|Home | About | Journals | Submit | Contact Us | Français|
MEC1 and TEL1 encode ATR- and ATM-related proteins in the budding yeast Saccharomyces cerevisiae, respectively. Phleomycin is an agent that catalyzes double-strand breaks in DNA. We show here that both Mec1 and Tel1 regulate the checkpoint response following phleomycin treatment. MEC1 is required for Rad53 phosphorylation and cell-cycle progression delay following phleomycin treatment in G1, S or G2/M phases. The tel1Δ mutation confers a defect in the checkpoint responses to phleomycin treatment in S phase. In addition, the tel1Δ mutation enhances the mec1 defect in activation of the phleomycin-induced checkpoint pathway in S phase. In contrast, the tel1Δ mutation confers only a minor defect in the checkpoint responses in G1 phase and no apparent defect in G2/M phase. Methyl methanesulfonate (MMS) treatment also activates checkpoints, inducing Rad53 phosphorylation in S phase. MMS-induced Rad53 phosphorylation is not detected in mec1Δ mutants during S phase, but occurs in tel1Δ mutants similar to wild-type cells. Finally, Xrs2 is phosphorylated after phleomycin treatment in a TEL1-dependent manner during S phase, whereas no significant Xrs2 phosphorylation is detected after MMS treatment. Together, our results support a model in which Tel1 contributes to checkpoint control in response to phleomycin-induced DNA damage in S phase.
The genomes of eukaryotic cells are under continuous assault from extracellular genotoxic agents and the byproducts of intracellular processes including DNA replication. The resulting genetic damage may cause cell death or genetic instability that can lead to cancer. To ensure the proper response to genetic damage, cells have employed a set of surveillance mechanisms termed checkpoint controls (1,2).
Checkpoint pathways are an evolutionarily conserved feature of eukaryotic cells. This conservation is exemplified by the family of genes encoding high molecular weight protein kinases: ATM (mammals), ATR (mammals), MEC1 (Saccharomyces cerevisiae), TEL1 (S.cerevisiae), rad3+ (Schizosaccharomyces pombe) (2,3). Each of these gene products falls into two family groups based on homology; ATM is homologous to Tel1, while ATR is more related to Mec1 and Rad3. This homology is not restricted to the kinase domain at the C-terminus, but rather extends over the length of the protein. The C-terminal kinase domain is structurally related to the catalytic domain of the PI-3 kinases. Despite this similarity, none of these proteins has been shown to phosphorylate lipids. Instead, all these proteins are capable of phosphorylating protein substrates (4–8).
In the budding yeast S.cerevisiae, MEC1 plays a critical role in checkpoint control (1,9), whereas TEL1 shares an overlapping role with MEC1 in maintaining survival after DNA damage (10,11). Mec1 physically interacts with Lcd1 (also called Ddc2 and Pie1), a protein that exhibits limited homology to the S.pombe Rad26 and mammalian ATRIP proteins (8,12–15). DNA damage responses have been well characterized in budding yeast, and consists of the G1-, S- and G2/M-phase damage checkpoints (9). Mec1 and Lcd1 function as a complex that is essential for all three DNA damage checkpoints as well as the DNA replication block checkpoint. Similar to Mec1, Rad53 plays an essential role in both the DNA damage and replication block checkpoints (1,9). RAD53 encodes a protein kinase and functions not only downstream of MEC1 but also downstream of TEL1 (11,16). Following DNA damage and replication block, the Rad53 protein is hyperphosphorylated and activated by a mechanism dependent on Mec1. Consistent with their upstream role, both Mec1 and Lcd1 are recruited to sites of DNA damage, suggesting that the Mec1–Lcd1 complex interacts with aberrant DNA structures or the DNA repair apparatus after DNA damage (17–19). Thus, the Mec1–Lcd1 complex appears to recognize DNA damage and stalled DNA replication, and transduces a checkpoint signal to Rad53.
Additional DNA damage checkpoint genes in budding yeast are DDC1, MEC3, RAD9, RAD17 and RAD24. Rad9 is hyperphosphorylated following DNA damage in a MEC1- and TEL1-dependent manner, and the phosphorylated Rad9 protein binds to Rad53 and modulates its kinase activity (20,21). Genetic evidence has suggested that RAD17, RAD24, MEC3 and DDC1 operate in a common checkpoint pathway. Indeed, Ddc1, Mec3 and Rad17 interact physically with each other and function in a complex to control the DNA damage checkpoints (22). It has been proposed that Ddc1, Mec3 and Rad17 are structurally related to PCNA (23). RAD24 encodes a protein structurally related to the subunits of replication factor C which consists of the one large subunit Rfc1 and the four small subunits Rfc2, Rfc3, Rfc4 and Rfc5 (24,25). Rad24 regulates the recruitment of the Ddc1–Mec3–Rad17 complex to sites of DNA damage (17,18).
