Protein phosphatase 2A (PP2A) promotes anaphase entry after DNA replication stress in budding yeast.

DNA replication stress activates the S-phase checkpoint that arrests the cell cycle, but it is poorly understood how cells recover from this arrest. Cyclin-dependent kinase (CDK) and protein phosphatase 2A (PP2A) are key cell cycle regulators, and Cdc55 is a regulatory subunit of PP2A in budding yeast. We found that yeast cells lacking functional PP2ACdc55 showed slow growth in the presence of hydroxyurea (HU), a DNA synthesis inhibitor, without obvious viability loss. Moreover, PP2A mutants exhibited delayed anaphase entry and sustained levels of anaphase inhibitor Pds1 after HU treatment. A DNA damage checkpoint Chk1 phosphorylates and stabilizes Pds1. We show that chk1Δ and mutation of the Chk1 phosphorylation sites in Pds1 largely restored efficient anaphase entry in PP2A mutants after HU treatment. In addition, deletion of SWE1, which encodes the inhibitory kinase for CDK or mutation of the Swe1 phosphorylation site in CDK (cdc28F19), also suppressed the anaphase entry delay in PP2A mutants after HU treatment. Our genetic data suggest that Swe1/CDK acts upstream of Pds1. Surprisingly, cdc55Δ showed significant suppression to the viability loss of S-phase checkpoint mutants during DNA synthesis block. Together, our results uncover a PP2A-Swe1-CDK-Chk1-Pds1 axis that promotes recovery from DNA replication stress.

INTRODUCTION

Accurate DNA replication is essential for genome stability, whereas genome instability contributes to cell death or tumorigenesis. In addition to the innate complexity associated with DNA replication, cells face numerous internal and external stressors that may slow or stall DNA replication (Zeman and Cimprich, 2014). External stressors include extreme temperatures, toxin exposure, and mechanical damage, while internal stressors include limited resources, chemical imbalance, dysfunctional DNA replication machinery, and replication errors. Given the importance of genomic integrity maintenance, cells have adapted numerous cell cycle responses to deal with stressful DNA replication (Fulda ).

The cell cycle is driven in part by a series of phosphorylation events. The main kinase involved in cell cycle progression is the cyclin-dependent kinase (CDK) (Hartwell ; Gómez-Escoda and Wu, 2017). Unlike mammalian cells, the budding yeast Saccharomyces cerevisiae has only one CDK, Cdk1 (Cdc28) (Beach ; Enserink and Kolodner, 2010). Progression through the cell cycle is driven by Cdk1 in association with different cyclins that are characteristic of each stage of the cell cycle (Andrews and Measday, 1998; Malumbres, 2014). In S. cerevisiae, cyclins Clb5 and Clb6 promote S-phase progression; Clb3 and Clb4 associate with early mitotic events; Clb1 and Clb2 are critical for mitosis (Kõivomägi ). One CDK inhibitor is Swe1, a protein kinase that phosphorylates tyrosine 19 (Y19) of Cdc28 to inhibit the mitotic activity of Cdk1 (Sia ; Liu and Wang, 2006; Hu ). The presence of Swe1 delays anaphase onset by phosphorylating Y19 of Cdk1 (Lianga ; Leitao ). Swe1 has also been implicated in the coordination of bud formation and cell division in budding yeast by delaying mitotic entry when bud formation is defective (Sia ; Barral ). Swe1, once phosphorylated by Cdk1, forms a stable complex with Cdk1, which leads to the inhibition of Cdk1 activity (Harvey ). However, Cdk1 dephosphorylation leads to further Swe1 phosphorylation by Cdk1, causing dissociation of Swe1 from Cdk1 and subsequent activation of mitotic CDK (Asano ).

Cell cycle checkpoints are vital for genome stability. Within a cell cycle, several checkpoints are in place to ensure accurate DNA replication and chromosome segregation (Barnum and O’Connell, 2014; Ovejero ). The S-phase checkpoint is a branch of the DNA damage checkpoint that is associated with DNA replication stress. This checkpoint is triggered by a block in DNA replication fork progression or other forms of replication perturbation that lead to accumulation of single-stranded DNA at replication forks (Willis and Rhind, 2009; Pardo ). The major components of the S-phase checkpoint in budding yeast include Mec1 and Rad53 kinases whose activation leads to stabilization of stalled DNA replication forks and inhibition of cell cycle progression (Tourrière and Pasero, 2007; Zou, 2013). Mec1, functioning through the mediator of the replication checkpoint (Mrc1), phosphorylates and activates the effector kinase Rad53 (Osborn and Elledge, 2003). Mec1 and Rad53 are also the essential components of the DNA damage checkpoint (Kiser and Weinert, 1996). It has been shown that Rad53 is required to maintain mitotic Cdk1 activity following the activation of the DNA damage checkpoint (Sanchez ).

Another well-conserved checkpoint, the spindle assembly checkpoint (SAC), functions to ensure proper kinetochore–microtubule interaction, which is fundamental for accurate chromosome segregation. The major components of the SAC include Mad1, Mad2, Mad3, Bub1, Bub3, and Mps1 (Wang ; Stukenberg and Burke, 2015). Once the SAC becomes activated, Mad2 disrupts the interaction between the anaphase promoting complex (APC) and its regulator Cdc20 (Ge ). This inhibits the degradation of anaphase inhibitor Pds1 by APCCdc20, thereby blocking anaphase onset. Anaphase entry is marked by sister chromatid separation. The cohesin complex holds the sister chromatids together until cleaved by separase Esp1. Before anaphase entry, Esp1 is inhibited by forming a complex with securin Pds1 (Ciosk ). The Pds1-Esp1 interaction requires Cdk1-mediated Pds1 phosphorylation (Agarwal and Cohen-Fix, 2002). Interestingly, Chk1, a kinase in the DNA damage checkpoint, phosphorylates and stabilizes Pds1 in response to DNA damage by preventing APCCdc20-mediated Pds1 degradation (Wang ).

