Interaction of checkpoint proteins Hus1/Rad1/Rad9 with DNA base excision repair enzyme MutY homolog in fission yeast, Schizosaccharomyces pombe.

The DNA glycosylase MutY homolog (MYH) is responsible for removing adenines misincorporated opposite DNA strands containing guanine or 7,8-dihydro-8-oxoguanine by base excision repair thereby preventing G:C to T:A mutations. MYH has been shown to interact with the proliferating cell nuclear antigen (PCNA) in both human and fission yeast Schizosaccharomyces pombe systems. Here we show that S. pombe (Sp) MYH physically interacts with all subunits of the PCNA-like checkpoint protein heterotrimer, SpRad9/SpRad1/SpHus1, in yeast extracts and when the individual subunits are expressed in bacteria. The SpHus1 and SpPCNA binding sites are located in discrete regions of SpMYH. Immunoprecipitation assays reveal that the interaction between SpHus1 and SpMYH increases dramatically after hydrogen peroxide treatment, and this increase in the SpHus1-SpMYH interaction correlates with the presence of SpHus1 phosphorylation. In contrast, the interaction between SpPCNA and SpMYH after hydrogen peroxide treatment remains nearly unchanged. SpMYH associates with SpHus1 in a complex of approximately 450 kDa, the reported native molecular mass of the SpRad9/SpRad1/SpHus1-MYC complex. A larger portion of SpMYH shifts to the 150-500-kDa regions after hydrogen peroxide treatment in comparison with untreated extracts. SpHus1 phosphorylation is substantially reduced in SpMYH Delta cells after hydrogen peroxide treatment. These data suggest that MYH may act as an adaptor to recruit checkpoint proteins to the DNA lesions.

as the proliferating cell nuclear antigen (PCNA) 1 -like 9-1-1 complex) and are related in structure to the PCNA sliding clamp (6 -8). The S. pombe (Sp) Rad17 protein is homologous to the largest subunit of replication factor C, the clamp loader. SpRad3, a phosphatidylinositol 3-kinase-related protein, is homologous to human ATM (ataxia telangiectasia-mutated protein) and ATR (ATM-and Rad3-related protein) kinases. SpRad3 plays a central role in cell cycle check point regulation. It acts to initiate cell cycle arrest by transducing the DNA damage signal through phosphorylation of SpChk1 and SpCds1 kinases in a SpRad9/Rad1/Hus1 and SpRad17-dependent manner. SpHus1 is also phosphorylated by SpRad3 in response to DNA damage (9). The SpRad26 protein is the homolog of human ATR-interacting protein and forms a complex with SpRad3.
Because myriad forms of DNA lesions are caused by endogenous and environmental factors, the mode through which checkpoint proteins recognize these various forms of DNA damage is still poorly understood. Human ATM/human ATR/ SpRad3 and Rad17 are proposed to act at an early step to sense DNA damage (10). It has been suggested that these checkpoint proteins may detect a common intermediate, such as singlestranded DNA coated by replication protein A, which is processed by various DNA repair pathways (11), or they may require a series of "adaptors" to recognize DNA damage. Potential candidates for such adaptor proteins are the DNA damage recognition proteins. Recently, a few DNA damage recognition proteins involved in mismatch repair, nucleotide excision repair, and double-strand break repair have been shown to interact with checkpoint proteins (12)(13)(14)(15). These reports support a hypothesis that DNA repair proteins recognize the lesions and provide signals for checkpoint factors. In this work, we investigate the possible association of S. pombe checkpoint proteins with the base excision repair protein MutY homolog (MYH).
Cells possess several DNA repair pathways for dealing with the many different types of DNA lesions. Reactive oxygen species are the most prevalent source of DNA lesions in aerobic organisms. Oxidative damage to DNA can result in mutagenesis, in some cases leading to degenerative diseases. One of the most abundant and highly mutagenic forms of oxidative damage to DNA is 7,8-dihydro-8-oxo-guanine (8-oxoG or GO), which causes G:C to A:T transversions. Although the eukaryotic mis-match repair protein MSH2/MSH6 (MutS␣) heterodimer can recognize A/GO mismatches (16,17), the GO lesions in DNA are repaired mainly by base excision repair pathways. The 8-oxoG glycosylase (OGG1) protein, a functional eukaryotic homolog of Escherichia coli MutM, can remove both ringopened purine lesions and mutagenic GO adducts if they are paired with cytosines (18 -21). However, no OGG1 or MutM homolog has been found in S. pombe. S. pombe possesses a MutY homolog (SpMYH) that is an adenine DNA glycosylase able to remove adenines misincorporated on the template DNA strand containing G or GO, thus preventing G:C to T:A mutations (22). Fission yeast deficient in SpMYH have a spontaneous mutation rate ϳ40-fold higher than the wild type cell and express higher sensitivity to oxidative agents (23). We have shown that MYH directly associates with PCNA in both S. pombe and human cells (24,25). We also provided direct evidence that the association between MYH and PCNA is important biologically for MYH function in mutation avoidance (24). Thus, it is suggested that the base excision repair pathway may be coupled with DNA replication through the MYH interaction with PCNA, targeting repair toward the daughter DNA strands (24 -28). A similar mechanism for mismatch repair to target the daughter DNA strands has also been proposed (29 -31), although PCNA has been demonstrated to modulate the polarity of excision steps during mismatch repair (32).
