RecQ DNA Helicase Rqh1 Promotes Rad3^ATR Kinase Signaling in the DNA Replication Checkpoint Pathway of Fission Yeast

Rad3 is the orthologue of ATR and the sensor kinase of the DNA replication checkpoint in Schizosaccharomyces pombe. Under replication stress, it initiates checkpoint signaling at the forks necessary for maintaining genome stability and cell survival. To better understand the checkpoint initiation process, we have carried out a genetic screen in fission yeast by random mutation of the genome, looking for mutants defective in response to the replication stress induced by hydroxyurea.

ABSTRACT

Rad3 is the orthologue of ATR and the sensor kinase of the DNA replication checkpoint in Schizosaccharomyces pombe. Under replication stress, it initiates checkpoint signaling at the forks necessary for maintaining genome stability and cell survival. To better understand the checkpoint initiation process, we have carried out a genetic screen in fission yeast by random mutation of the genome, looking for mutants defective in response to the replication stress induced by hydroxyurea. In addition to the previously reported mutant with a C-to-Y change at position 307 encoded by tel2 (tel2-C307Y mutant) (Y.-J. Xu, S. Khan, A. C. Didier, M. Wozniak, et al., Mol Cell Biol 39:e00175-19, 2019, https://doi.org/10.1128/MCB.00175-19), this screen has identified six mutations in rqh1 encoding a RecQ DNA helicase. Surprisingly, these rqh1 mutations, except for a start codon mutation, are all in the helicase domain, indicating that the helicase activity of Rqh1 plays an important role in the replication checkpoint. In support of this notion, integration of two helicase-inactive mutations or deletion of rqh1 generated a similar Rad3 signaling defect, and heterologous expression of human RECQ1, BLM, and RECQ4 restored the Rad3 signaling and partially rescued a rqh1 helicase mutant. Therefore, the replication checkpoint function of Rqh1 is highly conserved, and mutations in the helicase domain of these human enzymes may cause the checkpoint defect and contribute to the cancer predisposition syndromes.

INTRODUCTION

Multiple cellular mechanisms ensure the accurate transmission of genetic material from one generation to the next. These include proper repair of DNA damage, protection of perturbed DNA replication forks, and the checkpoints that coordinate DNA replication and repair with cell cycle progression. Chromosomes are particularly vulnerable to DNA damage during replication as they are decondensed and single stranded and highly accessible to various endogenous and exogenous damaging agents. In addition, other factors such as excessive trinucleotide repeats or insufficient supply of deoxynucleoside triphosphates (dNTPs) can also perturb DNA replication. Therefore, DNA replication is subject to exquisite regulations. One such regulation mechanism is the DNA replication checkpoint (DRC). The DRC senses the problems and activates cellular responses such as increased production of dNTPs, cell cycle delay, fork protection, and suppression of late-firing origins, which work in concert to minimize the mutation rate and ensure accurate and complete duplication of the genome before cell division. Consistent with its importance in genome stability, the DRC is conserved from yeasts to humans, and defects in the cell signaling pathway cause a wide range of developmental and cancer predisposition syndromes. Although the suggestion is debatable, mutations generated by errors in DNA replication followed by mistakes in repair likely contribute significantly to sporadic cancers (for reviews, see references 1 to 3).

The current model envisages that a related set of sensor proteins in all eukaryotes assemble at perturbed forks for the DRC signaling. Among the sensors, ATR is the protein kinase that works with its cofactor ATR-interacting protein (ATRIP) and the Rad9-Rad1-Hus1 (9-1-1) complex to initiate checkpoint signaling by phosphorylating various substrates, including the mediators and effector kinase (4–6). The activated mediators channel the signal to the effector kinase (7, 8). Once activated, the effector kinase diffuses away from the fork and relays the checkpoint signal to various cellular structures to stimulate the cellular responses mentioned above. However, the exact mechanism by which the checkpoint signaling is initiated at the perturbed forks remains incompletely understood (9–11).

To better understand the checkpoint initiation process (9, 10), we have searched for new DRC mutants in Schizosaccharomyces pombe by random mutation of the genome, looking for mutants with enhanced sensitivity to hydroxyurea (HU). HU perturbs DNA replication by inhibiting ribonucleotide reductase (RNR), a highly conserved enzyme that provides dNTPs for DNA replication and repair (12). HU slows down the polymerase movement at forks, and, consistent with this mechanism, HU resistance has been observed in cells overexpressing RNR or expressing a mutant RNR (13–15). Studies in various eukaryotic organisms have shown that the primary cellular response to HU treatment is the activation of the DRC (2, 16). We have previously described one of the screened mutants with a C-to-Y change at position 307 encoded by tel2 (tel2-C307Y mutant) that is highly sensitive to the replication stress induced by HU or DNA damage (17). Here, we report a group of rqh1 mutants.

Rqh1 is a member of the RecQ DNA helicase family that is conserved from bacteria to humans. There are five RecQ helicases in humans: RECQ1, BLM, WRN, RECQ4, and RECQ5. Each of the five human helicases plays critical roles in genome stability, and mutations in these helicases are linked to cancers or heritable cancer-prone syndromes such as Bloom’s, Werner’s, and Rothmund-Thomoson syndromes (for reviews, see references 18 to 20). In budding yeast, Sgs1 functions in homologous recombination-mediated DNA double-strand break repair, meiosis, telomere maintenance, and at the replication fork (for reviews, see references 21 and 22). It also functions in checkpoint signaling during G₁ through end resection (23) or in S phase as a checkpoint adaptor (24–26). Although Rqh1 has similar functions at replication forks, meiosis, and DNA damage repair (27–31), whether it has a checkpoint function has been controversial in fission yeast (32–34). Our genetic screen has identified six mutations in rqh1 from a group of nine mutants that are sensitive to HU and DNA damage. Except for a start codon mutation that depletes Rqh1, the remaining five mutations are all in the helicase domain of Rqh1. One truncation mutant lacking most of the helicase domain behaves similarly to rqh1 null cells. The remaining four mutations have substitutions in the amino acids that are all highly conserved in RecQ helicases. Further characterization of these mutants shows that Rqh1 functions in the DRC by promoting Rad3 kinase signaling under replication stress. Unlike the checkpoint mediator function of Sgs1 in Saccharomyces cerevisiae (24, 25), the DRC function of Rqh1 is likely mediated by the helicase activity in S. pombe.

RESULTS

Screening of an hus9 mutant with enhanced sensitivities to HU and DNA damage.

The hus (HU sensitive) screen in S. pombe was first described by T. Enoch and her coworkers that identified hus1 (Mec3) encoding a component of the 9-1-1 complex (35). To identify additional DRC mutants, we have carried out a large-scale hus screen and accumulated >1,000 primary mutants. After three backcrosses to remove bystander mutations and crossings with DRC mutants to remove known mutations, a small set of new hus mutants was screened. These hus mutants were then crossed with one another to establish the complementation groups. We have reported previously the tel2-C307Y mutant identified by this screen (17). Here, we report our characterization of a complementation group of nine hus mutants. Among these mutants, we chose hus9 for detailed analysis because of its significant sensitivity to HU and DNA damage. Preliminary results showed that, unlike in the tel2-C307Y mutant in which phosphorylation of the DRC mediator Mrc1 (claspin) by Rad3 (8, 11) was eliminated, the phosphorylation of Mrc1 was significantly compromised in this mutant, suggesting a DRC defect. We therefore decided to investigate the hus9 mutant further as well as the other mutants in the same group.

