- Email: [email protected]
mgs1 sgs1 double deletion mutants
DNA Repair 1 (2002) 671–682
Characterization of the slow-growth phenotype of S. cerevisiae whip/mgs1 sgs1 double deletion mutants Dana Branzei a,∗ , Masayuki Seki a , Fumitoshi Onoda a , Hideki Yagi a,b , Yoh-ichi Kawabe a , Takemi Enomoto a a
b
Molecular Cell Biology Laboratory, Graduate School of Pharmaceutical Sciences, Tohoku University, Aoba-ku, Sendai, Miyagi 980-8578, Japan Cell Biology Laboratory, School of Pharmaceutical Sciences, Kinki University, 3-4-1 Kowakae, Higashiosaka, Osaka 577-8502, Japan Received 5 February 2002; received in revised form 26 April 2002; accepted 8 May 2002
Abstract RecQ DNA helicases from many organisms have been indicated to function in the maintenance of genomic stability. In human cells, mutation in the WRN helicase, a RecQ-like DNA helicase, results in the Werner syndrome (WS), a genetic disorder characterized by genomic instability and premature ageing. Similarly, mutation in SGS1, the RECQ homologue in budding yeast, results in genomic instability and accelerated ageing. We previously demonstrated that mouse WRN interacts physically with a novel, highly conserved protein that we named WHIP, and that in budding yeast cells, simultaneous deletion of WHIP/MGS1 and SGS1 results in slow growth and shortened life span. Here we show by using genetic analysis in Saccharomyces cerevisiae that mgs1∆ sgs1∆ cells have increased rates of terminal G2/M arrest, and show elevated rates of spontaneous sister chromatid recombination (SCR) and rDNA array recombination. Finally, we report that complementation of the synthetic relationship between SGS1 and WHIP/MGS1 requires both the helicase and Top3-binding activities of Sgs1, as well as the ATPase activity of Mgs1. Our results suggest that Whip/Mgs1 is implicated in DNA metabolism, and is required for normal growth and cell cycle progression in the absence of Sgs1. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Whip; Mgs1; Sgs1; Genomic instability; Cell cycle arrest
1. Introduction Werner syndrome (WS) is a human autosomal recessive disorder characterized by the premature onset of a number of age related diseases. The gene defective in WS, WRN, encodes a RecQ-like DNA helicase [1]. Defects in other human RecQ-like DNA helicases, BLM and RECQL4 result in genetic disorders known as Bloom syndrome and Rothmund–Thomson syndrome, both characterized by increased genomic ∗ Corresponding author. Tel.: +81-22-217-6874; fax: +81-22-217-6873. E-mail address: [email protected] (D. Branzei).
instability and a predisposition for development of malignancies. At the cellular level, Bloom syndrome cells show increased frequencies of sister chromatid exchange and slowed DNA replication; Rothmund– Thomson syndrome cells show increased frequencies of chromosomal breaks and rearrangements; and WS cells display variegated translocation mosaicism and a decreased replicative lifespan in vitro (reviewed in [2]). Despite exciting advances in the biochemistry of WRN, its cellular function and the pathology of WS are yet to be fully understood. Of particular interest is to understand how WRN contributes to maintaining genomic stability and how, when mutated, leads to a
1568-7864/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 1 5 6 8 - 7 8 6 4 ( 0 2 ) 0 0 0 7 3 - 3
672
D. Branzei et al. / DNA Repair 1 (2002) 671–682
premature ageing phenotype. We hoped that by identifying WRN-interacting proteins we could delineate the principal pathways in which WRN participates, and the process that finally leads to the clinical manifestations of WS. We have screened for proteins that interact physically with mWRN and one of the proteins isolated in this way was a highly conserved, RFC-like protein that we named WHIP [3]. To investigate whether a functional relationship existed between WRN and WHIP, we turned to the yeast Saccharomyces cerevisiae where we constructed the corresponding double mutants and found that whip∆ sgs1∆ cells have a shortened lifespan and display a slow-growth phenotype [3]. The budding yeast homologue of WHIP was independently isolated and referred to as MGS1 [4]. The S. cerevisiae genome contains only one RecQ helicase gene, SGS1. Like its human homologues, SGS1 is required for maintaining chromosomal stability, and sgs1∆ mutants display hyper-recombination, and increased rates of chromosome loss and missegregation [5,6]. It has been demonstrated previously that, in a striking parallel with WS in humans, sgs1∆ mutants have a mean life span that is about 40% of that of wild-type strains [7]. Moreover, mutations in SGS1 result in hyper-sensitivity to various DNA damaging agents suggesting that Sgs1 functions in DNA repair. Further insight in the role of SGS1 comes from its interactions with topoisomerases. SGS1 was first identified as a suppressor of the slow-growth phenotype of top3 mutants [5], and sgs1 mutants show an epistatic relationship with top3 mutants, suppressing the elevated recombination rate and the sporulation block displayed by top3 mutants [8,9]. To help define the primary role of Mgs1, and to understand the causes of the slow-growth phenotype that mgs1∆ sgs1∆ cells display, we analyzed in detail the synthetic slow-growth phenotype of mgs1∆ sgs1∆ cells in regard to genomic instability and cell cycle defects.
2. Materials and methods 2.1. Strains The yeast strains used in this study are listed in Table 1. Most strains were constructed using standard
genetic methods. Complete oligonucleotide sequences of the primers used for disruption or disruption check will be provided upon request. To obtain a “START to STOP” MGS1deletion strain, the KANMX4 gene was amplified from pFA6a-KanMX4, a kind gift from Wach et al. [10], using MGS1-specific primers with 40–50 nt identity to MGS1 at their 5 ends. The resulting PCR fragment was transformed into the wild-type strain MR966, and G418-resistant colonies were selected. Correct replacement was verified by PCR amplification of the two replacement junction regions. To obtain additional MGS1 deletion mutants, genomic DNA from MR966w was amplified with two primers located about 600 nt upstream and downstream of the KANMX4 integration sites, and the strains were transformed with the PCR products, to give geneticin resistant cells. Proper integration and loss of the MGS1 gene were confirmed by PCR. SGS1 was disrupted as described previously [11]. The rad17::LEU2 strains were constructed by transforming yeast cells with BamH1-cut pWL8 plasmid, which was a gift from Dr. L.H. Hartwell. To obtain the rad52::LEU2 deletion strain 364r52, the LEU2 gene was amplified from pRS305, using RAD52 specific primers with 40-50 nt identity to RAD52 at their 5 ends, and to obtain additional rad52::LEU2 disruptants, genomic DNA from 364r52 was amplified with two primers located about 900 nt upstream and downstream of the LEU2 integration sites, and the strains were transformed with the PCR products, and Leu+ cells were selected. To obtain rad6::HIS3 disruptants, the rad6::HIS3 cassette was amplified by PCR using genomic DNA from W303␣ rad6 strains, provided generously by Dr. M. Saijo. The primers used for disruption were positioned about 400 nt upstream and downstream of the HIS3 integration sites. In all cases proper integration of the disruption cassettes was confirmed by PCR. 2.2. Plasmids The plasmids carrying full length SGS1 (Ycp1305) or mutations in SGS1 (K706A, 12-13G/S) were described in [12]. These plasmids were constructed by subcloning SGS1 into pRS314, a single copy vector containing the TRP1 marker, and then introducing mutations in SGS1 by using mutagenic DNA primers and the QuickChange Site-Directed Mutagenesis kit (Stratagene) [12]. For construction of the single
D. Branzei et al. / DNA Repair 1 (2002) 671–682
673
674
D. Branzei et al. / DNA Repair 1 (2002) 671–682
copy plasmid containing WHIP/MGS1, the ORF of WHIP/MGS1 was amplified from the yeast genome, and a HindIII–BamH1 fragment was subcloned into Ycplac22, a vector containing the TRP1 auxotrophic marker, to give Ycplac22-WHIP. The plasmid was checked with restriction enzymes and by PCR, and the in-frame insertion of WHIP/MGS1 was confirmed by DNA sequencing. The mutation in the Walker A site in Whip/Mgs1, K183A, was introduced by using mutagenic DNA primers and QuickChange Site-Directed Mutagenesis kit (Stratagene), and the mutation was confirmed by DNA sequencing.
2.5. Immunofluorescence
2.3. Growth curves, doubling-time and plating efficiency
2.6. Sister chromatid recombination
Overnight cultures were diluted to an OD600 nm of 0.1 and grown in YPAD for 12 h. Samples were taken every 1 h and their OD600 nm measured, and doublingtime calculated. During the time course, cell samples were taken, sonicated and counted. Appropriate dilutions were plated on several YPAD plates, and incubated at 30 ◦ C for 3 days. The number of colonies formed was counted and the plating efficiency determined. 2.4. Cell cycle analysis Cells were grown to early log phase when the cultures were harvested and resuspended in fresh YPAD medium to a concentration of 107 cells/ml. Cells were treated with ␣-factor at a final concentration of 5 g/ml or with nocodazole at a final concentration of 10 g/ml for 2 h. After confirmation of G1 or G2 arrest under the microscope, the ␣-factor or the nocodazole was washed away and cells were resuspended in fresh YPAD medium. Samples were taken at the indicated time points and fixed overnight at 4 ◦ C in 1 ml of 70% ethanol. The fixed cells were washed, sonicated, and resuspended in 400 l of 50 mM sodium citrate (pH 7.0), and then treated with ribonuclease A (0.25 mg/ml) for 1 h at 50 ◦ C, followed by treatment with proteinase K (0.5 mg/ml) for 1 h at 50 ◦ C. Propidium iodide was then added (16 g/ml) and cells were incubated for at least 12 h at 4 ◦ C. The DNA content of cells was analyzed by a BectonDickinson FACScan flow cytometer using the CellQuest system.