In this study we demonstrate for the first time that SpMYH physically interacts with all three subunits of the SpHus1/ SpRad1/pRad9 complex. SpHus1 interacts with SpMYH at a region that is separate from the SpPCNA-interaction motif. Moreover, the interaction of SpMYH and SpHus1 is enhanced after hydrogen peroxide treatment through which the resulting oxidative stress promotes phosphorylation of SpHus1. However, both unphosphorylated and phosphorylated SpHus1 interact with SpMYH. Hydrogen peroxide treatment and deletion of SpHus1 alter the elution profiles of SpMYH during gel filtration chromatography. SpHus1 phosphorylation is dependent on SpMYH expression after hydrogen peroxide treatment. These results support a model in which the DNA damage recognition protein MutY serves as an "adaptor" for checkpoint protein recognition of DNA lesions. Construction of hus1-MYC/SpMYH⌬ Strain-The his3 gene in the plasmid pSPMYH19 containing the SpMYH::his3 insert (pSPMYH-his) (23) was replaced by leu2 gene. The leu2 gene was amplified by PCR from pESP-3 (Stratagene, La Jolla, CA) by primers listed on Table I. The PCR product was purified, digested with BglII, and inserted into the BglII-cleaved plasmid pSPMYH-his. The plasmid with SpMYH::leu2 was amplified in bacterial cells, linearized with restriction enzyme AatII, and then transfected into S. pombe hus1-MYC (501: h Ϫ leu1-32, ura4-D18, ade6-706, hus1-MYC) by electroporation. The yeast genomic SpMYH was replaced with leu2 interrupted SpMYH cDNA (SpMYH::leu2) by homologous recombination. The transformed cells with leu ϩ phenotype were selected on the YNB plates (0.67% yeast nitrogen base without amino acids, 2% glucose, 1.5% agar) supplemented with 0.1 mg/ml adenine and 0.1 mg/ml of uracil. An interruption of SpMYH gene in the yeast chromosome was verified by PCR.
Transformation of hus1-MYC/SpMYH⌬ Cells with Wild-Type Sp-MYH cDNA-The entire open reading frame of SpMYH cDNA from pSPMYH19 (22) was amplified by PCR with primers listed on Table I. The product was digested with XhoI and BamHI and then inserted into pSCF172 (American Type Cell Culture). The obtained plasmid pSCF172-SpMYH was transformed into hus1-MYC/SpMYH⌬ cells by electroporation. Transformed cells were selected with ura ϩ phenotype on the YNB agar plates supplemented with adenine (33). The expression of SpMYH protein was confirmed by Western blotting analysis with polyclonal antibodies against SpMYH.
Construction of Truncated Glutathione S-Transferase (GST)-SpMYH Proteins-The plasmid pGEXSPMYH expressing full-length SpMYH tagged with GST has been previously described (24). The truncated SpMYH cDNA fragments were PCR-amplified using appropriate primers (listed in Table I) from the template pGEXSPMYH and were ligated into pGEX-4T-2 at the BamH1 site and transformed into E. coli BL21-Star cells (Invitrogen). The sequences of the cloned genes were confirmed through DNA sequencing. GST Pull-down Assay-GST-pull-down assays were performed in a manner similar to previously described procedures (25). E. coli (BL21Star/DE3) cells (Stratagene, La Jolla, CA) harboring the expression plasmids were cultured in Luria-Bertani broth containing 100 mg/ml ampicillin at 25°C. Protein expression was induced at an A 590 of 0.6 by the addition of isopropyl 1-thio-␤-D-galactopyranoside to a final concentration of 0.4 mM, and the cells were harvested 16 h later by centrifugation at 10,000 ϫ g for 20 min. The cell paste from a 1-liter culture was resuspended in 10 ml of phosphate-buffered saline (PBS, 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na 2 HPO 4 , and 1.4 mM KH 2 PO 4 ). After sonication, the solution was centrifuged at 10,000 ϫ g for 20 min. To the supernatant (10 ml), 1 ml of a 50% slurry of glutathione-Sepharose 4B (Amersham Bioscience) in PBS was added and incubated for 2 h at 4°C. The GST fusion proteins bound to the beads were pelleted at 1000 ϫ g for 5 min and then washed 5 times with 5 ml of PBS. The beads were suspended in PBS containing 0.1% sodium azide and 0.1% of a protease inhibitor mixture (Sigma-Aldrich) to form a 50% slurry and stored at 4°C. GST constructs (300 ng) immobilized on glutathione-Sepharose 4B were incubated with 5% bovine serum albumin in PBS for 1 h at 4°C, washed with PBS, then incubated with 500 g of E. coli or yeast extracts prepared as described by Chang et al. (23) overnight in 200 l of PBS at 4°C. After centrifugation at 1000 ϫ g for 2 min, the supernatant was saved, and the pellets were washed 5 times with 1 ml of PBS. The pellets and supernatants were fractionated on a 10% SDS-polyacrylamide gel followed by Western blot analysis (34). A control was run concurrently with immobilized GST alone.
