|Home | About | Journals | Submit | Contact Us | Français|
The Schizosaccharomyces pombe nip1+/ctp1+ gene was previously identified as an slr (synthetically lethal with rad2) mutant. Epistasis analysis indicated that Nip1/Ctp1 functions in Rhp51-dependent recombinational repair, together with the Rad32 (spMre11)-Rad50-Nbs1 complex, which plays important roles in the early steps of DNA double-strand break repair. Nip1/Ctp1 was phosphorylated in asynchronous, exponentially growing cells and further phosphorylated in response to bleomycin treatment. Overproduction of Nip1/Ctp1 suppressed the DNA repair defect of an nbs1-s10 mutant, which carries a mutation in the FHA phosphopeptide-binding domain of Nbs1, but not of an nbs1 null mutant. Meiotic DNA double-strand breaks accumulated in the nip1/ctp1 mutant. The DNA repair phenotypes and epistasis relationships of nip1/ctp1 are very similar to those of the Saccharomyces cerevisiae sae2/com1 mutant, suggesting that Nip1/Ctp1 is a functional homologue of Sae2/Com1, although the sequence similarity between the proteins is limited to the C-terminal region containing the RHR motif. We found that the RxxL and CxxC motifs are conserved in Schizosaccharomyces species and in vertebrate CtIP, originally identified as a cofactor of the transcriptional corepressor CtBP. However, these two motifs are not found in other fungi, including Saccharomyces and Aspergillus species. We propose that Nip1/Ctp1 is a functional counterpart of Sae2/Com1 and CtIP.
DNA double-strand breaks (DSBs), among the most critical types of DNA damage, can lead to chromosomal aberrations, disruption of genome integrity, and cancer. DSBs are caused not only by exogenous sources such as γ irradiation but also by endogenous factors such as free radicals generated by aerobic respiration. DSBs can also arise as a result of replication fork collapse. In addition, programmed DSBs are created at recombination hot spots during meiosis.
When DSBs arise in mitotic cells, DNA damage checkpoints are activated, resulting in cell cycle arrest and induction of appropriate repair machinery (reviewed in reference 27). DSBs can be repaired by two major DNA repair pathways: homologous recombination (HR) and nonhomologous end-joining (NHEJ) repair mechanisms (reviewed in references 35, 51, 58, and 60). In HR repair, the ends of a DSB are resected to produce recombinogenic DNA with 3′ single-strand overhangs. The resulting single-stranded DNA region pairs with homologous DNA of the intact sister chromatid, forming a D-loop structure. The repair process proceeds with DNA synthesis by using the sister chromatid as a template, essentially without loss of genetic information. In NHEJ, the broken ends are aligned, may or may not be processed, and are then directly rejoined. NHEJ repair is thus a potentially error-prone process. Both pathways are highly conserved from yeast to humans.
In the budding yeast Saccharomyces cerevisiae, the organism in which HR has been most extensively studied, and most likely in other organisms as well, meiotic HR is initiated by programmed, Spo11-dependent DSBs (31). These DSBs are formed by a topoisomerase-like trans-esterase mechanism mediated by Spo11. Spo11 covalently attached at the DNA ends has been identified as an intermediate in meiotic cell extracts (32). In addition, a large number of other gene products are involved in this step, including Mre11, Rad50, Xrs2, Mer2, Mei4, Rec102, Rec104, Rec114, and Ski8 (4, 37, 56, 63; reviewed in reference 31).
After the formation of DSBs, the ends are immediately processed to expose 3′ single-strand overhangs that are essential for homologous pairing and strand exchange. This process involves the removal of covalently attached Spo11 from the DNA ends and nucleolytic resection of the DNA strand whose end was attached to Spo11. Mre11, Rad50, and Xrs2 are required for processing of the DNA ends, in addition to being required for DSB formation (7, 32, 50). Sae2 (also called Com1) is also required for this resection step (33, 54, 55). Mutants lacking these proteins are sensitive to DNA-damaging agents and defective in strand resection of DSB ends during mitosis (5, 13, 42). Thus, Mre11, Rad50, Xrs2, and Sae2/Com1 are implicated in DSB end-processing mechanisms during both mitotic and meiotic cell cycles.
Mre11, Rad50, and Xrs2 form a protein complex called the MRX complex (14). Mre11 is a DNA-binding protein that contains four phosphodiesterase motifs, which are responsible for nuclease activity. Rad50 is a split ABC-type ATPase which contains two heptad repeats that are located at its center and that fold into a coiled-coil, bringing the two N- and C-terminal ATPase motifs, Walker A and B, into close proximity (17). These structural features are characteristic of SMC (structural maintenance of chromosome) family proteins. Rad50 was recently reported to have adenylate kinase activity, which is required for the efficient tethering of DNA molecules (6). Both Mre11 and Rad50 are conserved among eukaryotes. However, conservation of the Xrs2 amino acid sequence is quite low compared to that of Mre11 and Rad50 and only limited regions or short sequence motifs are conserved, as described below.
The fission yeast Schizosaccharomyces pombe, which is distantly related to S. cerevisiae, possesses the rad32 gene as its MRE11 homologue (61, 72, 73). rad32 was originally identified as a rad mutant showing sensitivity to various DNA-damaging treatments, and cloning and sequencing revealed that it corresponded to MRE11. The S. pombe rad50 gene was isolated by taking advantage of its sequence similarity to Rad50 homologues from S. cerevisiae, Caenorhabditis elegans, and humans (28).
