Scc1 returns after DSBs in late mitosis.

(a) Schematic of the experimental procedure. Cells in logarithmic growth phase were first arrested in telophase (Tel) by incubating them at 34ºC for 3 h (strains bear the cdc15-2 allele). Then, the culture was divided into three subcultures. One served as a mock control, and the other two were treated to generate DSBs, one with β-estradiol (to express the HO endonuclease) and the other with phleomycin. The strain is unable to repair the HO-DSB break by HR-driven gene conversion (Δhmr Δhml). (b) Western blot against Scc1-3myc. This experiment compares Scc1-3myc levels in an asynchronous, G2- and telophase-blocked cultures. The leftmost lane (No tag) is a control strain for Scc1 without the 3myc epitope tag. Pgk1 protein served as a housekeeping. Ponceau S staining is also shown as a loading control. Asyn.: Asynchronous. Tel: Telophase. +Phle: 10 mg·mL-1 phleomycin. +HO: 2 μM β-estradiol. *: Unspecific band detected by the α-myc antibody just over the Scc1-3myc signal. (c) Like in (b) but against Smc1-6HA. (d) Like in (c) but against Smc3-6HA. (e) Quantification of Scc1-3myc and Smc1-6HA levels in all conditions (mean ± sem, n=3). Statistical comparisons are shown for selected pairs (**, p<0.01; one-way ANOVA, Tukey’s post hoc).

Scc1 forms a reconstituted chromatin-bound cohesin complex after the HO DSB in late mitosis.

Cells were treated as in Figure 1a. (a) Co-IP of the core cohesin complex. Smc1-6HA was used as the bait protein; the leftmost lane includes a control with an untagged Smc1. (b) Chromatin fractionation of the strain expressing Smc1-6HA and Scc1-3myc. Relative protein levels (to the mock sample) are indicated under the blots. Pgk1 and histone H3 (HH3) were included as reporters of the cytosolic and chromatin-bound fractions, respectively. The asterisk (*) indicates an unspecific band in the whole cell extract (WCE), that is absent from the chromatin-bound fraction. (c) Cohesin does not bind the HO DSB in telophase. Chromatin immunoprecipitation (ChIP) of Smc3-6HA with (+HO) and without (-HO) HO induction (mean ± sem, n=3).

Homologous recombination repairs the HO-DSB in telophase.

(a,b) Resection in late mitosis is almost as efficient as in G2/M. Cells were arrested in G2/M or telophase (Tel), and then the HO DSB was generated by adding β-estradiol. Samples were taken at the indicated times to monitor the kinetics of resection. This strain is unable to repair the HO-DSB break by HR-driven gene conversion (Δhmr Δhml). (a) Charts depicting the resection kinetics for two different amplicons located at 726 bp and 5.7 kb downstream the HOcs break (mean ± sem, n=3; fresected is the proportion of resected DNA). (b) Representative western blots against Rad53 to follow the sensing of DNA damage detection in G2/M versus late mitosis. (c-f) Yeast cells use HR to repair the HO DSB in late mitosis. Cells were first blocked in telophase (Tel). Then, the HO DSB was generated by adding β-estradiol. After 1 hour, the β-estradiol was washed away, and samples were taken to monitor the repair for 3 hours. The strain can repair the HO-DSB break by HR-driven gene conversion (Δhmr HML). (c) Schematic of the fragments obtained after a StyI digestion for both MATa and MATα sequences. The probe to detect the fragments by Southern blot is in blue. When the MATa locus is intact, the digestion gives rise to a 0.9 kb fragment. The HO cutting site (HOcs) is located within the StyI-digested MATa locus. Thus, the HO-driven DSB shortens the fragment to 0.7 kb. HR leads to a gene conversion to MATα, which results in the loss of a StyI restriction site and a new fragment of 1.8 kb. (d) Representative Western blot analyses for HO induction and subsequent degradation (tagged with Flag epitope) and the DSB sensing through Rad53 hyperphosphorylation. A Pgk1 western blot served as a housekeeping, and Ponceau S staining of the membrane as a loading control for all lanes. The leftmost lane in the Ponceau S corresponds to the protein weight marker. (e) Representative Southern blot for the MAT switching assay in late mitosis. Alongside the MAT probe, a second probe against the ACT1 gene (1.1 kb fragment) was included for normalization. The MAT probe also recognizes an allele-independent MAT distal fragment (2.2 kb). (f) Quantification of relative band intensities in the MAT switching Southern blots (mean ± sem, n=3). Individual values were normalized to the ACT1 signals. Then, every lane was normalized to MATa at the arrest. Tel: Telophase. +βE: β-Estradiol addition.

Cohesin removal does not alter HR at the MAT locus in late mitosis.