Nickel-agarose Affinity Binding-The His-tagged SpHus1 expressed in BL21-star cells was bound to nickel-agarose (Qiagen Inc., Valencia, CA) according to the manufacturer's procedures. Purified SpMYH (200 ng) expressed in E. coli (22) were incubated with the beads at 4°C for 1 h. After washing with buffer N (50 mM potassium phosphate, pH 8.0, 300 mM NaCl) containing 50 mM imidazole, the bound proteins were eluted by buffer N containing 250 mM imidazole. The unbound and eluting fractions were fractionated by 10% SDS-polyacrylamide gel electrophoresis and transferred onto a nitrocellulose membrane. Affinity-purified SpMYH polyclonal antibodies (35) were used for Western blotting analysis.
Treatment of S. pombe with H 2 O 2 -Fission yeast (hus1-MYC) and (SpMYH⌬ hus1-MYC) cells (250 ml) were grown in YEPD medium (20 g of peptone, 10 g of yeast extract, 20 g of dextrose/liter) to an A 595 of ϳ0.6. Hydrogen peroxide was added to the culture at a final concentration of 6 mM. After shaking at 30°C for 1 h, cells were spun down and transferred to fresh YEPD medium without hydrogen peroxide. After various recovery time intervals, cells were harvested and resuspended in PBS for protein extraction via sonication.
Size Fractionation-Yeast cells were grown to log phase in 1 liter of YEPD medium. Hydrogen peroxide treatment was performed as described above followed by cell recovery for 2 h in fresh YEPD medium lacking hydrogen peroxide. After being harvested by centrifugation, cells were resuspended in 10 ml of PBS and sonicated for 6 cycles of 10 s sonication followed by 20 s of rest. After centrifuging at 10,000 rpm in a SS34 rotor for 30 min at 4°C, the supernatant (approximate 15 ml) was treated with 65% ammonium sulfate. The protein precipitant was resuspended in buffer S (20 mM potassium phosphate, pH 7.4, 150 mM KCl, 0.1% IGEPAL CA-630 (Nonidet P-40), 0.5 mM dithiothreitol, 0.1 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride, and 10% glycerol) and dialyzed against the same buffer for 3 h. The dialyzed protein sample was centrifuged, and the supernatant (1 mg of protein in 0.5 ml) was loaded onto a 24-ml Superose 12 HR 10/30 column (Amersham Biosciences) that had been equilibrated with buffer S. The flow rate was set to 0.25 ml per min, and 0.25-ml fractions were collected. The column was calibrated using size makers (blue dextran, thyroglobin, apoferritin, ␤-amylase, bovine serum albumin, oval albumin, and carbonic anhydrase) from Sigma/Aldrich.

Physical Interaction between SpMYH and SpHus1 in
S. pombe-The PCNA sliding clamp has been shown to interact with many proteins containing a common motif (36,37). We have previously shown that SpPCNA physically interacts with the C terminus of SpMYH (24). In light of the observation that the Hus1/Rad1/Rad9 heterotrimer is structurally related to the homotrimer PCNA (6 -8, 38 -40), we tested whether SpMYH is able to interact with SpHus1. The physical interaction between SpMYH and the SpHus1 protein was demonstrated using a GST pull-down assay. GST-SpMYH fusion protein bound to glutathione-Sepharose was incubated with extracts derived from fission yeast cells expressing HA-tagged SpHus1. Detection of bound SpHus1 was performed via Western blot analysis using an antibody against the HA peptide. As shown in Fig. 1A, HA-SpHus1 could be pulled down by GST-SpMYH (lane 4) but not by GST alone (lane 2). To determine a direct interaction between SpHus1 and SpMYH, His-tagged SpHus1, expressed in E. coli, was bound to Ni 2ϩ -agarose and incubated with purified SpMYH protein. SpMYH was found to bind to Ni 2ϩagarose-bound SpHus1 but not to Ni 2ϩ -agarose beads associ- ated with bovine serum albumin (BSA) (Fig. 1B). These results indicate that SpMYH can directly associate with SpHus1 in the absence of SpRad1 and SpRad9.