The vertebrate Nbs1 protein, which forms a complex with Mre11 and Rad50, is the functional counterpart of Xrs2. The overall sequence similarity between these two proteins is weak and is limited to an N-terminal forkhead-associated (FHA) domain and a small C-terminal conserved domain. Nbs1, but not Xrs2, contains a BRCA1 C-terminal (BRCT) domain in the N-terminal region next to the FHA domain (8, 41, 70). Based on this information, Russell, Rhind, and colleagues identified Nbs1 in S. pombe (10). We independently identified Nbs1 in a genetic screen of rad2Δ-synthetic lethality mutants, as described below (67).
Human and budding yeast Mre11 has a single-strand endonuclease activity, a 3′-5′ double-strand exonuclease activity, and a weak hairpin-opening activity. Both ATP and Rad50 stimulate the 3′-5′ exonuclease and hairpin-opening activities of Mre11, and ATP may regulate the DNA binding of Mre11 complexes via Rad50 (23, 30, 48, 52, 62, 69). Since point mutations that abolish the nuclease activity of Mre11 do not impair most forms of DSB repair in mitotic cells, the contribution of the Mre11/Rad50 complex to the DSB repair process is largely independent of its nuclease activity. Importantly, Mre11 removes Spo11 covalently bound to the break site by endonucleolytic cleavage a few bases from the site of attachment, leading to the release of a Spo11-oligonucleotide complex (50).
A specific class of separation-of-function mutants of S. cerevisiae RAD50, named rad50S, allow DSB formation but are totally defective in the processing of Spo11-induced meiotic DSBs, as observed in nuclease-deficient mre11 mutants, and thus they accumulate unprocessed DSBs with covalently attached Spo11 protein at the 5′ ends (2, 32). SAE2/COM1 gene deletion mutants possess a meiotic phenotype very similar to that of rad50S point mutants (13, 42, 54, 55), and Spo11-oligonucleotide complexes are not observed in rad50S or sae2Δ/Com1Δ mutants. These observations imply that Sae2/Com1 is involved in DSB end processing in collaboration with the MRX(N) complex. However, no clear Sae2/Com1 homologues with structural or functional similarity have been reported in organisms other than Saccharomyces species.
In addition to being involved in repair processes per se, the Mre11 complex is also involved in sensing DSB ends. Binding of Mre11 to break sites is a prerequisite for the recruitment and activation of protein kinases, including Tel1/SpTel1 and Mec1/SpRad3 (S. cerevisiae/S. pombe), which are homologues of mammalian ATM and ATR, respectively. Upon MRX(N)-dependent activation, ATM phosphorylates downstream substrates, including Mre11 and Nbs1 (19, 36, 49). S. cerevisiae Sae2/Com1 is also phosphorylated in a Tel1- and Mec1-dependent manner, and the Mec1 dependency is much stronger than that of Tel1 (5). Importantly, Sae2/Com1 negatively regulates the DNA damage checkpoint in a phosphorylation-dependent manner (12).
The S. pombe rad2+ gene encodes a structure-specific endonuclease homologous to mammalian Fen-1 and S. cerevisiae Rad27, which is required for Okazaki fragment maturation during DNA replication (47). Loss of the nuclease activity in combination with some single mutations that cause defects in HR in S. cerevisiae and S. pombe is lethal (16, 59). We used this relationship to isolate previously unidentified S. pombe genes involved in HR. We constructed two slr (synthetic lethality with rad2Δ) libraries consisting of mutants that were inviable in a rad2Δ background (64, 67). Among these mutants, we identified the rhp57+, rad60+, rad62+, fbh1+, nbs1+, rad32+, and mcl1+ genes, mutations of which are all epistatic to rhp51Δ mutations, with the exception of mcl1, which is colethal with the rhp51Δ mutation (44-46, 64, 65, 67; unpublished results).
In this study, we characterized the slr9 mutant, which was isolated in our second screen (67). Slr9 expressed from a multicopy plasmid suppressed the DNA repair defect of an Nbs1 mutant containing a mutation in its FHA domain, a phosphopeptide-binding domain. From this genetic interaction, we renamed Slr9 Nbs1-interacting protein 1 (Nip1). Further, epistasis analysis suggested that Nip1 functions in Rhp51-dependent recombinational repair, together with the Rad32 (spMre11)-Rad50-Nbs1 (MRN) complex. Nip1 is phosphorylated in asynchronous exponentially growing cells and is more extensively phosphorylated in response to bleomycin treatment. Interestingly, meiotic DNA DSBs accumulate in the nip1Δ mutant. Therefore, we propose that Nip1 functions to regulate the nuclease activity of the MRN complex. Some, but not all, of these mutant properties are similar to those of S. cerevisiae sae2/com1 mutants. Nip1 and Sae2/Com1 have a sequence similarity that is limited to the C-terminal region, containing an RHR motif, implying that Nip1 is a functional counterpart of Sae2/Com1. Possible functional homologues of Sae2/Com1 and Nip1 in higher eukaryotes include the human CtIP protein and the Arabidopsis thaliana gamma response 1 protein (AtGR1). Very recently, Russell's group has identified the S. pombe ctp1+ gene as a cell cycle-regulated gene required for HR (39). nip1+ is identical to ctp1+.
