2-84M	46.7	37	14	25	3	51
xrs2-228M	89.2	94	20	12	4	0
xrs2-314M	89.0	93	19	16	2	0
xrs2-630	0.18	0	0	0	1	129
xrs2-664	94.6	107	19	3	1	0
xrs2-SH	95.4	115	8	5	2	0
xrs2-GE	95.6	110	18	3	0	0
xrs2-AA	5.96	2	2	5	7	114
xrs2Δ	0.58	0	0	1	1	128
Wild typed	95.0	107	20	3	0	0
xrs2-84Md	7.31	4	1	7	5	113

^aNKY1551 derivative diploid strains of indicated homozygous genomic background were used.

^bSpore viability of various xrs2 mutant strains. Each diploid strain was transferred onto a sporulation medium plate. After 48 hr of incubation at 30°, 130 asci were dissected and spore viability and distribution of tetrad types were measured.

^cThe numbers of asci with four, three, two, one, and zero viable spores are shown. A total of 130 asci were dissected.

^dSpore viability of wild-type and xrs2-84M mutant strains from liquid SPM after 24 hr of incubation from the meiosis time-course experiment (Figure 6).

We then examined the commitment of meiotic recombination in various xrs2 mutants using the return-to-growth assay. In the wild type, a 1000- and 500-fold induction of His⁺ and Arg⁺ prototrophs, respectively, was observed after 5 hr incubation in SPM (Figure 5). For the xrs2 null mutant, induction of the His⁺ and Arg⁺ prototrophs was not observed and frequency in survival of the cells decreased during further incubation in SPM. Also, the xrs2 null mutant showed a 10-fold higher frequency of mitotic (0 hr of incubation in SPM) intragenic recombination than wild type. This hyperrecombination phenotype at the same his4 heteroalleles was reported in a mre11 null mutant, in some mre11 mutant alleles (Ajimura et al. 1993; Ivanov et al. 1994; Tsubouchi and Ogawa 1998), and in a xrs2 null mutant at another locus (Ivanov et al. 1994).

Figure 5.—

Return-to-growth assay in each xrs2 mutant allele. The assay was performed as described in materials and methods. Fraction of surviving cells determined as colony-forming units (CFUs) on SC plates at the time indicated divided by CFUs on SC medium at t = 0. Fractions of surviving cells in indicated strains are shown as open diamonds. Fractions of recombinants at the HIS4 (open circles) or the ARG4 (solid circles) locus were determined by dividing the number of His⁺ (Arg⁺) CFUs on SD-His (SD-Arg) medium by the number of CFUs on SC medium. Strains used in this assay were wild type (NKY1151), xrs2-GE (MSY1509), -SH (MSY1867), -84M (MSY1494), -228M (MSY1524), -314M (MSY1992), -AA (MSY1511), -630 (MSY1628), and -664 (MSY2015).

The xrs2-630 and -AA mutants showed a defect in commitment of meiotic recombination and cell viability and hypermitotic recombination, which was similar to the xrs2 null mutant (Figure 1). The xrs2-84M mutant showed induction of the His⁺ prototroph but after 24 hr incubation the total amount was eightfold lower than that in wild type. The xrs2-228M and -314M mutations conferred no defects in meiotic recombination and cell viability in this assay and the xrs2-84M, -228M, and -314M mutants showed no hypermitotic recombination phenotype. Finally, there were no significant differences in meiotic phenotypes among xrs2-GE, -SH, -664, and wild type.

