Regulatory features of Srs2 and corresponding mutants.

(A) Schematic of Srs2 protein domains, regulatory features, and mutations affecting each of its features. Cdk1 phosphorylation sites depicted here include T604, S698, S879, S893, S938, S950, and S965 (Chiolo et al., 2005). Potential Mec1 phosphorylation sites include S890 and S933 (Albuquerque et al., 2015; Faca et al., 2020). Sumoylation sites mapped previously include K1081, K1089, and K1142 (Saponaro et al., 2010). Protein-interaction domains or motifs include the Rad51 binding domain (BD), PCNA Interaction Motif (PIM), and SUMO Interaction Motif (SIM) (Colavito et al., 2009; Kolesar et al., 2016; Kolesar et al., 2012). The residues for these domains are indicated. Mutant alleles disabling each of these features are included in the parentheses.

(B) Mutations affecting Srs2 features do not affect Srs2 protein level. Protein extracts were prepared from asynchronous cells. Srs2 was examined using anti-Srs2 antibody by immunoblotting. Srs2 levels in each mutant were first normalized to the Pgk1 loading control and then to those in wild-type cells. Mean of two biological isolates per genotype is graphed with error bars representing standard deviation (SD).

Srs2 PCNA binding and sumoylation are involved in suppressing rfa1 mutants

(A) Schematic to show that Srs2’s roles in checkpoint dampening (anti-checkpoint for simplicity) and DNA repair both contribute to genotoxin resistance.

(B) Schematic to highlight the mutual suppression between srs2Δ and rfa1 mutants and the underlying effects on the RPA-ssDNA levels in cells.

(C) srs2-ΔPIM and -3KR suppress rfa1-zm2 sensitivity to CPT. A 10-fold serial dilution of cells of the indicated genotypes were spotted and growth was assessed after incubation at 30°C for 40 hours unless otherwise noted.

(D) Several srs2 mutants did not suppress rfa1-zm2 sensitivity to CPT. Experiments were performed as in panel 2C.

(E) srs2-ΔPIM and -3KR suppress rfa1-t33 sensitivity to CPT. Top: schematic of the Rfa1 protein domains and rfa1 mutant alleles. Bottom: srs2-ΔPIM and -3KR improved CPT sensitivity of rfa1-t33 cells. Experiments were performed as in panel 2C.

srs2-ΔPIM and -3KR correct checkpoint abnormalities in rfa1-zm2 cells.

(A) srs2-ΔPIM and -3KR reduce the levels of active Rad53 in rfa1-zm2 cells. Protein extracts were examined after G1 cells were released into cycling in the presence of 16 ug/ml CPT for 2 hours. Activated Rad53 was detected by the F9 antibody by immunoblotting. Active Rad53 signals in each indicated strain were compared to the Pgk1 loading control and normalized to those of rfa1-zm2. Two biological isolates per genotype were examined in two technical repeats and results are graphed with error bars representing SD. T-test generated p value is indicated.

(B) srs2-ΔPIM and -3KR allow better G1 entry of rfa1-zm2 cells. Experiments were performed as in panel A and samples were collected at indicated time points. FACS profiles of the samples are shown at the top and percentages of G1 cells after 2 hours of CPT treatment (CPT 2 h) are plotted at the bottom. Two biological isolates per genotype were examined in two technique repeats and results are graphed with error bars representing SD. T-test generated p value is indicated.

Srs2 sumoylation depends on Srs2 PIM motif and the Mec1 kinase

(A) Srs2 sumoylation level is reduced in srs2-ΔPIM cells. Samples were collected after CPT treatment as in Figure 3A and sumoylated proteins were enriched as described in the text and methods. Sumoylated form of Srs2 (Srs2-Su) migrated slower than unmodified Srs2 on gel and both were detected using anti-Srs2 antibody in immunoblotting. Equal loading is indicated by Ponceau-S stain (Stain). The levels of sumoylated Srs2 relative to those of unmodified Srs2 were normalized to wild-type cells and plotted. Two biological isolates per genotype were examined in two technique repeats and results are graphed with error bars representing SD. T-test generated p value is indicated. ns, not significant.

(B) Srs2 sumoylation level during CPT treatment. Samples were collected as described in Figure 3B. Experiments were conducted, and immunoblotting data is presented as those in panel A. Time 0 samples are G1 cells before CPT treatment. FACS analysis at corresponding time points are included below the plot.

(C) Rad9 phosphorylation level during CPT treatment. Samples used in panel B were examined for Rad9 phosphorylation using an anti-Rad9 antibody. Unmodified and phosphorylated (Rad9-p) Rad9 forms are indicated. FACS analysis at corresponding time points are indicated below blots.

(D) Quantification of Srs2 sumoylation and Rad9 phosphorylation level changes during CPT treatment. Srs2 sumoylation signals were normalized based on unmodified form. Values at each timepoint were relative to that of G1 sample (Time 0) before CPT treatment, and fold changes were indicated on left y-axis. Rad9 signals were normalized to non-modified form and percentage of phosphorylated form were indicated on right y-axis. Two biological isolates per genotype were examined in two technique repeats and results are graphed with error bars representing SD.

(E) Mec1 is required for Srs2 sumoylation. Experiments were performed and data are presented as in panel A.

Mec1 and RPA phosphorylation did not link to Srs2 anti-checkpoint function.