The S. pombe strains used in this study are listed in Table Table1.1. All of the original slr mutants had the genotype h− Msmt-0 slr rad2+ leu1-32 ura4-D18 ade6-704 (67). Standard procedures and media were used for propagation and genetic manipulations, as previously described (43). The sensitivity of S. pombe cells to γ and UV irradiation was analyzed as previously described (46).
slr9-1 cells were transformed with an S. pombe genomic library (67) constructed with the vector pSP102 (64), and methyl methanesulfonate (MMS)-resistant transformants were selected on plates containing 0.004% MMS (64). Plasmids conferring MMS resistance to slr9-1 mutants were isolated, and their inserts were sequenced. The 5′ and 3′ regions of slr9+/nip1+ cDNAs were amplified by PCR with the S. pombe cDNA library constructed in the vector pGADGH (Clontech) as a template and the primer sets 5′ AD/IWA590 and 3′ AD/IWA217, respectively. The sequences of 5′ AD, 3′ AD, IWA217, and IWA590 are 5′-CTATTCGATGATGAAGATACCCCACCAAACCC-3′, 5′-GTGAACTTGCGGGGTTTTTCAGTATCTACGAT-3′, 5′-TAAAGCACTTCCTGCTTACG-3′, and 5′-TTTCGTCATTCCAAGTAGGC-3′, respectively.
Genomic DNA containing the slr9+/nip1+ region was cloned into the SmaI site of pBluescript II SK (+). An NdeI recognition site was created at the nip1 stop codon by site-direct mutagenesis with the QuikChange site-directed mutagenesis kit (Stratagene). The 3×FLAG sequence created by annealing of an oligonucleotide set, IWA533 and IWA534, was inserted into the NdeI site of nip1, yielding pSK-nip1F. The sequences of IWA533 and IWA534 are 5′-TATGGACTACAAAGACCATGACGGTGATTATAAAGATCATGACATCGATTACAAGGATGACGATGACAAGCA-3′ and 5′-TATGCTTGTCATCGTCATCCTTGTAATCGATGTCATGATCTTTATAATCACCGTCATGGTCTTTGTAGTCCA-3′. The ura4+ marker was inserted into the PstI site of pSK-nip1F. The plasmid was digested at the unique EcoO65I site within the nip1 sequence and used for transformation of S. pombe YA119. Pop-out clones having lost the ura4+ marker were selected on the basis of resistance to 5-fluoroorotic acid. To generate nip1 point mutants, site-directed mutagenesis was performed with pSK-nip1F and the following primer sets: IWA963 (5′-GCTAGACTTAAAGCTCAATTGGTCTTGG-3′) and IWA964 (5′-CCAAGACCAATTGAGCTTTAAGTCTAGC-3′) for the S35A mutation, IWA965 (5′-GCTACGTGAAGCACAACCATTAGCTCC-3′) and IWA966 (5′-GGAGCTAATGGTTGTGCTTCACGTAGC-3′) for the T175A mutation, IWA1097 (5′-AATCCCTCCGCATGCACCCACTCTTCCTG-3′) and IWA1098 (5′-CAGGAAGAGTGGGTGCATGCGGAGGGATT-3′) for the S114A mutation, and IWA1099 (5′-CATTTACGAGCAAGGGCACCTGAAGACATG-3′) and IWA1100 (5′-CATGTCTTCAGGTGCCCTTGCTCGTAAATG-3′) for the S165A mutation. The kanMX6 sequence was inserted at the Bsp1407I site, which was located 67 bp distal from the nip1 stop codon. The DNA fragments of the nip1 genomic region containing kanMX6 were transformed into S. pombe. All mutants were confirmed by PCR, Southern blotting, and sequencing analysis.
Cells were grown in YES or EMM medium (43) at 30°C to ~107 cells/ml. Bleomycin (10 mU/ml; Sigma) or mock solution (distilled water) was added to cultures, and incubation was continued for 4.5 h. Cells were collected by low-speed centrifugation, washed with distilled water, resuspended in 400 μl HB buffer [25 mM morpholinepropanesulfonic acid (MOPS; pH 7.2), 100 mM NaCl, 15 mM MgCl2, 15 mM EGTA, 60 mM β-glycerophosphate, 15 mM para-nitrophenylphosphate, 0.5 mM sodium vanadate, 10 μg/ml pepstatin, 10 μg/ml chymostatin, 120 μg/ml 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride, 1.57 mg/ml benzamidine, 5 μg/ml leupeptin, 5 μg/ml bestatin, 0.01% Triton X-100), and lysed by vortexing with glass beads. Clarified lysates obtained by centrifugation were used as cell extracts.
Antiserum against the recombinant S. pombe Nbs1 protein was raised in rabbits. The antibody was affinity purified from the immune serum as previously described (53). The anti-FLAG M2 monoclonal antibody was purchased from Sigma.
The protein concentration of cell extracts was measured with a protein assay kit (Bio-Rad), and aliquots of cell extracts containing 50 μg protein were used for Western blot analysis. The anti-FLAG M2 monoclonal antibody was used as the primary antibody to detect Nip1-3×FLAG, and the rabbit anti-Nbs1 antibody was used to detect untagged Nbs1. The ECL Plus system (GE Healthcare Bio-Science) was used for detection of immunocomplexes.
Aliquots (200 μl, ~2 mg total protein) of the cell extracts described above were incubated with 20 μl anti-FLAG M2 agarose (~50% suspension) (Sigma) at 4°C for 3 h to precipitate the Nip1-3×FLAG fusion protein. The beads were washed with HB buffer, and the precipitates were subjected to Western blotting. When phosphatase treatment was performed, lambda phosphatase (NEB) was added to a reaction mixture containing precipitated beads in 10 μl phosphatase buffer (NEB) at 30°C for 30 min, followed by Western blotting analysis.