Physical analysis of meiotic DSBs and recombinants in xrs2 alleles:

To identify regions of the Xrs2 protein that are important in meiotic DSB formation and/or repair of DSB, we analyzed meiotic DSBs and crossover recombination in various xrs2 mutant alleles by Southern blotting (Figure 6A). We analyzed the formation of meiotic DSBs at the HIS4-LEU2 hot spot on chromosome III. DSBs are induced at two sites in this locus and can be detected as two distinct bands separated from a parental band in Southern blotting analysis. In the wild type, the DSBs reached a maximum after 3 hr incubation in SPM and the bands appeared smeared due to resection of the 5′-ends of the DSBs. Thereafter the DSBs disappeared after 6 hr incubation, indicating that the breaks had been turned over. Crossover recombinants, which show different mobility from that of the parental molecules due to restriction site polymorphisms, appear to start after 4 hr incubation and include 18% of parental bands. In the xrs2 null mutant, meiotic DSBs were not detected, indicating that Xrs2 is necessary for DSB formation during meiosis, and crossover recombinants were barely detectable. A similar deficiency was previously reported in meiotic DSB formation for mre11 and rad50 (Alani et al. 1990; Johzuka and Ogawa 1995).

Figure 6.—

Physical detection of meiotic DSBs and crossover products at the his4-LEU2 hot spot. (A) Meiotic DSB and its repair (top) and recombinants (middle) in indicated strain. Genomic DNAs from cells harvested at different times after incubation in SPM were analyzed for DSB or recombinants as described in materials and methods. The position of full-length fragments that do not contain meiotic DSBs is indicated by “P.” The positions of fragments generated by meiotic DSBs at the two major DSB sites in the HIS4 locus are shown as DSB I and DSB II. (Bottom) The amounts of DSBs (open diamonds) or recombinants (solid diamonds) were quantified. (B) Meiosis progression and spore formation frequency in the xrs2 mutant alleles. Progression of meiosis I (MI) and meiosis II (MII) was examined by staining cells with DAPI at each time point in SPM. Percentages of post-MI (solid circles) and post-MII (open circles) cells are shown. Each graph shows percentages of matured spores per total cells at 24 hr in SPM. Strains used in this assay are wild type (NKY1151), xrs2-GE (MSY1509), -SH (MSY1867), -84M (MSY1494), -228M (MSY1524), -314M (MSY1992), -AA (MSY1511), -630 (MSY1628), and -664 (MSY2015).

The xrs2-GE, -SH, and -664 mutants showed no apparent defects in timing of the appearance of meiotic DSBs compared to wild type. Also the total amounts of crossover products for these mutants were the same as those for wild type.

The xrs2-AA and -630 mutants showed defects in formation of meiotic DSBs and crossover products similar to those of the xrs2 null mutant (Figure 6A). While the xrs2-314M mutant showed no defects in the formation of DSBs and the total amounts of crossover products, the xrs2-228M mutant, which has a shorter deleted region than xrs2-314M, produced fainter DSB bands that were still detectable after 10 hr incubation. The formation of the recombinants was delayed 3 hr in the xrs2-228M compared to wild type. In addition, crossover products were slightly decreased (76.5% of wild type after 12 hr incubation) in the xrs2-228M mutant. These suggest a role of the N-terminal region of Xrs2 in the conversion of the DSB into crossovers. The xrs2-84M mutant produced a reduced level of DSBs and showed a defect in formation of crossover products similar to the xrs2 null mutant. And the xrs2-GE, -SH, -84M, -228M, and -314M showed a slight delay not only in timing of DSB disappearance and appearance of crossover products (Figure 6A) but also in transition to meiosis I (Figure 6B). These findings suggest that the N-terminal region as well as Mre11 binding domain is involved in meiotic function of Xrs2.