In mammals, the ATM gene has an established role in controlling the cellular responses to double-strand breaks (DSBs) (3). ATM is mutated in patients with the genetic disorder ataxia-telangiectasia (AT), and mice carrying mutations in ATM exhibit phenotypes similar to those of AT patients. AT cells exhibit chromosomal instability and defects in the cellular response to ionizing radiation at the G1, S and G2 phases. Ionizing radiation causes various types of DNA damage, among which DSBs are the most lethal. In addition, AT cells are sensitive to the anticancer drug bleomycin, which causes DSBs (26,27). To activate the checkpoint response, ATM collaborates with the Mre11–Rad50–Nbs11 complex which has been implicated in cellular responses to DSBs. Cells defective in the Nbs1 and Mre11 function shows phenotypes similar to AT cells (28). Moreover, ATM phosphorylates Nbs1 to activate the S-phase checkpoint response (29–32). Thus, ATM and the Mre11–Rad50–Nbs1 complex appear to constitute a DNA damage response pathway. In budding yeast, the Mre11–Rad50–Xrs2 complex, which is related to the mammalian Mre11–Rad50–Nbs1 complex, plays an essential role in DSB processing (33). Several lines of evidence have suggested that Tel1 and the Mre11–Rad50–Xrs2 complex also constitute a DNA damage response pathway. Cells carrying mre11Δ, rad50Δ or xrs2Δ mutations are sensitive to DNA damage, and are defective in checkpoint responses specifically to DSBs (34). Although the tel1Δ mutation alone does not confer sensitivity to DNA damage, it does enhance sensitivity of mec1 mutants (10,11). Interestingly, mre11Δ and xrs2Δ mutations exacerbate the sensitivity of mec1 mutants to similar extents as a tel1Δ mutation (35,36). Furthermore, Mre11 and Xrs2 are phosphorylated after DNA damage in a TEL1-dependent manner (35,36). However, whether tel1Δ mutations confer a defect in checkpoint responses remains to be fully addressed.
The observation that ATM is specifically required for cellular responses to DSBs raises a possibility that Tel1 might have a similar role in the checkpoint responses following treatment with DSB-inducing agents. Phleomycin, like bleomycin, creates DSBs in DNA, but causes more significant DNA lesions than bleomycin in budding yeast (37,38). In this study, we investigated whether TEL1 is involved in the checkpoint responses to phleomycin-induced DNA damage. We show that tel1Δ mutants are defective in delaying cell-cycle progression after phleomycin treatment in S phase. Correspondingly, TEL1 is required for the phleomycin-induced Rad53 phosphorylation in S phase. In contrast, MEC1 is required for the checkpoint responses to phleomycin treatment in G1, S and G2/M phases. Consistent with the previous finding that tel1Δ mutants are not defective in the S-phase regulation following methyl methanesulfonate (MMS) treatment, Rad53 is normally phosphorylated in tel1Δ mutants after MMS treatment during S phase. Phleomycin treatment induces Xrs2 phosphorylation in a TEL1-dependent manner during S phase, whereas MMS treatment does not cause significant Xrs2 phosphorylation. These studies provide evidence that both Mec1 and Tel1 control the checkpoint responses to phleomycin-induced DNA damage.
Yeast strains used in this study are isogenic and listed in Table Table1.1. The mec1-81 mutation was isolated by a synthetic lethal screen with a rnr1 mutation (T.Shimomura and K.Sugimoto, manuscript in preparation). Standard genetic techniques were used for manipulating yeast strains. Synthetic complete medium containing 0.5% casamino acids and the appropriate supplements was used to maintain selection of URA3 or TRP1 plasmids. All the culture was grown at 30°C.