SpPCNA and SpHus1 Bind to Separate Regions of Sp-MYH-It has been reported that the Rad9/Rad1/Hus1 complex resembles the PCNA homotrimer but contains unique features likely adapted to its role in checkpoint control (39). We have shown that the SpPCNA binding site is located at the C terminus of SpMYH, involving residues 434 -448 (24). To test whether SpPCNA and SpHus1 bind to the same region of SpMYH, we analyzed the ability of SpHus1 to bind to a GSTfused truncated SpMYH (N433 containing residues 1-433), which lacks the SpPCNA binding motif. The results reveal an association of N433 with SpHus1 in extracts derived from fission yeast cells expressing HA-tagged SpHus1 (Fig. 1A, lane 6), indicating that SpHus1 and SpPCNA bind to separate regions of SpMYH and interact in unique ways with SpMYH. To map the SpHus1 binding domain in the SpMYH protein, two truncated forms of SpMYH fused to GST were assayed. The C-terminal half of SpMYH (C245 containing residues 245-461) was shown to associate with SpHus1 in the pull-down experiment (Fig. 1, lane 10). A mild interaction was also detected with the N-terminal half of SpMYH (N230 containing residues 1-230) (Fig. 1, lanes 8). This suggests that SpHus1 may contact the central region of SpMYH or that SpHus1 has one major binding site at the C-terminal half and another minor binding site at N-terminal half of SpMYH.
SpRad1 and SpRAD9 Also Interacts with SpMYH-In light of the fact that SpHus1 forms a complex with SpRad1 and SpRad9, we next assayed the ability of SpMYH to bind SpRad1 and SpRad9. Protein extracts were prepared from yeast cells expressing an HA-tagged SpRad1. GST-SpMYH pull-down assays revealed an association between SpRad1 and SpMYH ( Fig. 2A, lane 2). In addition, the major binding site of SpRad1 resides at the C-terminal domain of SpMYH (Fig. 2A, lane 6), similar to that seen in SpHusI. To determine direct interactions between SpMYH-SpRad1 and SpMYH-SpRad9, Histagged SpRad1 and SpRad9 expressed in E. coli were assayed for their ability to bind the GST-tagged C-terminal domain of SpMYH. Both SpRad1 and SpRad9 were shown to bind GST-SpMYH-C245 but not to GST alone (Fig. 2, B and C). The binding affinities of SpMYH to SpHus1, SpRad1, and SpRad9 are weak; a maximal 5% of input proteins were pulled down (Figs. 1A and 2A). Thus, all three individual PCNA-related checkpoint proteins can directly interact with SpMYH and do so in the absence of the other two partners.
Oxidative Stress Enhances SpMYH-SpHus1 Interaction and Induces SpHus1 Phosphorylation-Because SpMYH is involved in repair of oxidative damage, we tested whether the SpHus1-SpMYH interaction was altered after hydrogen peroxide-induced oxidative stress. For this in vivo interaction, we used S. pombe hus1-MYC in which the SpHus1 is C-terminally tagged with 13 MYC epitopes (6). Like wild type cells, hus1-MYC cells arrest progression through cell cycle upon ionizing radiation (6). Cells were grown to log phase, treated with hydrogen peroxide for 1 h, and then cultured in media lacking hydrogen peroxide at various time intervals. Protein extracts from hydrogen peroxide-treated cells were immunoprecipitated using antibodies directed against SpMYH, and the immunoblot was detected with c-Myc antibody. In untreated cells, the Sp-MYH-SpHus1 interaction could be detected; however, it was weak (Fig. 3A, lane 1). The amount of SpHus1 bound to SpMYH increased slightly immediately after hydrogen peroxide treatment (Fig. 3A, lane 2). When treated cells were allowed to recover in fresh media for 2 h, the amount of SpHus1 precipitated by SpMYH antibody increased dramatically (Fig. 3A, lane  3). The levels of SpHus1 in the immunoprecipitant peaked at 2 h and then decreased after 6 h of recovery. A quantitative analysis of the Western blot showed 16-, 12-, and 6-fold increases of SpMYH-SpHus1 interactions after 2, 4, and 6 h of recovery over untreated cells, respectively. In contrast, the interaction between SpPCNA and SpMYH in the H 2 O 2 -treated cells remained almost unchanged, as assayed by co-immunoprecipitation (Fig. 3B). There was a slight decrease of SpMYH-SpPCNA interaction at 6 h after hydrogen peroxide treatment (Fig. 3B, lane 5). Therefore, the SpMYH-SpHus1 interaction is altered, but the SpMYH-SpPCNA interaction remains almost constant in fission yeast cells under oxidative stress.