Quantification of the total amount of meiotic DSBs in the rad50S xrs2 double mutants:

To confirm that the total amount of DSBs induced at the HIS4-LEU2 locus did indeed decrease in the xrs2-84M, xrs2-228M, and xrs2-314M mutant cells, we analyzed meiotic DSBs in various xrs2 mutants with the rad50S mutation (Figure 7A). From previous studies it is known that in rad50S background cells, meiotic DSBs are generated with normal timing but the ends are never processed (Alani et al. 1990). Therefore, it is possible to quantify the actual amount of meiotic DSBs induced in this mutant background. In the rad50S xrs2-84M and rad50S xrs2-228M double mutants, the amount of DSBs decreased 14 and 40% (Figure 7B), respectively, from that in the rad50S single mutant after 8 hr in SPM. However, the rad50S xrs2-314M double mutant showed no reduction in the amount of DSBs. Therefore, the rad50S mutation was epistatic to the xrs2 N-terminal truncated alleles, which showed partial defects in the formation of meiotic DSBs and in the frequency of meiotic DSB formation, but were proficient in processing the DSB ends. While the rad50S xrs2-630 and rad50S xrs2 null mutant cells generated no meiotic DSBs, although greatly reduced, a significant level of DSBs was detected in the rad50S xrs2-AA double mutant.

Figure 7.—

Meiotic DSBs in rad50S xrs2 double mutants. (A) Meiotic DSBs in indicated strains were detected by Southern blotting. Genomic DNAs from harvested cells at each time point indicated were analyzed as described in materials and methods. The positions of fragments with or without meiotic DSBs are shown in Figure 6. (B) The amounts of meiotic DSB I at each time point were quantified. The strains used in this assay are shown in Table 1.

Multi-copy suppression of the xrs2-84M mutation by itself:

Curiously, the xrs2-84M mutant cells showed severe defects, especially in meiosis, compared with the xrs2-228M or xrs2-314M mutants, which contained larger N-terminal truncations in Xrs2 than xrs2-84M did. In the xrs2-84M mutant cell, the total amount of Xrs2 protein was significantly reduced compared with that in xrs2-228M or xrs2-314M (Figure 2A). To confirm whether the defects in xrs2-84M were caused by a reduction in mutant Xrs2 protein, we analyzed the complex formation with Mre11 and spore viability in the xrs2-84M mutant cells, which contain multi-copy plasmids or single-copy plasmids with the xrs2-84M gene. From Western blot analysis, the total amount of Xrs2-84M mutant protein increased due to overexpression (Figure 8A). The co-immunoprecipitation assay revealed that elevation of the total amount of Xrs2-84M mutant protein increased the total amount of Mre11 protein in the co-immunoprecipitates as determined using an anti-Xrs2 antibody.

Figure 8.—

Effect of overexpression of Xrs2 and Mre11. (A) Suppression of defects in xrs2-84M by overexpressed Xrs2-84M protein. Western blotting of whole-cell extract (W) and Xrs2 IP from the xrs2-84M mutant strain (MSY1767) or wild-type strain (MSY1179) in log phase with each plasmid indicated. YCp, pRS313; YEp, pRS423; YCp-xrs2-84M, pMS402; YEp-xrs2-84M, pMS403. (B) Overexpression of Mre11 suppresses the MMS sensitivity of the xrs2-AA. The MMS sensitivity of the xrs2-AA and xrs2 null mutants with various plasmids was analyzed on SD (Ura) plates containing 0.02% of MMS. Vector, pRS316; YCp-XRS2, YCp50-XRS2; YEp-MRE11, pMS421.

We examined spore viability in the xrs2-84M mutant cell with and without overexpression of the Xrs2-84M mutant protein. The spore viability of xrs2-84M with vector plasmid was 9.6% (Table 3); however, spore viability increased from 38.1 to 77.7% by overexpression of the Xrs2-84M protein using either single-copy or high-copy-number plasmids, respectively, in a dose-dependent manner. We also found that overexpression of the Xrs2-84M protein suppressed a defect in the telomere length in xrs2-84M in a dose-dependent manner (data not shown). These results indicate that the severe defect of the xrs2-84M mutant in meiotic recombination and telomere metabolism was caused by a reduced level of Xrs2 protein, but not by deletion of the N-terminal domain per se. However, a high copy of the mutant allele could not recover spore viability completely, suggesting a minor role of this region in meiotic recombination, as described above.