To construct the N-terminal HA-tagged TEL1, the 5′-noncoding and N-terminal region of TEL1 gene were amplified by polymerase chain reaction (PCR) with the 5′-noncoding HA-TEL1 primers KS152: 5′-TACGCGTAATCTAC-3′ and KS007: 5′-TAATACGACTCACTATAGGGCGA-3′ or the N-terminal HA-TEL1 primers KS153: 5′-TACGCGTATCCTTATGACGTACCAGATTATGCGGAGGATCATGGGATTGTAGAAAC-3′ and KS154: 5′-CTCGT CGACTCATGTGAGCTGTTCTCC-3′. YEp-TEL1-HA plasmid was constructed by a three-part ligation of the SacI–MluI treated 5′-noncoding fragment and the RsrII–MluI treated N-terminal fragment with the SacI–RsrII-linearized YEp-TEL1 (pDM198) (10). These PCRs add the same HA epitope to Tel1 at the same amino acid position from the initiation methionine as Mec1-HA (8). To construct the kinase-negative version of TEL1-HA (tel1-KN-HA), in vitro mutagenesis was performed substituting aspartic acid to alanine at 2612 and asparagine to serine at 2616 by PCR using YEp-TEL1 with the oligonucleotide primers KS807: 5′-GCAGCTAGCTCTAT CGTTGGATACATATTAGGCCTCGGTGCTAGGCACTT AAGCAATATC-3′ and KS680: 5′-AGAGTCGACTTTATT CCATTAGACAACTTT-3′. The NheI–SalI fragment of YEp-TEL1-HA was replaced by a 2.1 kb NheI–SalI fragment from the PCR product, generating YEp-TEL1-KN-HA. The BstXI–PvuII fragment from YEp-TEL1-HA, which contains the N-terminal part of Tel1-HA, was cloned into the BstXI–SmaI sites of pRS306 (39), generating YIp-TEL1-HA. To construct the TEL1-HA strain, YIp-TEL1-HA was cleaved with RsrII and transformed into cells. The precise integration, which destroys the endogenous TEL1 gene, was confirmed by PCR. The tagged constructs (TEL1-HA and tel1-KN-HA) expressed appropriate-sized proteins from their own promoter. The TEL1-HA construct suppressed the mec1Δ tel1Δ double mutants similar to the wild-type TEL1 gene whereas the tel1-KN-HA construct did not. To construct the C-terminal HA-tagged XRS2, the 5′-noncoding and N-terminal region of XRS2 gene were amplified by PCR with the primers KS759: 5′-ATCGGTACCAGAGGACACC AAAGTAAATTA-3′ and KS760: 5′-ATCGGATCCCTT TTCTTCTTTTGAACGTAAACT-3′. The KpnI–BamHI- treated PCR fragment and a BamHI–SalI fragment containing sequence encoding HA epitopes were cloned into the KpnI–SalI-linearized YCplac22 (40), generating YCpT-XRS2-HA. The tagged XRS2 construct expressed appropriate- sized proteins from its own promoter and complemented the xrs2 null mutation with respect to sensitivity to DNA damage. YCp-RAD53-HA, pGST-Rad53C, YEp-MEC1-HA and YEp-MEC1-KN-HA were described previously (8). The tel1Δ::KanMX strain was constructed using pFA6a-kanMX4 as described (41).
To determine colony formation ability in the presence of phleomycin, cell cultures were serially diluted, spotted on YEPD plates with or without 5 µg/ml phleomycin (Sigma). To determine cell viability in high concentrations of phleomycin, yeast cells were incubated with 50 µg/ml phleomycin. At timed intervals, cells were withdrawn and spread on YEPD medium. After incubation for 3 days, the number of colonies was counted.
Protein extracts for immunoblotting were prepared and resolved by electrophoresis on SDS–polyacrylamide gels as previously described (8). To examine phosphorylation of Rad53 and Xrs2 in S phase, cells carrying YCp-RAD53-HA and YCpT-XRS2-HA, respectively, were arrested with 6 µg/ml α-factor for 120 min to synchronize cells in G1. Cells were then washed to remove α-factor and released into fresh YEPD or YEPD containing 25 µg/ml phleomycin or 0.1% MMS. To examine Rad53 phosphorylation in G1 or G2/M phase, cells carrying YCp-RAD53-HA were arrested with 6 µg/ml α-factor or 15 µg/ml nocodazole for 120 min in G1 or G2/M, respectively. Cells were then treated with or without 50 µg/ml phleomycin maintaining the cell-cycle arrest.