We first examined the sensitivities of hus9 cells to HU and DNA damage by using a spot assay, described in Materials and Methods. In S. pombe, Rad3 is the master checkpoint kinase responsible for activation of both the DRC and the DNA damage checkpoint (DDC) pathways (36) (Fig. 1A). The ATM orthologue Tel1 contributes minimally to the checkpoints. Deletion of rad3 sensitizes S. pombe to both HU and DNA damage due to a lack of the checkpoint functions. As shown in the top panels of Fig. 1B, while S. pombe lacking rad3, mrc1, or cds1 was highly sensitive to HU, the cells lacking chk1, the effector kinase of the DDC, were less sensitive, suggesting that the replication stress induced by HU is dealt with mainly by the DRC. Under similar conditions, the hus9 mutant was found to be sensitive to HU, and the sensitivity was comparable to that of the mrc1 and cds1 mutants. We then tested the DNA-damaging agents methyl methanesulfonate (MMS) and UV (Fig. 1B, middle panels). Unlike results with the HU treatment, the chk1 mutant was more sensitive to MMS and UV than the cds1 mutant, indicating that the DNA damage that occurs at G₂, the major cell cycle stage in S. pombe, is mainly dealt with by the DDC. The hus9 mutant was also sensitive to MMS and UV, and the sensitivity was comparable to that of the rad3 mutant. We then examined the sensitivity of the hus9 mutant to bleomycin (Fig. 1B, bottom panels), an antibiotic that generates single-strand and double-strand breaks in chromosomal DNA (37). The hus9 mutant was sensitive to bleomycin with a sensitivity even slightly higher than that of the rad3 mutant.

FIG 1.

Sensitivity of the newly identified hus9 mutant to HU and DNA damage. (A) Rad3 kinase signaling in the DRC (left) and the DDC (right) pathways of S. pombe. The serine and threonine residues phosphorylated by Rad3 are indicated. In the presence of perturbed replication, Rad3 phosphorylates Mrc1 and Cds1 to activate the DRC (8, 11, 50). When DNA damage occurs outside S phase or when forks collapse, such as occurs in HU-treated mrc1 or cds1 cells, Rad3 phosphorylates Crb2 (53BP1/Rad9) and Chk1 to activate the DDC (53, 55, 56). Phosphorylation of Rad9 in the 9-1-1 complex is required for activation of both the DRC and DDC (11, 52). In the DDC pathway, phosphorylated Rad9 recruits Rad4 (TopBP1/Dpb11) to promote Crb2 and Chk1 activation. However, the function of phosphorylated Rad9 in promoting Cds1 activation remains unknown (dashed line). (B) Sensitivities of the hus9 mutant (NF198) to HU, MMS, UV, and bleomycin (Bleo) were examined by the spot assay described in Materials and Methods. The wild type (TK48) and rad3 (NR1826), mrc1 (YJ15), cds1 (GBY191), and chk1 (TK197) mutants were used as controls. (C) Sensitivity of the hus9 mutant to acute HU treatment in liquid medium. Cells used in the experiment shown in panel B were incubated in YE6S medium containing 15 mM HU. At the indicated time points, cells were spread onto YE6S plates to recover for 3 days. Colonies were counted, and results are presented as percentages. Error bars represent means and standard deviations of triplicates. (D) Overexpression of the RNR small subunit Suc22 rescues the hus9 mutant. Suc22 was expressed in S. pombe on a vector under the control of its own promoter. V, vector. (E) The hus mutant phenotype of hus9 cells is caused by a single missense mutation in rqh1. The identified rqh1-G804D mutation was integrated at the genomic locus in wild-type S. pombe (see Fig. S2 in the supplemental material for details). As a control, wild-type rqh1 was integrated by the same method. HU sensitivities of the representative integrants of the wild type (NF82) and the rqh1 (NF83) mutant were determined by spot assay. For comparison, rad3 and hus9 mutants were included. WT, wild type.

We have recently screened a set of hus mutants of various metabolic pathways. These metabolic mutants, such as erg11-1 and hem13-1 mutants, are highly sensitive to chronic exposure to HU, as determined by a spot assay (38, 39). However, they are not sensitive to DNA damage and the acute treatment of HU in liquid cultures. Upregulation of Suc22, the small subunit of RNR and the major regulatory target of the DRC in S. pombe, can rescue all DRC mutants but not the metabolic mutants. We then treated the hus9 mutant with HU in liquid culture and found that it was also sensitive and that the sensitivity was less than the levels for the mrc1 and rad3 mutants (Fig. 1C). The higher HU sensitivity observed in the spot assay was likely due to oxidative stress (see Discussion). We also tested Suc22 and found that its upregulation rescued the hus9 mutant, similar to the result with the rad3 mutant (Fig. 1D). Tetrad dissection showed that crosses of the hus9 mutant with all known DRC mutants generated spores with wild-type HU resistance (see Fig. S1 in the supplemental material), showing that the hus9 mutant is not allelic to the checkpoint mutants.

hus9 carries two missense mutations in rqh1.

To identify the mutated gene, we transformed the hus9 mutant with genomic DNA expression libraries marked with ura4. Colonies grown on medium lacking uracil were replicated onto plates containing HU to screen those with conferred HU resistance. Plasmids recovered from the yeast colonies were subjected to restriction enzyme digestion to remove those with suc22. Subsequent DNA sequencing identified rqh1 and two G-to-A missense mutations at the rqh1 genomic locus of hus9 which replace Gly⁷⁶⁷ and Gly⁸⁰⁴ with Asp in the helicase domain of Rqh1 (Fig. S2A). Expression of wild-type rqh1 on a vector fully rescued the hus9 strain while the mutant rqh1 did not (Fig. S2B, middle part). Separation of the two mutations showed that while rqh1 carrying the G767D mutation rescued hus9, rqh1 carrying G767D plus G804D or G804D alone did not. To confirm the mutations, we integrated them at the genomic locus in wild-type S. pombe by using the method diagrammed in Fig. S2C. As the control, wild-type rqh1 was integrated by the same method. The integrants were screened by colony PCRs, backcrossed once to ensure single-copy integration, and then confirmed by PCRs (Fig. S2D) and Western blotting (Fig. S2E). The HU sensitivity of the resulting integrants was assessed by spot assay (Fig. S2B, bottom part). The result showed that while the G767D integrant, like wild-type cells, was resistant to HU, the integrants of G804D or G767D G804D were sensitive, like the hus9 strain. We also deleted rqh1 by replacing the whole open reading frame (ORF) with ura4 and found that hus9 cells showed similar or slightly higher HU sensitivity (see below) than the rqh1 deletion cells (Fig. S2B, top part). Figure 1E is a representative of the repeated spot assay, showing that a single G804D mutation in rqh1 (rqh1-G804D) sensitizes S. pombe to HU and DNA damage, similar to observations in the hus9 mutant. We here renamed hus9 as rqh1-G804D in the following studies.