We investigated whether the increased interaction between SpMYH-SpHus1 is due to differences in protein expression levels. Thus, we determined the total protein levels of SpMYH, SpHus1, and SpPCNA in cell extracts directly through Western blotting. The protein levels of SpMYH decreased slightly, whereas those of SpPCNA did not change in response to hydrogen peroxide treatment (Fig. 3, C and D). As shown in Fig.  3E, SpHus1 protein levels were unchanged, but a minor band with slightly lower mobility was observed in H 2 O 2 -treated extracts. This upper band could be observed at 1 h after hydrogen peroxide treatment (Fig. 3E, lane 2), was maximal at 2 h of recovery (lane 2), and remained after 6 h of recovery (lane 5). The upper SpHus1 band is the phosphorylated form, as evidenced by conversion of this band to a lower band after -phosphatase treatment (Fig. 3F, lane 2). SpHus1 has been shown to be phosphorylated by SpRad3 in response to hydroxyurea replication block and bleomycin treatment (9). Our result demonstrates that SpHus1 also becomes phosphorylated in response to oxidative stress. Interestingly, the presence of SpHus1 phosphorylation correlated with the increase in SpHus1-SpMYH interaction after hydrogen peroxide treatment. However, the phosphorylation of SpHus1 was not necessary for SpMYH-SpHus1 interaction, since both phosphorylated and non-phosphorylated SpHus1 could be immunoprecipitated by SpMYH antibodies (Fig. 3A, lane 3). Redistribution (6). To test whether SpMYH and SpHus1 co-eluted on gel filtration chromatography, we loaded yeast extracts from the hus1-MYC strain on a Superose 12 column and then performed Western blotting with the respective antibodies. The majority of SpHus1-MYC eluted as two peaks on this column; peak I of larger than 670 kDa (fractions 32-34) and peak II of ϳ450 kDa (fractions 40 -44) (Fig. 4, A and B, and Fig. 5A). Peak I eluted in the void volume, where blue dextran and thyroglobin also co-eluted. The 450-kDa peak II of SpHus1-MYC is similar to that reported by Caspari et al. (6); however, peak I was not observed by the same authors. The elution profiles of SpHus1 were similar for H 2 O 2 -treated and untreated cell extracts. This is consistent with the findings of Caspari et al. (6) that the elution profile of SpHus1 does not change after ionizing radiation or hydroxyurea treatment. The phosphorylated SpHus1 form was not readily detected in these fractions from H 2 O 2treated extracts due to poor gel resolution.
Three peaks of SpMYH in the H 2 O 2 -untreated cell extract eluted from the Superose 12 column; peak I consisted of larger than 670-kDa (fractions 30 -34 in the void volume), peak II consisted of 150-kDa (fractions 46 -50), and peak III consisted of less than 50 kDa (fractions 54 -64) (Fig. 4C and Fig. 5B, open  circles). The purified SpMYH (calculated M r of 51) expressed in bacteria eluted from the same column at a position of ϳ45 kDa (data not shown). Thus, peak III from yeast extract represents the native monomeric SpMYH. This suggests that peaks I and II of SpMYH may be associated with other proteins. The elution profiles of SpHus1 (Fig. 5A) and SpMYH (Fig. 5B) do not coincide, indicating they do not exist as a strong complex. SpMYH in the H 2 O 2 -treated cell extract also eluted from the Superose 12 column as three peaks ( Fig. 4D and Fig. 5B, filled diamonds). However, the distribution of the three peaks was different from that of SpMYH in the H 2 O 2 -untreated cell extract. An increase in peak II accompanied by a decrease of peak III was observed for SpMYH in the H 2 O 2 -treated cell extract (Fig. 5B, filled diamonds). There is also an increase of SpMYH at fractions 38 -44 that corresponds to molecular mass of 400 -500 kDa. These elution profiles were reproducible in two independent experiments. Thus, the portion of SpMYH (Peak II) co-eluting with the SpHus1 increases as a result of H 2 O 2 treatment (compare Fig. 5, A and B, filled diamonds).