To examine regulation of the S phase, log-phase cultures were arrested with 6 µg/ml α-factor for 120 min to synchronize cells in G1 (25). Cells were then washed to remove α-factor and released into fresh YEPD or YEPD containing 25 µg/ml phleomycin. To examine regulation of the G1-phase progression, cells were similarly arrested at G1 and incubated with 50 µg/ml phleomycin for 60 min. Cells were then washed to remove α-factor and phleomycin and released into fresh YEPD. Cells were withdrawn at different times and subjected to DNA flow cytometry analysis (42). To analyze the cell-cycle delay in G2/M, log-phase cultures were arrested with 15 µg/ml nocodazole for 120 min to synchronize cells in G2/M. Cells were treated with 50 µg/ml phleomycin for 60 min after arrest, and then washed to remove nocodazole and phleomycin and released into fresh YEPD. At timed intervals, cells were withdrawn and stained with 4′,6-diamidino-2-phenylindole (DAPI) for microscopic examination (8).
Immunoprecipitation was performed as described previously (8). For the kinase assay, immunoprecipitates were separated into equal portions for immunoblotting and kinase reaction. Kinase assay was initiated in 40 µl of kinase buffer by the addition of 10 mCi [γ-32P]ATP (3000 Ci/mmol, Amersham Pharmacia Biotech), substrate (2 µg of GST-Rad53C and 0.6 µg of PHAS-1) and ATP to 100 µM. Reactions were terminated by addition of 5× sample buffer and boiling for 5 min. The eluted proteins were separated by SDS–PAGE and gels were dried and autoradiographed.
We investigated how Mec1 and Tel1 control cellular responses to treatment with the DSB-inducing agent phleomycin. Rad53 is modified by phosphorylation in response to DNA damage or replication block, and its phosphorylation correlates with activation of the checkpoint pathways (11,16). We first monitored Rad53 phosphorylation after phleomycin treatment in wild-type, mec1Δ and tel1Δ mutant cells. Because sml1 mutations suppress the lethality of mec1Δ but not the checkpoints (43), all the genetic experiments were performed hereafter in an sml1Δ background. Cells expressing HA-tagged Rad53 were incubated with 5 or 50 µg/ml phleomycin for 120 min, and Rad53 modification was examined by immunoblotting analysis (Fig. (Fig.1).1). After exposure to phleomycin, Rad53 modification was detected in wild-type and tel1Δ cells in a dosage-dependent manner. In mec1Δ cells, Rad53 modification was observed after treatment with 50 µg/ml phleomycin, but not 5 µg/ml phleomycin. The expression level of HA-tagged Rad53 was not altered in wild-type, mec1Δ or tel1Δ mutant cells (data not shown). These results suggest that mec1Δ mutants retain a residual activity to phosphorylate Rad53 in response to phleomycin treatment, and that TEL1 might be involved in activation of the checkpoint pathways after phleomycin treatment.
In budding yeast, DNA damage activates the checkpoint pathways in G1, S and G2/M phases. We first investigated the roles of Mec1 and Tel1 in the S-phase checkpoint response to phleomycin treatment. The S-phase checkpoint was analyzed by monitoring the DNA content of phleomycin-treated cells after release from the G1 block (Fig. (Fig.2A).2A). When released from α-factor arrest and treated with 25 µg/ml phleomycin, the S-phase progression was delayed in wild-type cells as evidenced by lower rates of DNA synthesis. Compared with wild-type cells, mec1Δ mutants were slow in the S-phase progression in the absence of phleomycin, suggesting that the mec1Δ defect is not completely suppressed by the sml1Δ mutation. Phleomycin treatment further delayed the S-phase progression of mec1Δ mutants (see the cytometry profiles at 40 min after release). However, mec1Δ cells progressed through S phase faster than wild-type cells in the presence of phleomycin. Similarly, tel1Δ mutants delayed the S-phase progression, but progressed through S phase faster than wild-type cells in the presence of phleomycin. These results show that both mec1Δ and tel1Δ mutants are partially defective in the S-phase checkpoint after phleomycin treatment.
We next compared induction of Rad53 phosphorylation by phleomycin treatment in wild-type, mec1Δ and tel1Δ cells. Cells expressing Rad53-HA were synchronized by α-factor and released into medium containing 25 µg/ml phleomycin (Fig. (Fig.2B).2B). We found that Rad53 became phosphorylated within 30 min in wild-type cells. However, in mec1Δ mutants, no Rad53 phosphorylation was detected at 30 min after release. Although Rad53 phosphorylation became apparent in mec1Δ mutants at the later time points, its phosphorylation was clearly decreased as evidenced by the appearance of more weakly smeared bands in mec1Δ cells than wild-type cells. In tel1Δ cells, Rad53 phosphorylation was not detected at 30 min after release, but became apparent at 50 min after release at levels much greater than in mec1Δ cells. The phosphorylation observed in tel1Δ cells may result from the progression of cells into G2/M phase, because tel1Δ cells are proficient in Rad53 phosphorylation in response to phleomycin treatment in G2/M phase whereas mec1Δ cells are not (see below). These results indicate that both MEC1 and TEL1 are required for the S-phase checkpoint following phleomycin treatment.