Cells were grown to midlog phase and fixed for 1 h in 3.7% formaldehyde. The cells were treated with 0.3 mg/ml zymolyase-100T for 30 min at 30 ◦ C. DNA was visualized by staining with 4 ,6-diamidino-2-phenylindole (DAPI), and microtubules were visualized using the anti-TUB (Y0l1/34) antibody (Serotec) diluted 1:500 in PBS/1% BSA/0.1% Triton X-100 and a 1:100 diluted Cy3-conjugated anti-rat secondary antibody (Jackson, ImmunoResearch Laboratories, INC). Digital images were obtained using a Zeiss system.
Cells in mid-log phase were washed and plated at appropriate dilutions on YPAD (or SC-Trp when the strains were carrying plasmids) and SC/-His (or SC/-His, -Trp when the strains were carrying plasmids) plates to determine sister chromatid recombination (SCR) frequencies [13]. The experiments were done three times with eight cultures for each mutant strain when strains were not carrying plasmids, and twice with seven cultures when strains were carrying plasmids. The recombination frequencies were calculated according to the median method described by Lea and Coulson [14]. 2.7. rDNA recombination Cells in stationary phase were washed and plated at appropriate dilutions on SC-Ura (or SC-Ura, -Trp when the strains were carrying plasmids) and SC + 5-FOA (or SC-Trp + 5-FOA when the strains were carrying plasmids) plates for detection of rDNA URA3 marker loss. The experiments were done three times with six cultures for each strain when strains were not carrying plasmids, and twice with seven and six cultures, respectively, when strains were carrying plasmids. The recombination frequencies were calculated according to the median method described by Lea and Coulson [14]. 2.8. Drop assay Log phase grown cells were harvested, washed once in distilled water, counted and diluted. Ten-fold serial dilutions of cells (105 , 104 , 103 , and 102 cells)
D. Branzei et al. / DNA Repair 1 (2002) 671–682
were spotted onto YPAD. The plates were incubated at the indicated temperatures for 2 days, and the cells were photographed.
675
3. Results
bud neck, the spindle was short and located at the bud neck, characteristic of cells in metaphase, prior to or at the beginning of chromosome segregation (Fig. 1B). Thus, the phenotype of mgs1∆ sgs1∆ cells is similar to that observed in cdc mutants with defects in DNA replication or mitosis.
3.1. Slow-growth phenotype displayed by mgs1∆ sgs1∆ cells
3.2. Characterization of the mitotic arrest of mgs1∆ sgs1∆ cells
We found that mWRN interacts in vivo and in vitro with a novel protein, mWHIP, and that simultaneous deletion of SGS1 and WHIP/MGS1 in S. cerevisiae results in a slow-growth phenotype and a dramatic decrease in the life span of yeast cells [3]. We wished to further characterize the slow-growth phenotype of mgs1∆ sgs1∆ mutants and determine its causes. To accomplish this, we constructed mgs1∆ sgs1∆ double mutants in different backgrounds and observed that these mutants grew much slower than wild-type or single mutant strains (Table 2). Furthermore, the plating efficiency of mgs1∆ sgs1∆ cells was reduced to about 40% of that of wild-type cells, while that of sgs1∆ cells was found to be about 70% of that of wild-type cells (Table 2). We observed that mgs1∆ sgs1∆ double mutants show an increased proportion of dumbbell shaped cells, and consistent with our observation, the flow cytometry profiles of log phase mgs1∆ sgs1∆ cells displayed a major peak of G2 content (Fig. 1A). To determine the stage of mitosis at which the cells arrested, we stained wild-type and mutant cells with DAPI and anti-tubulin antibodies to visualize the mitotic spindle. Approximately 18% of mgs1∆ sgs1∆ cells were large budded, with a single nucleus at the neck of the mother cell. This phenomenon was also observed to a smaller extent in sgs1∆ cells (6%), but was almost absent in wild-type and mgs1∆ cells (Fig. 1B). In the cells that had the nucleus at the
The G2 peak accumulation observed in mgs1∆ sgs1∆ cells could be due to DNA damage or failure to complete DNA synthesis. We monitored cell cycle progression by synchronizing cells in G1 or G2 by exposure to ␣-factor or nocodazole, respectively. We found that mgs1∆ sgs1∆ mutants were able to complete S-phase, but showed great difficulties in progressing through mitosis (data not shown). If mgs1∆ sgs1∆ cells undergo cell cycle arrest because of DNA lesions, the G2/M arrest may be mediated by a functional DNA damage checkpoint. Six genes, RAD9, RAD17, RAD24, MEC1, MEC2, and MEC3 have been shown to be required for DNA damage arrest of cdc mutants in G2/M [15]. To determine the role of the RAD17 checkpoint in cell cycle arrest in mgs1∆ sgs1∆ mutants, we constructed mgs1∆ sgs1∆ rad17∆ triple mutants. Flow cytometric analysis of exponentially growing cells showed that deletion of RAD17 in mgs1∆ sgs1∆ strains reduced the level of arrest at G2/M (Fig. 2). We also observed that the aberrant morphology distribution of mgs1∆ sgs1∆ cells was partially suppressed by the rad17∆ mutation, but synchronization experiments with ␣-factor showed that the G2 accumulation of mgs1∆ sgs1∆ cells is only little suppressed by RAD17 deletion (data not shown). Viability analysis of mgs1 sgs1 rad17 cells showed that their viability is similar to that of mgs1 sgs1 cells, and not lower as it might have been expected. Perhaps most of the DNA damage that accumulates in mgs1∆ sgs1∆ cells is unrepairable and leads to cell death. That also explains the short life span of mgs1∆ sgs1∆ cells, which is only of about 4–5 generations [3].
Table 2 Effects of mgs1∆ and sgs1∆ mutations on doubling time and plating efficiencies Genotype
Doubling time (min)
Wild type mgs1 sgs1 mgs1 sgs1
94.6 94.8 105.8 170.4
± ± ± ±
2.82 4.25 3.92 6.61
The strains used were MR966 derivatives.
Plate efficiency (%) 97.6 87.9 69.7 39.2
± ± ± ±
4.2 2.9 3.9 4.46
3.3. Increased mitotic recombination in mgs1∆ sgs1∆ cells We have seen that mgs1∆ sgs1∆ cells display a cell cycle and morphology phenotype similar to that
676
D. Branzei et al. / DNA Repair 1 (2002) 671–682
Fig. 1. Cell cycle and morphological analysis of mgs1∆ sgs1∆ cells. (A) Cell cycle analysis of midlog-phase cells. (B) Analysis of nuclear and tubulin spindle morphology. Indirect immunofluorescence was performed on midlog-phase wild-type, mgs1∆, sgs1∆ and mgs1∆ sgs1∆ cells. At least 200 cells from each strain were scored. Percentages of cells that displayed nucleus at the bud neck with a short mitotic spindle were as follows: wild-type, 0.1%; mgs1∆, 1%; sgs1∆, 6%; mgs1∆ sgs1∆, 18%. The strains used in A and B were MR966 and KJY374 derivatives.
observed in replication and segregation mutants (Fig. 1). It has been reported that mutations in genes involved in DNA replication lead to increased mitotic recombination frequencies, a feature that distinguishes them from mitotic defects that show only elevated chromosome loss and missegregation frequencies [16]. We assayed the frequency of SCR, which is thought to be mainly the outcome of recombinational repair of DSBs that arise mainly in the S-phase as a result of DNA replication [13,17,18]. We found that SCR frequencies were increased in mgs1∆ sgs1∆
mutants as compared to the single mutants alone (Fig. 3A). Previous reports have shown that Sgs1 might play an important role in the rDNA and at telomeres, to suppress homologous recombination. sgs1∆ cells display an about seven-fold increase in rDNA marker loss (Fig. 3B, Table 3, and [5]), and mgs1∆ cells have a 3.5-fold increase in rDNA recombination frequencies (Fig. 3B, Table 3, and [4]). We found that mgs1∆ sgs1∆ cells display a 20-fold increase in rDNA array recombination frequencies (Fig. 3B and Table 3).
D. Branzei et al. / DNA Repair 1 (2002) 671–682
677
Fig. 2. Flow cytometric analysis of the mitotic arrest that occurs in mgs1∆ sgs1∆ cells. Flow cytometric analysis of midlog-phase cells. The strains used were MR966 derivatives.
Fig. 3. Effect of mgs1 and sgs1 mutations on spontaneous mitotic recombination. (A) Histograms of the sister chromatid recombination (SCR) frequencies. The value indicates the number of recombinants per 106 viable cells. The strains used were yMP10381 derivatives. (B) Histograms of the rDNA recombination frequencies. The value indicates the number of recombinants per 104 viable cells. The strains used were KJY374 derivatives.
678
D. Branzei et al. / DNA Repair 1 (2002) 671–682
Table 3 Effects of mutations in the helicase and Top3-binding activities of Sgs1 on complementation of increased recombination frequencies displayed by mgs1 sgs1 cells Relevant genotype
Plasmid transformed
SCE frequency (SD-His,Trp/105 cells)a
wild-type mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1
pRS314 pRS314 pRS314 pRS314 Ycp1305 K706A 12-13G/S
4.02 4.55 10.21 24.22 4.94 15.04 16.4
± ± ± ± ± ± ±
0.65 0.32 1.65 2.9 0.9 1.42 2.1
rDNA recombination frequency (SC-Trp+5-FOA/103 cells)b 5.6 15.6 41.7 110.6 17.92 113.78 119.84
± ± ± ± ± ± ±
0.9 2.5 9.2 10.2 3.2 8.2 6.09
pRS314 is a single copy vector containing the TRP1 marker, Ycp1305 is pRS314 carrying full length SGS1, K706A is Ycp1305 with mutations in SGS1 at K706 (K706A), and 12-13G/S is Ycp1305 with mutations in SGS1 at amino acids E12 and H13 (E12G/H13S) [12]. a The strains used were yMP10381 derivatives. b The strains used were KJY374 derivatives.
3.4. RAD52 deletion does not suppress the slow-growth phenotype of mgs1∆ sgs1∆ cells It was previously reported that knocking out genes encoding homologous recombination proteins such as Rad51 and Rad52 suppresses the slow-growth phenotype of sgs1∆ srs2∆ cells [19,20] and the
synthetic lethality of sgs1 mus81 cells (see [21]), suggesting that the lethal mitotic structures that accumulate in these cells are the product of recombination. Since mgs1∆ sgs1∆ cells display increased levels of mitotic recombination (Fig. 3, Table 3), we analyzed the possibility that the mitotic arrest of these cells is the result of promiscuous homologous recombination.