Protein phosphorylation and dephosphorylation play a key role in cell cycle progression. One of the primary phosphatases critical for the cell cycle is the holoenzyme protein phosphatase 2A (PP2A). In budding yeast, PP2A consists of three subunits: a scaffolding subunit (Tpd3), a regulatory subunit (Cdc55 or Rts1), and a catalytic subunit (Pph21, Pph22, or Pph3) (Sneddon ). As a regulatory subunit, Cdc55 governs cell cycle progression through the timed dephosphorylation of downstream targets (Godfrey ). One demonstrated function of PP2ACdc55 is to inhibit mitotic exit, a process that inactivates CDK by freeing phosphatase Cdc14 from the nucleolus. PP2ACdc55 dephosphorylates Net1, a nucleolar protein that anchors Cdc14 to the nucleolus, and loss of function of PP2ACdc55 results in premature release of Cdc14 from the nucleolus (Queralt ; Wang and Ng, 2006; Yellman and Burke, 2006). Recent evidence also indicates that PP2ACdc55 reverses Cdk1-mediated phosphorylation of Pds1, which is expected to abolish Pds1-Esp1 interaction for anaphase progression (Khondker ). PP2A has also been shown to be required for mitotic arrest in response to defective kinetochore–microtubule interaction (Minshull ; Wang and Burke, 1997). In addition, PP2ACdc55 promotes Swe1 degradation to facilitate the cell cycle transition from G2 to M phase (Yang ). Although the accumulation of Swe1 and enhanced Cdk1 phosphorylation at Y19 by Swe1 contribute to the abnormal bud morphology and cold sensitivity cdc55Δ mutant, the mitotic checkpoint defect in cdc55Δ seems to be independent of Swe1-mediated Y19 phosphorylation of Cdk1 (Wang and Burke, 1997). Therefore, PP2ACdc55 plays multiple roles in cell cycle progression, but the detailed molecular mechanisms are not fully understood yet.

While the cell cycle arrest induced by DNA replication stress is well documented, it remains poorly defined how cells recover from this arrest. In this study, we found that loss of function of PP2ACdc55 in budding yeast led to slow growth in the presence of hydroxyurea (HU), a DNA synthesis inhibitor, but the slow growth is not a result of viability loss. Moreover, PP2A mutants exhibited sustained levels of anaphase inhibitor Pds1 along with delayed anaphase entry after treatment with HU, indicating that PP2ACdc55 is required for recovery from DNA replication stress. Interestingly, chk1Δ and mutations of the Chk1 phosphorylation sites in Pds1 largely restored efficient anaphase entry in PP2A mutants after HU treatment. In addition, our results indicate that PP2ACdc55 promotes Swe1 degradation and the reversal of Swe1-mediated Cdk1 phosphorylation, which further facilitates Pds1 degradation and anaphase onset. Surprisingly, loss of the S-phase checkpoint failed to suppress the anaphase entry delay in PP2A mutant cells, but the viability loss of S-phase checkpoint mutants in response to DNA synthesis stress is largely suppressed by PP2A mutants, and this suppression likely acts through Swe1 and Pds1. Together, our results uncover the PP2A-Swe1-CDK-Chk1-Pds1 axis that promotes the recovery from cell cycle arrest induced by DNA synthesis inhibition.

RESULTS

PP2A mutants are sensitive to DNA replication stress without viability loss

PP2A is a primary phosphatase critical for cell cycle progression. Previous works have shown that yeast PP2A mutants cdc55Δ and pph21Δ pph22Δ are sensitive to microtubule depolymerizing agents, such as benomyl and nocodazole (Minshull ; Wang and Burke, 1997; Wang and Ng, 2006). We wanted to further determine whether loss of function of PP2A in budding yeast also led to sensitivity to other stresses that disrupt cell cycle progression. One such stressor is HU, which impedes DNA replication by inhibiting ribonucleotide reductase and depleting free nucleotides (Koç ). Interestingly, yeast mutants in several subunits within the PP2A complex showed slow growth on YPD (yeast extract peptone dextrose) plates containing 100 mM HU. Deletion of the scaffold subunit (tpd3Δ) of PP2A caused dramatic HU sensitivity, but this mutant also exhibited obvious slow growth in the absence of HU. Pph21, Pph22, and Pph3 are the three catalytic subunits of PP2A. The pph21Δ mutant was not sensitive to HU, but the loss of both Pph21 and Pph22 led to HU sensitivity. For the regulatory subunits, cdc55Δ, but not the rts1Δ mutant, showed obvious slow growth on HU plates (Figure 1A), indicating that the HU sensitivity is specific to the loss of function of PP2ACdc55.

FIGURE 1:

PP2A mutants are sensitive to DNA replication stress. (A) PP2A mutants show slow growth in the presence of HU. Saturated cells with the indicated genotypes were 10-fold serially diluted and spotted onto YPD plates with or without 100 mM HU. Growth was analyzed after 2 d on YPD and 3 d on HU plates at 30°C. (B) PP2A mutants do not show significant viability loss when treated with HU. WT and PP2A mutants were grown at 30°C in YPD media until mid–log phase, and then 200 mM HU was added. Samples were collected at 0, 2, 4, and 8 h and spread onto YPD plates to determine the plating efficiency after overnight growth at 25°C (n ≥ 200). (C) DNA replication kinetics is normal in cdc55Δ cells after release from HU arrest. WT and cdc55Δ cells were grown at 30°C until mid–log phase and then arrested at G1 phase with α-factor. G1-arrested cells were released into YPD with 200 mM HU for 100 min before release into YPD. Cells were collected every 20 min after HU release. All samples were embedded into 1.5% agarose and treated with zymolyase and proteinase K to release genetic material. Samples were analyzed via PFGE and visualized after ethidium bromide staining.

To determine why PP2A mutants show HU sensitivity, we examined whether these mutants showed a checkpoint defect in response to DNA replication stress, which would result in viability loss. We first analyzed the viability of PP2A mutant cells grown in liquid YPD media containing 200 mM HU. Despite the dramatic slow growth of cdc55Δ and pph21Δ pph22Δ mutant cells on HU plates, their viability loss was not significant even after prolonged HU exposure (Figure 1B). This indicates competent checkpoint function in these mutants, which are able to recover from cell cycle arrest after HU is removed. Next, we determined whether the PP2A mutants experienced any DNA replication defect following HU exposure. We analyzed the replication kinetics of cdc55Δ cells using pulsed-field gel electrophoresis (PFGE), which separates chromosomes based on their size. Owing to pore size constraints, replicating chromosomes are excluded, allowing only duplicated chromosomes to be visualized by PFGE. Therefore, this method has been used to monitor the kinetics of DNA replication (Liu and Wang, 2006). Strikingly, cdc55Δ cells exhibited normal DNA replication kinetics compared with wild-type (WT) cells following release from HU arrest (Figure 1C). On the basis of these results, we conclude that the HU sensitivity of PP2A mutants is not due to the defect in checkpoint arrest or DNA replication.