To monitor meiosis-associated DSBs, pat1-114 haploid cells were synchronously induced to undergo meiosis and then processed for DNA plug preparation as previously described (9). Agarose-embedded chromosomal DNA was fractionated by pulsed-field gel electrophoresis (PFGE) on a 0.8% chromosomal-grade agarose gel in 1× Tris-acetate-EDTA buffer with the CHEF-MAPPER system (Bio-Rad) under the following conditions: temperature, 14°C; run time, 48 h; voltage, 2 V/cm; included angle, 106°; initial and final switching times, 20 and 30 min, respectively. The DNA was stained with ethidium bromide.
The nucleotide sequence data obtained in this study, including exon and intron information, have been deposited with the DDBJ under accession number AB363065.
In previous studies, we constructed two slr mutant libraries in which we collected mutants with mutations that caused synthetic lethality in the rad2Δ background and that also conferred sensitivity to MMS (64, 67). In this study, we analyzed slr9, which corresponds to mutant no. 45 from the second screen (67). To clone the slr9+ gene, the slr9-1 mutant was transformed with an S. pombe genomic library. Transformants which MMS sensitivity complemented to the wild-type level were isolated, and plasmid DNA was recovered. Sequence analysis revealed that one plasmid contained a 2.2-kb genomic fragment in which a putative open reading frame (ORF), SPCC338.08, and a very questionable ORF, SPCC338.09, were annotated. The 5′ and 3′ regions of the SPCC338.08 cDNA were isolated from the S. pombe cDNA library by PCR with the primer sets 5′ AD/IWA590 and 3′ AD/IWA217, and the amplified regions were sequenced. The gene structure is shown in Fig. Fig.1A.1A. SPCC338.08 is transcribed from 429 bp upstream to 958 bp downstream of the putative initiation codon. Since there were several stop codons and no ATG upstream from the putative initiation codon, this appeared to be the actual initiation codon. An intron that had not been predicted in the database was found in the 3′ region. Thus, the SPCC338.08 gene product is encoded by two exons and consists of 294 amino acids, although the database had predicted 285 amino acids. For SPCC338.09, cDNA was not amplified by PCR with the primer set from the cDNA library, and thus, the SPCC338.09 ORF may not be transcribed, at least during vegetative growth. Therefore, we concluded that SPCC338.08, but not SPCC338.09, corresponds to the slr9 gene.
To confirm the identity of the slr9 gene, we replaced the SPCC338.08 ORF with the ura4+ marker. The resulting deletion mutant (YA1097) exhibited a phenotype very similar to that of the originally isolated slr9-1 mutant, including slow growth, sensitivity to MMS, and synthetic lethality with rad2Δ (data not shown). Sequencing of the genomic DNA revealed that the slr9-1 mutant had a CG-to-TA transition mutation that converted Arg162 (CGA) to a stop codon (TGA) in SPCC338.08. Thus, SPCC338.08 contains the slr9 gene. Since slr9 interacts genetically with nbs1, as described below, we named it Nbs1-interacting protein 1, nip1+.
BLAST and PSI-BLAST searches allowed the identification of potentially important Nip1 motifs (Fig. 1B to D). A summary of these motifs is schematically presented in Fig. Fig.1B.1B. The RHR motif in the C-terminal region is the representative signature of Sae2/Com1 homologues (Fig. (Fig.1C).1C). SAE2/COM1 is an S. cerevisiae gene that was originally identified in mutants defective in meiotic recombination (42, 54). Although the region containing the RHR motif is limited (approximately 30 amino acids), this region is well conserved from Saccharomyces Sae2/Com1 to human CtIP. CtIP was originally discovered as a cofactor of the transcriptional corepressor CtBP (C terminus-binding protein), a coregulator that binds to the C terminus of adenovirus E1A (57), BRCA1 (74, 81), and the retinoblastoma tumor suppressor protein RB (24) (for reviews, see references 11 and 75). The Arabidopsis thaliana gamma response 1 (AtGR1) protein also exhibits sequence similarity to the RHR motif in the C terminus. The gene that encodes it was originally identified as induced by γ irradiation with no known function (18). Very recently, it has been identified as a functional and structural homologue of Sae2/Com1 (66).
We identified the RxxL and CxxC motifs approximately 30 to 50 amino acids upstream of the RHR motif in Nip1 homologues from vertebrates, plants, nematodes, protists, and fission yeast, but not in Nip1 homologues from other fungi, including Saccharomyces and Aspergillus (Fig. (Fig.1D).1D). The RxxL motif is also called the D-box, which is a representative signature of APC/C substrates (3, 25, 34). The presence of this motif suggests that Nip1 family proteins are proteolytically regulated during the cell cycle by APC/C-mediated ubiquitination.
The conserved CxxC sequences are found in proteins with a D-box, but not in proteins lacking this motif (Fig. (Fig.1D).1D). The CxxC motif in Rad50 is involved in zinc chelation (29). In addition, the N-terminal region of Nip1, like those of CtIP and AtGR1, is predicted to form a coiled-coil structure (Fig. (Fig.1B).1B). The CtIP N terminus is involved in homodimerization (20). Therefore, Nip1 family proteins have two potentially important domains enabling them to dimerize, the CxxC motif and the N-terminal coiled-coil region.
Nip1 contains two potential CDK-dependent phosphorylation sites and two Rad3/Tel1-dependent phosphorylation sites. CtIP has one CDK-dependent and two ATM-dependent phosphorylation sites (38, 79). However, these sites do not seem to be conserved between Nip1 and CtIP.