Fig. 4. Effect of RAD52 and RAD6 deletions on the growth ability of mgs1 sgs1 cells. Ten-fold serial dilutions of log phase grown cells (105 , 104 , 103 , and 102 cells) were spotted onto YPAD, and photographed after 2 days. (A) Growth of mgs1 sgs1 rad52 cells as compared to that of mgs1 sgs1 strains. (B) Growth of rad6 sgs1 cells as compared to that of single mutants. The strains used in A and B were in the A364 background.
D. Branzei et al. / DNA Repair 1 (2002) 671–682
We reasoned that if the lesions that form in mgs1∆ sgs1∆ cells are processed by the homologous recombination proteins into chromosomal intermediates that cannot be separated prior to mitosis and lead to mitotic arrest, then deletion of RAD52 should suppress the slow growth and mitotic arrest of these cells. We constructed mgs1∆ sgs1∆ rad52 cells, but neither the growth was improved nor the mitotic arrest reduced in these cells as compared to mgs1∆ sgs1∆ cells (Fig. 4A). In fact, the growth of mgs1∆ sgs1∆ rad52 cells was slightly slower than that of mgs1∆ sgs1∆ cells, suggesting that mgs1∆ sgs1∆ cells use the RAD52 homologous recombination pathway in repairing DSBs and DNA lesions. However, since the mgs1∆ sgs1∆ rad52 combination was not synthetic lethal, it is conceivable that additional repair pathways can be recruited for the repair of DNA lesions that accumulate in mgs1∆ sgs1∆ cells. One such pathway could be the RAD6/RAD18 post-replication repair pathway. This assumption is suggested by the discovery that mgs1∆ is synthetic lethal with mutations in RAD6 and RAD18 [22]. In addition, we have found that rad6 sgs1∆ cells have a synthetically sick phenotype (Fig. 4B). These results indicate that in the absence of either MGS1 or SGS1 cells rely on the RAD6 pathway for repair of spontaneously arising DNA lesions. 3.5. Specificity of interaction between MGS1 and SGS1 Sgs1 is a multifunctional protein, which was shown to interact genetically with proteins involved in various processes of DNA metabolism. It was suggested that Sgs1 and Srs2, another yeast DNA helicase that mediates DNA repair and recombination, perform overlapping functions and could partly substitute for each other. This assumption is supported by the fact that simultaneous deletion of SGS1 and SRS2 leads to slow grow or synthetic lethality [19,20,23], and the fact that SGS1 is a multicopy suppressor of MMS and HU sensitivities displayed by srs2∆ strains [24]. We have analyzed whether MGS1 affects srs2∆ strains, but MGS1 deletion or overexpression did not affect the growth rate of srs2∆ cells (data not shown). Also, sgs1∆ was found to exhibit synthetic slow growth when combined with top1 [25,26] and synthetic lethality with mms4 [27], but no synthetic growth defect was observed when mgs1∆ was combined with
679
top1 or mms4 mutations (data not shown), suggesting that MGS1 deletion affects sgs1∆ cells specifically. 3.6. Domains in Mgs1 and Sgs1 required to complement the slow grow and hyper-recombination phenotypes of mgs1∆ sgs1∆ cells SGS1 was first identified as a suppressor of the slow-growth phenotype of top3 mutants [5]. Interestingly, genes found to be synthetic lethal with sgs1∆ were also lethal in combination with top3 [27], and a top3 mgs1∆ double mutant was demonstrated to show extreme slow growth [4]. To further analyze the interaction between MGS1 and the Sgs1/Top3 complex, we assayed the phenotypes of mgs1∆ sgs1∆ cells carrying a helicase defective SGS1 (K706A) gene or missense mutations in Sgs1 (12-13G/S) that renders it unable to interact with Top3 [12]. We found that both the helicase activity of Sgs1 and a functional Sgs1-Top3 interaction were absolutely required to complement the hyper-recombination phenotype of mgs1∆ sgs1∆
Table 4 Effects of mutations in the helicase and Top3-binding activities of Sgs1, and in the NTP-binding motif of Mgs1 on complementation of the slow-growth phenotype of mgs1∆ sgs1∆ cells Relevant genotype
Plasmid transformed
Doubling time (min)
wild-type mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1
pRS314 pRS314 pRS314 pRS314 Ycp1305 K706A 12-13G/S Ycplac22 Ycplac22-WHIP Ycp22W-WalkerA
95.5 94.00 105.9 160.2 107.00 140.5 158.5 168.6 112.8 144.8
The strains used were yMP10381 derivatives. pRS314 is an ARS-CEN single copy vector containing the auxotrophic TRP1 marker, Ycp1305 is pRS314 carrying full length SGS1, K706A is Ycp1305 with mutations in SGS1 at K706 (K706A), and 12-13G/S is Ycp1305 with mutations in SGS1 at amino acids E12 and H13 (E12G/H13S) [12]. Ycplac22 is an ARS-CEN single copy vector containing the TRP1 auxotrophic marker, Ycplac22-WHIP is Ycplac22 carrying full length WHIP/MGS1, and Ycp22W-WalkerA is Ycplac22-WHIP carrying a mutation in the Walker A site of Mgs1 at K183 (K183A).
680
D. Branzei et al. / DNA Repair 1 (2002) 671–682
cells (Table 3). Furthermore, we found that deficiency in these activities in Sgs1 were unable to complement the synthetic slow-growth phenotype of mgs1∆ sgs1∆ cells (Table 4). We also analyzed whether a point mutation in the NTP-binding motif of Mgs1, K183A, which was shown to abolish the ATPase activity of Mgs1 [4] could complement the slow-growth phenotype of mgs1∆ sgs1∆ cells. We found that an ATPase defective Mgs1 protein did not retain the ability to complement the slow-growth phenotype of mgs1∆ sgs1∆ cells (Table 4).
4. Discussion In this study we analyzed the synthetic sick phenotype of mgs1∆ sgs1∆ cells. mgs1 sgs1 double mutants have a high percentage of mitotic arrest and display an accumulation of large budded cells with a single nucleus at the neck of the bud (Fig. 1B). The flow cytometric profile of these cells suggests that DNA replication is essentially complete (Fig. 1A), at least at the level of detection provided by flow cytometry. These two phenotypes indicate that while the bulk of DNA replication is completed, mitosis is blocked in mgs1 sgs1 cells. It was previously shown that DNA lesions that accumulate in different replication mutants lead to cell cycle arrest in G2 prior to mitosis. Although we could not detect DNA double strand breaks (DSBs) in mgs1 sgs1 cells by pulse field gel electrophoresis (data not shown), DNA damage is detected in these cells at the level of elevated mitotic recombination frequencies (Fig. 3 and [16]). Our data suggest that mgs1∆ sgs1∆ cells might accumulate their lesions during DNA replication. We have seen that mgs1∆ sgs1∆ cells display increased recombination frequencies, and previous studies have shown that elevated mitotic recombination frequencies are signals of replication problems [16]. Furthermore, mgs1∆ sgs1∆ cells show increased frequencies in SCR, which are thought to be mainly the result of homologous recombination mediated repair of DSBs that arise as a consequence of DNA replication, or the result of homologous recombination-directed reestablishment of inactivated replication forks (see [18] for a review). Therefore, we interpret our results as to suggest that in the absence of both Mgs1 and Sgs1, replication problems occur more frequently leading to
replication fork collapse and DSBs that are converted to sister chromatid exchanges and increased recombination frequencies. According to this model, RAD52 homologous recombination pathway functions in repair of the DSBs or DNA gaps that arise in mgs1 sgs1 cells during replication, and probably that is why deletion of RAD52 does not improve the growth of mgs1∆ sgs1∆ cells. However, if mgs1∆ sgs1∆ cells were entirely dependent on RAD52 function for repair of spontaneously arising DNA lesions, one would expect the mgs1 sgs1 rad52 combination to be a lethal one or at least significantly sicker than mgs1 sgs1 cells. We have seen that this is not the case (Fig. 4A), and we take this result to indicate that additional pathways are involved in repair of the lesions that accumulate in mgs1∆ sgs1∆ cells. One pathway that might be implicated in this process is that defined by RAD6/RAD18, since this pathway brings an important contribution to the repairing of the lesions that occur during replication, and it was found that the combination mgs1 rad6 is synthetic lethal [22], and sgs1 rad6 strains are synthetic sick (Fig. 4B). In regard to additional proteins that might be involved in the repair of DNA lesions arising during replication, it was found that DNA replication proteins are implicated in DSB repair and recombinationmediated restoration of replication forks [28]. We found that triple mutants between mgs1 sgs1 and polymerase ␦ mutants grow slower than mgs1 sgs1 double mutants (D. Branzei et. al, unpublished results), suggesting that the replication machinery itself might be involved in processing of DSBs that arise in mgs1 sgs1 cells. Alternatively, this result may suggest that Mgs1 and Sgs1 are implicated in the processing of DNA lesions that arise during replication, in cells having a defective polymerase ␦ or cells challenged with DNA damage or other type of stress that would lead to fork collapse or replication arrest. Interestingly, a mammalian homologue of Sgs1, WRN, was found to interact with DNA polymerase ␦ [29], and to stimulate polymerase ␦-mediated replication past secondary or aberrant DNA structures [30,31]. Furthermore, we identified the mammalian homologue of Mgs1, WHIP, as a protein that interacts with WRN [3]. Although we could not detect any physical interaction between Mgs1 and Sgs1 by standard methods, by the yeast two-hybrid method we found that both
D. Branzei et al. / DNA Repair 1 (2002) 671–682
Sgs1 and Mgs1 interact with Pol3, and Pol31 (D. Branzei et. al, unpublished results), which are the catalytic and the second subunits of the DNA polymerase ␦. These results suggest that Mgs1 and Sgs1 may be components of a replication complex that contains DNA polymerase ␦, and may be involved in preventing DSBs and inappropriate recombination from occurring during DNA replication. We found that the ATPase activity of Mgs1 is necessary to complement the slow-growth phenotype of mgs1∆ sgs1∆ cells (Table 4). In regard to Sgs1, it was previously shown that the functions of Sgs1 could be explained by two different activities of this protein: the helicase activity, and the Top3-binding “activity”, which presumably lies in the N-terminal domain of Sgs1, and seems to be necessary to target topoisomerases to protein complexes at stalled replication forks or other aberrant DNA structures in preparation for DNA repair [12,26,32]. We found that both the helicase activity and the Top3-binding activity of Sgs1 are needed to complement the slow growth (Table 4) and hyper-recombination phenotype (Table 3) of mgs1∆ sgs1∆ cells. The fact that MGS1mutation interacts with both Sgs1’s activities indicates that MGS1 gene functions in a parallel pathway to that defined by SGS1/TOP3. If Mgs1 were to perform a function related to that of Sgs1, or Top3, overexpression of MGS1 should be sufficient to suppress defects seen in the absence of Sgs1 or Top3. However, Mgs1 overproduction cannot suppress the growth defects or other phenotypes of these strains (D. Branzei et. al, unpublished results), indicating that Mgs1 is incapable of performing the function of Sgs1/Top3. These results suggest that the synthetic slow-growth phenotype of mgs1∆ sgs1∆ cells is primarily caused by the abolishment of different rather than redundant functions of these proteins. In conclusion, the findings presented by this study argue that WHIP/MGS1 is a gene involved in processes of DNA metabolism, which interacts genetically with SGS1 to promote genomic stability. The slow-growth phenotype displayed by mgs1∆ sgs1∆ mutants is due to mitotic arrest, which is most likely the result of DNA lesions accumulated during DNA replication. Further analyses of SGS1 and SGS1 interacting genes should continue to illuminate the role of RecQ helicases in vivo, and help elucidate the mechanism of ageing in humans.