We further investigated the additive effect of the mec1 and tel1 mutations on the cell-cycle progression. For this purpose, we used a weak mec1 (mec1-81) allele in combination with the tel1Δ mutation, because the mec1Δ tel1Δ double mutation confers a senescent phenotype resulting in a growth defect (7,44). In contrast to mec1Δ tel1Δ double mutants, mec1-81 tel1Δ double mutants grow as well as wild-type cells (see Fig. Fig.7).7). We found that mec1-81 single and mec1-81 tel1Δ double mutants were defective in the S-phase checkpoint; these mutants progressed faster than wild-type cells in the presence of phleomycin (Fig. (Fig.2A).2A). However, it was difficult to determine the additive defect in the cell-cycle delay. Similar to mec1Δ mutants, these mec1-81 mutants were slow in the S-phase progression in the absence of phleomycin, whereas tel1Δ mutants were not. We then examined Rad53 phosphorylation in mec1-81 and mec1-81 tel1Δ double mutants. While residual Rad53 phosphorylation was observed in both mec1-81 and tel1Δ cells, no phosphorylation was detectable in mec1-81 tel1Δ double mutants. Together, these results indicate that Mec1 and Tel1 act in parallel to activate the S-phase checkpoint responses following phleomycin treatment.
We next investigated the G1-phase checkpoint of mec1 and tel1 mutants cells after phleomycin treatment. We first examined the delay in cell-cycle progression at the G1- to S-phase transition after DNA damage by monitoring cellular DNA content (Fig. (Fig.3A).3A). Phleomycin treatment induced a G1-phase arrest in wild-type cells and the arrest continued for 120 min. Although mec1Δ cells exhibited delayed progression into S phase, these cells apparently progressed from G1 phase much faster than wild-type cells. We found that tel1Δ mutants were a little defective in delaying progression into S phase. We also examined the cell-cycle progression in mec1-81 and mec1-81 tel1Δ double mutants, and found that these mutants were defective in the G1 checkpoint. However, it was not determined whether the introduction of the tel1Δ mutation enhances the defect of the mec1-81 mutation. As discussed above, mec1-81 and mec1-81 tel1Δ mutants were slow in cell-cycle progression through S phase in the absence of phleomycin, whereas tel1Δ mutants were not.
We then monitored phleomycin-induced Rad53 phosphorylation in G1-arrested cells (Fig. (Fig.3B).3B). Rad53 phosphorylation became visible in the G1-arrested mec1Δ cells, but its phosphorylation was significantly decreased compared with the wild-type cells. Although Rad53 phosphorylation was detectable in tel1Δ mutants as in wild-type cells at 120 min after phleomycin treatment, it was clearly decreased in tel1Δ mutants at 60 min after treatment. We also examined the Rad53 phosphorylation in mec1-81 and mec1-81 tel1Δ double mutants. Similar to mec1Δ cells, the phleomycin-induced Rad53 phosphorylation was detectable, but was significantly decreased in mec1-81 mutants arrested in G1. No phosphorylation was observed in the G1-arrested mec1-81 tel1Δ double mutants. Thus, as observed for the S-phase checkpoint, Mec1 and Tel1 affect Rad53 modification and cell-cycle delay in a parallel manner in G1 phase, but the role of Tel1 in delaying cell-cycle progression seems not to be pronounced in G1 phase.
We further investigated the G2/M-phase DNA damage checkpoint in mec1 and tel1 mutant cells after phleomycin treatment. We examined the cell-cycle arrest of mec1Δ and tel1Δ cells following DNA damage by monitoring mitotic division (Fig. (Fig.4A).4A). When cell cultures were released from nocodazole arrest after phleomycin treatment, wild-type cells showed delayed nuclear division, whereas mec1Δ proceeded through mitosis faster than wild-type cells. In contrast to mec1Δ mutants, tel1Δ mutants underwent mitosis at the same rate as wild-type cells. We also monitored Rad53 phosphorylation status in mec1Δ and tel1Δ mutant cells arrested with nocodazole in G2/M (Fig. (Fig.4B).4B). Upon exposure to phleomycin, Rad53 became highly phosphorylated in tel1Δ mutants as well as wild-type cells. In contrast, no apparent phosphorylation was detected in mec1Δ mutants. These results indicate that MEC1 is essential for the G2/M-phase response to phleomycin-induced DNA damage, whereas TEL1 contributes little to this checkpoint.