To study the association of SpMYH and SpHus1, we then performed coimmunoprecipitation utilizing the SpMYH antibody to precipitate SpHus1 from the Superose 12 fractions. In the H 2 O 2 -untreated cell extract, SpHus1precipitated by the SpMYH antibody mainly distributed in fractions [37][38][39][40][41][42][43] and 5C, open circles). Fractions 37-43 contain peak II of SpHus1 and correspond to a M r of ϳ400 -600, which has been previously reported as the native M r of the 9-1-1 complex (6). However, the precipitable SpHus1 peaking at fraction 39 does not coincide to peak II of SpHus1. Using the H 2 O 2 -treated cell extract, the pattern of SpHus1 coimmunoprecipited by SpMYH antibody differed from that of the untreated cell extract; more SpHus1 was detected in fraction 31 and fractions 45-51 (com- pare Figs. 4, E and F, and Fig. 5C, filled diamonds). The distribution pattern of the precipitable SpHus1 (Fig. 5C, filled  diamonds) is very similar to the pattern seen in the Western blot of SpHus1 (Fig. 5A, filled diamonds). This suggests that the redistribution of SpMYH in H 2 O 2 -treated cells (Fig. 5B) may be due to the increased SpMYH-SpHus1 interaction.
To analyze the influence of SpHus1 on SpMYH association with other proteins, extracts were prepared from hus1-deleted cells and subjected to gel filtration chromatography. As can be seen in Figs. 4, G and H, and 5D, the majority of SpMYH eluted in the void volume from the Superose 12 column (fractions 30 -36, M r larger than 670), and the profiles of SpMYH distribution are very similar for those of H 2 O 2 -treated and untreated cell extracts.
SpHus1 Phosphorylation Is Dependent on SpMYH Expression after Hydrogen Peroxide Treatment-Kostrub et al. (9) reported that SpHus1 is phosphorylated by SpRad3 in response to DNA damage. To test a model that SpMYH senses DNA damage and activates cell cycle checkpoint pathways after oxidative stress, we measured the phosphorylation level of SpHus1 in H 2 O 2 -treated SpMYH⌬ cells. We noted a 10-fold reduction of SpHus1 phosphorylation level in SpMYH⌬/hus1-MYC cells as compared with SpMYH-proficient cells after H 2 O 2 -induced oxidative stress and recovery for 2 h (Fig. 6B,  lanes 1 and 2, Fig. 6C). To further investigate the correlation of SpMYH expression levels and SpHus1 phosphorylation, an expression vector pSCF172 containing SpMYH cDNA was incorporated into the SpMYH⌬ hus1-MYC cells. The expression of the SpMYH protein in the transformed cells was detected by Western analysis (Fig. 6A, lanes 3-7). Because the nmt1 promoter controls the expression of SpMYH cDNA in pSCF172, SpMYH protein expression can be regulated by the concentration of thiamine in the minimal medium. As shown in Fig. 6A,  lanes 3 and 4, when cells  g/ml thiamine, the expression of SpMYH was almost completely suppressed (Fig. 6A, lane 7). As shown in Fig. 6, the percentages of SpHus1 phosphorylation are correlated with the SpMYH expression levels. These results clearly indicate that phosphorylation of SpHus1 in response to oxidative stress is dependent on SpMYH.  Fig. 4 were quantitated, and the percentages of each band are calculated by dividing the intensity of each band by the total intensity (presented as the relative intensity on the y-axis). A, quantitative analyses of the Western blots from Fig. 4, A and B. B, quantitative analyses of the Western blots from Fig. 4, C and D. C, quantitative analyses of the Western blots from Fig. 4, E and F. IP, immunoprecipitate. D, quantitative analyses of the Western blots from Fig. 4, G and H. Open circles and filled squares represent H 2 O 2 untreated and treated cell extracts, respectively. DISCUSSION DNA base lesions induced by oxidative damage are repaired mainly by base excision repair pathways. In light of an absence of OGG1 in S. pombe, SpMYH may be the important player in reduction of GO mutagenesis. We have shown that SpMYHdeficient fission yeast has an elevated mutation frequency (23). The most important known biological function of SpMYH is the recognition of A/GO mismatches by removing misincorporated A from the template GO. The high affinity SpMYH for its reaction product apurinic/apyrimidinic/GO may extend to G/GO, T/GO, and C/GO mismatches, similar to that seen for E. coli MutY (41). It has been suggested that the base excision repair pathway may involve highly coordinated processes governed by protein-protein and protein-DNA interactions, possibly involving the transient "handing off" of intermediates (42)(43)(44). Human MYH has been shown to interact with apurinic/ apyrimidinic endonuclease, PCNA, replication protein A, and MSH6 (25,27). We have shown that SpMYH can interact with SpPCNA and hPCNA (24,25). The interaction of SpMYH with SpPCNA is important for the ability of SpMYH to repair oxidative DNA damages (24). In this study we show that three fission yeast PCNA-related checkpoint proteins (SpHus1, SpRad1, and SpRad9) can also interact with SpMYH through GST pull-down and co-immunoprecipitation assays and that SpMYH interaction with recombinant SpHus1, SpRad1, and SpRad9 occurs even in the absence of other yeast proteins. In addition, as assayed by co-immunoprecipitation of fractions from a gel filtration column, SpMYH associates with SpHus1 in a complex of ϳ450 kDa, the reported native molecular mass of the 9-1-1 complex. This is the first demonstration that a DNA base excision protein directly interacts with the PCNA-like 9-1-1 complex.