681
Acknowledgements We thank Dr. L.H. Hartwell, and Dr. K. Johzuka for yMP1081 and KJY374 wild-type yeast strains. We thank Dr. M. Saijo for the W303 rad6 strains. We thank Dr. P Philippsen and Dr. L.H. Hartwell for the plasmids pFA6a-KanMX4, and pWL8, respectively. This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, by Takeda Science Foundation of Japan, and by Health Sciences Research Grant from the Ministry of Health, Labor, and Welfare of Japan. References [1] C.E. Yu, J. Oshima, Y.H. Fu, E.M. Wijsman, F. Hisama, R. Alisch, S. Matthews, J. Nakura, T. Miki, S. Ouais, G.M. Martin, J. Mulligan, G.D. Schellenberg, Positional cloning of the Werner’s syndrome gene, Science 272 (1996) 258–262. [2] J.C. Shen, L.A. Loeb, The Werner syndrome gene: the molecular basis of RecQ helicase-deficiency diseases, Trends Genet. 16 (2000) 213–220. [3] Y. Kawabe, D. Branzei, T. Hayashi, H. Suzuki, T. Masuko, F. Onoda, S.J. Heo, H. Ikeda, A. Shimamoto, Y. Furuichi, M. Seki, T. Enomoto, A novel protein interacts with the Werner’s syndrome gene product physically and functionally, J. Biol. Chem. 276 (2001) 20364–20369. [4] T. Hishida, H. Iwasaki, T. Ohno, T. Morishita, H. Shinagawa, A yeast gene, MGS1, encoding a DNA-dependent AAA(+) ATPase is required to maintain genome stability, Proc. Natl. Acad. Sci. U.S.A. 98 (2001) 8283–8289. [5] S. Gangloff, J.P. McDonald, C. Bendixen, L. Arthur, R. Rothstein, The yeast type I topoisomerase Top3 interacts with Sgs1, a DNA helicase homolog: a potential eukaryotic reverse gyrase, Mol. Cell Biol. 14 (1994) 8391–8398. [6] P.M. Watt, I.D. Hickson, R.H. Borts, E.J. Louis, SGS1, a homologue of the Bloom’s and Werner’s syndrome genes, is required for maintenance of genome stability in Saccharomyces cerevisiae, Genetics 144 (1996) 935–945. [7] D.A. Sinclair, K. Mills, L. Guarente, Accelerated ageing and nucleolar fragmentation in yeast sgs1 mutants, Science 277 (1997) 1313–1316. [8] S. Gangloff, B. de Massy, L. Arthur, R. Rothstein, F. Fabre, The essential role of yeast topoisomerase III in meiosis depends on recombination, EMBO J. 18 (1999) 1701–1711. [9] J.W. Wallis, G. Chrebet, G. Brodsky, M. Rolfe, R. Rothstein, A hyper-recombinant mutation in S. cerevisiae identifies a novel eukaryotic topoisomerase, Cell 58 (1989) 409–419. [10] A. Wach, A. Brachat, R. Pohlmann, P. Philippsen, New heterologous modules for classical or PCR-based gene disruptions in Saccharomyces cerevisiae, Yeast 10 (1994) 1793–1808.
682
D. Branzei et al. / DNA Repair 1 (2002) 671–682
[11] F. Onoda, M. Seki, A. Miyajima, T. Enomoto, Involvement of SGS1 in DNA damage-induced heteroallelic recombination that requires RAD52 in Saccharomyces cerevisiae, Mol. Gen. Genet. 264 (2001) 702–708. [12] A. Ui, Y. Satoh, F. Onoda, A. Miyajima, M. Seki, T. Enomoto, The N-terminal region of Sgs1, which interacts with Top3, is required for complementation of MMS sensitivity and suppression of hyper-recombination in sgs1 disruptants, Mol. Genet. Genom. 265 (2001) 837–850. [13] A.G. Paulovich, C.D. Armour, L.H. Hartwell, The Saccharomyces cerevisiae RAD9, RAD17, RAD24 and MEC3 genes are required for tolerating irreparable, ultraviolet-induced DNA damage, Genetics 150 (1998) 75–93. [14] D.E. Lea, E.C.A. Coulson, The distribution of the numbers of mutants in bacterial populations, J. Genet. 49 (1949) 264–285. [15] T.A. Weinert, G.L. Kiser, L.H. Hartwell, Mitotic checkpoint genes in budding yeast and the dependence of mitosis on DNA replication and repair, Genes Dev. 8 (1994) 652–665. [16] L.H. Hartwell, D. Smith, Altered fidelity of mitotic chromosome transmission in cell mutants of S. cerevisiae, Genetics 110 (1985) 381–395. [17] B.J. Merrill, C. Holm, A requirement for recombinational repair in Saccharomyces cerevisiae is caused by DNA replication defects of mec1 mutants, Genetics 153 (1999) 595–605. [18] K.J. Marians, Replication and recombination intersect, Curr. Opin. Genet. Dev. 10 (2000) 151–156. [19] S. Gangloff, C. Soustelle, F. Fabre, Homologous recombination is responsible for cell death in the absence of the Sgs1 and Srs2 helicases, Nat. Genet. 25 (2000) 192–194. [20] M. McVey, M. Kaeberlein, H.A. Tissenbaum, L. Guarente, The short life span of Saccharomyces cerevisiae sgs1 and srs2 mutants is a composite of normal ageing processes and mitotic arrest due to defective recombination, Genetics 157 (2001) 1531–1542. [21] H.L. Klein, K.N. Kreuzer, Replication, recombination, and repair: going for the gold, Mol. Cell 9 (2002) 471–480.
[22] T. Hishida, T. Ohno, H. Iwasaki, H. Shinagawa, Saccharomyces cerevisiae MGS1 is essential in strains deficient in the RAD6-dependent DNA damage tolerance pathway, EMBO J. 21 (2002) 1–11. [23] S.K. Lee, R.E. Johnson, S.L. Yu, L. Prakash, S. Prakash, Requirement of yeast SGS1 and SRS2 genes for replication and transcription, Science 286 (1999) 2339–2342. [24] H.W. Mankouri, T.J. Craig, A. Morgan, SGS1 is a multicopy suppressor of srs2: functional overlap between DNA helicases, Nucleic Acids Res. 30 (2002) 1103–1113. [25] J. Lu, J.R. Mullen, S.J. Brill, S. Kleff, A.M. Romeo, R. Sternglanz, Human homologues of yeast helicase, Nature 383 (1996) 678–679. [26] J.R. Mullen, V. Kaliraman, S.J. Brill, Bipartite structure of the SGS1 DNA helicase in Saccharomyces cerevisiae, Genetics 154 (2000) 1101–1114. [27] J.R. Mullen, V. Kaliraman, S.S. Ibrahim, S.J. Brill, Requirement for three novel protein complexes in the absence of the Sgs1 DNA helicase in Saccharomyces cerevisiae, Genetics 157 (2001) 103–118. [28] A.M. Holmes, J.E. Haber, Double-strand break repair in yeast requires both leading and lagging strand DNA polymerases, Cell 96 (1999) 415–424. [29] A.M. Szekely, Y.H. Chen, C. Zhang, J. Oshima, S.M. Weissman, Werner protein recruits DNA polymerase ␦ to the nucleolus, Proc. Natl. Acad. Sci. U.S.A. 97 (2000) 11365–11370. [30] A.S. Kamath-Loeb, E. Johansson, P.M. Burgers, L.A. Loeb, Functional interaction between the Werner syndrome protein and DNA polymerase ␦, Proc. Natl. Acad. Sci. U.S.A. 97 (2000) 4603–4608. [31] A.S. Kamath-Loeb, L.A. Loeb, E. Johansson, P.M. Burgers, M. Fry, Interactions between the Werner syndrome helicase and DNA polymerase ␦ specifically facilitate copying of tetraplex and hairpin structures of the d(CGG)n trinucleotide repeat sequence, J. Biol. Chem. 276 (2001) 16439–16446. [32] R.K. Chakraverty, I.D. Hickson, Defending genome integrity during DNA replication: a proposed role for RecQ family helicases, Bioessays 21 (1999) 286–294.