The above results indicate that Mec1 and Tel1 are both required for the S-phase checkpoint response after phleomycin treatment. However, it has been demonstrated that mec1Δ mutants are defective in the S-phase checkpoint induced by MMS treatment, whereas tel1Δ mutants are not (45,46). MMS is an alkylating agent and generates methylated purines (N3-methyladenine, N7-methylguanine and O6-methylguanine) in DNA (47). Thus, MMS produces primary DNA lesions that are different from those produced by phleomycin, although MMS could create DSBs secondarily at high concentrations. To further assess the function of Mec1 and Tel1 in the MMS-induced S-phase checkpoint response, we monitored Rad53 phosphorylation in mec1Δ and tel1Δ mutant cells (Fig. (Fig.5).5). Cells were arrested with α-factor, and then released from α-factor into medium containing 0.1% MMS. To monitor Rad53 phosphorylation, we here used a higher concentration of MMS than previously used (45,46). Under these conditions, all the cells remained in S phase for 60 min after release from the G1 arrest (Fig. (Fig.5).5). However, consistent with the previous findings (45,46), mec1Δ mutants underwent S-phase progression a little faster than wild-type and tel1Δ mutant cells. MMS treatment induced Rad53 phosphorylation in tel1Δ mutants as observed in wild-type cells. In contrast, no apparent phosphorylation was detectable in mec1Δ cells after the same MMS treatment. Thus, Mec1 and Tel1 respond differently to DNA damage induced by phleomycin and MMS in S phase.
Mre11, Rad50 and Xrs2 form a complex which is required for DSB repair in budding yeast (33). The Mre11–Rad50–Xrs2 complex appears to be targeted by Tel1 after DNA damage (35,36). Mre11 and Xrs2 are phosphorylated after DNA damage in a TEL1-dependent manner, and their phosphorylation is detectable as slowly migrating forms by immunoblotting analysis (35,36). Because Tel1 is required for the S-phase checkpoint response after phleomycin treatment but not MMS treatment, we examined the phosphorylation status of Xrs2 during S phase in mec1Δ and tel1Δ mutants. Cells expressing HA-tagged Xrs2 were arrested with α-factor, and then released from α-factor into medium containing 25 µg/ml phleomycin or 0.1% MMS as described above (see Figs Figs22 and and5).5). Xrs2 phosphorylation was observed after phleomycin treatment in wild-type cells, whereas no significant phosphorylation was detected after MMS treatment (Fig. (Fig.6).6). The phleomycin-induced Xrs2 phosphorylation was largely dependent on Tel1, because its phosphorylation was abolished by the introduction of the tel1Δ mutation but not the mec1Δ mutation (Fig. (Fig.6).6). These results are consistent with the model in which Tel1 is required for the S-phase checkpoint response after phleomycin treatment.
Because both mec1 and tel1 mutants are defective in phleomycin-induced checkpoints, we examined whether the mec1 and tel1 mutations confer sensitivity to phleomycin. Cells carrying mec1Δ and mec1-81 mutations were hypersensitive to phleomycin, whereas tel1Δ mutants were no more sensitive than wild-type cells (Fig. (Fig.7).7). The tel1Δ mutation had no effect on cell viability after short exposure to high concentrations of phleomycin (data not shown). However, introduction of the tel1Δ mutation enhanced sensitivity to phleomycin in mec1-81 mutant cells (Fig. (Fig.7).7). These results indicate that MEC1 has an essential role in phleomycin-induced DNA damage repair and that TEL1 has an overlapping role with MEC1.