Anaphase entry is delayed in PP2A mutants following DNA replication stress

One explanation for the HU sensitivity of PP2A mutants might be the delayed recovery from the cell cycle arrest induced by DNA replication stress. To test this idea, we first examined levels of the anaphase inhibitor Pds1 in WT and cdc55Δ mutant cells during the cell cycle, because Pds1 degradation marks anaphase entry (Cohen-Fix ). cdc55Δ and pph21Δ pph22Δ mutant cells exhibited a clear delay in Pds1 degradation during undisturbed cell cycle, and this is consistent with the persistent appearance of large-budded cells as indicated by the budding index (Figure 2A). Next, we compared Pds1 levels in WT and PP2A mutant cells after release from the cell cycle arrest induced by 200 mM HU. We noticed a significant delay in Pds1 degradation in cdc55Δ cells (Figure 2B), indicating that Cdc55 promotes anaphase entry after replication stress. pph21Δ pph22Δ mutants also showed delayed Pds1 degradation, but the delay was not as dramatic as in cdc55Δ mutant cells. Moreover, the delayed Pds1 degradation is consistent with a high percentage of large-budded cells (Figure 2B). These results suggest that the HU sensitivity of PP2A mutants is likely due to the delayed Pds1 degradation that blocks anaphase entry.

FIGURE 2:

Anaphase entry is delayed in PP2A mutants after HU exposure. (A) Pds1 levels during the cell cycle in WT and PP2A mutant cells. WT, cdc55Δ, and pph21Δ pph22Δ cells with Pds1-18myc were grown at 30°C in YPD media until mid–log phase and then they were arrested at G1 with α-factor. G1 cells were released into YPD at 30°C with samples collected every 20 min. After G1 release for 40 min, α-factor was added back to block the following cell cycle. Budding index and Pds1 protein levels are shown. Pgk1, loading control. (B) Pds1 protein remains stabilized in PP2A mutants following HU exposure. G1-arrested cells were released into YPD with 200 mM HU for 2 h at 30°C. After HU was washed off, the cells were released into YPD, and α-factor was added to block the following cell cycle. Samples were collected every 20 min after HU release. Budding index and Pds1 protein levels are shown. Pgk1, loading control. (C) PP2A mutants show significant delay in nuclear division following HU treatment. G1-arrested WT, cdc55Δ, and pph21Δ pph22Δ cells containing mApple-tagged H2A were released into YPD with 200 mM HU for 2 h. Then cells were washed and released into YPD at 30°C, and α-factor was added to block the following cell cycle. Every 20 min, cells were collected and counted for budding index and percentage of cells with undivided DNA using fluorescence microscopy. Budding index and the kinetics of nuclear division are shown in the right panel. Nuclear divisions (H2A-mApple) in some representative cells are shown in the left panel. Arrows: the nucleus. Scale bar, 5 μm. (D) The SAC mutant mad1Δ fails to rescue the delayed recovery of cdc55Δ cells following HU arrest. Cells with the indicated genotypes were treated as described in panel B. Budding index and Pds1 protein levels are shown. Pgk1, loading control. (E) mad1Δ and mad2Δ mutants do not suppress the HU sensitivity of cdc55Δ mutant cells. Saturated cells with the indicated genotypes were 10-fold serially diluted and spotted onto YPD plates with or without 100 mM HU. Growth was analyzed after 2 d on YPD and 3 d on HU plates at 30°C.

We also analyzed nuclear separation kinetics in WT and cdc55Δ mutants under normal growth conditions and after HU treatment. Strains containing mApple-tagged histone H2A (HTA1-mApple) allow for the fluorescence visualization of chromosome segregation. As expected, cdc55Δ and pph21Δ pph22Δ mutant cells exhibited slightly delayed nuclear division during the normal cell cycle compared with WT cells (Supplemental Figure S1). When comparing nuclear separation kinetics in these cells following release from HU arrest, we noticed a significant nuclear division delay in cdc55Δ cells. The delay was also obvious in pph21Δ pph22Δ cells, although it was less significant than in cdc55Δ cells (Figure 2C). Additionally, using a GFP-marked centromere of chromosome IV (CEN4-GFP), we observed a significant delay in sister chromatid segregation into two daughter cells in cdc55Δ cells following release from HU arrest (Supplemental Figure S2). Together, these results indicate that PP2ACdc55 is required for efficient anaphase entry after DNA replication stress.

The delayed degradation of Pds1 in PP2A mutants might be the result of SAC activation, which prevents anaphase onset in response to kinetochore attachment defects (Wang ). Previous research has shown that PP2ACdc55 plays a role in metaphase arrest in response to the disruption of kinetochore–microtubule interaction (Minshull ; Wang and Burke, 1997). Interestingly, deletion of MAD1, a SAC gene, in cdc55Δ cells did not suppress the delayed Pds1 degradation and cell cycle progression after HU treatment (Figure 2D). Moreover, neither mad1Δ nor mad2Δ suppressed the slow growth phenotype of cdc55Δ mutants on HU plates (Figure 2E). The failure to restore the growth of the cdc55Δ mutant in the presence of HU indicates that SAC activation is unlikely to play a role in the delayed recovery from DNA replication stress in PP2A mutants.

Mutation of Chk1 phosphorylation sites in Pds1 suppresses the HU sensitivity of PP2A mutants

In response to DNA damage, Pds1 becomes stabilized after phosphorylation by Chk1, which blocks anaphase entry (Wang ). DNA replication stress induced by HU treatment might also trigger Pds1 phosphorylation by Chk1 to prevent anaphase entry. Previous research has mapped Chk1 phosphorylation sites within Pds1, resulting in the generation of pds1-m8 mutants (S37A, S121A, S132A, S139A, S158A, S170A, S213A, and S212A) with mutated Chk1 phosphorylation sites (Wang ). pds1-m9 contains an additional mutated threonine (T289A), but both pds1-m8 and pds1-m9 mutants exhibit a defective DNA damage checkpoint. Interestingly, we found that the elimination of Chk1 phosphorylation sites in Pds1 rescued the HU sensitivity of cdc55Δ mutants based on the growth on YPD plates containing 100 mM HU (Figure 3A).

FIGURE 3:

Mutation of Chk1 phosphorylation sites in Pds1 suppresses the HU sensitivity of cdc55Δ mutants. (A) Elimination of Chk1 phosphorylation sites in Pds1 (m8 and m9) restores the growth of cdc55Δ cells on HU plates. Saturated cells with the indicated genotypes were 10-fold serially diluted onto YPD plates with or without 100 mM HU. Growth was analyzed after incubation for 2 d (YPD) or 3 d (HU plates) at 30°C. (B) The pds1-m8 mutation suppresses the nuclear division delay in cdc55Δ mutants after HU release. G1-arrested WT, cdc55Δ, and cdc55Δ pds1-m8 cells containing H2A-mApple were released into YPD with 200 mM HU for 2 h. After HU washout, the cells were released into YPD medium at 30°C, and α-factor was added to block the following cell cycle. Samples were collected every 20 min. Budding index and the percentage of cells with undivided DNA were determined using fluorescence microscopy. Nuclear division in some representative cells is shown in the left panel. Arrows: the nucleus. Scale bar, 5 μm. (C) The chk1Δ mutation partially suppresses the HU sensitivity of PP2A mutants. Saturated cells with the indicated genotypes were 10-fold serially diluted onto YPD plates with or without 100 mM HU. Growth was analyzed after 2 d on YPD and 3 d on HU plates at 30°C. (D) The chk1Δ mutant partially suppresses the delayed anaphase entry in cdc55Δ cells after HU release. G1-arrested cells were released into YPD with 200 mM HU for 2 h at 30°C before release into YPD. After release from HU, α-factor was added to block the following cell cycle. Samples were collected every 20 min after HU release. Budding index and Pds1 protein levels are shown. Pgk1, loading control. (E) The chk1Δ mutation partially rescues the nuclear division defect in cdc55Δ cells after HU release. WT, cdc55Δ, and cdc55Δ chk1Δ cells containing H2A-mApple were treated as described in panel B. Cells were imaged via fluorescence microscopy to follow nuclear division (left). Scale bar, 5 μm. Cells were counted at each time point for budding index and the percentage of cells with undivided DNA (right).