Characterization of the slow-growth phenotype of S. cerevisiae whip/mgs1 sgs1 double deletion mutants Dana Branzei a,∗ , Masayuki Seki a , Fumitoshi Onoda a , Hideki Yagi a,b , Yoh-ichi Kawabe a , Takemi Enomoto a a
b
Molecular Cell Biology Laboratory, Graduate School of Pharmaceutical Sciences, Tohoku University, Aoba-ku, Sendai, Miyagi 980-8578, Japan Cell Biology Laboratory, School of Pharmaceutical Sciences, Kinki University, 3-4-1 Kowakae, Higashiosaka, Osaka 577-8502, Japan Received 5 February 2002; received in revised form 26 April 2002; accepted 8 May 2002
Abstract RecQ DNA helicases from many organisms have been indicated to function in the maintenance of genomic stability. In human cells, mutation in the WRN helicase, a RecQ-like DNA helicase, results in the Werner syndrome (WS), a genetic disorder characterized by genomic instability and premature ageing. Similarly, mutation in SGS1, the RECQ homologue in budding yeast, results in genomic instability and accelerated ageing. We previously demonstrated that mouse WRN interacts physically with a novel, highly conserved protein that we named WHIP, and that in budding yeast cells, simultaneous deletion of WHIP/MGS1 and SGS1 results in slow growth and shortened life span. Here we show by using genetic analysis in Saccharomyces cerevisiae that mgs1∆ sgs1∆ cells have increased rates of terminal G2/M arrest, and show elevated rates of spontaneous sister chromatid recombination (SCR) and rDNA array recombination. Finally, we report that complementation of the synthetic relationship between SGS1 and WHIP/MGS1 requires both the helicase and Top3-binding activities of Sgs1, as well as the ATPase activity of Mgs1. Our results suggest that Whip/Mgs1 is implicated in DNA metabolism, and is required for normal growth and cell cycle progression in the absence of Sgs1. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Whip; Mgs1; Sgs1; Genomic instability; Cell cycle arrest
1. Introduction Werner syndrome (WS) is a human autosomal recessive disorder characterized by the premature onset of a number of age related diseases. The gene defective in WS, WRN, encodes a RecQ-like DNA helicase [1]. Defects in other human RecQ-like DNA helicases, BLM and RECQL4 result in genetic disorders known as Bloom syndrome and Rothmund–Thomson syndrome, both characterized by increased genomic ∗ Corresponding author. Tel.: +81-22-217-6874; fax: +81-22-217-6873. E-mail address: [email protected] (D. Branzei).
instability and a predisposition for development of malignancies. At the cellular level, Bloom syndrome cells show increased frequencies of sister chromatid exchange and slowed DNA replication; Rothmund– Thomson syndrome cells show increased frequencies of chromosomal breaks and rearrangements; and WS cells display variegated translocation mosaicism and a decreased replicative lifespan in vitro (reviewed in [2]). Despite exciting advances in the biochemistry of WRN, its cellular function and the pathology of WS are yet to be fully understood. Of particular interest is to understand how WRN contributes to maintaining genomic stability and how, when mutated, leads to a
1568-7864/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 1 5 6 8 - 7 8 6 4 ( 0 2 ) 0 0 0 7 3 - 3
672
D. Branzei et al. / DNA Repair 1 (2002) 671–682
premature ageing phenotype. We hoped that by identifying WRN-interacting proteins we could delineate the principal pathways in which WRN participates, and the process that finally leads to the clinical manifestations of WS. We have screened for proteins that interact physically with mWRN and one of the proteins isolated in this way was a highly conserved, RFC-like protein that we named WHIP [3]. To investigate whether a functional relationship existed between WRN and WHIP, we turned to the yeast Saccharomyces cerevisiae where we constructed the corresponding double mutants and found that whip∆ sgs1∆ cells have a shortened lifespan and display a slow-growth phenotype [3]. The budding yeast homologue of WHIP was independently isolated and referred to as MGS1 [4]. The S. cerevisiae genome contains only one RecQ helicase gene, SGS1. Like its human homologues, SGS1 is required for maintaining chromosomal stability, and sgs1∆ mutants display hyper-recombination, and increased rates of chromosome loss and missegregation [5,6]. It has been demonstrated previously that, in a striking parallel with WS in humans, sgs1∆ mutants have a mean life span that is about 40% of that of wild-type strains [7]. Moreover, mutations in SGS1 result in hyper-sensitivity to various DNA damaging agents suggesting that Sgs1 functions in DNA repair. Further insight in the role of SGS1 comes from its interactions with topoisomerases. SGS1 was first identified as a suppressor of the slow-growth phenotype of top3 mutants [5], and sgs1 mutants show an epistatic relationship with top3 mutants, suppressing the elevated recombination rate and the sporulation block displayed by top3 mutants [8,9]. To help define the primary role of Mgs1, and to understand the causes of the slow-growth phenotype that mgs1∆ sgs1∆ cells display, we analyzed in detail the synthetic slow-growth phenotype of mgs1∆ sgs1∆ cells in regard to genomic instability and cell cycle defects.
2. Materials and methods 2.1. Strains The yeast strains used in this study are listed in Table 1. Most strains were constructed using standard
genetic methods. Complete oligonucleotide sequences of the primers used for disruption or disruption check will be provided upon request. To obtain a “START to STOP” MGS1deletion strain, the KANMX4 gene was amplified from pFA6a-KanMX4, a kind gift from Wach et al. [10], using MGS1-specific primers with 40–50 nt identity to MGS1 at their 5 ends. The resulting PCR fragment was transformed into the wild-type strain MR966, and G418-resistant colonies were selected. Correct replacement was verified by PCR amplification of the two replacement junction regions. To obtain additional MGS1 deletion mutants, genomic DNA from MR966w was amplified with two primers located about 600 nt upstream and downstream of the KANMX4 integration sites, and the strains were transformed with the PCR products, to give geneticin resistant cells. Proper integration and loss of the MGS1 gene were confirmed by PCR. SGS1 was disrupted as described previously [11]. The rad17::LEU2 strains were constructed by transforming yeast cells with BamH1-cut pWL8 plasmid, which was a gift from Dr. L.H. Hartwell. To obtain the rad52::LEU2 deletion strain 364r52, the LEU2 gene was amplified from pRS305, using RAD52 specific primers with 40-50 nt identity to RAD52 at their 5 ends, and to obtain additional rad52::LEU2 disruptants, genomic DNA from 364r52 was amplified with two primers located about 900 nt upstream and downstream of the LEU2 integration sites, and the strains were transformed with the PCR products, and Leu+ cells were selected. To obtain rad6::HIS3 disruptants, the rad6::HIS3 cassette was amplified by PCR using genomic DNA from W303␣ rad6 strains, provided generously by Dr. M. Saijo. The primers used for disruption were positioned about 400 nt upstream and downstream of the HIS3 integration sites. In all cases proper integration of the disruption cassettes was confirmed by PCR. 2.2. Plasmids The plasmids carrying full length SGS1 (Ycp1305) or mutations in SGS1 (K706A, 12-13G/S) were described in [12]. These plasmids were constructed by subcloning SGS1 into pRS314, a single copy vector containing the TRP1 marker, and then introducing mutations in SGS1 by using mutagenic DNA primers and the QuickChange Site-Directed Mutagenesis kit (Stratagene) [12]. For construction of the single
D. Branzei et al. / DNA Repair 1 (2002) 671–682
673
674
D. Branzei et al. / DNA Repair 1 (2002) 671–682
copy plasmid containing WHIP/MGS1, the ORF of WHIP/MGS1 was amplified from the yeast genome, and a HindIII–BamH1 fragment was subcloned into Ycplac22, a vector containing the TRP1 auxotrophic marker, to give Ycplac22-WHIP. The plasmid was checked with restriction enzymes and by PCR, and the in-frame insertion of WHIP/MGS1 was confirmed by DNA sequencing. The mutation in the Walker A site in Whip/Mgs1, K183A, was introduced by using mutagenic DNA primers and QuickChange Site-Directed Mutagenesis kit (Stratagene), and the mutation was confirmed by DNA sequencing.
2.5. Immunofluorescence
2.3. Growth curves, doubling-time and plating efficiency
2.6. Sister chromatid recombination
Overnight cultures were diluted to an OD600 nm of 0.1 and grown in YPAD for 12 h. Samples were taken every 1 h and their OD600 nm measured, and doublingtime calculated. During the time course, cell samples were taken, sonicated and counted. Appropriate dilutions were plated on several YPAD plates, and incubated at 30 ◦ C for 3 days. The number of colonies formed was counted and the plating efficiency determined. 2.4. Cell cycle analysis Cells were grown to early log phase when the cultures were harvested and resuspended in fresh YPAD medium to a concentration of 107 cells/ml. Cells were treated with ␣-factor at a final concentration of 5 g/ml or with nocodazole at a final concentration of 10 g/ml for 2 h. After confirmation of G1 or G2 arrest under the microscope, the ␣-factor or the nocodazole was washed away and cells were resuspended in fresh YPAD medium. Samples were taken at the indicated time points and fixed overnight at 4 ◦ C in 1 ml of 70% ethanol. The fixed cells were washed, sonicated, and resuspended in 400 l of 50 mM sodium citrate (pH 7.0), and then treated with ribonuclease A (0.25 mg/ml) for 1 h at 50 ◦ C, followed by treatment with proteinase K (0.5 mg/ml) for 1 h at 50 ◦ C. Propidium iodide was then added (16 g/ml) and cells were incubated for at least 12 h at 4 ◦ C. The DNA content of cells was analyzed by a BectonDickinson FACScan flow cytometer using the CellQuest system.
Cells were grown to midlog phase and fixed for 1 h in 3.7% formaldehyde. The cells were treated with 0.3 mg/ml zymolyase-100T for 30 min at 30 ◦ C. DNA was visualized by staining with 4 ,6-diamidino-2-phenylindole (DAPI), and microtubules were visualized using the anti-TUB (Y0l1/34) antibody (Serotec) diluted 1:500 in PBS/1% BSA/0.1% Triton X-100 and a 1:100 diluted Cy3-conjugated anti-rat secondary antibody (Jackson, ImmunoResearch Laboratories, INC). Digital images were obtained using a Zeiss system.