We first examined the DNA repair activity of a nip1 deletion strain (Fig. (Fig.2).2). The nip1Δ mutant is sensitive to various DNA-damaging treatments, such as UV and γ irradiation, bleomycin, and MMS. Bleomycin is a radiomimetic reagent that causes both single-strand breaks and DSBs in DNA. To determine whether Nip1 is involved in the Rhp51-dependent recombinational repair pathway, a nip1Δ rhp51Δ double mutant was constructed and its UV repair phenotype was compared to that of the single mutants. As shown in Fig. Fig.2A,2A, the nip1Δ rhp51Δ double mutant was as sensitive to UV as the rhp51Δ single mutant. Thus, Nip1 is involved in an Rhp51 recombinase-dependent recombinational repair pathway.
Mutant no. 16 isolated in our previous second screen (67) was named slr10 in this study. We identified the slr10+ gene by the same procedures used to clone the slr9+/nip1+ gene. Two transformants of slr10, clones 2-36 and 7, displayed MMS resistance similar to that of wild-type cells (Fig. (Fig.3A).3A). Plasmids were isolated from clones 2-36 and 7 and analyzed. DNA sequencing revealed that clone 2-36 contained a genomic fragment identical to that of previously isolated pNT101, which contains the nbs1+ gene (67). On the basis of this result, the genomic sequence of the nbs1 locus in the slr10 mutant was determined. We found that the isolated slr10 mutant had a GC-to-AT transition mutation that converted Gly103 (GGT) to Asp (GAT). Genetic mapping also showed that slr10 was very close to nbs1 on chromosome II (data not shown). A newly reconstructed strain carrying the same mutation as that in nbs1 exhibited a phenotype indistinguishable from that of the original slr10 mutant (data not shown). Therefore, we concluded that slr10 is an allele of nbs1. Hereafter, the slr10 mutation is referred to as nbs1-s10.
Gly103 is within the FHA domain and is a conserved amino acid among Nbs1 homologues (Fig. (Fig.3B).3B). The nbs1-s10 mutant was slightly less sensitive to UV, bleomycin, hydroxyurea (HU), and MMS than was the nbs1Δ mutant (Fig. (Fig.3D3D and data not shown). The amount of cellular Nbs1-s10 protein was similar to that of wild-type Nbs1 (Fig. (Fig.3C),3C), and overexpression of Nbs1-s10 from a plasmid vector did not suppress the sensitivity to various DNA-damaging treatments (Fig. (Fig.3E).3E). These results suggest that the nbs1-s10 mutation does not cause instability of Nbs1, but rather that the defect in DNA repair in Nbs1-s10 is probably due to defects in FHA domain function. Overexpression of nbs1-s10 from a multicopy plasmid did not affect DNA repair in wild-type cells, indicating that nbs1-s10 is a recessive mutation.
Clone 7 plasmid contained genomic DNA encoding the nip1+ gene, and sequencing analysis revealed that the original slr10 mutant did not have a nip1 mutation. Therefore, nip1+ is a multicopy suppressor of nbs1-s10 with respect to UV, bleomycin, HU, and MMS sensitivity (Fig. (Fig.3F).3F). Importantly, plasmid-expressed nip1+ did not suppress nbs1Δ sensitivity to DNA-damaging agents (Fig. (Fig.3F),3F), indicating that nip1+ is an allele-specific multicopy suppressor of the nbs1-s10 mutation.
The allele-specific multicopy suppression of nbs1-s10 by nip1+ described above suggested a functional interaction between Nip1 and Nbs1. We therefore analyzed the epistatic relationships of these genes with regard to DNA repair (Fig. 2B to D). The nip1Δ single mutant displayed sensitivities similar to that of the nbs1Δ mutant in response to γ or UV irradiation, MMS, and HU. The sensitivities of the nip1Δ mutant to these DNA-damaging agents were comparable to those of a rad32Δ single mutant (Fig. 2B and C). We then constructed a nip1Δ nbs1Δ double mutant and compared its sensitivities to those of the single mutants. As shown in Fig. 2C and D, the double mutant was as sensitive as each single mutant to UV, bleomycin, and MMS. Therefore, Nip1 is likely to function in the same repair pathway as Nbs1.
The observed genetic interaction between Nip1 and Nbs1 in the DNA damage response led us to examine whether there is a direct physical interaction between the two proteins. We constructed a yeast strain expressing C-terminally FLAG-tagged Nip1 at the original genomic locus that showed normal repair activity (Fig. (Fig.4B).4B). Nip1 was immunoprecipitated from cell extracts prepared from either normal or bleomycin-treated cells with an anti-FLAG antibody. However, Nbs1 was not detected by Western blotting. Moreover, Rad32 was not detected (data not shown). When immunoprecipitation was performed with Rad32 tagged with hemagglutinin epitopes as bait, Nip1 signals were not detected in the Rad32-containing immunocomplexes, although Nbs1 coprecipitated with Rad32 (data not shown).
In these experiments, we found that expression of Nip1 protein was induced by severalfold upon bleomycin treatment, consistent with a previous report (71). We also detected a more slowly migrating band above the normal Nip1 signal in samples from cells treated with bleomycin (Fig. (Fig.4).4). The slower signal was also detected in samples from MMS- or HU-treated cells (Fig. (Fig.4A),4A), suggesting that Nip1 was modified in response to DNA damage. As protein phosphorylation is one of the major modifications occurring in response to DNA damage, we treated Nip1 immunoprecipitates with protein phosphatase to determine whether Nip1 was phosphorylated. As shown in Fig. Fig.4B,4B, the more slowly migrating signal disappeared after phosphatase treatment and the Nip1 band migrated faster than that from untreated immunoprecipitates from normal cells. We treated Nip1-containing immunoprecipitates from normal cells with phosphatase and found that Nip1 in normally growing cells is also phosphorylated (Fig. (Fig.4B).4B). These results suggest that Nip1 is phosphorylated during at least two phases, normal growth and in response to DNA damage. Hereafter, the fully phosphorylated form of Nip1 is designated Nip1-S (slow mobility) and basally phosphorylated Nip1 is designated Nip1-F (fast mobility).