We further analyzed the nuclear division kinetics of cdc55Δ and cdc55Δ pds1-m8 mutants containing H2A-mApple under normal growth conditions and after HU exposure. Under normal growth conditions, the slight delay of nuclear division in cdc55Δ cells was suppressed by pds1-m8 (Supplemental Figure S3A). After release from treatment with 200 mM HU, a dramatic nuclear division delay was observed in cdc55Δ mutants, but cdc55Δ pds1-m8 exhibited a nuclear division kinetics similar to that of WT cells (Figure 3B). We also analyzed the viability of cdc55Δ pds1-m8 cells after treatment with HU, but no discernible viability loss was observed for both cdc55Δ single and cdc55Δ pds1-m8 double mutants (Supplemental Figure S3B). These results indicate that Pds1 phosphorylation by Chk1 likely contributes to the anaphase entry delay in PP2A mutants after HU treatment.

Given the dramatic suppression of the HU sensitivity of PP2A mutants by pds1-m8 and pds1-m9 mutants that lack Chk1 phosphorylation sites, we hypothesized that the elimination of Chk1 would have a similar suppression. Interestingly, CHK1 deletion in cdc55Δ mutants only partially suppressed the HU sensitivity (Figure 3C). Similarly, chk1Δ showed partial suppression of the HU sensitivity of pph21Δ pph22Δ mutants. A partial rescue by chk1Δ was also observed with regard to cell cycle progression, nuclear separation, and Pds1 degradation in cdc55Δ mutant cells after release from HU arrest (Figure 3, D and E). Because pds1-m8 and pds1-m9 mutants exhibit stronger suppression of the HU sensitivity of cdc55Δ cells than chk1Δ mutants, it is likely that the phosphorylation of Pds1 by a kinase other than Chk1 also contributes to the HU sensitivity of PP2A mutants.

Our results suggest that Chk1-independent Pds1 phosphorylation may also regulate replication stress recovery. In addition to Chk1, Cdk1 also phosphorylates Pds1, which facilitates the interaction of Pds1 with separase (Esp1) for its nuclear accumulation (Agarwal and Cohen-Fix, 2002). A recent study indicates that PP2A directly reverses Cdk1-mediated Pds1 phosphorylation (Khondker ). Therefore, Cdk1-dependent Pds1 phosphorylation might play a role in replication stress recovery. To test this idea, we acquired the pds1-38 strain from the Cohen-Fix laboratory, wherein the three Cdk1 phosphorylation sites (S277A, S292, and T301A) were mutated to alanine. However, the phosphodeficient pds1-38 mutant did not show any rescue of the HU sensitivity of cdc55Δ mutants; instead, more dramatic slow growth on HU plates was observed for cdc55Δ pds1-38 double mutants (Supplemental Figure S4A). Like Pds1, separase Esp1 is also a substrate of both Cdk1 and PP2ACdc55, and Cdk1-dependent Esp1 phosphorylation promotes anaphase entry (Lianga ). To test whether Esp1 phosphorylation plays a role in replication stress recovery in PP2A mutants, we constructed double mutants of cdc55Δ in combination with either the phosphomimetic esp1-3D or the phosphodeficient esp1-3A. Although esp1-3A showed a slight rescue for the HU sensitivity of cdc55Δ mutants, esp1-3D exhibited a similar rescue (Supplemental Figure S4, B and C), indicating that the reversion of Cdk1-dependent Esp1 phosphorylation is unlikely to play a major role during the recovery from HU arrest. Taken together, our results indicate that Chk1 and an additional kinase act together to phosphorylate Pds1 and prevent anaphase onset after DNA replication stress, but the reversal of this phosphorylation is required for stress recovery.

Defective S-phase checkpoint fails to suppress the delayed recovery from DNA replication stress in PP2A mutants

One major function of a checkpoint is to delay cell cycle progression after its activation, allowing cells more time to fix cell cycle problems. Thus, the S-phase checkpoint might be required for Pds1 stabilization, which contributes to the anaphase entry delay after DNA replication stress in PP2A mutants. If that is the case, a defective S-phase checkpoint would abolish this delay. Within the S-phase checkpoint pathway, Mec1 and Rad53 kinases are the key components, and their function is essential for cell cycle delay as well as the stabilization of replication forks in response to DNA replication stress (Kiser and Weinert, 1996; Pardo ).

Because both kinases are essential for viability, we utilized mec1-1 and rad53-21 point mutants to test their potential suppression of the stress recovery defect in PP2A mutants (Desany ). cdc55Δ, mec1-1, rad53-21 single mutants and double mutants carrying both cdc55Δ and mec1-1/rad53-21 were first arrested in G1 and then released into HU to examine Pds1 protein levels. Pds1 accumulation was observed in all of these single and double mutant cells after release into HU medium (Figure 4, A and B), which is consistent with a previous study (Palou ). For these mutants, we further examined their recovery process following HU treatment. Neither mec1-1 nor rad53-21 was able to restore Pds1 degradation in cdc55Δ mutants after HU release. In contrast, a higher Pds1 level was detected in mec1-1, rad53-21 single mutants as well as double mutants with cdc55Δ after HU release (Figure 4, C and D). Our explanation is that the collapsed replication forks in S-phase checkpoint mutants after HU treatment likely block DNA replication and anaphase entry that is marked by Pds1 degradation. Therefore, we conclude that a defective S-phase checkpoint does not abolish the accumulation of anaphase inhibitor Pds1 in PP2A mutants after DNA replication stress.