Cells in mid-log phase were washed and plated at appropriate dilutions on YPAD (or SC-Trp when the strains were carrying plasmids) and SC/-His (or SC/-His, -Trp when the strains were carrying plasmids) plates to determine sister chromatid recombination (SCR) frequencies [13]. The experiments were done three times with eight cultures for each mutant strain when strains were not carrying plasmids, and twice with seven cultures when strains were carrying plasmids. The recombination frequencies were calculated according to the median method described by Lea and Coulson [14]. 2.7. rDNA recombination Cells in stationary phase were washed and plated at appropriate dilutions on SC-Ura (or SC-Ura, -Trp when the strains were carrying plasmids) and SC + 5-FOA (or SC-Trp + 5-FOA when the strains were carrying plasmids) plates for detection of rDNA URA3 marker loss. The experiments were done three times with six cultures for each strain when strains were not carrying plasmids, and twice with seven and six cultures, respectively, when strains were carrying plasmids. The recombination frequencies were calculated according to the median method described by Lea and Coulson [14]. 2.8. Drop assay Log phase grown cells were harvested, washed once in distilled water, counted and diluted. Ten-fold serial dilutions of cells (105 , 104 , 103 , and 102 cells)
D. Branzei et al. / DNA Repair 1 (2002) 671–682
were spotted onto YPAD. The plates were incubated at the indicated temperatures for 2 days, and the cells were photographed.
675
3. Results
bud neck, the spindle was short and located at the bud neck, characteristic of cells in metaphase, prior to or at the beginning of chromosome segregation (Fig. 1B). Thus, the phenotype of mgs1∆ sgs1∆ cells is similar to that observed in cdc mutants with defects in DNA replication or mitosis.
3.1. Slow-growth phenotype displayed by mgs1∆ sgs1∆ cells
3.2. Characterization of the mitotic arrest of mgs1∆ sgs1∆ cells
We found that mWRN interacts in vivo and in vitro with a novel protein, mWHIP, and that simultaneous deletion of SGS1 and WHIP/MGS1 in S. cerevisiae results in a slow-growth phenotype and a dramatic decrease in the life span of yeast cells [3]. We wished to further characterize the slow-growth phenotype of mgs1∆ sgs1∆ mutants and determine its causes. To accomplish this, we constructed mgs1∆ sgs1∆ double mutants in different backgrounds and observed that these mutants grew much slower than wild-type or single mutant strains (Table 2). Furthermore, the plating efficiency of mgs1∆ sgs1∆ cells was reduced to about 40% of that of wild-type cells, while that of sgs1∆ cells was found to be about 70% of that of wild-type cells (Table 2). We observed that mgs1∆ sgs1∆ double mutants show an increased proportion of dumbbell shaped cells, and consistent with our observation, the flow cytometry profiles of log phase mgs1∆ sgs1∆ cells displayed a major peak of G2 content (Fig. 1A). To determine the stage of mitosis at which the cells arrested, we stained wild-type and mutant cells with DAPI and anti-tubulin antibodies to visualize the mitotic spindle. Approximately 18% of mgs1∆ sgs1∆ cells were large budded, with a single nucleus at the neck of the mother cell. This phenomenon was also observed to a smaller extent in sgs1∆ cells (6%), but was almost absent in wild-type and mgs1∆ cells (Fig. 1B). In the cells that had the nucleus at the
The G2 peak accumulation observed in mgs1∆ sgs1∆ cells could be due to DNA damage or failure to complete DNA synthesis. We monitored cell cycle progression by synchronizing cells in G1 or G2 by exposure to ␣-factor or nocodazole, respectively. We found that mgs1∆ sgs1∆ mutants were able to complete S-phase, but showed great difficulties in progressing through mitosis (data not shown). If mgs1∆ sgs1∆ cells undergo cell cycle arrest because of DNA lesions, the G2/M arrest may be mediated by a functional DNA damage checkpoint. Six genes, RAD9, RAD17, RAD24, MEC1, MEC2, and MEC3 have been shown to be required for DNA damage arrest of cdc mutants in G2/M [15]. To determine the role of the RAD17 checkpoint in cell cycle arrest in mgs1∆ sgs1∆ mutants, we constructed mgs1∆ sgs1∆ rad17∆ triple mutants. Flow cytometric analysis of exponentially growing cells showed that deletion of RAD17 in mgs1∆ sgs1∆ strains reduced the level of arrest at G2/M (Fig. 2). We also observed that the aberrant morphology distribution of mgs1∆ sgs1∆ cells was partially suppressed by the rad17∆ mutation, but synchronization experiments with ␣-factor showed that the G2 accumulation of mgs1∆ sgs1∆ cells is only little suppressed by RAD17 deletion (data not shown). Viability analysis of mgs1 sgs1 rad17 cells showed that their viability is similar to that of mgs1 sgs1 cells, and not lower as it might have been expected. Perhaps most of the DNA damage that accumulates in mgs1∆ sgs1∆ cells is unrepairable and leads to cell death. That also explains the short life span of mgs1∆ sgs1∆ cells, which is only of about 4–5 generations [3].
Table 2 Effects of mgs1∆ and sgs1∆ mutations on doubling time and plating efficiencies Genotype
Doubling time (min)
Wild type mgs1 sgs1 mgs1 sgs1
94.6 94.8 105.8 170.4
± ± ± ±
2.82 4.25 3.92 6.61
The strains used were MR966 derivatives.
Plate efficiency (%) 97.6 87.9 69.7 39.2
± ± ± ±
4.2 2.9 3.9 4.46
3.3. Increased mitotic recombination in mgs1∆ sgs1∆ cells We have seen that mgs1∆ sgs1∆ cells display a cell cycle and morphology phenotype similar to that
676
D. Branzei et al. / DNA Repair 1 (2002) 671–682
Fig. 1. Cell cycle and morphological analysis of mgs1∆ sgs1∆ cells. (A) Cell cycle analysis of midlog-phase cells. (B) Analysis of nuclear and tubulin spindle morphology. Indirect immunofluorescence was performed on midlog-phase wild-type, mgs1∆, sgs1∆ and mgs1∆ sgs1∆ cells. At least 200 cells from each strain were scored. Percentages of cells that displayed nucleus at the bud neck with a short mitotic spindle were as follows: wild-type, 0.1%; mgs1∆, 1%; sgs1∆, 6%; mgs1∆ sgs1∆, 18%. The strains used in A and B were MR966 and KJY374 derivatives.
observed in replication and segregation mutants (Fig. 1). It has been reported that mutations in genes involved in DNA replication lead to increased mitotic recombination frequencies, a feature that distinguishes them from mitotic defects that show only elevated chromosome loss and missegregation frequencies [16]. We assayed the frequency of SCR, which is thought to be mainly the outcome of recombinational repair of DSBs that arise mainly in the S-phase as a result of DNA replication [13,17,18]. We found that SCR frequencies were increased in mgs1∆ sgs1∆
mutants as compared to the single mutants alone (Fig. 3A). Previous reports have shown that Sgs1 might play an important role in the rDNA and at telomeres, to suppress homologous recombination. sgs1∆ cells display an about seven-fold increase in rDNA marker loss (Fig. 3B, Table 3, and [5]), and mgs1∆ cells have a 3.5-fold increase in rDNA recombination frequencies (Fig. 3B, Table 3, and [4]). We found that mgs1∆ sgs1∆ cells display a 20-fold increase in rDNA array recombination frequencies (Fig. 3B and Table 3).
D. Branzei et al. / DNA Repair 1 (2002) 671–682
677
Fig. 2. Flow cytometric analysis of the mitotic arrest that occurs in mgs1∆ sgs1∆ cells. Flow cytometric analysis of midlog-phase cells. The strains used were MR966 derivatives.
Fig. 3. Effect of mgs1 and sgs1 mutations on spontaneous mitotic recombination. (A) Histograms of the sister chromatid recombination (SCR) frequencies. The value indicates the number of recombinants per 106 viable cells. The strains used were yMP10381 derivatives. (B) Histograms of the rDNA recombination frequencies. The value indicates the number of recombinants per 104 viable cells. The strains used were KJY374 derivatives.
678
D. Branzei et al. / DNA Repair 1 (2002) 671–682
Table 3 Effects of mutations in the helicase and Top3-binding activities of Sgs1 on complementation of increased recombination frequencies displayed by mgs1 sgs1 cells Relevant genotype
Plasmid transformed
SCE frequency (SD-His,Trp/105 cells)a
wild-type mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1
pRS314 pRS314 pRS314 pRS314 Ycp1305 K706A 12-13G/S
4.02 4.55 10.21 24.22 4.94 15.04 16.4
± ± ± ± ± ± ±
0.65 0.32 1.65 2.9 0.9 1.42 2.1
rDNA recombination frequency (SC-Trp+5-FOA/103 cells)b 5.6 15.6 41.7 110.6 17.92 113.78 119.84
± ± ± ± ± ± ±
0.9 2.5 9.2 10.2 3.2 8.2 6.09
pRS314 is a single copy vector containing the TRP1 marker, Ycp1305 is pRS314 carrying full length SGS1, K706A is Ycp1305 with mutations in SGS1 at K706 (K706A), and 12-13G/S is Ycp1305 with mutations in SGS1 at amino acids E12 and H13 (E12G/H13S) [12]. a The strains used were yMP10381 derivatives. b The strains used were KJY374 derivatives.
3.4. RAD52 deletion does not suppress the slow-growth phenotype of mgs1∆ sgs1∆ cells It was previously reported that knocking out genes encoding homologous recombination proteins such as Rad51 and Rad52 suppresses the slow-growth phenotype of sgs1∆ srs2∆ cells [19,20] and the
synthetic lethality of sgs1 mus81 cells (see [21]), suggesting that the lethal mitotic structures that accumulate in these cells are the product of recombination. Since mgs1∆ sgs1∆ cells display increased levels of mitotic recombination (Fig. 3, Table 3), we analyzed the possibility that the mitotic arrest of these cells is the result of promiscuous homologous recombination.