TABLE 3.

Suppression of spore viability in thexrs2-84M mutant

		Viable spores per ascus (%)c
Plasmida	Spore viability (%)b	4	3	2	1	0	P-value
Vector	9.6	5	5	11	3	132	—
YCp-XRS2	82.7	116	13	6	1	20	<0.001
YEp-XRS2	86.4	116	18	8	5	9	<0.001
YCp-xrs2-84M	38.1	39	11	23	3	80	<0.001
YEp-xrs2-84M	77.7	96	27	9	3	21	<0.001

^aThe xrs2-84M haploid (MSY2223) with the various plasmids indicated was mated with MSY2225 on a SD-Trp plate for 6 hr and then transferred onto a sporulation plate. Vector, pRS314; YCp-XRS2, pMS398; YEp-XRS2, pMS399; YCp-xrs2-84M, pMS387; YEp-xrs2-84M, pMS388.

^bAfter 48 hr of incubation, 156 asci were dissected, and spore viability and distribution of tetrad types were measured.

^cThe numbers of asci with four, three, two, one, and zero viable spores are shown. A total of 156 asci were dissected. P-values indicate the significance of difference from the vector control.

Dissection of the Mre11-interaction domain of Xrs2:

The above results reveal the importance of the Mre11-interaction domain, region B, in Xrs2 functions. The xrs2-AA mutant seems to have a reduced, but significant, amount of the MRX complex. Then we examined the effect of the overexpression of Mre11 on the xrs2-AA. The overexpression of Mre11 suppresses the MMS sensitivity of the xrs2-AA mutant, supporting a residual Mre11-Xrs2 interaction in the mutant. We further studied the role of each amino acid residue in this region. The xrs2-AA allele was separated into two alleles, xrs2-K641A and xrs2-K645A, and analyzed for its abilities to bind to Mre11, to repair DNA damage [MMS, CPT, HU, and phleomycin (PLM) sensitivity], to maintain telomeres, and to form viable spores. In addition, we constructed the xrs2-F640A allele, in which a conserved phenylalanine at 640 is substituted by alanine. The two-hybrid assay showed that xrs2-K641A, xrs2-K645A, and xrs2-F640A are partially deficient in the interaction with Mre11 (Figure 2D). However, all three mutants showed similar sensitivity to DNA-damaging agents such as CPT, HU, and phleomycin as in wild type (Figure 3D) and maintained a wild-type length of telomeres (Figure 4B). And xrs2-K641A, xrs2-K645A, and xrs2-F640A mutations did not affect spore viability (94.2, 94.6, and 94.2%, respectively) (data not shown). Thus, each amino acid residue in this conserved region B seems to play an overlapping role in the interaction with Mre11.

DISCUSSION

Functional domains of Xrs2 protein in S. cerevisiae:

In this study, we constructed several novel mutations of the XRS2 gene and analyzed them in complex formation with Mre11 protein, DNA damage repair, telomere control, spore viability, meiotic DSB formation, and meiotic recombination. The phenotypes of these xrs2 mutants are summarized in Figure 9. On the basis of the results, we discuss the functional domains of the Xrs2 protein and the role of Xrs2 in the MRX complex. Xrs2 protein has three distinct domains: the N-terminal region containing the FHA domain, the conserved C-terminal domain, and the less conserved C-terminal ends of the protein.

Figure 9.—

Summary of phenotypes of various xrs2 mutants.

Our results indicate that the 630–664 region (region B) of Xrs2 is essential for Xrs2-Mre11 interaction. This region is well conserved among the species (Figure 1) (Chahwan et al. 2003) and, consistent with this, previous works have shown that a Mre11 interaction domain is located in the C-terminal region of human Nbs1 (Desai-Mehta et al. 2001; Maser et al. 2001; Tauchi et al. 2001). Interestingly, this Mre11-interaction domain plays a more important role in telomere maintenance and meiotic recombination than in DNA repair (see below).