To investigate the genetic requirements for the DNA damage-dependent modification of Nip1, we monitored Nip1 phosphorylation in various mutant backgrounds (Fig. (Fig.5).5). In rad32 or nbs1 deletion mutants, a small amount of Nip1-S signal was observed during normal growth. However, the induced accumulation of Nip1-S signal in response to bleomycin treatment was not observed. In contrast, deletion of rhp51, rhp57, or ku70 did not substantially affect the Nip1 phosphorylation pattern, compared to that of wild-type cells, regardless of bleomycin treatment (Fig. (Fig.5A5A and data not shown). These results suggest that the MRN complex is important for Nip1 phosphorylation.
Next, we examined the involvement of DNA damage checkpoint proteins in Nip1-S formation (Fig. (Fig.5B).5B). We first monitored the effect of deleting rad3+, which encodes an ATR kinase homologue, a major kinase in the DNA damage response in S. pombe. Bleomycin-dependent phosphorylation of Nip1 was not altered in the rad3 deletion mutant. Similar results were obtained with a mutant containing a deletion of rad26, which encodes an ATRIP homologue, an essential interactor with ATR. However, phosphorylation was greatly decreased by deletion of tel1+, which encodes a minor kinase in S. pombe, the ATM kinase. This kinase plays an important role in vertebrate cells, including human cells. Bleomycin-dependent Nip1 phosphorylation was almost completely abolished in a rad3 tel1 double-deletion mutant.
Requirements for the checkpoint clamp (rad1 and rad9), the clamp loader (rad17), downstream components of the DNA damage checkpoint (crb2 or mrc1), and the downstream kinases chk1 and/or cds1 were also examined (Fig. (Fig.5B).5B). Single deletion mutations and the chk1Δ cds1Δ double mutation did not affect Nip1 phosphorylation, suggesting that the Tel1 kinase, but not other downstream components, is required for DNA damage-dependent Nip1 phosphorylation. In the absence of Tel1, Rad3 (ATR) can partially substitute for Tel1 kinase activity in Nip1 phosphorylation.
Next, we examined potential Nip1 phosphorylation sites. Nip1 phosphorylation is dependent on Tel1 (and Rad3), as described above, and these kinases phosphorylate serine and threonine in SQ and TQ consensus sequences. CDK-dependent phosphorylation occurs at SP consensus sequences. Wild-type Nip1 has four potential phosphorylation sites fitting these criteria: Ser-35, Ser-114, Ser-165, and Thr-175 (Fig. (Fig.1B1B and and4C).4C). The SQ TQ double mutation (S35A and T175A) was expected to eliminate putative phosphatidylinositol 3-kinase (PI3-kinase)-dependent phosphorylation sites. The SP1 SP2 double mutation (S114A and S165A) was expected to eliminate putative CDK-dependent phosphorylation sites. The SQ TQ SP1 SP2 quadruple mutation eliminated all four potential sites. We constructed these three nip1 mutants and examined bleomycin-induced phosphorylation. As shown in Fig. Fig.4C,4C, the SQ TQ mutant displayed a phosphorylation pattern similar to that of the wild-type Nip1 protein. In contrast, the Nip1-F signal of the SP1 SP2 mutant was not detected in cell extracts prepared from undamaged normal cells. In these samples, a more rapidly migrating band, Nip1-FF, appeared. We hypothesized that Nip1-FF corresponds to the fully dephosphorylated form of wild-type Nip1 (Fig. (Fig.4B),4B), which was confirmed by phosphatase treatment (Fig. (Fig.4D).4D). Since Nip1 is phosphorylated in asynchronous cultures during normal growth, we conclude that Ser-114 and/or Ser-165 are the site of this basal phosphorylation, which may be CDK dependent. However, this SP1 SP2 mutant became phosphorylated upon bleomycin treatment and the modified signal migrated at the same position as wild-type Nip1-F, although Nip1-S was not detected (Fig. 4C and D). The SQ TQ SP1 SP2 quadruple mutant showed a phosphorylation pattern very similar to that of the SP1 SP2 mutant. Therefore, it is most likely that the site(s) of bleomycin-induced phosphorylation is an amino acid residue(s) other than these four S/T residues.
The three nip1mutants showed no detectable defect in DNA repair in response to UV irradiation, MMS, or bleomycin (Fig. (Fig.4E).4E). These results also indicate that none of the four candidate sites is necessary for DNA repair. Therefore, the observed phosphorylation at Ser-114 and/or Ser-165 may not play a direct role in DNA repair, and other, unidentified, phosphorylation sites may be critically important for DNA repair.