The PCNA sliding clamp is a homotrimer that interacts with many replication and repair proteins (29,(45)(46)(47)(48). So far no one has been able to demonstrate that one PCNA homotrimer can bind three proteins simultaneously. Although Hus1, Rad1, and Rad9 have structures similar to that of PCNA (6 -8), it has been shown that more than one subunit of the 9-1-1 complex can associated with the same protein. Bermudez et al. (49) has shown that hRad17 interacts predominately with hRad9 and to a much lesser degree with hRad1 but not with hHus1. Giannattasio et al. (13) demonstrated that ScRad14 interacts strongly with ScDdc1 (SpRad9 homolog) and more weakly with ScMec3 (SpHus1 homolog) but did not test the interaction of ScRad14 with ScRad17 (SpRad1 homolog). It appears that one of the three Hus1/Rad1/Rad9 subunits is involved in major interactions with other proteins. We show here that SpMYH can interact with each subunit of the 9-1-1 complex and with individual subunits in the absence of the other two subunits.
Our data indicate that SpMYH may undergo an asymmetrical interaction with the 9-1-1 complex in the decreasing affinity order of SpHus1, SpRad1, SpRad9 (compare Fig. 1, lanes 9 and 10; Fig. 2A, lanes 5 and 6; Fig. 2B, lanes 1 and 2; Fig. 2C, lanes  1 and 2). We also observed similar asymmetrical interactions between hMYH and the human 9-1-1 complex. 2 Our binding data suggest that the SpHus1 binding domain in SpMYH protein mainly resides at the C-terminal half of SpMYH and is separate from its PCNA binding site (residues 434 -448) (24). This substantiates the notion that the Rad9/Rad1/ Hus1complex possesses some unique features, likely adapted to its role in checkpoint control (39). Several PCNA-binding proteins have been shown to bind the interdomain connector loop (residues 118 -135) of the PCNA trimer (50,51). Because SpHus1 and SpPCNA bind to disparate sites on SpMYH, Sp-MYH may not bind to the interdomain connector loop of SpHus1 as in the modeled structure.
The DNA damage response to genotoxic insults is essential for genome stability. To date, little is known about the relationship between DNA checkpoint proteins and enzymes involved in base excision repair. It has been suggested that base excision repair appears as invisible to the checkpoint system (52,53). Leroy et al. (53) reported that treatment of Saccharomyces cerevisiae cells with H 2 O 2 induces a ScMec1-dependent (human ATR/SpRad3 homolog) phosphorylation of ScRad53 (hChk2/SpCds1 homolog) during S phase but not during G 1 and G 2 phases. They also showed that the response to DNA damage after H 2 O 2 treatment can occur during the G 1 and G 2 phases in yeast cells defective in the apurinic/apyrimidinic endonucleases ScApn1 and ScApn2. This may be caused by the accumulation of the intermediates and/or the repair is operated by alterative repair pathways. In human cells both OGG1 and MYH glycosylases are involved in GO lesions base excision repair, whereas S. cerevisiae and S. pombe do not contain MYH and OGG1, respectively. Therefore, DNA checkpoint activation in response to oxidative stress is likely unique in these two organisms.
We have shown that fission yeast cells deficient in the functions of SpMYH are more sensitive to hydrogen peroxide (23). Here, we show that hydrogen peroxide can induce the phosphorylation of SpHus1 and can enhance the SpMYH-SpHus1 interaction. Moreover, SpHus1 phosphorylation is dependent on SpMYH expression after hydrogen peroxide treatment. Even though the phosphorylation of SpHus1 is not essential for SpHus1-SpMYH interaction, the increase in the in vivo SpHus1-SpMYH interaction correlates with the presence of SpHus1 phosphorylation after hydrogen peroxide treatment. Our result is similar to the finding of Brown et al. (12) who showed that interaction between MSH2 and CHK2 and interaction between MLH1 and ATM are enhanced after ionizing radiation. However, Giannattasio et al. (13) demonstrated that the interaction between ScRad14 and ScDdc1 is not affected by a UV mimetic agent. In contrast, the SpMYH-SpPCNA interaction demonstrated minimal changes after oxidative stress. The data of Fig.  3, A and B, show that when SpMYH-SpHus1 interaction is enhanced, the SpMYH-SpPCNA does not decrease. In addition, SpHus1 and SpPCNA bind to separate regions of SpMYH. Thus, SpMYH-SpHus1 interaction seems independent of Sp-MYH-SpPCNA interaction. We have shown that the association between MYH and PCNA is important in vivo for MYH function in mutation avoidance (24). It will be interesting to see the phenotype of a functional SpMYH with a defective interaction with the 9-1-1 complex. Our working model is that PCNA acts as the coordinator of base excision repair directing the repair on the newly synthesized DNA strands (24,25,27), whereas the MYH-Hus1 interaction occurs to induce the DNA damage response after oxidative stress. It remains to be deter-mined whether MYH is able to bind to PCNA and Hus1 simultaneously.