Mec1 and Tel1 play a redundant role in activation of the checkpoint pathway, but Mec1 plays a more significant role than Tel1. One explanation could be that Mec1 and Tel1 have different substrate preference. It has been shown that both Mec1 and Tel1 phosphorylate PHAS-1 in vitro (7), and that Mec1 phosphorylates the C-terminal half of Rad53 in vitro (8). We first asked whether Tel1 phosphorylates the C-terminal half of Rad53 in vitro as Mec1 does. For this purpose, we generated an HA-tagged TEL1 construct (TEL1-HA). Tel1 contains the motif DXXXXN at positions 2612–2616, which plays a critical role in protein kinase catalysis. We mutated this conserved motif to AXXXXS and constructed a kinase-negative version of TEL1-HA (tel1-KN-HA). Rad53 has 16 SQ/TQ motifs; eight of them are located in the N-terminus including the kinase domain, and eight are in the C-terminus. To examine Rad53 phosphorylation, GST fusion proteins with the C-terminal half of Rad53 (GST-Rad53C) were expressed and purified from bacteria (8). PHAS-1 also contains SQ/TQ motifs (7). Extracts were prepared from cells expressing Tel1-HA, Tel1-KN-HA or no HA-tagged protein, and were immunoprecipitated with anti-HA antibodies. Immuno precipitates were then subjected to an in vitro kinase assay. GST-Rad53C and PHAS-1 were phosphorylated by the immunoprecipitates containing Tel1-HA but not by the immunoprecipitates containing Tel1-KN-HA or no HA-tagged protein (Fig. (Fig.8A).8A). No specific phosphorylation of GST alone was observed with Tel1-HA (data not shown). Consistent with the previous findings (7,8), Mec1 phosphorylated GST-Rad53C and PHAS-1 (Fig. (Fig.8A).8A). Our analysis, however, does not exclude the possibility that Mec1 and Tel1 phosphorylate others than SQ/TQ motifs in GST-Rad53C and PHAS-1.
We then performed the Mec1 and Tel1 kinase assay in the presence of the same amounts of PHAS-1 and GST-Rad53C substrates. In these reactions, Mec1 phosphorylated GST-Rad53C more efficiently than Tel1 did, whereas Tel1 phosphorylated PHAS-1 more efficiently than Mec1 did (Fig. (Fig.8B).8B). Both Tel1 and Mec1 were fused to the same HA tag at their N-termini (see Materials and Methods) and the amounts of each protein in the immunoprecipitates were estimated to be similar by immunoblotting analysis (Fig. (Fig.8B).8B). Together, these results suggest that Mec1 and Tel1 have different substrate preferences in vitro.
In mammalian cells, both ATM and ATR have been implicated in the response to DSB-inducing ionizing radiation. Both ATR-related and ATM-related proteins are also found in budding yeast, encoded by MEC1 and TEL1, respectively. Genetic and biochemical studies have demonstrated that Mec1 plays a central role in the DNA damage checkpoint, but the function of Tel1 in checkpoints has not been well characterized. Since ATM is specifically required for checkpoint responses to DSBs, we hypothesized that Tel1 might have a similar role. In this report, we investigated the roles of Mec1 and Tel1 in the checkpoint responses to phleomycin, an agent that causes DSBs in DNA. Phleomycin, like other DNA-damaging agents, induces Rad53 phosphorylation in cells. MEC1 is required for Rad53 phosphorylation and cell-cycle arrest following phleomycin treatment in all the G1, S and G2/M phases. In contrast, TEL1 is required for Rad53 phosphorylation and cell-cycle arrest in S phase. TEL1 is as well partially required for Rad53 phosphorylation and cell-cycle arrest in G1 phase, but not in G2/M phase. Consistent with the role of Tel1 in the S-phase regulation, Xrs2 is phosphorylated after phleomycin treatment in a TEL1-dependent manner in S phase. Moreover, mec1-81 tel1Δ double mutants are more defective than either single mutants in Rad53 phosphorylation in S phase. Finally, both Mec1 and Tel1 phosphorylate the Rad53 and PHAS-1 proteins in vitro. Taken together, these results strongly suggest that both Mec1 and Tel1 control checkpoint responses to phleomycin-induced DNA damage.
Similar to phleomycin treatment, MMS treatment activates checkpoints and induces Rad53 phosphorylation in S phase. However, mec1 and tel1 mutants behave differently in response to MMS treatment; mec1 mutants are defective in the S-phase checkpoint after MMS treatment whereas tel1Δ mutants are not (45,46). Correspondingly, MMS treatment induces Rad53 phosphorylation in tel1Δ mutants as observed in wild-type cells, but the phosphorylation is not detected in mec1Δ mutants. Together, these results support a model in which Mec1 and Tel1 regulate pathways in parallel and are activated by different types of DNA damage.