We examined the meiotic functions of Nip1. In meiosis, transient DSB formation is induced and subsequent repair is observed. These events are closely associated with an elevated frequency of HR. To easily assess the formation and repair of meiotic DSBs, we induced pat1-114 haploid cells to undergo highly synchronous meiosis and analyzed their chromosomal DNA by PFGE. PFGE allows visualization of intact chromosomes and meiotically broken DNA as three discrete bands and smeared bands, respectively. Thus, as shown in wild-type cells, we can detect transient DSB formation by the appearance of smeared DNA bands (Fig. (Fig.6,6, wild type at 5 h) and successive repair by restoration of the three chromosomal bands (Fig. (Fig.6,6, wild type at 7 h). In nip1Δ cells, a high level of broken DNA was observed at 5 h in sporulation culture (Fig. (Fig.6,6, nip1Δ mutant at 5 h), and it persisted for at least 2 h more (Fig. (Fig.6,6, nip1Δ mutant at 7 h). In contrast, the progression of premeiotic DNA synthesis (data not shown) and the time points at which DSBs appeared in nip1Δ cells were very similar to those in wild-type cells (Fig. (Fig.6,6, wild type at 5 h and nip1Δ mutant at 5 h). These features are reminiscent of the rad50S mutant, in which meiotic DSBs accumulate without being repaired (9, 78). Next, we compared nip1Δ cells with a mutant impaired in an MRN complex component, Rad32, and a mutant defective in the catalytic engine that produces meiotic DSBs, Rec12 (the S. pombe Spo11 homologue), to determine whether their phenotypes differed. In rad32Δ cells, smeared DNA remained at 7 h but the smeared signal was much weaker than in nip1Δ cells, while breaks were not observed in rec12Δ cells. These meiotic phenotypes of rec12Δ and rad32Δ cells are consistent with the results of Young et al. (77, 78). Importantly, the accumulation of DSBs in the nip1Δ mutant was dependent on Rad32 and Rec12, since nip1Δ rad32Δ and nip1Δ rec12Δ double mutants showed phenotypes very similar to those of rad32Δ and rec12Δ single mutants, respectively. It also should be pointed out that the migration of chromosome III was faster in samples from cells of the nip1Δ and rad32Δ backgrounds than in samples from wild-type and rec12Δ mutant cells. Although the reasons for this alteration are unknown, the difference might be attributable to different copy numbers of the ribosomal DNA genes carried in chromosome III.
In this study, we identified the previously uncharacterized slr9 gene in fission yeast, which we isolated as a rad2Δ-synthetic lethal mutant (64, 67). We named the gene nip1 on the basis of its genetic interaction with the nbs1-s10 allele. The nip1 null mutation elevates sensitivity to various DNA-damaging agents, including UV and γ irradiation, bleomycin, MMS, and HU. Although these agents affect DNA differently, they all directly or indirectly cause DSBs, which are predominantly repaired by HR in yeast. Additionally, nip1 belongs to the same epistasis group as rhp51 (Fig. (Fig.2).2). These findings suggest that nip1 is involved in Rhp51-dependent recombinational repair. The nip1 null mutant exhibited DNA repair defects very similar to those of the rad32Δ and nbs1Δ mutants. These DNA repair defects were not exacerbated by mutations in these MRN complex genes. Therefore, we conclude that Nip1 functions in recombinational repair together with the MRN complex. Importantly, meiotic DSBs accumulated in nip1Δ cells (Fig. (Fig.6),6), similar to the phenotype of rad50S mutants. These phenotypes, together with epistasis analysis and sequence conservation, suggest that Nip1 is a functional counterpart of the S. cerevisiae Sae2/Com1 protein and that these proteins are involved in the processing of DSB ends in collaboration with the MRN(X) complex (13, 33, 54, 55).
Although Sae2/Com1 and Nip1 are expected to play similar roles in vivo, there are some important differences between the two proteins. SAE2/COM1 null mutants are only slightly sensitive to MMS, while MRX null mutants show severe sensitivity to MMS (42). In contrast, nip1Δ mutants and mutants with deletions of MRN genes are in the same epistasis group with regard to DNA repair (Fig. (Fig.2).2). In addition, fungal Sae2/Com1 homologues do not have the D-box or CxxC motif. The Nip1 level in cells is cell cycle regulated (39) and may be regulated by proteolysis involving the D-box and APC/C-mediated ubiquitination. Expression of nip1+ is induced by DNA damage (Fig. (Fig.4)4) (71). In contrast, the cellular Sae2/Com1 level is largely unaffected by the cell cycle or DNA damage (5).
The C-terminal regions of human CtIP and A. thaliana AtGR1 share significant homology with Nip1. AtGR1 is a gene that was identified in a screen for mRNAs that accumulate in response to ionizing radiation and is cell cycle regulated (18). The AtGR1 gene expression pattern is similar to that of nip1+. Very recently, it has been identified as a functional and structural homologue of Sae2/Com1 (66).
CtIP was originally identified as a cofactor of the transcriptional corepressor CtBP (for reviews, see references 11 and 75). DNA damage-induced interactions in the BRCA1/BARD1-containing supercomplex include CtIP and the MRN complex (26), and CtIP is ubiquitinated in a BRCA-dependent manner in response to DNA damage (38, 79, 80). These reports, together with sequence homology, suggest that CtIP is a functional homologue of Nip1.
The genetic interaction we demonstrated between nip1 and nbs1 is allele specific for nbs1-s10 (Fig. (Fig.3).3). The mutation in nbs1-s10 alters the evolutionarily conserved glycine in the most C-terminal part of the FHA domain to an aspartic acid (Fig. (Fig.3B).3B). Since the FHA domain is a phosphopeptide-binding motif (22) and Nip1 is a phosphorylated protein (Fig. (Fig.44 and and5),5), it is likely that Nip1 interacts physically with Nbs1 via its FHA domain. However, we did not detect a direct physical interaction between Nip1 and Nbs1 by immunoprecipitation or by yeast two-hybrid analysis (data not shown). Thus, this interaction may be very weak and/or transient.