Several interesting findings emerge from the analyses of SpMYH and SpHus1 by gel filtration chromatography. (i) We observed different elution profiles of SpMYH on a gel filtration column between the H 2 O 2 -untreated and treated cell extracts. Particularly, SpMYH in the H 2 O 2 -untreated extract is shifted to the region (fractions 38 -50) with a molecular mass of 150 -500 kDa, the same position of second peak of SpHus1. (ii) Although the interaction between SpMYH and SpHus1 increases after H 2 O 2 treatment, as assayed by coimmunoprecipitation (Fig. 3A), both proteins do not co-elute completely on a gel filtration column (Fig. 5, A and B, filled diamonds). Similar findings were reported by Caspari et al. (6), that the majority of SpRad1 and SpRad9 do not co-elute with the 450-kDa native 9-1-1 complex. This suggests that SpMYH may form SpHus1independent protein complexes. (iii) It is interesting to note that peak II of SpMYH elutes at a M r of 150 -200 (fractions 46 -50), which coincides with the expected trimeric size of the 9-1-1 complex and with the reported SpRad9 peak (6). In addition, more SpHus1 was precipitated by SpMYH antibody in this region in the H 2 O 2 -treated extract as compared with the untreated extract (fractions 45-51 in Fig. 5C). (iv) Coimmunoprecipitation of SpHus1 with SpMYH antibody demonstrates SpMYH is associated with the 450-kDa native 9-1-1 complex (Fig. 5C). The distribution pattern of the precipitable SpHus1 from the H 2 O 2 -treated cell extract is very similar to the pattern of SpHus1 by direct Western blot analysis (compare Figs. 5, A and C, filled diamonds). (v) The redistribution of SpMYH in H 2 O 2 -treated cells is dependent on SpHus1, because the profiles of SpMYH distribution are not altered by the H 2 O 2 treatment in hus1-deleted cells. These results suggest that an intact 9-1-1 complex may be critical for the enhanced SpMYH-SpHus1 interaction. Further testing the SpMYH-SpHus1 interaction in SpRad1 and SpRad9 mutants remains to be investigated.
Our findings support a model that MYH is one of the adaptors for checkpoint proteins in recognition of DNA lesions. First, the physically association between SpMYH and the 9-1-1 complex is enhanced after oxidative stress. Second, there is a similar kinetics between SpHus1 phosphorylation and SpHus1-SpMYH interaction after hydrogen peroxide treatment. Third, SpHus1 phosphorylation is dependent on SpMYH expression after hydrogen peroxide treatment. In this model, MYH functions upstream of Rad9/Rad1/Hus1 in the DNA damage signaling pathway. MYH first recognizes the lesions and then recruits Rad9/Rad1/Hus1 to the sites of DNA damage. After binding with SpMYH at the lesion site, SpHus1 is phosphorylated by SpRad3 kinase, signaling the DNA damage response. Several reports also support this model. S. cerevisiae Rad14 and Rad1 as well as human XPA, which are all involved in nucleotide excision repair, are required for the damage response (54,55). A direct interaction between nucleotide excision repair enzymes and checkpoint proteins has been demonstrated by Giannattasio et al. (13) showing that ScRad14 physically and functionally interacts with the S. cerevisiae 9-1-1 complex. In addition, mismatch repair proteins have been suggested to function as a sensor in signaling apoptosis, based on the findings that mismatch repair deficient cells are more resistant to a variety of chemotherapeutic agents (56 -59). Wang and Qin (15) showed that human mismatch repair enzyme MSH2 interacts with the ATR kinase to form a signaling module in response to alkylating agents. Brown et al. (12) showed that MSH2 interacts with CHK2, MLH1 associates with ATM, and the mismatch repair system is required for S-phase checkpoint activation. A series of recent reports pro-vides support that the Mre11/Rad50/Nbs1 complex senses double strand breaks in DNA and activates cell cycle checkpoint pathways after exposure to radiation (for review, see Ref. 14). It will be interesting to discover whether other damage-recognition proteins also interact with PCNA-like proteins.