Cells carrying the tel1Δ mutation are defective in delaying cell-cycle progression in S phase when DNA is damaged by phleomycin. In contrast, tel1Δ mutants are only weakly defective in the G1-phase responses and proficient in the G2/M-phase responses following phleomycin treatment. Thus, the checkpoint defect in tel1Δ mutants upon the occurrence of DSBs appears to be mainly caused by the inability to cope with DSBs in S phase. One possible explanation was that the checkpoint defect observed in tel1Δ mutants could be attributed, at least in part, to the cell-cycle-dependent expression of Mec1 and/or Tel1. However, the expression levels of both Mec1 and Tel1 are not altered during the cell cycle (data not shown). Alternatively, DSBs might be more improperly processed during S phase than G1 or G2/M phase, resulting in activation of a TEL1-dependent checkpoint response. Consistent with this possibility, Usui et al. (36) have demonstrated that sae2 and rad50s mutations, which cause a defect in DSB processing, enhance the Tel1-mediated checkpoint response.
Rad53 is phosphorylated in response to DNA damage, and its phosphorylation is correlated with activation of the Rad53 kinase. We showed that phleomycin treatment induces Rad53 phosphorylation in a MEC1- and TEL1-dependent manner. Recent evidence has suggested that Rad53 phosphorylation following DNA damage might result entirely from increased autophosphorylation activity (20). However, DNA damage induces hyperphosphorylation of Rad53 mutant proteins which possess little autophosphorylation activity (48). Thus, after DNA damage Rad53 becomes phosphorylated in trans in a Mec1- and Tel1-dependent manner. Consistently, both Mec1 and Tel1 were found to phosphorylate the Rad53 protein in vitro.
It has been suggested that DNA damage accumulates more in checkpoint defective mutants than wild-type cells. Although tel1Δ mutants are defective in the S-phase checkpoint, they do not show hypersensitivity to phleomycin. One explanation could be that a functional G2/M checkpoint compensates for defects in S phase. Perhaps if the G2/M checkpoint is fully functional and activated, DNA damage might be repaired sufficiently to allow cells to remain viable. Indeed, similar phenotypes have been observed in several mec1 mutants (49); these mutants are as defective as mec1Δ in both the G1- and S-phase checkpoints following treatment with UV light and MMS, although they do not show hypersensitivity to UV or MMS. Alternatively, it is possible that in tel1Δ mutants DNA damage accumulates but might be processed efficiently by the MEC1-dependent repair pathway. Supporting this, the tel1Δ mutation was found to enhance sensitivity of mec1 mutants to phleomycin.
The human genes mutated in AT and Nijmegen breakage syndrome, ATM and NBS1, respectively, are both involved in the cellular response to DSBs (2,3). The Nbs1 protein forms a complex with the human Mre11 and Rad50 proteins, which are implicated in DSB repair (50). Mutations in the MRE11 gene were also found in individuals with another AT-like disorder (51). Moreover, ATM directly phosphorylates Nbs1 on several sites required for its checkpoint function in vivo (30–32). These results have demonstrated a functional and biochemical link between ATM and the Mre11–Rad50–Nbs1 complex in human cells. The budding yeast Mre11– Rad50–Xrs2 complex has an established role in both non-homologous end-joining and homologous recombination repair of DSBs (33). Recently, several groups showed that the Mre11–Rad50–Xrs2 complex is required for checkpoint responses after treatment of DSB-inducing agents (34–36). It has been also shown that mre11Δ and xrs2Δ mutations enhance the DNA-damage sensitivity of mec1 mutants to similar extents as a tel1Δ mutation (35,36). Moreover, Mre11 and Xrs2 are phosphorylated after DNA damage in a TEL1-dependent manner (35,36). These findings have suggested the model in which Tel1 and the Mre11–Rad50–Xrs2 complex constitute a DNA damage response pathway (35,36). We have shown that Tel1 is required for Xrs2 phosphorylation as well as checkpoint responses after phleomycin treatment in S phase. Our results extend the notion that the Mre11–Rad50–Nbs1/Xrs2 complex and ATM family protein are structurally and functionally conserved in eukaryotes.
We thank Phil Hieter and Tomoko Ogawa for materials, Seiko Ando for plasmid construction, and Marc Lamphier for critical readings of the manuscript. This work was supported by Grant-in-Aid for Scientific Research on Priority Areas and General Research from the Ministry of Education, Science, Sports and Culture of Japan.