(A) Mec1 autophosphorylation at S1964 during CPT treatment. Samples were collected as described in Figure 3B. Ddc2-myc was immunoprecipitated and the associated Mec1 was co-purified. The anti-Mec1-S1964-p antibody detected the phosphorylated form of Mec1 and a non-specific band (*). Quantification below the blot was based on results from two biological isolates.

(B) mec1-S1964A did not rescue rfa1-zm2 sensitivity toward CPT. Experiments were performed as in Figure 2C.

(C) Examination of the rfa1-S178A and -S178D mutants. Top: schematic of the Rfa1 protein domains and position of Ser178. ssDNA binding domain (DBD) A-C are shown. Bottom, rfa1-S178A and -S178D had no effect on srs2Δ’s sensitivity to CPT. Experiments were performed as in Figure 2C.

(D) Examination of the rfa2-3SA and -3SD mutants. Top, schematic of the Rfa2 protein domains and position of Ser122, Ser187 and Ser189 mutated in rfa2-3SA and -3SD. ssDNA binding domain (DBD)-D is indicated. Bottom, rfa2-3SA has no effect on srs2Δ’s sensitivity toward CPT while rfa2-3SD showed additive effect. Experiments were performed as in Figure 2C.

A working model for the regulation of Srs2-mediated checkpoint dampening.

Genomic sites containing ssDNA gaps can be generated during disrupted DNA replication and DNA repair under genotoxin treatment. The resultant ssDNA gaps can be coated by RPA, which then recruit the Mec1/Ddc2 kinase complex. Mec1 is subsequently activated by DDC proteins, such as the 9-1-1 complex that demarcates the 5’ end junctions of ssDNA gaps (not drawn).

After DDC is activated for a prolonged period, Srs2 recruited to PCNA that demarcates the 3’ end junction flanking the ssDNA gaps can be sumoylated. Srs2 sumoylation level increases as Mec1 activation heightens and can result in more efficient RPA removal, leading to Mec1 downregulation.

srs2 mutants examined for CPT sensitivity and interactions with rfa1 mutant.

(A) srs2 mutants exhibit different levels of CPT sensitivity. srs2-ΔPIM showed slow growth on 8 µg/ml CPT. Sensitivity srs2-SIMMUT toward CPT was seen at three CPT concentrations. srs2-7AV showed slow growth on media containing 4 and 8 µg/ml CPT. srs2Δ and rfa1-zm2 are mutually suppressive for CPT sensitivity. Experiments are performed as in Figure 2C.

(B) srs2-ΔPIM confers better suppression of rfa1-zm2’s CPT and MMS sensitivity than srs2-3KR. We note that mild suppression of rfa1-zm2 by srs2-3KR was only seen at 0.03% MMS plates. Experiments are performed as in Figure 2C, and incubation times are noted. Dashed lines indicate the removal of superfluous rows.

(C) srs2-ΔPIM suppresses rfa1-t33 sensitivity toward MMS. Dashed lines indicate the removal of superfluous rows.

srs2-ΔPIM and slx4RIM are additively defective in DDC down-regulation.

(A) The srs2-ΔPIMΔ slx4RIM double mutant shows stronger CPT sensitivity than either single mutant. Experiments are performed as in Figure 2C.

(B) srs2-ΔPIM and slx4RIM are additive for increasing the level of active Rad53 when cells are treated with CPT. Experiments were conducted and quantifications are presented as those described in Figure 3A, except that three biological isolates are included.

(C) srs2-ΔPIM and slx4RIM are additive for reducing cells exiting from G2/M arrest at 2 hours post CPT treatment. Experiments were conducted and quantifications are presented as those described in Figure 3B, except that three biological isolates are included.

Confirmation of Srs2 sumoylation and Mec1-S1964 phosphorylation.

(A) Detection of Srs2 sumoylation. Protein extracts from asynchronous cells were bound to Ni-NTA beads and the elutes were examined by immunoblotting using an anti-Srs2 antibody. Sumoylated Srs2 form was detected in wild-type cells expressing 8His-tagged SUMO (Smt3) but not in cells lacking this construct or in srs2Δ cells. Unmodified Srs2 showed unspecific Ni-NTA bead binding, thus being detectable in cells containing Srs2 regardless of the SUMO status.

(B) Detection of Mec1-S1964 phosphorylation. Cells were treated with CPT for 2 hours before immunoprecipitating Ddc2-myc was conducted. Ddc2 and the co-immunoprecipitated Mec1 were examined by immunoblotting. Phosphorylation of Ser1964 of the Mec1 protein was detected using the anti-Mec1-S1964-p antibody. This antibody also detected a non-specific band (*). Cells containing untagged Ddc2 or the mec1-S1964A mutation were used as controls.

Examination of Mec1-S964 phosphorylation.

(A) mec1-S1964A does not affect the CPT sensitivity of srs2-3KR or srs2-3KR rfa1-zm2 cells. Experiments were performed as in Figure 2C.

(B) The effects of mec1-S1964E on the CPT sensitivity of srs2-3KR and srs2-ΔPIM cells with or without rfa1-zm2. Experiments were performed as in Figure 2C.

(C) mec1-S1964A or -S1964E does not affect Srs2 sumoylation. Experiments were performed as in Figure 4A. Mean values of three biological isolates per genotype are graphed with error bars representing SD. ns, not significant.

Strains used in this study.

All strains are derived from W1588-4C, a RAD5 derivative of W303 (MATa ade2-1 can1-100 ura3-1 his3-11,15, leu2-3, 112 trp1-1 rad5-535). Only one strain is listed per each genotype, but at least two independent isolates of each genotype were used in the experiments.