Fig. 4. Effect of RAD52 and RAD6 deletions on the growth ability of mgs1 sgs1 cells. Ten-fold serial dilutions of log phase grown cells (105 , 104 , 103 , and 102 cells) were spotted onto YPAD, and photographed after 2 days. (A) Growth of mgs1 sgs1 rad52 cells as compared to that of mgs1 sgs1 strains. (B) Growth of rad6 sgs1 cells as compared to that of single mutants. The strains used in A and B were in the A364 background.
D. Branzei et al. / DNA Repair 1 (2002) 671–682
We reasoned that if the lesions that form in mgs1∆ sgs1∆ cells are processed by the homologous recombination proteins into chromosomal intermediates that cannot be separated prior to mitosis and lead to mitotic arrest, then deletion of RAD52 should suppress the slow growth and mitotic arrest of these cells. We constructed mgs1∆ sgs1∆ rad52 cells, but neither the growth was improved nor the mitotic arrest reduced in these cells as compared to mgs1∆ sgs1∆ cells (Fig. 4A). In fact, the growth of mgs1∆ sgs1∆ rad52 cells was slightly slower than that of mgs1∆ sgs1∆ cells, suggesting that mgs1∆ sgs1∆ cells use the RAD52 homologous recombination pathway in repairing DSBs and DNA lesions. However, since the mgs1∆ sgs1∆ rad52 combination was not synthetic lethal, it is conceivable that additional repair pathways can be recruited for the repair of DNA lesions that accumulate in mgs1∆ sgs1∆ cells. One such pathway could be the RAD6/RAD18 post-replication repair pathway. This assumption is suggested by the discovery that mgs1∆ is synthetic lethal with mutations in RAD6 and RAD18 [22]. In addition, we have found that rad6 sgs1∆ cells have a synthetically sick phenotype (Fig. 4B). These results indicate that in the absence of either MGS1 or SGS1 cells rely on the RAD6 pathway for repair of spontaneously arising DNA lesions. 3.5. Specificity of interaction between MGS1 and SGS1 Sgs1 is a multifunctional protein, which was shown to interact genetically with proteins involved in various processes of DNA metabolism. It was suggested that Sgs1 and Srs2, another yeast DNA helicase that mediates DNA repair and recombination, perform overlapping functions and could partly substitute for each other. This assumption is supported by the fact that simultaneous deletion of SGS1 and SRS2 leads to slow grow or synthetic lethality [19,20,23], and the fact that SGS1 is a multicopy suppressor of MMS and HU sensitivities displayed by srs2∆ strains [24]. We have analyzed whether MGS1 affects srs2∆ strains, but MGS1 deletion or overexpression did not affect the growth rate of srs2∆ cells (data not shown). Also, sgs1∆ was found to exhibit synthetic slow growth when combined with top1 [25,26] and synthetic lethality with mms4 [27], but no synthetic growth defect was observed when mgs1∆ was combined with
679
top1 or mms4 mutations (data not shown), suggesting that MGS1 deletion affects sgs1∆ cells specifically. 3.6. Domains in Mgs1 and Sgs1 required to complement the slow grow and hyper-recombination phenotypes of mgs1∆ sgs1∆ cells SGS1 was first identified as a suppressor of the slow-growth phenotype of top3 mutants [5]. Interestingly, genes found to be synthetic lethal with sgs1∆ were also lethal in combination with top3 [27], and a top3 mgs1∆ double mutant was demonstrated to show extreme slow growth [4]. To further analyze the interaction between MGS1 and the Sgs1/Top3 complex, we assayed the phenotypes of mgs1∆ sgs1∆ cells carrying a helicase defective SGS1 (K706A) gene or missense mutations in Sgs1 (12-13G/S) that renders it unable to interact with Top3 [12]. We found that both the helicase activity of Sgs1 and a functional Sgs1-Top3 interaction were absolutely required to complement the hyper-recombination phenotype of mgs1∆ sgs1∆
Table 4 Effects of mutations in the helicase and Top3-binding activities of Sgs1, and in the NTP-binding motif of Mgs1 on complementation of the slow-growth phenotype of mgs1∆ sgs1∆ cells Relevant genotype
Plasmid transformed
Doubling time (min)
wild-type mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1 mgs1 sgs1
pRS314 pRS314 pRS314 pRS314 Ycp1305 K706A 12-13G/S Ycplac22 Ycplac22-WHIP Ycp22W-WalkerA
95.5 94.00 105.9 160.2 107.00 140.5 158.5 168.6 112.8 144.8
The strains used were yMP10381 derivatives. pRS314 is an ARS-CEN single copy vector containing the auxotrophic TRP1 marker, Ycp1305 is pRS314 carrying full length SGS1, K706A is Ycp1305 with mutations in SGS1 at K706 (K706A), and 12-13G/S is Ycp1305 with mutations in SGS1 at amino acids E12 and H13 (E12G/H13S) [12]. Ycplac22 is an ARS-CEN single copy vector containing the TRP1 auxotrophic marker, Ycplac22-WHIP is Ycplac22 carrying full length WHIP/MGS1, and Ycp22W-WalkerA is Ycplac22-WHIP carrying a mutation in the Walker A site of Mgs1 at K183 (K183A).
680
D. Branzei et al. / DNA Repair 1 (2002) 671–682
cells (Table 3). Furthermore, we found that deficiency in these activities in Sgs1 were unable to complement the synthetic slow-growth phenotype of mgs1∆ sgs1∆ cells (Table 4). We also analyzed whether a point mutation in the NTP-binding motif of Mgs1, K183A, which was shown to abolish the ATPase activity of Mgs1 [4] could complement the slow-growth phenotype of mgs1∆ sgs1∆ cells. We found that an ATPase defective Mgs1 protein did not retain the ability to complement the slow-growth phenotype of mgs1∆ sgs1∆ cells (Table 4).
4. Discussion In this study we analyzed the synthetic sick phenotype of mgs1∆ sgs1∆ cells. mgs1 sgs1 double mutants have a high percentage of mitotic arrest and display an accumulation of large budded cells with a single nucleus at the neck of the bud (Fig. 1B). The flow cytometric profile of these cells suggests that DNA replication is essentially complete (Fig. 1A), at least at the level of detection provided by flow cytometry. These two phenotypes indicate that while the bulk of DNA replication is completed, mitosis is blocked in mgs1 sgs1 cells. It was previously shown that DNA lesions that accumulate in different replication mutants lead to cell cycle arrest in G2 prior to mitosis. Although we could not detect DNA double strand breaks (DSBs) in mgs1 sgs1 cells by pulse field gel electrophoresis (data not shown), DNA damage is detected in these cells at the level of elevated mitotic recombination frequencies (Fig. 3 and [16]). Our data suggest that mgs1∆ sgs1∆ cells might accumulate their lesions during DNA replication. We have seen that mgs1∆ sgs1∆ cells display increased recombination frequencies, and previous studies have shown that elevated mitotic recombination frequencies are signals of replication problems [16]. Furthermore, mgs1∆ sgs1∆ cells show increased frequencies in SCR, which are thought to be mainly the result of homologous recombination mediated repair of DSBs that arise as a consequence of DNA replication, or the result of homologous recombination-directed reestablishment of inactivated replication forks (see [18] for a review). Therefore, we interpret our results as to suggest that in the absence of both Mgs1 and Sgs1, replication problems occur more frequently leading to
replication fork collapse and DSBs that are converted to sister chromatid exchanges and increased recombination frequencies. According to this model, RAD52 homologous recombination pathway functions in repair of the DSBs or DNA gaps that arise in mgs1 sgs1 cells during replication, and probably that is why deletion of RAD52 does not improve the growth of mgs1∆ sgs1∆ cells. However, if mgs1∆ sgs1∆ cells were entirely dependent on RAD52 function for repair of spontaneously arising DNA lesions, one would expect the mgs1 sgs1 rad52 combination to be a lethal one or at least significantly sicker than mgs1 sgs1 cells. We have seen that this is not the case (Fig. 4A), and we take this result to indicate that additional pathways are involved in repair of the lesions that accumulate in mgs1∆ sgs1∆ cells. One pathway that might be implicated in this process is that defined by RAD6/RAD18, since this pathway brings an important contribution to the repairing of the lesions that occur during replication, and it was found that the combination mgs1 rad6 is synthetic lethal [22], and sgs1 rad6 strains are synthetic sick (Fig. 4B). In regard to additional proteins that might be involved in the repair of DNA lesions arising during replication, it was found that DNA replication proteins are implicated in DSB repair and recombinationmediated restoration of replication forks [28]. We found that triple mutants between mgs1 sgs1 and polymerase ␦ mutants grow slower than mgs1 sgs1 double mutants (D. Branzei et. al, unpublished results), suggesting that the replication machinery itself might be involved in processing of DSBs that arise in mgs1 sgs1 cells. Alternatively, this result may suggest that Mgs1 and Sgs1 are implicated in the processing of DNA lesions that arise during replication, in cells having a defective polymerase ␦ or cells challenged with DNA damage or other type of stress that would lead to fork collapse or replication arrest. Interestingly, a mammalian homologue of Sgs1, WRN, was found to interact with DNA polymerase ␦ [29], and to stimulate polymerase ␦-mediated replication past secondary or aberrant DNA structures [30,31]. Furthermore, we identified the mammalian homologue of Mgs1, WHIP, as a protein that interacts with WRN [3]. Although we could not detect any physical interaction between Mgs1 and Sgs1 by standard methods, by the yeast two-hybrid method we found that both
D. Branzei et al. / DNA Repair 1 (2002) 671–682
Sgs1 and Mgs1 interact with Pol3, and Pol31 (D. Branzei et. al, unpublished results), which are the catalytic and the second subunits of the DNA polymerase ␦. These results suggest that Mgs1 and Sgs1 may be components of a replication complex that contains DNA polymerase ␦, and may be involved in preventing DSBs and inappropriate recombination from occurring during DNA replication. We found that the ATPase activity of Mgs1 is necessary to complement the slow-growth phenotype of mgs1∆ sgs1∆ cells (Table 4). In regard to Sgs1, it was previously shown that the functions of Sgs1 could be explained by two different activities of this protein: the helicase activity, and the Top3-binding “activity”, which presumably lies in the N-terminal domain of Sgs1, and seems to be necessary to target topoisomerases to protein complexes at stalled replication forks or other aberrant DNA structures in preparation for DNA repair [12,26,32]. We found that both the helicase activity and the Top3-binding activity of Sgs1 are needed to complement the slow growth (Table 4) and hyper-recombination phenotype (Table 3) of mgs1∆ sgs1∆ cells. The fact that MGS1mutation interacts with both Sgs1’s activities indicates that MGS1 gene functions in a parallel pathway to that defined by SGS1/TOP3. If Mgs1 were to perform a function related to that of Sgs1, or Top3, overexpression of MGS1 should be sufficient to suppress defects seen in the absence of Sgs1 or Top3. However, Mgs1 overproduction cannot suppress the growth defects or other phenotypes of these strains (D. Branzei et. al, unpublished results), indicating that Mgs1 is incapable of performing the function of Sgs1/Top3. These results suggest that the synthetic slow-growth phenotype of mgs1∆ sgs1∆ cells is primarily caused by the abolishment of different rather than redundant functions of these proteins. In conclusion, the findings presented by this study argue that WHIP/MGS1 is a gene involved in processes of DNA metabolism, which interacts genetically with SGS1 to promote genomic stability. The slow-growth phenotype displayed by mgs1∆ sgs1∆ mutants is due to mitotic arrest, which is most likely the result of DNA lesions accumulated during DNA replication. Further analyses of SGS1 and SGS1 interacting genes should continue to illuminate the role of RecQ helicases in vivo, and help elucidate the mechanism of ageing in humans.