Nip1 phosphorylation is puzzling (Fig. (Fig.44 and and5).5). Mutagenesis of the potential CDK phosphorylation target residues abolished Nip1 phosphorylation during normal growth, suggesting that Nip1 regulation is cell cycle dependent. In addition, Nip1 is further phosphorylated in response to bleomycin, MMS, and HU treatments. Cells carrying rad3Δ, rad26Δ, rad1Δ, rad9Δ, and rad17Δ mutations exhibited the same level of bleomycin-induced phosphorylation as wild-type cells, but those carrying a tel1Δ mutation exhibited severely reduced phosphorylation levels. A rad3Δ tel1Δ double mutation abolished almost all damage-induced phosphorylation of Nip1 (Fig. (Fig.5).5). A mutant Nip1 protein containing mutations of all of the canonical phosphorylation sites still migrated more slowly following bleomycin treatment (Fig. (Fig.4).4). DNA damage-induced phosphorylation completely depended on the PI3-kinases Rad3 and Tel1 (Fig. (Fig.5).5). Therefore, Tel1 and/or Rad3 may phosphorylate residues other than the consensus sites on Nip1. Alternatively, an unknown downstream kinase activated by Tel1 and/or Rad3 may directly phosphorylate Nip1.
In S. cerevisiae, Mec1 is largely responsible for DNA damage-induced Sae2/Com1 phosphorylation in coordination with the checkpoint signaling proteins Ddc1, Rad17, Mec3, and Rad24 (5). However, as shown in Fig. Fig.5,5, cells carrying rad3Δ, rad26Δ, rad1Δ, rad9Δ, and rad17Δ mutations exhibited levels of bleomycin-induced phosphorylation similar to that of wild-type cells, and cells carrying the tel1Δ mutation exhibited severely decreased phosphorylation. Therefore, unlike S. cerevisiae Sae2/Com1, Nip1 is primarily phosphorylated by a Tel1-dependent pathway upon bleomycin treatment. The Crb2, Mrc1, Chk1, and Cds1 proteins, downstream factors in the Rad3- and Tel1-dependent checkpoints, are dispensable for Nip1 phosphorylation, suggesting that Nip1 is phosphorylated early in the checkpoint pathway.
The rad32Δ and nbs1Δ mutants exhibited reduced levels of Nip1 phosphorylation, similar to that of the tel1Δ mutant upon bleomycin treatment. The MRN(X) complex is required for efficient activation of ATM (Tel1) in S. cerevisiae and vertebrates (15, 21, 68). Fission yeast Tel1 is also reported to bind conserved motifs in the C terminus of Nbs1 (76). Therefore, the DNA damage-induced phosphorylation of Nip1 may proceed in a Tel1-MRN-dependent manner in the early steps of the DNA damage checkpoint pathway. The fact that DNA damage-induced phosphorylation depends on the PI3-kinases is consistent with this model.
Human CtIP is also phosphorylated on three serine residues, S327, S664, and S745, in cell cycle-regulated and DNA damage-induced manners (38, 79, 80). Phosphorylation on S327 in the putative CDK phosphorylation motif reaches a maximum in G2 phase and is important for the CtIP interaction with BRCA1 and for the G2/M transition checkpoint in response to γ irradiation. S664 and S745 are phosphorylated in an ATM-dependent manner in response to γ irradiation. BRCA1 is also required for CtIP phosphorylation in response to irradiation. There is no distinct orthologue of Brca1 in fission yeast. Nevertheless, the regulation of Nip1 phosphorylation is likely conserved in fission yeast, although the function of phosphorylation remains to be elucidated.
In rad32Δ and nbs1Δ mutants, but not in the rhp51Δ mutant, a minor amount of Nip1 is phosphorylated to levels observed in damage-induced cells in the absence of exogenous DNA damage (Fig. (Fig.5A).5A). This state may reflect the accumulation of a specific DNA structure that induces Nip1 phosphorylation. Nip1 is required to repair Rec12 (spSpo11)-induced DSBs in meiosis (Fig. (Fig.6).6). Importantly, the nip1Δ mutant accumulates more abundant broken DNAs than does the rad32Δ single mutant. This result suggests that the formation of both Rad32-dependent and Rad32-independent DSBs is processed by Nip1. Taking the results of the epistasis analysis and meiotic DSB experiments together, we hypothesize that Nip1 regulates the nuclease activity of the Mre11 complex. Alternatively, Nip1 per se might be a specific nuclease that works with the MRN complex to produce recombinogenic single-stranded DNA. Further biochemical studies are needed to delineate Nip1 activity.
We thank Akira Shinohara and Shunichi Takeda for valuable information concerning the homology among Sae2, Nip1, and CtIP; Kanji Furuya for Nip1 sequence analysis in S. japonicus; and Kouji Hirota for preliminary experiments of meiotic DSBs. We also thank Tetsuro Kokubo for helpful discussions and encouragement.
This study was supported in part by Grants-in-Aid for Scientific Research on Priority Areas from the Ministry of Education, Culture, Sports, Science, and Technology (MECSST) of Japan to K.O. and H.I. and for Scientific Research (B) and for Young Scientists (Startup) from the Japan Society for the Promotion of Science (JSPS) to H.I. and to T.Y., respectively, and by grants from the 2007 Strategic Research Project (No. K19011) of Yokohama City University to H.I. and from the Bio-oriented Technology Research Advancement Institution to K.O.
Published ahead of print on 31 March 2008.