rqh1 was originally identified in S. pombe as hus2 (27, 35) and rad12 (32). Although studies in S. cerevisiae have shown that the Rqh1 homolog Sgs1 has checkpoint functions (24, 25), earlier studies in fission yeast showed that Rqh1 functions mainly in DNA repair, recombination, and fork recovery (27, 28, 31–33, 40, 41). Rqh1 is a member of the superfamily 2 (SF2) RecQ DNA helicase family that is conserved from bacteria to humans. There are five RecQ helicases in humans, RECQ1, BLM, WRN, RECQ4, and RECQ5, that play distinct and sometimes overlapping roles in maintaining genomic integrity through extensive involvement in DNA metabolism (18, 19, 42). Defects in BLM, WRN, or RECQ4 cause the monogenic Bloom’s, Werner’s, and Rothmund-Thomson syndromes that are characterized by cancer predisposition or premature aging. Recent studies have also associated mutations of RECQ1 and RECQ5 with a cancer predisposition or increased resistance to chemotherapies (43–47). All RecQ helicases share ATP-dependent 3′-to-5′ DNA unwinding activity. Although S. pombe possesses additional homologs (48, 49), Rqh1 is generally believed to be the only RecQ helicase that functions at perturbed forks, which promotes unambiguous studies. In budding yeast, Sgs1 functions in the DRC as a mediator that recruits Rad53 (CHK2/Cds1) to be activated by Mec1 (ATR/Rad3) (24, 25). Since our preliminary data suggest that Rqh1 promotes Rad3 phosphorylation of Mrc1, the mediator that recruits Cds1 for its activation by Rad3, we decided to further investigate rqh1-G804D and its potential effect on the DRC by using a set of phospho-specific antibodies that we made available in previous studies (8, 11, 50).

HU arrests rqh1-G804D cells in S phase.

As mentioned above, HU arrests the metabolic mutants at G₂/M, not S phase, as they are killed by HU via the mechanisms unrelated to replication stress (38, 39). We then examined the cell cycle progression by flow cytometry (Fig. 2A). In the presence of 15 mM HU, wild-type cells were increasingly arrested at S phase in 1 to 3 h. Further incubation did not completely stop DNA synthesis and eventually finished the bulk of DNA synthesis in ∼7 to 8 h. rad3 and mrc1 cells were also arrested in S phase after 3 h in HU. However, these checkpoint mutants, and particularly the rad3 mutant, could not properly synthesize DNA in HU. Under similar conditions, HU arrested the majority of rqh1-G804D cells at S phase although a small number of cells remained at G₂/M after 3 to 4 h of incubation. Similar to the rad3 and mrc1 mutants, the rqh1-G804D mutant showed a minor defect of the continued DNA synthesis in HU.

FIG 2.

The DRC defect in the rqh1-G804D mutant. (A) Cell cycle analysis of the rqh1-G804D mutant. Wild-type, rad3, mrc1, and rqh1-G804D cells used in the experiment shown in Fig. 1C were incubated with 15 mM HU and analyzed by flow cytometry every hour during the incubation. Red lines indicate 1C and 2C DNA contents. (B) rqh1-G804D cells were elongated and showed the cut phenotype in HU. Cells were treated with HU for 6 h as described for panel A, fixed onto glass slides by brief heating, and stained with Hoechst and Blankophor for microscopic examination. Arrows indicate the cut cells. (C) A small number of rqh1-G804D cells undergo cell division in HU. Cells were incubated with HU as described for panel A and fixed in 2.5% glutaraldehyde at the indicated time points. The fixed cells were washed and then stained as described for panel B. A total of ≥150 cells were counted under microscope for each sample, which was repeated three times. The quantity of cells with a septum is shown as a percentage of the total number of cells. (D) rqh1-G804D cells did not fully recover from HU arrest. The S. pombe strains used in the experiment shown in panel A were arrested in HU for 4 h and then released in fresh medium. Cell cycle progression was monitored every hour during the release. Asyn, asynchronous.

Defects in recovery from replication stress.

In the presence of HU, the DRC is activated to stimulate dNTP production and a delay in mitosis so that the cells can complete DNA synthesis before the division. The DRC mutants, however, proceed into mitosis, generating the “cut” (cell untimely torn) phenotype with two daughters having unequal amounts or lacking genomic DNAs. We stained the cells with Hoechst for genomic DNAs and Blankophor for septum and examined cell septation during the HU treatment (Fig. 2B and C). After wild-type cells were treated with HU for 3 h, cell division was almost completely suppressed and remained suppressed during the rest of HU treatment (Fig. 2C), suggesting an activated DRC. In contrast, the rad3 mutant showed robust cell division activity with HU treatment. As a result, most of the cells were short and showed the cut phenotype (Fig. 2B, arrows). The mrc1 mutant initially shown slowed cell division during the first 1 to 3 h. After 3 to 4 h, the cells began to divide, generating the cut cells (Fig. 2B and C). However, unlike the HU-treated rad3 cells that were short due to the lack of both the DRC and DDC, the mrc1 cells elongated because the collapsed forks activated the DDC (8, 51). The rqh1-G804D mutant also showed the cut cells with HU treatment, and the septation index was lower than that of mrc1 cells during the HU treatment (Fig. 2B and C), which is consistent with the milder HU sensitivity shown in Fig. 1C. Since the HU-treated rqh1-G804D cells, including the cut cells, were elongated, the DDC is likely activated as in mrc1 cells (see below). Consistent with our result, the elongated and cut cells have also been described in HU-treated S. pombe lacking rqh1 (27).

We then monitored cell recovery from HU treatment by flow cytometry (Fig. 2D). After 4 h of treatment with HU, the majority of wild-type and mutant cells were arrested at S phase. When released in fresh medium, wild-type cells fully recovered and returned to a normal cell cycle in ≤4 h. In contrast, rqh1-G804D cells, similar to rad3 and mrc1 cells, failed to fully recover. We also examined the cell recovery every 20 min after release (Fig. S3), and the results are consistent with those shown in Fig. 2D, suggesting a defect in DRC or cell recovery (27) from the S-phase arrest by HU.

Compromised Rad3 signaling in the DRC.

We next examined the potential DRC defect by monitoring Rad3 phosphorylation of Rad9, Mrc1, and Cds1 using phospho-specific antibodies (8, 11, 50). Under replication stress, Rad3 phosphorylates two functionally redundant residues, Thr⁶⁴⁵ and Thr⁶⁵³, in the Thr-Gln (TQ) motifs of Mrc1 and Thr⁴¹² in Rad9 of the 9-1-1 complex (Fig. 1A, left). Phosphorylation of Mrc1 and Rad9 facilitates the phosphorylation of Thr¹¹ in Cds1 (CHK2/Rad53) by Rad3, which promotes the autophosphorylation of Cds1-Thr³²⁸ and autoactivation of the effector kinase (8, 11, 50). When DNA damage occurs outside S phase or when forks collapse, such as in HU-treated mrc1 or cds1 cells, Rad3 phosphorylates Rad9 and Crb2 (53BP1/Rad9) (52–54), which in turn facilitates the phosphorylation of Chk1-Ser³⁴⁵ by Rad3 (55, 56), leading to the activation of the DDC (Fig. 1A, right).