FIGURE 4:

Genetic interactions between cdc55Δ and S-phase checkpoint mutants. (A) Pds1 levels in cdc55Δ, rad53-21, and cdc55Δ rad53-21 mutants treated with HU. G1-arrested cells were released into YPD with 200 mM HU at 30°C. Samples were collected every 20 min after G1 release. Budding index and Pds1 protein levels are shown. Pgk1, loading control. (B) Pds1 levels in cdc55Δ, mec1-1, and cdc55Δ mec1-1 mutants treated with HU. Cells were treated as described above. Budding index and Pds1 protein levels are shown. Pgk1, loading control. (C) The rad53-21 mutant fails to suppress the anaphase entry delay in cdc55Δ cells after HU release. G1-arrested cells were released into YPD with 200 mM HU for 2 h before release into YPD medium at 30°C. After release from HU arrest, α-factor was added to block the following cell cycle. Samples were collected every 20 min after HU release. Budding index and Pds1 protein levels are shown. Pgk1, loading control. (D) The mec1-1 mutant fails to suppress the anaphase entry delay in cdc55Δ cells after HU release. The cells were treated as described above. Budding index and Pds1 levels are shown. (E)The cdc55Δ mutant suppresses the viability loss of S-phase checkpoint mutants after HU treatment. WT and mutant cells were grown at 30°C in YPD media until mid–log phase. Cultures were split four ways with HU concentrations at 0, 50, 100, and 200 mM. Samples were collected after 2-h treatment and spread onto YPD plates to determine the plating efficiency after overnight growth at 25°C (n ≥ 200).

Although S-phase checkpoint mutants do not suppress the anaphase entry delay in cdc55Δ cells after HU treatment, surprisingly, cdc55Δ drastically increased the viability of mec1-1 and rad53-21 mutants following exposure to HU at different concentrations, 50, 100, and 200 mM (Figure 4E). A further question is how loss of function of PP2A rescues the viability of S-phase checkpoint mutants treated with HU.

Swe1 stabilization contributes to the recovery defect from HU arrest in PP2A mutants

Three branches of the DNA damage/replication checkpoint have previously been identified, including the Chk1, Rad53, and Swe1/Cdk1 branches (Sanchez ; Palou ). Therefore, we further explored whether Swe1-dependent Cdk1 phosphorylation played role in the delayed recovery following HU exposure in PP2A mutants. Previous studies indicate that PP2ACdc55 is required for the degradation of Swe1, the inhibitory kinase for Cdk1 (Yang ; Liu and Wang, 2006). We first examined the Swe1 protein levels in WT and cdc55Δ and pph21Δ pph22Δ mutants during the normal cell cycle and confirmed that Swe1 levels remained relatively high in PP2A mutants compared with WT cells, with cdc55Δ cells exhibiting a more dramatic phenotype. Moreover, both cdc55Δ and pph21Δ pph22Δ mutants exhibited higher level of slow-migrating Swe1 compared with WT cells, and this phenotype was also more pronounced in cdc55Δ cells (Figure 5A). Swe1 phosphorylation likely contributes to this slow migration (Sreenivasan and Kellogg, 1999). We also examined Swe1 levels in these cells after their release from HU arrest. Increased Swe1 stability was detected in cdc55Δ and pph21Δ pph22Δ mutant cells. Similarly, cdc55Δ cells showed more dramatic Swe1 stabilization than pph21Δ pph22Δ cells after HU release (Figure 5B).

FIGURE 5:

Swe1 stabilization contributes to the delayed recovery from HU arrest in PP2A mutants. (A) Swe1 levels are higher in PP2A mutants during the normal cell cycle. WT, cdc55Δ, and pph21Δ pph22Δ cells with Swe1-myc were grown at 30°C in YPD media until mid–log phase and then arrested at G1 with α-factor. Cells in G1 were released into YPD at 30°C with samples collected every 20 min. α-factor was added back after release for 40 min to block the following cell cycle. Budding index and Swe1 protein levels are shown. Pgk1, loading control. (B) PP2A mutants show defective Swe1 degradation following release from HU arrest. G1-arrested WT, cdc55Δ, and pph21Δ pph22Δ cells expressing Swe1-myc were released into YPD with 200 mM HU for 2 h at 30°C before release into YPD. After release from HU arrest, α-factor was added to block the following cell cycle. Samples were collected every 20 min after HU release. Budding index and Swe1 protein levels are shown. Pgk1, loading control. (C) swe1Δ suppresses the HU sensitivity of cdc55Δ cells. Saturated cells with the indicated genotypes were 10-fold serially diluted onto YPD plates with or without 100 mM HU. Growth was analyzed after 2 d on YPD and 3 d on HU plates at 30°C. (D) swe1Δ rescues the nuclear division defect in cdc55Δ mutants after HU release. WT, cdc55Δ, and swe1Δ cdc55Δ cells containing H2A-mApple were synchronized at G1 with α-factor and then released into YPD with 200 mM HU for 2 h. The cells were released into YPD at 30°C with samples collected every 20 min. After HU release, α-factor was added to block the following cell cycle. Cells were imaged via fluorescence microscopy to follow nuclear division (left). Arrows: the nucleus. Scale bar, 5 μm. Cells were counted at each time point for budding index and the percentage of cells with undivided DNA (right).

After establishing the stabilization of Swe1 in PP2A mutants following HU release, we next tested whether the elimination of Swe1 was able to suppress the HU sensitivity of PP2A mutants. We found that swe1Δ showed clear suppression of the slow growth phenotype of cdc55Δ mutants on plates containing 100 mM HU, although to a lesser extent compared with pds1-m8 and pds1-m9 (Figure 5C). Furthermore, we used the H2A-mApple strains to examine the nuclear division kinetics in cdc55Δ and cdc55Δ swe1Δ cells after release from HU arrest. The nuclear division kinetics was completely restored in the double mutants (Figure 5D). Together, these results indicate that Swe1 stabilization also contributes to the delayed recovery from DNA replication stress in PP2A mutants.

Our results show that swe1Δ and pds1-m8 mutants suppress the anaphase entry delay in cdc55Δ mutants after release from HU arrest. cdc55Δ mutants are also sensitive to microtubule depolymerizing agents, such as nocodazole and benomyl, because of the failure of mitotic arrest (Minshull ; Wang and Burke, 1997; Yellman and Burke, 2006). Thus, we further tested whether swe1∆ or pds1-m8 was able to suppress the sensitivity of cdc55Δ to spindle poison benomyl, but no suppression was observed (Supplemental Figure S5), indicating that PP2A likely regulates the response to stressful DNA replication and disrupted kinetochore attachment through different mechanisms.

We noticed that the pds1-m8 mutant shows stronger suppression for the HU sensitivity of the cdc55Δ mutant than chk1Δ and swe1Δ. Then, we tested whether swe1Δ chk1Δ double mutants exhibited stronger suppression than each single mutant. Surprisingly, swe1Δ chk1Δ cdc55Δ triple mutants showed slow growth on HU plates, like the cdc55∆ single mutant (Supplemental Figure S6A). It is possible that Swe1 and Chk1 play additional roles in response to DNA replication stress, and the slight HU sensitivity of swe1Δ chk1Δ mutants supports this speculation. We also compared the viability of cdc55Δ single and double/triple mutants in combination with chk1Δ, swe1Δ, or swe1Δ chk1Δ after HU exposure. All these mutants did not show significant viability loss (Supplemental Figure S6B). Therefore, swe1Δ, chk1Δ, and pds1-m8 mutants restore the efficient recovery of cdc55Δ mutants from HU arrest without causing viability loss.