The xrs2 mutant lacking the 664–854 region (xrs2-664) is defective only in telomere maintenance, but is proficient in DNA repair and meiotic recombination. It has already been reported that Tel1 associates with the C-terminal (693–854 aa) region of Xrs2 and is recruited at the DSB site through this region of the Xrs2 protein (Nakata et al. 2003). The xrs2-11 mutation lacking the C-terminal region, which is required for association with Tel1 (Nakata et al. 2003), produced the same phenotype as the xrs2-664 mutation did (H. Shima and M. Suzuki, unpublished result; K. Sugimoto, personal communication). Thus, the C-terminal Tel1-binding domain of Xrs2 is required only for telomere maintenance.

Our results described here could not assign any functions for a large N-terminal domain of Xrs2, although it has a minor role in meiotic recombination and DNA damage tolerance (see below). However, our recent analyses showed that the N-terminal truncation mutants as well as the FHA mutants exhibit a defect in the resealing of linearized plasmids (M. Suzuki, unpublished results). Thus, at least an N-terminal region of Xrs2, possibly the FHA domain, is likely to be responsible for nonhomologous end joining (NHEJ). Since the FHA domain is a binding motif for the protein, this region might be involved in the interaction with the component(s) of the NHEJ. Indeed, in yeast, Xrs2 binds to a complex containing Dnl4 and its interactor, Lif1 (Chen et al. 2001). Further studies of this domain may shed light on the role of Xrs2 or the MRX complex in NHEJ.

In addition, our results suggest a minor role for the N-terminal domain of Xrs2, which cannot be detected analyzing xrs2 null mutants. Spore viability in the xrs2-314M mutant was slightly lower than that in the wild type, and the ratio of the 2sv tetrads to total tetrads was also significantly high. The physical analysis of meiotic recombination revealed a slight defect in the xrs2-314M mutant in the appearance of crossover product and meiosis progression, but not in meiotic DSB formation. This cannot be explained by the reduced amount of the MRX in the xrs2-314M mutant, since the mutant has a wild-type amount of the mutant MRX complex. Thus, this suggests a role for the N terminus of Xrs2 in meiosis.

The role of the MRX in a late step of recombination:

The analysis of the xrs2-84M mutant suggests that meiotic DSBs were produced at a reduced level and were not repaired efficiently in this mutant, suggesting that a function of Xrs2 is required for the steps after meiotic DSB formation. Interestingly, the physical analysis revealed that the DSB bands in the xrs2-84M single mutant were smeared. This result indicates that Spo11 proteins are removed from the DSB ends in the xrs2-84M mutant. Thus, the defect in xrs2-84M may involve a step in meiotic recombination other than processing of DSB ends.

The interaction of Xrs2 with Mre11 is a critical determinant for the MRX functions:

Our analyses reveal the correlation of the degree of the binding of Xrs2 protein to Mre11 with the functions of the MRX. A weak interaction, and thus a small amount of the complex, is sufficient for the ability of the complex to repair DNA damage, while a strong interaction or more of the MRX complex is required for meiotic recombination and telomere maintenance. Indeed, we showed that the xrs2-AA mutation, which makes the interaction between the Xrs2 and Mre11 very unstable, still retain functions in DNA repair, but is defective in the formation of meiotic DSBs and telomere maintenance. While we cannot deny the possibility that the xrs2-AA mutation is a separation-of-function mutation, the mutation is similar to a situation observed for the xrs2-84M. The xrs2-84M mutant, which has a reduced level of the MRX, showed few defects in repairing damaged DNA, but showed a defect in meiotic recombination and telomere length control. Indeed, the overexpression of Xrs2-84M mutant protein suppressed the defects of xrs2-84M in spore viability. These results suggest that the MRX complex for DNA damage repair should somehow be quantitatively or qualitatively different from the MRX complex for telomere elongation and meiotic DSB formation.