We first examined Rad3-specific phosphorylation of Mrc1-Thr⁶⁴⁵, a representative of the two redundant sites in Mrc1 (Fig. 3A) (8). After 3 h of HU treatment, while the phosphorylation was undetectable in rad3Δ cells, it significantly increased in wild-type cells. Because the activated DRC upregulates Mrc1 (57, 58), the level of Mrc1 in HU-treated rad3 cells was lower than that in wild-type cells. Under similar conditions, Mrc1 phosphorylation in rqh1-G804D cells was reduced to 35% (±1.0%; n = 3) of the wild-type level (Fig. 3A and Fig. S4A). We then treated the cells with 0.01% MMS for 90 min, and a similar reduction of Mrc1 phosphorylation was observed in rqh1-G804D cells (Fig. 3B and Fig. S4B). To eliminate the cell cycle effect, we examined Mrc1 phosphorylation every hour during the HU treatment (Fig. S5A and B). While the phosphorylation was significantly increased in wild-type cells during the first 3 h, it eventually decreased during the remaining time of the treatment, which is consistent with the flow cytometry data shown in Fig. 2A. In rqh1-G804D cells, Mrc1 phosphorylation was significantly lower at all time points examined although it followed kinetics similar to those in wild-type cells.

FIG 3.

Compromised Rad3 kinase signaling in the DRC pathway. (A) HU-induced phosphorylation of Mrc1 by Rad3 was reduced in rqh1-G804D cells. Wild-type and mutant strains used in the experiment shown in Fig. 1B were treated with 15 mM HU for 3 h (+) or left untreated (−). Phosphorylation of Mrc1 (top panel) was detected in whole-cell lysates prepared by the TCA method using phospho-specific antibodies. The same blot was stripped and reprobed with anti-Mrc1 antibodies (middle panel). A section of the Ponceau S-stained membrane is shown (bottom panel). The phosphorylation bands were quantified, and band intensities compared to those in HU-treated wild-type cells are shown at the bottom. (B) MMS-induced phosphorylation of Mrc1 was also reduced in rqh1-G804D cells. The cells were treated with 0.01% MMS for 90 min and then lysed for Western blotting and subsequent quantification as described for panel A. (C) HU-induced phosphorylation of Rad9 by Rad3 was moderately affected in rqh1-G804D cells. Wild-type and the indicated mutant cells were treated with 15 mM HU for 3 h or left untreated. Rad9-HA was immunoprecipitated and separated by SDS-PAGE for Western blotting. The blot was first probed with anti-HA antibody (lower panel), stripped, and then reprobed with a phospho-specific antibody (upper panel). (D) Rad9 phosphorylation in the presence of MMS was also moderately affected in rqh1-G804D cells. The cells were treated with 0.01% MMS for 90 min and analyzed by Western blotting as described for panel C. (E) Phosphorylation of Cds1 by Rad3 was significantly reduced in HU-treated rqh1-G804D cells. The cells were treated with HU as described for panel A. Cds1-HA was immunoprecipitated and then analyzed by Western blotting using anti-HA antibody (bottom panel). The same membrane was stripped and then blotted with phospho-specific antibody (top two panels). (F) MMS-induced phosphorylation of Cds1 was also significantly reduced in rqh1-G804D cells. The cells were treated with MMS as described for panels B and D and analyzed as described for panel E. (G) MMS-induced phosphorylation of Chk1 by Rad3 was minimally affected in rqh1-G804D cells. Wild-type, rad3, and rqh1-G804D cells were treated with MMS as described for panels B, D, and F. Chk1-HA was immunoprecipitated from an equal number of cells for Western blotting using anti-HA antibody. All experiments in this figure were repeated ≥3 times, and the quantification results are shown in Fig. S4 in the supplemental material.

Under normal conditions, Rad9 is phosphorylated at a basal level mainly by Rad3 in S. pombe (11) (Fig. 3C and Fig. S4C). After HU treatment, the phosphorylation was significantly increased in wild-type cells. In untreated rqh1-G804D cells, the basal phosphorylation was slightly higher. After HU treatment, however, the phosphorylation was also increased to 70.5% (±1.5%; n = 3) of the wild-type level, suggesting that Rad9 phosphorylation was moderately affected. This result was confirmed by MMS treatment (Fig. 3D and Fig. S4D) as well as by the time course study in HU (Fig. S5C and D). Finally, we examined Rad3 phosphorylation of Cds1 and found that the phosphorylation was significantly decreased to only 24.8% (±2.0%; n = 3) of the wild-type level in HU (Fig. 3E and Fig. S4E) and to 26.5% (±5.0%; n = 3) in MMS (Fig. 3F and Fig. S4F). This result was also confirmed by the time course study in HU (Fig. S5E and F). These results show collectively that under replication stress, although Rad9 phosphorylation was moderately affected, the rqh1-G804D mutation significantly compromised Rad3 phosphorylation of Mrc1 and Cds1.

Minimal reduction of Rad3 signaling in the DDC.

Next, we examined the Rad3 signaling in the DDC pathway (Fig. 1A, right). As described above, in the presence of MMS, Rad9 phosphorylation by Rad3 was only moderately affected in rqh1-G804D cells (Fig. 3D and Fig. S4D). Under similar conditions, Rad3 phosphorylation of Chk1 was minimally affected and remained at 90% (±1.5%; n = 3) of the wild-type level (Fig. 3G and Fig. S4G). This result was confirmed by the time course study (Fig. S5G and H) showing that although the DRC was significantly compromised, the DDC remained functional, which explains the cell elongation observed in rqh1-G804D cells (Fig. 2B). Though more severe, the similar defects in the DRC and the DDC have also been described in the previously reported tel2-C307Y mutant (17). To confirm the specific DRC defect, we made the cdc10-129 rqh1∆ and cdc10-129 rqh1-G804D double mutants and synchronized the cells at G₁ by culturing at the restrictive temperature of 37°C. As shown in Fig. S6A, after being cultured at 37°C for 5 h, the majority of cdc10-129 cells and cdc10-129 rqh1∆ double mutant cells were arrested at G₁. When cells were released in HU at 25°C, Cds1 phosphorylation peaked at 2 h after the release in cdc10-129 cells (Fig. S6B). In the double mutant, however, the phosphorylation was significantly compromised during the 5-h release in HU. Similar results were observed with the cdc10-129 rqh1-G804D double mutant (Fig. S6C).

To further investigate the checkpoint defects, we crossed rqh1-G804D cells into checkpoint mutants and assessed the drug sensitivities of the resulting double mutants by spot assay. The result showed that while the rad3Δ rqh1-G804D double mutant was slightly more sensitive than the single mutants in the presence of HU or DNA damage; the sensitivity of the double mutants containing mrc1Δ or cds1Δ mutations is comparable to that of the rqh1-G804D mutant (Fig. S7A), suggesting that Rqh1 plays an important role in the DRC. We also made a chk1Δ rqh1-G804D double mutant and found that it showed a similar or slightly higher sensitivity than the rqh1-G804D mutant in the presence of HU, MMS, and bleomycin, but in the UV treatment the double mutant was significantly more sensitive than the rqh1-G804D mutant (Fig. S7B). This result is consistent with the multiple functions of Rqh1 in genome stability at both S and G₂ phases (27, 31, 33). To provide further evidence for the DRC function, we examined the HU sensitivity of mrc1Δ rqh1-G804D cells in liquid culture and found that the double mutant showed sensitivity similar to that of mrc1Δ cells although a slightly higher sensitivity was observed during the initial stage of HU treatment (Fig. S7C). These results are consistent with the DRC function of Rqh1.