Elimination of Swe1-dependent Cdk1 phosphorylation suppresses the HU sensitivity of the cdc55Δ mutant

Swe1 inhibits Cdk1 kinase activity by phosphorylating tyrosine 19 (Y19) on Cdc28 in budding yeast (Booher ). Thus, Cdc28-Y19 phosphorylation might play a role in the delayed recovery from HU-induced cell cycle arrest in PP2A mutants. To test this idea, we used the phosphodeficient cdc28F19 mutant, in which tyrosine 19 was replaced with phenylalanine to abolish Cdk1-Y19 phosphorylation by Swe1 kinase (Amon ). We found that cdc28F19 largely suppressed the slow growth phenotype of cdc55∆ cells on HU plates (Figure 6A). Following release from HU-induced arrest, cdc28F19 cdc55Δ cells exhibited kinetics for Pds1 degradation and nuclear segregation similar to those of WT cells (Figure 6, B and C). In addition, cdc28F19 cdc55Δ cells did not show increased viability loss after HU treatment compared with cdc55Δ cells (Figure 6D). Together, our results suggest that Swe1-dependent Cdk1-Y19 phosphorylation plays a major role in the delayed recovery from HU-induced arrest in PP2A mutants.

FIGURE 6:

Abolishment of Swe1-mediated Cdk1 phosphorylation suppresses the HU sensitivity of cdc55Δ mutants. (A) The cdc28F19 mutation suppresses the HU sensitivity of cdc55Δ. Saturated cells with the indicated genotypes were 10-fold serially diluted and spotted onto YPD plates with or without 100 mM HU. Growth was analyzed after incubation for 2 d (YPD) or 3 d (HU plates) at 30°C. (B) cdc28F19 largely suppresses the delayed anaphase entry in cdc55Δ cells after HU release. G1-arrested cells were released into YPD with 200 mM HU for 2 h at 30°C before release into YPD. After HU release, α-factor was added to block the following cell cycle. Samples were collected every 20 min after HU release. Budding index and Pds1 protein levels are shown. Pgk1, loading control. (C) cdc28F19 rescues the nuclear division defect in cdc55Δ mutants after HU release. WT, cdc55Δ, cdc28F19, and cdc28F19 cdc55Δ cells containing H2A-mApple were synchronized at G1 and then released into YPD with 200 mM HU for 120 min. Cells were then released into YPD at 30°C with samples collected every 20 min. After HU release, α-factor was added to block the following cell cycle. Cells were counted at each time point for budding index and the percentage of cells with undivided DNA (top). Cells were also imaged via fluorescence microscopy to follow nuclear division (bottom). Arrows: the nucleus. Scale bar, 5 μm. (D) cdc28F19 cdc55Δ mutants do not show viability loss after HU exposure. WT, cdc55Δ, cdc28F19, and cdc28F19 cdc55Δ cells were grown at 30°C in YPD media until mid–log phase, and then 200 mM HU was added. Samples were collected at 0, 2, 4, and 8 h and spread onto YPD plates to determine the plating efficiency after overnight growth at 25°C (n ≥ 200).

Swe1 acts upstream of Pds1 to regulate the recovery from DNA replication stress

Our results indicate that Swe1 and Pds1 act downstream of PP2A in the recovery from cell cycle arrest induced by HU. Next, we assessed the relationship between Swe1 and Pds1 in this recovery process. For this purpose, we first examined whether Swe1 stabilization in PP2A mutants was dependent on Pds1 phosphorylation. WT, cdc55Δ, and cdc55Δ pds1-m8 mutants with Swe1-myc were arrested with HU, and Swe1 protein levels were determined after release from HU arrest. As expected, cdc55Δ cells showed delayed Swe1 degradation. Strikingly, we observed an even stronger Swe1 stabilization in cdc55Δ pds1-m8 cells compared with cdc55Δ cells after HU release, although the accumulation of large-budded cells was abolished by pds1-m8 (Figure 7A). Therefore, cdc55Δ pds1-m8 cells still show impaired Swe1 protein degradation despite the restoration of efficient anaphase entry in these cells after HU treatment. To further analyze the relationship between Swe1 and Pds1, we performed a reciprocal experiment, in which Pds1 levels were tracked in cdc55Δ swe1Δ mutant cells. Cells were first arrested with HU and then released. Pds1 levels were stabilized in cdc55Δ cells after HU release, but this stabilization was largely diminished in cdc55Δ swe1Δ cells. swe1Δ also restored cell cycle progression in cdc55Δ mutants after HU release based on the budding index (Figure 7B). The abolished Pds1 stabilization in cdc55Δ swe1Δ cells after HU release suggests that Pds1 likely acts downstream of Swe1 in the process of recovery from HU arrest.

FIGURE 7:

As the downstream targets of PP2A, Swe1 acts upstream of Pds1 to regulate cell cycle recovery from HU arrest. (A) Swe1 levels remain high in cdc55Δ mutant cells expressing phosphodeficient Pds1 (pds1-m8). G1-arrested cells were released into YPD medium containing 200 mM HU for 2 h at 30°C before release into YPD. After HU release, α-factor was added to block the following cell cycle. Samples were collected every 20 min after HU release. Budding index and Swe1 protein levels are shown. Pgk1, loading control. (B) The swe1Δ mutant abolishes delayed Pds1 degradation in cdc55Δ cells after HU release. Cells with the indicated genotypes were treated as described above. Budding index and Pds1 protein levels are shown. Pgk1, loading control. (C) swe1Δ and pds1-m8 mutants restore the HU-induced viability loss in rad53-21 cdc55Δ. Mid–log phase cell cultures of WT and mutants were split four ways with HU concentrations at 0, 50, 100, and 200 mM. Samples were collected after 2-h treatment and spread onto YPD plates to determine the plating efficiency after overnight growth at 25°C (n ≥ 200). (D) The pds1-m8 mutant restores the HU-induced viability loss in mec1-1 cdc55Δ mutants. Cells were treated as described above. The plating efficiency is shown. (E) Proposed model for PP2ACdc55-Swe1-CDK-Chk1-Pds1 pathway that promotes the recovery from DNA replication stress.