681
Acknowledgements We thank Dr. L.H. Hartwell, and Dr. K. Johzuka for yMP1081 and KJY374 wild-type yeast strains. We thank Dr. M. Saijo for the W303 rad6 strains. We thank Dr. P Philippsen and Dr. L.H. Hartwell for the plasmids pFA6a-KanMX4, and pWL8, respectively. This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, by Takeda Science Foundation of Japan, and by Health Sciences Research Grant from the Ministry of Health, Labor, and Welfare of Japan. References [1] C.E. Yu, J. Oshima, Y.H. Fu, E.M. Wijsman, F. Hisama, R. Alisch, S. Matthews, J. Nakura, T. Miki, S. Ouais, G.M. Martin, J. Mulligan, G.D. Schellenberg, Positional cloning of the Werner’s syndrome gene, Science 272 (1996) 258–262. [2] J.C. Shen, L.A. Loeb, The Werner syndrome gene: the molecular basis of RecQ helicase-deficiency diseases, Trends Genet. 16 (2000) 213–220. [3] Y. Kawabe, D. Branzei, T. Hayashi, H. Suzuki, T. Masuko, F. Onoda, S.J. Heo, H. Ikeda, A. Shimamoto, Y. Furuichi, M. Seki, T. Enomoto, A novel protein interacts with the Werner’s syndrome gene product physically and functionally, J. Biol. Chem. 276 (2001) 20364–20369. [4] T. Hishida, H. Iwasaki, T. Ohno, T. Morishita, H. Shinagawa, A yeast gene, MGS1, encoding a DNA-dependent AAA(+) ATPase is required to maintain genome stability, Proc. Natl. Acad. Sci. U.S.A. 98 (2001) 8283–8289. [5] S. Gangloff, J.P. McDonald, C. Bendixen, L. Arthur, R. Rothstein, The yeast type I topoisomerase Top3 interacts with Sgs1, a DNA helicase homolog: a potential eukaryotic reverse gyrase, Mol. Cell Biol. 14 (1994) 8391–8398. [6] P.M. Watt, I.D. Hickson, R.H. Borts, E.J. Louis, SGS1, a homologue of the Bloom’s and Werner’s syndrome genes, is required for maintenance of genome stability in Saccharomyces cerevisiae, Genetics 144 (1996) 935–945. [7] D.A. Sinclair, K. Mills, L. Guarente, Accelerated ageing and nucleolar fragmentation in yeast sgs1 mutants, Science 277 (1997) 1313–1316. [8] S. Gangloff, B. de Massy, L. Arthur, R. Rothstein, F. Fabre, The essential role of yeast topoisomerase III in meiosis depends on recombination, EMBO J. 18 (1999) 1701–1711. [9] J.W. Wallis, G. Chrebet, G. Brodsky, M. Rolfe, R. Rothstein, A hyper-recombinant mutation in S. cerevisiae identifies a novel eukaryotic topoisomerase, Cell 58 (1989) 409–419. [10] A. Wach, A. Brachat, R. Pohlmann, P. Philippsen, New heterologous modules for classical or PCR-based gene disruptions in Saccharomyces cerevisiae, Yeast 10 (1994) 1793–1808.
682
D. Branzei et al. / DNA Repair 1 (2002) 671–682
[11] F. Onoda, M. Seki, A. Miyajima, T. Enomoto, Involvement of SGS1 in DNA damage-induced heteroallelic recombination that requires RAD52 in Saccharomyces cerevisiae, Mol. Gen. Genet. 264 (2001) 702–708. [12] A. Ui, Y. Satoh, F. Onoda, A. Miyajima, M. Seki, T. Enomoto, The N-terminal region of Sgs1, which interacts with Top3, is required for complementation of MMS sensitivity and suppression of hyper-recombination in sgs1 disruptants, Mol. Genet. Genom. 265 (2001) 837–850. [13] A.G. Paulovich, C.D. Armour, L.H. Hartwell, The Saccharomyces cerevisiae RAD9, RAD17, RAD24 and MEC3 genes are required for tolerating irreparable, ultraviolet-induced DNA damage, Genetics 150 (1998) 75–93. [14] D.E. Lea, E.C.A. Coulson, The distribution of the numbers of mutants in bacterial populations, J. Genet. 49 (1949) 264–285. [15] T.A. Weinert, G.L. Kiser, L.H. Hartwell, Mitotic checkpoint genes in budding yeast and the dependence of mitosis on DNA replication and repair, Genes Dev. 8 (1994) 652–665. [16] L.H. Hartwell, D. Smith, Altered fidelity of mitotic chromosome transmission in cell mutants of S. cerevisiae, Genetics 110 (1985) 381–395. [17] B.J. Merrill, C. Holm, A requirement for recombinational repair in Saccharomyces cerevisiae is caused by DNA replication defects of mec1 mutants, Genetics 153 (1999) 595–605. [18] K.J. Marians, Replication and recombination intersect, Curr. Opin. Genet. Dev. 10 (2000) 151–156. [19] S. Gangloff, C. Soustelle, F. Fabre, Homologous recombination is responsible for cell death in the absence of the Sgs1 and Srs2 helicases, Nat. Genet. 25 (2000) 192–194. [20] M. McVey, M. Kaeberlein, H.A. Tissenbaum, L. Guarente, The short life span of Saccharomyces cerevisiae sgs1 and srs2 mutants is a composite of normal ageing processes and mitotic arrest due to defective recombination, Genetics 157 (2001) 1531–1542. [21] H.L. Klein, K.N. Kreuzer, Replication, recombination, and repair: going for the gold, Mol. Cell 9 (2002) 471–480.
[22] T. Hishida, T. Ohno, H. Iwasaki, H. Shinagawa, Saccharomyces cerevisiae MGS1 is essential in strains deficient in the RAD6-dependent DNA damage tolerance pathway, EMBO J. 21 (2002) 1–11. [23] S.K. Lee, R.E. Johnson, S.L. Yu, L. Prakash, S. Prakash, Requirement of yeast SGS1 and SRS2 genes for replication and transcription, Science 286 (1999) 2339–2342. [24] H.W. Mankouri, T.J. Craig, A. Morgan, SGS1 is a multicopy suppressor of srs2: functional overlap between DNA helicases, Nucleic Acids Res. 30 (2002) 1103–1113. [25] J. Lu, J.R. Mullen, S.J. Brill, S. Kleff, A.M. Romeo, R. Sternglanz, Human homologues of yeast helicase, Nature 383 (1996) 678–679. [26] J.R. Mullen, V. Kaliraman, S.J. Brill, Bipartite structure of the SGS1 DNA helicase in Saccharomyces cerevisiae, Genetics 154 (2000) 1101–1114. [27] J.R. Mullen, V. Kaliraman, S.S. Ibrahim, S.J. Brill, Requirement for three novel protein complexes in the absence of the Sgs1 DNA helicase in Saccharomyces cerevisiae, Genetics 157 (2001) 103–118. [28] A.M. Holmes, J.E. Haber, Double-strand break repair in yeast requires both leading and lagging strand DNA polymerases, Cell 96 (1999) 415–424. [29] A.M. Szekely, Y.H. Chen, C. Zhang, J. Oshima, S.M. Weissman, Werner protein recruits DNA polymerase ␦ to the nucleolus, Proc. Natl. Acad. Sci. U.S.A. 97 (2000) 11365–11370. [30] A.S. Kamath-Loeb, E. Johansson, P.M. Burgers, L.A. Loeb, Functional interaction between the Werner syndrome protein and DNA polymerase ␦, Proc. Natl. Acad. Sci. U.S.A. 97 (2000) 4603–4608. [31] A.S. Kamath-Loeb, L.A. Loeb, E. Johansson, P.M. Burgers, M. Fry, Interactions between the Werner syndrome helicase and DNA polymerase ␦ specifically facilitate copying of tetraplex and hairpin structures of the d(CGG)n trinucleotide repeat sequence, J. Biol. Chem. 276 (2001) 16439–16446. [32] R.K. Chakraverty, I.D. Hickson, Defending genome integrity during DNA replication: a proposed role for RecQ family helicases, Bioessays 21 (1999) 286–294.