Association of Rqh1 with perturbed forks.

RecQ DNA helicases bind to various fork DNA structures and replisome proteins (for a review, see reference 19), suggesting that Rqh1 associates with the forks to promote the DRC signaling in fission yeast. To investigate this possibility, we examined the potential interactions between Rqh1 and the replisome proteins Mrc1 and replication protein A (RPA). For this purpose, cells expressing Mrc1 and Rpa1 tagged with a hemagglutinin (HA) epitope were crossed into S. pombe with myc-tagged Rqh1. As shown in Fig. 4A, when Rqh1 was immunoprecipitated, Rpa1 was coimmunoprecipitated with Rqh1 (Fig. 4A and E). In the reciprocal coimmunoprecipitation (co-IP), Rqh1 was coimmunoprecipitated with Rpa1 (Fig. 4B and E). Since the co-IPs under similar conditions did not detect any interactions between Rqh1 and Cds1 (Fig. S11C and D) and since addition of ethidium bromide to eliminate the potential indirect binding via DNAs (59) did not affect the interaction, Rqh1 may physically interact with RPA in a manner similar to that of the budding yeast Sgs1 (24). However, the rqh1-G804D mutation did not affect the interaction (Fig. 4E). As mentioned above, Mrc1 is upregulated by HU treatment (Fig. 4C and D, inputs). Like Rpa1-Rqh1, Rqh1 and Mrc1 were found coimmunoprecipitated with each other (Fig. 4C, D, and F). Furthermore, the interactions between Rqh1 and Rpa1 and Mrc1 were enhanced by the HU treatment, and unlike the Rqh1-Rpa1 interaction, the rqh1-G804D mutation slightly enhanced the Rqh1-Mrc1 interaction (Fig. 4E and F), indicating that Rqh1 as well as the mutant Rqh1 is enriched at the HU-treated forks. Although the co-IPs cannot clearly show whether the interactions of Rqh1 with RPA and Mrc1 are direct or indirect, nonetheless, Rqh1 is likely recruited to the perturbed forks via interactions with RPA and Mrc1 or its DNA substrates associated with the forks.

FIG 4.

Interaction of Rqh1 with RPA and Mrc1. (A) Co-IP of Rpa1 with Rqh1 was enhanced in a HU-dependent manner. S. pombe with myc-tagged wild-type or mutant rqh1 was crossed into cells expressing HA-tagged Rpa1. Wild-type (NF147) or mutant rqh1 (NF164) cells were treated 15 mM HU for 3 h (+) or left untreated (−). Rqh1 was immunoprecipitated using anti-myc antibody (right) from cell extracts to detect Rpa1 coimmunoprecipitated with Rqh1 as described in Materials and Methods. An untagged rqh1 strain (YJ1586) was used as the control. Cell extracts (2.4%) were loaded on the same gel as the inputs (left). A section of the Ponceau S-stained inputs is shown. (B) Rqh1 was coimmunoprecipitated with Rpa1. Rpa1 was immunoprecipitated under conditions similar to those described for panel A with anti-HA antibody to detect coimmunoprecipitated Rqh1. (C) HU treatment enhanced the co-IP of Mrc1 with Rqh1. The myc-rqh1 strain was crossed into the mrc1-HA strain as described in panel A to generate the wild-type (NF157) and rqh1-G804D (NF168) strains. Rqh1 was immunoprecipitated with anti-myc antibody as described for panel A to detect Mrc1. (D) Reciprocal co-IP of Rqh1 with Mrc1 was carried out in wild-type and rqh1-G804D cells as described for panel C. (E and F) Co-IPs shown in panels A, B, C, and D were repeated ≥3 times. The coimmunoprecipitated bands were quantified and normalized with the inputs. After nonspecific bindings were removed, intensities of the co-IP bands were quantified and are shown in percentages relative to levels in HU-treated wild-type cells in the presence or absence of HU. Values are means and standard deviations.

In addition to Rqh1, fission yeast expresses additional RecQ helicase homologs (48, 49), particularly the RecQ4 orthologue Hrq1 that plays important roles in DNA repair and genome stability. To see whether Hrq1 or other 3′-to-5′ helicases contribute to Rad3 signaling at the forks, we examined S. pombe lacking hrq1, fml1, or fml2. The spot assay showed that while the rqh1-G804D mutant was highly sensitive to HU, hrq1Δ, flm1Δ, and fml2Δ cells were not or only minimally sensitive (Fig. S8A). Although the hrq1Δ mutant was sensitive to bleomycin and the fml1Δ mutant was sensitive to MMS, the sensitivities were lower than the level for the rqh1-G804D mutant. Consistent with the drug resistance, Rad3 phosphorylation of Mrc1 in HU-treated hrq1Δ, fml1Δ, and fml2Δ cells was at or near the wild-type level (Fig. S8B and C), which excluded the possibility that these helicases function in the DRC.

The DRC defect in other newly screened rqh1 mutants.

The helicase domain of Rqh1 is flanked by the long N- and C-terminal regions with less defined functions. As mentioned above, in addition to the rqh1-G804D mutation, our hus screen identified eight other mutants in the same linkage group (Fig. 5A and B). To see whether mutations in the N- or C-terminal regions of rqh1 can also sensitize S. pombe to HU, we sequenced the rqh1 genomic locus of these mutants and identified five more mutations (Fig. 5B). Among the mutations, hus86 carries a start codon mutation, which likely affects translation, and our N-terminal tagging confirmed this notion (data not shown). The remaining five mutations, to our surprise, are all located in the helicase domain. While hus12 contains a truncation mutation that deletes the C-terminal 765 amino acids, the other four mutations replace the amino acids that are all highly conserved in RecQ DNA helicases (Fig. S9). To confirm the mutations, we expressed the mutant Rqh1 proteins in their respective mutants on a vector under its own promoter (Fig. 5C). The results showed that while wild-type Rqh1 fully rescued all mutants, the mutants expressing the respective mutant rqh1 did not, which confirmed the mutations. Furthermore, we found that while hus16 (T622M), hus19 (F814L), and hus242 (C755Y) cells were slightly less sensitive, hus12 (R563STOP) cells lacking the C-terminal half of Rqh1 were slightly more sensitive than the rqh1-G804D mutant, suggesting that unlike the truncation mutation in hus12, the missense mutations in rqh1-G804D, hus12, hus19, and hus242 did not fully abolish the function of Rqh1 (see below). We then examined the phosphorylation of Mrc1 (Fig. 5D and F) and Cds1 (Fig. 5E and G) in these mutants. The results showed clearly that in the presence of HU, all mutations significantly compromised Rad3 phosphorylation of Mrc1 and Cds1, similar to the result with the rqh1-G804D mutant. Since these mutants were screened by random mutation of the genome, the helicase domain of Rqh1 must play an important role in promoting Rad3 signaling and cell survival in HU treatment.

FIG 5.