An interesting finding is that the cdc55Δ mutant largely suppresses the viability loss of the S-phase mutants rad53-21 and mec1-1 following HU exposure (Figure 4E). Because our data indicate that Swe1 and Pds1 act downstream of PP2A, we further examined whether swe1Δ or pds1-m8 could abolish this suppression by cdc55Δ. Given that cdc55Δ, mec1-1, and rad53-21 mutants are all HU sensitive, and swe1Δ or pds1-m8 suppresses the HU sensitivity of cdc55Δ, the strain construction would be challenging. Therefore, we first inserted a Sphis5 marker in a chromosome locus close to either rad53-21 or mec1-1 to generate rad53-21-Sphis5 and mec1-1-Sphis5 strains. With these strains, we first generated rad53-21-Sphis5 and rad53-21-Sphis5 strains. Interestingly, introduction of the pds1-m8 or swe1Δ mutant into rad53-21-Sphis5 cells caused viability loss after HU treatment (Figure 7C). We were unable to construct a mec1-1-Sphis5 strain, indicating synthetic lethality. This lethality is likely a result of disruption of many cell cycle events caused by the loss of function of PP2ACdc55, Mec1, and Swe1 (Palou ). However, the mec1-1-Sphis5 mutant was constructed successfully, and viability loss induced by HU treatment was restored in this triple mutant (Figure 7D). All these results suggest that the suppression of HU-induced viability loss in S-phase checkpoint mutants by cdc55Δ depends on the PP2A downstream targets Swe1 and Pds1. Taken together, these results support the conclusion that PP2ACdc55 promotes anaphase entry after DNA replication stress through the sequential reversal of Swe1-mediated Cdk1 phosphorylation and Chk1-dependent Pds1 phosphorylation. Therefore, our data revealed the PP2A-Swe1-CDK-Chk1-Pds1 axis that is critical for the recovery from DNA replication stress (Figure 7E).

DISCUSSION

DNA replication stress activates the S-phase checkpoint that delays cell cycle progression and stabilizes stalled replication forks. This checkpoint has been studied extensively because of its critical role in genome stability. However, much less is known about the recovery process following the removal of DNA replication stress. Here we report that a protein phosphatase, PP2A, plays a critical role in replication stress recovery. First, we found that the PP2A holoenzyme containing B-regulatory subunit Cdc55 is required for efficient anaphase entry after DNA replication stress. Moreover, our results indicate that PP2ACdc55-dependent degradation of anaphase inhibitor Pds1 and CDK inhibitory kinase Swe1 is important for cell cycle progression after DNA replication stress. We further found that the phosphorylation of Cdk1 at Y19 by Swe1 and the phosphorylation of Pds1 by Chk1 contribute to anaphase entry delay in PP2A mutants after DNA replication stress. Finally, Swe1 degradation likely acts upstream of Pds1 to regulate the recovery from HU arrest. Interestingly, the loss of function of PP2A restores the viability of S-phase checkpoint mutants under DNA replication stress, and this restoration is abolished by either swe1Δ or pds1-m8. These results reveal an unappreciated PP2A-Swe1-CDK-Chk1-Pds1 pathway required for the recovery from DNA replication stress (Figure 7E).

Although our results support the critical role of PP2ACdc55 in the recovery from DNA replication stress, it remains unclear which protein is directly dephosphorylated by PP2A to promote this recovery process. The known PP2ACdc55 substrates are involved in cellular processes such as mitosis and cytokinesis (Moyano-Rodriguez and Queralt, 2020). Previous studies indicate that PP2A along with Cdc14 counteracts CDK-imposed phosphorylation (Godfrey ). Interestingly, a recent work showed increased Pds1 phosphorylation in cdc55Δ cells and that PP2ACdc55 directly dephosphorylates Pds1 to reverses Cdk1-mediated phosphorylation (Khondker ). However, it is unlikely that PP2ACdc55 promotes cell cycle recovery from replication stress by directly reversing CDK-dependent Pds1 phosphorylation, because the pds1-38 mutant lacking the CDK phosphorylation sites was unable to suppress the HU sensitivity of PP2A mutants (Supplemental Figure S4A). Similarly, it is unlikely that PP2A promotes anaphase entry after DNA replication stress by counteracting CDK-dependent phosphorylation of separase Esp1 based on our observation that the suppression of the HU sensitivity of cdc55Δ by phosphodeficient esp1-3A was not significant (Supplemental Figure S4B).

In addition to Pds1, another downstream target of PP2ACdc55 is Swe1, which phosphorylates and inhibits CDK (Sia ). PP2A-dependent Swe1 degradation has been shown to facilitate the transition from G2 to M phase (Yang ). We detected delayed Swe1 degradation in PP2A mutant cells (Figure 5A). Because Cdc5, Cla4, and other kinases phosphorylate Swe1 to promote its degradation (Sreenivasan and Kellogg, 1999; Sakchaisri ; Liu and Wang, 2006), PP2A may promote Swe1 degradation by up-regulating these kinases. In addition, Mec1 kinase phosphorylates the SQ motif of Swe1 after S-phase checkpoint activation (Palou ); thus it will be interesting to test whether PP2A reverses Mec1-dependent Swe1 phosphorylation for its degradation. Swe1 is also phosphorylated by Cdk1 to regulate Cdk1-Swe1 association (Harvey ), but PP2ACdc55 opposes Swe1 phosphorylation by Cdk1, which limits Swe1 activity and allows mitotic CDK activation (Harvey ). Because we observed that cdc28-F19 suppresses the delayed recovery from HU arrest in PP2A mutants, we speculate that one important role of PP2ACdc55 in the anaphase entry after DNA replication stress is to activate mitotic CDK by reversing Swe1-mediated Cdk1 phosphorylation. However, we cannot exclude the possibility that PP2ACdc55 directly dephosphorylates other CDK substrates to promote the recovery from DNA replication stress.

We found that deletion of SWE1 largely abolished the delay in Pds1 degradation in cdc55Δ mutant cells after HU release. In contrast, the pds1-m8 mutant lacking Chk1 phosphorylation sites had no effect on Swe1 degradation in cdc55Δ mutant cells after HU release (Figure 7). Therefore, we conclude that Swe1 acts upstream of Chk1-dependent Pds1 phosphorylation to regulate anaphase entry. Because Swe1 phosphorylates Cdk1 to inhibit its mitotic activity (Harvey ; Lianga ), it is likely that mitotic CDK negatively regulates Chk1-mediated Pds1 phosphorylation to promote anaphase entry (Figure 7E). In support of this possibility, evidence from human cells shows that CDK phosphorylates Chk1 for its nuclear export, which promotes mitotic entry (Enomoto ). In budding yeast, DNA damage checkpoint protein Rad9 is phosphorylated by CDK, and this phosphorylation exhibits dual regulation of the Rad9-Chk1 interaction (Abreu ). One possibility is that S-phase CDK-mediated Rad9 phosphorylation enhances Rad9-Chk1 interaction for checkpoint activation, but further Rad9 phosphorylation by mitotic CDK decreases Rad9-Chk1 interaction, which down-regulates checkpoint activity for anaphase onset. Further experiments are needed to test these ideas. We also showed that mutations of Chk1 sites on Pds1 (pds1-m8 and pds1-m9) resulted in a nearly complete rescue of the HU sensitivity of cdc55Δ mutants, but the rescue by chk1Δ was only partial. This indicates that some of the identified Chk1 phosphorylation sites on Pds1 are also phosphorylated by another kinase, and it will be our future interest to identify this kinase.