(a) Serial dilution spot assay for the smc3:aid* strains. The fitness effect of Smc3-aid* degradation was tested in the smc3:aid* OsTIR1 strain in the presence of 8 mM IAA. The same strain without the ubiquitin-ligase OsTIR1 was also used as the control. (b-d) Cells of four strains with all possible combination of the smc3:aid* and OsTIR1 pairs were treated as in Figure 1a but including a 1 h IAA step between the cdc15-2 arrest and the induction of HO. Then, HO was induced for 1h with β-estradiol (1h+βE sample), after which it was washed off and the cells were allowed to recover from the DSB for 2h (3h-βE sample). IAA and the Tel arrest were maintained throughout the experiment. (b) Representative Western blot of the four strains at the time of the arrest (Tel) and 1h after IAA addition (+IAA). Note the confirmation of the different genotypic combinations, as well as the Smc3-aid* decline after IAA addition. (c) Representative Southern blot for the MAT switching assay of the four strains in the presence of 8 mM IAA. Note that HO cutting is less efficient than without IAA (see Figures 3e and S6b,c) but gene conversion is still possible after degrading Smc3-aid* in late mitosis. (d) Quantification of relative MAT switching (mean ± sem, n=3). Gene conversion to MATα in the 3h-βE sample was normalized to the cut HOcs in the 1h+βE sample. Then, gene conversion yield was normalized to the wild type (no tagged Smc3, no OsTir1) strain. Tel: Telophase. +βE: β-Estradiol addition. −βE: β-Estradiol removal.+IAA: Indole-acetic acid. (e) Model of how DSBs reconstitute cohesin in late mitosis. The cohesin complex (Smc1-Smc3-Scc1) entraps sister chromatids before anaphase. (1) At anaphase onset, Esp1 (separase) is activated to cleave Scc1 and release sister chromatids for segregation. (2) If DSBs occur before cytokinesis, Scc1 returns and reconstitutes the cohesin complex. At least a fraction of the complex binds to chromatin. The reappearance of Scc1 likely occurs through the inhibition of Esp1 and/or post-translational protection of Scc1 by the DNA damage checkpoint.

Strains used in this work.

Co-IP of the core cohesin copmlex.

This is a biological replicate of the experiment shown in Figure 2a.

Assessment of chromatin-bound cohesin after DSBs in late mitosis.

Related to Figure 2b. (a) Optimization of the chromatin fractionation using the chromatin-bound pool of Smc3 (acetylated Smc3; acSmc3) and the histone H3 (HH3). The best results were obtained after three successive rounds of fractionation (pellet #3, P3; see Material and Methods). S, supernatant; P, pellet. (b) The strain expressing Smc3-6HA was subjected to the same experiment and chromatin fractionation performed in Figure 2b for Smc1-6HA and Scc1-3myc. Note that there are more chromatin-bound acSmc3 after DSBs, especially after the HO DSB.

Chromatin imumnoprecipitation (ChIP) of Smc3-6HA in G2/M.

Related to Figure 2c. The ChiP was performed upon cells arrested in G2/M with nocodazole and further subdivided into two subcultures, one in which the HO DSB was generated (+HO) and one without the DSB (-HO). The strain is unable to repair the break by gene conversion (Δhmr Δhml) and the HO induction lasted 2 h. Note the increase in Smc3 binding in the vicinity of HOcs after the DSB. This experiment was performed in parallel to the second repetition for Tel arrested cells shown in Figure 2c.

The principle of the qPCR resection assay.

Related to Figure 3a,b. (a) Schematic representation of the HO resection assay. Primers (blue arrows) are designed to amplify a sequence that contains a StyI target site adjacent to the HO cutting site (HOcs). When this restriction enzyme is used on the extracted genomic DNA, amplification is inhibited. If resection extends beyond the StyI site, StyI does not cut and the primers can amplify. (b) Summary table of the amplification yield obtained after the StyI digestion during HOcs resection.

Features of HR to repair the HOcs DSB in late mitosis.

Related to Figure 3c-f. Cells were first blocked in the cdc15-2 arrest at 34 ºC for 3 hours (Tel). Then, the HO DSB was generated by adding β-estradiol. After 1 hour, the β-estradiol was washed away, and samples were taken to monitor the repair for 3 hours. (a-c) Kinetics of the HOcs DSB repair in yku70Δ. Yku70 is part of a conserved complex that recognizes a DSB and drives its repair towards NHEJ. (d-f) Kinetics of the HOcs DSB repair in rad9Δ. Rad9 is a mediator in the DDC that promotes activation of the effector kinase Rad53. (g-i) Kinetics of the HOcs DSB repair in mre11 Δ. Mre1 is part of the MRX complex, which tethers the DSB ends and facilitates end resection. (a,d,g) Representative Western blot analyses for HO induction and subsequent degradation (tagged with Flag epitope) and the DSB sensing through Rad53 hyperphosphorylation (note that Rad53 does not get hyperphosphorylated in rad9Δ). A PGK1 western blot served as a housekeeping and Ponceau S staining of the membrane as a loading control for all lanes. The leftmost lane in the Ponceau S corresponds to the protein weight marker. (b,e,h) Representative Southern blots for the MAT switching assays in late mitosis. Alongside the MAT probe, a second probe against the ACT1 gene (1.1 kb fragment) was included for normalization. The MAT probe also recognizes an allele-independent MAT distal fragment (2.2 kb). (c,f,i) Quantification of relative band intensities in the MAT switching Southern blots (mean ± sem, n=3). Individual values were normalized to the ACT1 signals. Then, every lane was normalized to MAT a at the arrest. Note that repair kinetics is similar to that of the wild type in Figure 3e,f. Tel: Telophase. +βE: β-estradiol addition.

MAT switching yield of Smc3-aid* OsTIR1 upon IAA addition.

Cells were treated as in Figure 3d-f but including a 1 h IAA step between the cdc15-2 arrest and the induction of HO. IAA was then maintained throughout the experiment. (a) Representative Western blot as in Figure 3d but also including Smc3-aid* decline after IAA addition. Note that the chromatin-bound acSmc3 is degraded as well. (b) Representative Southern blot for the MAT switching assay as in Figure 3e. (c) Quantification of relative band intensities for MAT switching Southern blots (mean ± sem, n=3). Values were normalized as in Figure 3f. The MAT switching in the wild type strain without IAA is represented alongside for comparison. Note that there is a decrease in the cutting efficiency of HO (MAT a levels remained higher) as well as in the gene conversion into MAT α. However, these phenotypic changes are due to IAA rather than cohesin depletion (see Figure 4c). Tel: Telophase. +βE: β-Estradiol addition. +IAA: Indole-acetic acid.