Drug sensitivity and the compromised DRC in other newly screened rqh1 helicase mutants. (A) DNA sequencing of the nine newly screened rqh1 mutants identified six uncharacterized mutations (B). Sensitivities of the representative rqh1 mutants to HU and the DNA damage were examined by spot assay. The dashed line indicates discontinuity. (B) Schematics of Rqh1 with the conserved helicase, RQC (RecQ conserved C-terminal), and HRDC (helicase and RNase D C-terminal) domains. Relative locations of the mutations identified from the screened hus mutants are indicated. The asterisk associated with hus12 indicates a truncation mutation. (C) Wild-type rqh1 and mutant rqh1 were expressed in the indicated rqh1 mutants on a vector under its own promoter. HU sensitivity was assessed by spot assay. (D) Rad3 phosphorylation of Mrc1 was compromised in all newly screened rqh1 helicase mutants. The wild type and the representative mutants used for the experiment shown in panel A were treated with 15 mM HU for 3 h or left untreated. Mrc1 phosphorylation was detected in whole-cell lysates using phospho-specific antibodies and quantified as described in the legend to Fig. 3A. (E) Rad3 phosphorylation of Cds1 was also compromised in the newly screened rqh1 mutants. The cells used in the experiment shown in panel D were treated with 15 mM HU for 3 h. Cds1-HA was immunoprecipitated and then analyzed by using a phospho-specific antibody and anti-HA antibody as described in the legend to Fig. 3E. (F and G) Experiments shown in panels D and E were repeated three times. The quantification results are shown for Mrc1 and Cds1. Values are means and standard deviations.

The rqh1-G804D mutation likely inactivates the enzyme.

All RecQ helicases have seven conserved motifs in their helicase domains (42) (Fig. S9). The mutated glycine in the rqh1-G804D mutant is highly conserved in motif V. Earlier studies have shown that mutation of this residue in other SF2 helicases eliminated the helicase and NTP hydrolysis activities (60, 61), suggesting that the mutation abolishes the helicase activity of Rqh1. To investigate this possibility, we integrated two previously reported helicase-inactive T543I and K547A mutations (31) at the genomic locus. Western blotting showed that they were expressed at the same levels as in wild-type and rqh1-G804D cells (Fig. S10). We found that in the presence of HU, MMS, UV, and bleomycin, the rqh1-G804D mutant, although slightly less sensitive, behaved in the same way as the two helicase-inactive mutants and S. pombe lacking rqh1 (Fig. 6A). We then compared the phosphorylation levels of Mrc1 (Fig. 6B) and Cds1 (Fig. 6C) in these mutants. As expected, phosphorylation was reduced in rqh1Δ cells and in the two helicase-inactive mutants, which is similar to the results in the rqh1-G804D mutant. These results show that the rqh1-G804D mutation as well as the other newly identified rqh1 mutations likely eliminates most of the helicase activity of Rqh1, leading to the compromised Rad3 signaling.

FIG 6.

Deletion or helicase-inactive mutations of rqh1 cause a Rad3 signaling defect similar to that of the rqh1-G804D mutant. (A) The helicase-inactive rqh1-K547A and rqh1-T543I mutations were integrated at the genomic locus as described for the rqh1-G804D mutant (see Fig. S2C and S10 in the supplemental material). Sensitivities of the indicated cells to HU and the DNA damaging agents were examined by spot assay. The wild type, the checkpoint mutants, and S. pombe lacking rqh1 (NF121) were used as controls. The dashed line indicates discontinuity. (B) The reduction of Rad3 phosphorylation observed in the rqh1Δ mutant and the two helicase-inactive K547A and T543I mutants was similar to that in the rqh1-G804D mutant. Cells used in the experiment shown in panel A were treated with 15 mM HU for 3 h or left untreated before the analysis of Mrc1 phosphorylation. The experiment was repeated three times, and the quantification results are shown in the bottom panel. Values are means and standard deviations. (C) Reduced Rad3 phosphorylation of Cds1 in the rqh1Δ mutant and the helicase-inactive mutants. Cds1 phosphorylation was examined in cells carrying the indicated mutations, either untreated or treated with HU as described in the legend to Fig. 3. Quantification results from three independent experiments are shown in bottom panel.

Rqh1 does not function as a checkpoint mediator at the perturbed forks.

In budding yeast, Sgs1 is phosphorylated by Mec1, which functions as a mediator to recruit Rad53 to the forks mainly via the phosphorylated TQ motif containing Thr⁴⁵¹, which binds to the FHA1 domain of Rad53 (24–26). Our results show, however, that the compromised phosphorylation of Mrc1 and Cds1 in the rqh1 mutants is likely due to the inactivated helicase activity. As mentioned above, in fission yeast, Mrc1 recruits Cds1 via its two functional redundant Cds1 docking sites of eight amino acids embedded with TQ motifs in the middle of Mrc1 that are phosphorylated by Rad3 (8). Similar to the FHA1 domain of Rad53, the N-terminal forkhead-associated (FHA) domain of Cds1 binds specifically to the phosphorylated TQ motifs, not Ser-Gln (SQ) motifs (8). There are only two TQ motifs in Rqh1 containing Thr³⁴⁹ and Thr¹²⁶⁵. None of the TQ motifs reside in the consensus Cds1 docking site, suggesting that Rqh1 may not function as a mediator for Cds1 recruitment. To investigate this possibility, we replaced the two threonines with alanines and examined HU sensitivity and Mrc1 phosphorylation in S. pombe expressing the two TQ mutants (Fig. S11A and B). The results clearly showed that the TQ mutations did not significantly affect HU resistance and Mrc1 phosphorylation. To investigate further, we carried out co-IP experiments for Rqh1 and Cds1 by using a method that detected the interactions of Rqh1 with Rpa1 and Mrc1. After multiple repeated experiments, we found that Cds1 was not coimmunoprecipitated with Rqh1 in the presence or absence of HU (Fig. S11C and D). These results show that unlike Sgs1, Rqh1 does not function as a mediator for recruiting Cds1 for its phosphorylation by Rad3.

Human BLM, RECQ1, and RECQ4 restore Rad3 signaling and rescue rqh1-G804D cells.

S. pombe is an established model for studying the cellular mechanisms that are conserved in higher eukaryotes (62), and cross-species complementation of yeast mutations by human genes has been successfully utilized to elucidate the functional homology between human and yeast proteins (63, 64). We then tested whether expression of the mammalian enzymes can rescue the rqh1-G804D mutation. To this end, human BLM, RECQ1, RECQ4, and RECQ5 and mouse WRN were expressed from a vector under the control of the nmt1 promoter in rqh1-G804D cells (Fig. 7A). Western blotting confirmed the thiamine-regulated expression of all mammalian proteins in rqh1-G804D cells (Fig. 7B). A spot assay showed that while WRN or RECQ5 did not show any rescuing effect, BLM, RECQ1, and RECQ4 partially rescued rqh1-G804D cells in the presence of HU, UV, and MMS. We then examined Rad3 phosphorylation of Mrc1 and found that, consistent with the rescuing effect, phosphorylation was significantly restored in rqh1-G804D cells expressing RECQ1, BLM, or RECQ4. Consistent with our data, an earlier study showed that while expression of BLM suppressed the HU sensitivity of a budding yeast sgs1 mutant, WRN did not (65). Importantly, since human RECQ1 is the smallest homolog that contains mainly the helicase domain, its rescuing effect observed in rqh1-G804D cells strongly supports our notion that it is the helicase activity of Rqh1, not the protein per se, that plays an important role in promoting Rad3 signaling at the perturbed forks in S. pombe.