One interesting observation from this study is that cdc55Δ mutants drastically rescue the viability of S-phase checkpoint mutants, mec1-1 and rad53-21, after exposure to HU (Figure 4E). These two mutants are extremely sensitive to even low concentrations of HU because of their roles in dNTP synthesis and DNA replication fork stabilization (Desany ; Zhao ; Gupta ; Rodriguez and Tsukiyama, 2013). One explanation is that the delayed anaphase onset in cdc55Δ mutants allows S-phase checkpoint mutants to fix the problems caused by DNA replication stress. Another explanation for this rescue could be the increased dNTP pool or stabilized replication forks in cdc55Δ cells. Indeed, in human cells, active PP2A induces Cdc45 to decouple from the replisome, resulting in the collapse of replication forks (Perl ). Interestingly, we observed that swe1Δ and pds1-m8 restored the viability loss in mec1-1 cdc55Δ and rad53-21 cdc55Δ (Figure 7, C and D). Therefore, an important open question is how the PP2A-Swe1-CDK-Chk1-Pds1 axis modulates the DNA replication process that contributes to the viability loss in S-phase checkpoint mutants after HU exposure.

In summary, we show that PP2A facilitates the recovery from HU-induced cell cycle arrest and the likely downstream targets of PP2A involved in this recovery process include Swe1 and Pds1. Another interesting observation in this research is that defective PP2A suppresses the viability loss of S-phase checkpoint mutants after DNA replication stress. Given that PP2A is a well-conserved holoenzyme across eukaryotic species, it will be important to understand whether the function of PP2A in response to replication stress is also conserved (Orgad ). In the fission yeast Schizosaccharomyces pombe, PP2A regulates the phosphorylation of Wee1, the Swe1 homologue (Lucena ). As described above, human PP2A destabilizes the replisome by dephosphorylating Cdc45 (Perl ). Given that PP2A dysregulation has been found in many solid tumors and leukemia (Cristóbal ; Seshacharyulu ), the functional studies of PP2A in the DNA replication stress response may provide new therapeutic strategies.

MATERIALS AND METHODS

Request a protocol through Bio-protocol.

Yeast strains and growth conditions

The relevant genotypes and the sources of the strains used in this study are listed in Supplemental Table S1. All of the strains listed are isogenic to Y300, a derivative of W303. The pds1-38 strain was a gift from the Cohen-Fix laboratory. The pds1-m8 and pds1-m9 strains were from the Elledge laboratory. The esp1-3A and esp1-3D strains were a gift from the Rudner laboratory. Yeast cells were grown in YPD medium. To arrest cells in G1 phase, 5 μg/ml α-factor was added into cell cultures. After incubation for 120 min, the G1-arrested cells were washed twice with water and then released into fresh YPD medium to start the cell cycle. For HU treatment, G1-arrested cells were released into YPD containing 200 mM HU (Sigma). After incubation for 120 min, the S-phase arrested cells were washed twice with water and then released into fresh YPD medium to start the cell cycle.

H2A-mApple strain construction

The H2A-mApple strain was constructed using a PCR-based method. The template plasmid pHG72 was provided by Hong-Guo Yu, and primers are listed in Supplemental Table S2. PCR products were transformed into Y300 WT cells, and the cells were spread onto histidine dropout plates and incubated at 30°C. Colonies were confirmed via colony PCR and fluorescence microscopy.

Construction of mec1-1 and rad53-21 strains conjugated with Sphis5

A PCR-based method was used to insert the Sphis5 marker into a chromosome locus close to either mec1-1 or rad53-21. A template plasmid (pFA6aHis3MX6) containing Sphis5 was used for the PCR (Longtine ). The primers used for this PCR are listed in Supplemental Table S3. The PCR products were transformed into either AY202 (rad53-21) or AY203 (mec1-1) yeast strains. Cells were spread onto HIS dropout plates and incubated at 30°C to select transformants. Colonies were confirmed via colony PCR.

Western blot analysis

We collected 1.5 ml of yeast cell culture, the cell pellets were resuspended in 100 ml of H2O, and then 100 ml of 0.2 M NaOH was added. The mixture was left at room temperature for 5 min. After centrifugation, the pellets were resuspended in the 1× loading buffer. For Pds1-18myc protein detection, we used 10% acrylamide gels for SDS–PAGE. For Swe1-myc protein detection, we used 8% acrylamide gels. The anti-myc antibody (9E10) (Covance Research Products) was used at 1:1000 dilution. Phosphoglycerate kinase 1 (Pgk1) antibody (Molecular Probes, Eugene, OR) was used at 1:5000 dilution. Proteins were detected with ECL (Perkin-Elmar-Cetus, Norwalk, CT). After enhanced chemiluminescence (ECL), the Western blot membranes were imaged using Bio-Rad ChemiDoc.

PFGE analysis

Collected yeast cells were washed once with water and then fixed with 70% ethanol for 1 h at room temperature. To remove cell walls, cells were resuspended in LiSorb buffer (100 mM lithium acetate, 50 mM Tris-HCl, pH 8.0, 1 mM EDTA, 1 mM sorbitol) and treated with zymolyase at 37°C for 1 h. Then cells were resuspended in TE buffer (10 mM Tris, 10 mM EDTA, pH 8.0). An equal volume of 1% melted agarose (after cooling down to 50°C) was added into cells. After agarose was solidified, cells embedded in agarose blocks were subject to digestion with lysis buffer (100 mM EDTA, 10 mM Tris, 1% sarkosyl, 100 μg/ml proteinase K, pH 8.0) overnight at 50°C. After that, the agarose blocks were washed with TE buffer twice and were ready for PFGE analysis. The CHEF-DR II pulsed-field electrophoresis system (Bio-Rad, Richmond, CA) was used. The running time was 20 h at 6 V/cm with a 60–120-s switch time ramp (14°C).

Fluorescent signal analysis

Strains containing mApple-labeled H2A (HTA1-mApple) or the GFP-marked centromere of chromosome IV (CEN4-GFP) and Tub1-mCherry were collected and fixed with 3.7% formaldehyde at room temperature for 5 min. The cells were then washed with water and resuspended in 1× phosphate-buffered saline. The fluorescent signal was analyzed over time after G1 or S-phase release to examine nuclear division using a fluorescence microscope (Keyence).

Click here for additional data file.

Click here for additional data file.