FIG 7.

Complementation of the rqh1-G804D mutant by human RecQ DNA helicases. (A) Heterologous expression of human RECQ1, BLM, or RECQ4, not human RECQ5 or mouse WRN, rescued rqh1-G804D cells. The mammalian helicases were expressed in rqh1-G804D cells from a vector under control of the nmt1 promoter. The sensitivities of rqh1-G804D cells expressing the human enzymes to HU, MMS, and UV were examined by spot assay. The wild type, the checkpoint mutants, and the rqh1-G804D and rqh1-G804D mutants carrying an empty vector were used as the controls. (B) Thiamine-regulated expression of the mammalian helicases in rqh1-G804D cells. The myc-tagged mammalian helicases were expressed under control of the nmt1 promoter. The expression levels were analyzed in rqh1-G804D cells cultured in the presence (+) or absence (−) of thiamine by Western blotting using anti-myc antibodies. A section of Ponceau S-stained membrane is shown as the loading control (bottom panel). *, full-length enzymes; **, degradation product. (C) Rad3 phosphorylation of Mrc1 was restored in rqh1-G804D cells expressing human RECQ1, BLM, or RECQ4. Cells used in the experiment shown in panel A were treated with 15 mM HU for 3 h. Phosphorylation of Mrc1 (top panel) and Mrc1 (middle panel) was sequentially detected by using phospho-specific antibodies and the anti-Mrc1 antibodies as described in the legend to Fig. 3A. Quantification results from three repeated experiments are shown on the bottom. Values are means and standard deviations. A t test was performed for each mammalian enzyme to compare results to those of the control (***, P < 0.05; ns, not significant). (D) Proposed checkpoint function of Rqh1 at the perturbed forks in fission yeast. When DNA polymerases are slowed or paused by lesions on the leading-strand templates (red dot), Rqh1 unwinds and releases nascent DNAs on the lagging strand in a 3′-to-5′ direction to generate long-stretch single-stranded DNAs for Rad3 recruitment and checkpoint activation. Rqh1 may also regulate fork reversal for fork stabilization, repair, or restart after the lesions are removed or bypassed by translesion DNA synthesis.

DISCUSSION

Our genetic screen in fission yeast has identified a new set of hus mutants with DRC defects. Characterization of nine hus mutants in the same complementation group identified six mutations in rqh1 (Fig. 5B). Except for the start codon mutation in hus86, all remaining five mutations are in the helicase domain. One of the mutants carries a truncation mutation that behaves similarly to rqh1 null cells under all tested conditions. The other four mutations replace the amino acids that are all highly conserved in the RecQ DNA helicases (see Fig. S9 in the supplemental material). Although subtle differences are observed in drug sensitivities and the DRC signaling, these four missense mutants behave quite similarly. Consistent with their slightly milder HU sensitivity, the defect in Rad3 kinase signaling was also slightly less severe than that in rqh1 null cells. These results, together with experiments on rqh1-G804D cells, show that Rqh1 promotes Rad3 kinase signaling at the perturbed forks in S. pombe. At the molecular level, all newly screened mutations compromised Rad3 phosphorylation of Mrc1 and Cds1 in the presence of HU or MMS. At the cellular level, HU generates cut cells, and the number of cut cells correlates with the cell killing effect of HU in the rqh1-G804D mutant, the representative of the newly screened nine hus mutants. We also found that the DDC is minimally affected in the rqh1-G804D mutant, which is consistent with the cell elongation and the compromised DRC observed in this mutant. Although our data provide an explanation for the sensitivity of rqh1 mutants to replication stress induced by HU or DNA damage, they do not exclude the involvement of Rqh1 in fork restart and recombination repair (27, 31, 33, 34). However, whether Rqh1 has a DRC function similar to that of S. cerevisiae Sgs1 has been controversial (32–34). Because Cds1 undergoes autoactivation when local concentrations are high (8, 50), it is possible that the Cds1 kinase assays used in earlier studies prevent accurate measurement of a partial DRC defect in the rqh1 mutants. Alternatively, this discrepancy is caused by allele-specific effects.

With respect to mrc1Δ cells lacking the DRC, there is a difference between the HU sensitivity of the rqh1-G804D mutant determined by spot assay (Fig. 1B) and that in liquid cultures (Fig. 1C). This difference has been observed in the metabolic hus mutants mentioned above that are killed by HU via oxidative stress or other mechanisms (38, 39). It has also been observed in our previous study on several mutations of Rad4 (TOPBP1/Dpb11), a protein that functions in both replication initiation and the DDC (9). It is likely that the oxidative stress generated by perturbed DNA replication in S. pombe (66) may exacerbate the cell-killing effect of HU, particularly with chronic HU exposure (67). Rqh1, though not essential for cell growth, may contribute to normal DNA replication and fork stability, particularly at the hard-to-replicate chromosomal regions. In support of this possibility, slightly increased Rad9 and Chk1 phosphorylation (Fig. 3D and G) and longer G₂/M delays (Fig. 2A and D) have been observed in the rqh1-G804D mutant under normal conditions. Therefore, rqh1 mutations may minimally perturb normal DNA replication, which generates oxidative stress and thus enhances the chronic cell-killing effects of HU.

In S. cerevisiae, Sgs1 is phosphorylated by Mec1 (ATR/Rad3) to recruit Rad53 (CHK2/Cds1) as a checkpoint mediator for Rad53 activation (24–26). Our data show that Rqh1 likely promotes Rad3 kinase signaling at the forks by its helicase activity, not as a mediator in fission yeast. First, except for the mutations truncating or depleting Rqh1, the missense mutations identified by our random genome-wide screen are all in the helicase domain, not the other parts of the protein. Second, the DRC is similarly compromised in an rqh1Δ mutant and the two helicase-inactive mutants as well as in the newly screened rqh1 mutants. Third, cut cells, a typical phenotype associated with DRC mutants, are observed in the rqh1-G804D mutant (Fig. 2B and C) as well as in rqh1Δ cells (27). Fourth, unlike Sgs1, Rqh1 lacks the consensus docking sites for Cds1 via the phospho-peptide binding FHA domain of Cds1 (8, 68). Consistent with this, mutation of the only two TQ motifs on Rqh1, the potential binding sites for Cds1 FHA domain, did not significantly affect HU resistance and Mrc1 phosphorylation (Fig. S11A and B). Furthermore, our co-IPs did not detect the interaction of Rqh1 with Cds1 (Fig. S11C and D) although the interaction of Sgs1 with Rad53 (24, 26) and that of Rqh1 with Rpa1 and Mrc1 can be detected by the same method (Fig. 4). Finally, heterologous expression of human RecQ helicases BLM, RECQ4, and, particularly, the smallest RecQ helicase, RECQ1, which mainly contains the helicase domain, rescues and restores Rad3 signaling in the rqh1-G804D mutant. Based on these results, we propose a model in Fig. 7D for the DRC function of Rqh1 in fission yeast.

In addition to the DRC function of budding yeast Sgs1 mentioned above (24, 25), RecQ, the founding member of the RecQ helicase family, initiates the SOS response at perturbed forks by its helicase activity in bacteria (69). The DRC functions have also been described for human RECQ1 (70), RECQ4 (71, 72), and WRN (73–75). Although not directly involved in the DRC signaling, BLM is regulated by ATR phosphorylation for cell recovery from S-phase arrest (76). It is likely that the DRC function of Rqh1 is conserved from bacteria and yeasts to humans, and the defects associated with the helicase mutations in the human enzymes contribute to the disease syndromes mentioned above. Future studies are needed to investigate the enzymatic activities of Rqh1 that promote Rad3 signaling and the possible redundant factor(s) responsible for the remaining 20 to 30% of Mrc1 phosphorylation in fission yeast. Furthermore, our humanized S. pombe may also provide an in vivo platform to understand DRC defects and to screen for specific inhibitors of the human RecQ DNA helicases.

MATERIALS AND METHODS

Yeast strains and plasmids.

The S. pombe strains were cultured at 30°C in YE6S medium (0.5% yeast extract, 3% dextrose, supplemented with adenine, uracil, leucine, lysine, histidine, and arginine) or in EMM(6S) medium (Edinburgh minimal medium with the six supplements) lacking an appropriate supplement, according to standard methods (77). Yeast strains, plasmids, and PCR primers used in this study are listed in Tables S1, S2, and S3, respectively, in the supplemental material. Mammalian RecQ helicases RECQL, pTRIP-CMV-puro-2A-BLM (where CMV is cytomegalovirus), pCMV-WS-Myc-WRN, and TAL2220 were obtained from Addgene (catalog no. 38890, 127641, 19273, and 36679, respectively). RECQ5 plasmid was kindly provided by Nicola Burgess-Brown of the University of Oxford. These mammalian RecQ helicases were amplified by Phusion polymerase and cloned between NotI and XmaI sites in a yeast expression vector under the control of the nmt1 promoter. Cloned genes and the mutations were confirmed by DNA sequencing (Retrogen, San Diego, CA).

The hus screen.

The method for screening new hus mutants was described in our previous study (17).

Integration of rqh1 mutations.

The rqh1 expression cassette was cloned into the pIRT-2U vector between SacI and XmaI sites. An myc epitope and a Kan^r marker were sequentially inserted in frame at the N terminus of rqh1 and after the terminator, respectively (see Fig. S2C in the supplemental material). After digestion with NheI and PciI, a 7,131-bp integration fragment was gel purified and transformed into the wild-type TK7 strain. After an overnight recovery, cells were replicated onto plates containing 100 μg/ml G418 (Sigma). Colonies were screened by colony PCRs to confirm correct integrations at both the 5′ and 3′ ends. The integrants were backcrossed once to ensure single-copy integration and confirmed again by PCRs that cover the whole regions of the integration locus as well as Western blotting with anti-myc antibody, as shown in Fig. S2D and E.

Drug sensitivity.

Sensitivities to HU and various DNA damaging agents were determined by spot assay or in liquid medium as described in our previous studies (8, 11, 38, 39). For the spot assay, such as that shown in Fig. 1B, 2 × 10⁷ cells/ml of logarithmically growing S. pombe were diluted in 5-fold steps and spotted in 3 μl onto YE6S plates or YE6S plates containing the drugs at the indicated concentrations. The plates spotted with the cells were dried before the UV treatment (Stratalinker 2400). The plates were incubated at 30°C for 3 days and then photographed. All spot assays were repeated at least once. For acute HU sensitivity (35), liquid cell cultures were incubated with 15 mM HU. Every hour during the treatment, an equal number of cells was removed, diluted 1,000-fold, spread onto three YE6S plates, and incubated at 30°C for 3 days for cell recovery. Colonies were counted and presented as percentages of counts of the untreated cells.

IP and co-IP.

A total of 1 × 10⁸ logarithmically growing cells were harvested and saved at −20°C in a 1.5-ml screw-cap tube. The frozen cell pellets were lysed by mini-bead beater in the buffer containing 25 mM HEPES-NaOH (pH 7.5), 50 mM NaF, 1 mM NaVO₄, 10 mM NaP₂O₇, 40 mM β-glycerophosphate, 0.1% Tween 20, 0.5% NP-40, and protease inhibitors aprotinin, bestatin, pepstatin, leupeptin, benzamidine, and bacitracin. The lysates were clarified by centrifugation at 16,000 × g at 4°C for 5 min. Anti-HA or anti-myc antibody–agarose resin (Santa Cruz) was prewashed three times with Tris-buffered saline containing 0.05% Tween 20 (TBS-T) and incubated with 5% bovine serum albumin (BSA) in TBS-T for ≥30 min at 4°C. The cell extract was treated with 100 μg/ml ethidium bromide (Sigma) (59) for 1 h before incubation with the prewashed antibody resins by rotation in 2-ml tubes at 4°C for 2 h. The resins were washed three times with TBS-T at 4°C for 20 min and then separated by SDS-PAGE, followed by Western blotting with mouse monoclonal anti-HA (12CA5; Sigma) or anti-myc (9E10; Thermo Scientific) antibody.

Western blotting.

The method for examining Rad3 phosphorylation of Mrc1-Thr⁶⁴⁵, Rad9-Thr⁴¹², and Cds1-Thr¹¹ using the phospho-specific antibodies from equal numbers of cells has been described in our previous studies (11, 17, 50). The myc- or HA-tagged proteins were examined by Western blotting by using mouse monoclonal antibodies. For the Western analyses, 1 × 10⁸ logarithmically growing cells were fixed in 15% trichloroacetic acid (TCA) on ice for ≥3 h and then lysed by mini-bead beater. The lysates from 2 × 10⁶ to 4 × 10⁶ cells were separated by SDS-PAGE before being transferred to nitrocellulose membrane and stained with Ponceau S as the loading control (78). The blotting signal was detected by electrochemiluminescence using a ChemiDoc XRS Imaging system (Bio-Rad). Intensities of the specific bands were quantified and analyzed by ImageLab (Bio-Rad).

Flow cytometry.

A total of 1 × 10⁷ logarithmically growing cells were collected, fixed in ice-cold 70% ethanol, and then analyzed by an Accuri C6 flow cytometer as described in our previous studies (38, 39).

Microscopy.

The cells were fixed directly onto uncoated glass slides by briefly heating them at 75°C for 30 s or in medium containing 2.5% glutaraldehyde at 4°C for ≥3 h. The glutaraldehyde-fixed cells were washed with phosphate-buffered saline (PBS) by centrifugation at 2,300 × g for 30 s, stained in the same buffer with 5 μg/ml Hoechst 33258 (Sigma-Aldrich) and a 1:100 dilution of the Blankophor working solution (MP Biochemicals). The stained cells were examined using an Olympus EX41 fluorescence microscope. Images were captured with an IQCAM camera (Fast1394) using Qcapture Pro, version 6.0, software. Approximately 150 cells were counted under microscope for each sample, and counts were repeated three times. Images were also exported into Photoshop (Adobe) to generate Fig. 2B.