(A) Assay for DSB-induced cohesion (see experimental procedures). Briefly, cells establish cohesion during S phase using thermo-sensitive mcd1-1p (referred to as S phase cohesin, and diagramed as a yellow circle). These cells are then arrested in G2/M using nocodazole, and then wild-type Mcd1p (Mcd1p^6HA) is expressed along with the HO endonuclease, which makes two DSBs on chromosome III (Chr III) at HO cleavage sites. To test whether the newly loaded cohesin containing wild-type Mcd1p (indicated as a red circle) has become cohesive, the S phase cohesin is specifically inactivated by a temperature shift to 37.5° C. Cohesin that is loaded but not cohesive is diagramed outside of the center axis of the chromosomes. Cohesion of sister chromatids is visualized by LacI-GFP fusion proteins bound to a tandem array of Lac operators integrated at LYS4 on Chr IV, indicated on the diagram by a green bar. The presence of a single GFP spot in the cell indicates cohesion between two sister chromatids whereas two GFP spots indicate loss of cohesion. (B) Example of experimental data for the DSB-induced cohesion assay diagramed in (A). The first column of data represents the basal level of cohesion loss at the permissive temperature when the thermo-sensitive S phase cohesin is active (a). The second column represents the maximum amount of separation of sister chromatids that is observed when S phase cohesion is inactivated (i) and no cohesion has been generated in G2/M. The third and fourth column represent the experimental condition where we address whether cohesion has been generated in G2/M. Without a DSB (indicated by a – sign), no cohesion is generated in G2/M, so the sister chromatids fall apart upon inactivation of S phase cohesion (third column). Alternatively, the presence of a DSB (indicated by a + sign), along with expression of wild-type Mcd1p (indicated by a + sign), results in robust cohesion generation (fourth column) equivalent to the level generated during S phase. (C) DSB-induced cohesion assay in EU3376 (galactose-inducible mcd1-S83D^6HA, eco1-R222G,K222G). The control strain EU3336 (galactose-inducible Mcd1^6HA, eco1-R222G,K222G) was used to obtain the data for the third column. N=3 (with a total of 300-600 cells counted per experimental condition. For more details see experimental procedures). Error bars indicate standard deviation.

(A) G2/M-induced cohesion assay in JH4145 (galactose-inducible Mcd1-K84Q, K210Q^6HA). N=3. Cohesin containing mcd1p-K84Q, K210Q mutant generates cohesion in G2/M, but does not rescue cohesion to the levels established with S phase cohesion (p=.003 Student’s T test) (B) DSB-induced cohesion assay in strain JH4148, which contains the eco1-R222G, K223G allele and has galactose-inducible mcd1p-K84Q, K210Q^6HA. N=3 (C) Experimental design for DSB-induced cohesion assay in JH4153 (galactose-inducible mcd1-K84Q, K210Q^6HA, eco1-203, mcd1-1) and EU3326 (galactose-inducible Mcd1^6HA, eco1-203, mcd1-1). The DSB cohesion assay was modified to allow temporal inactivation of Eco1p prior to inactivation of Mcd1p. The protocol was modified as follows: cells were arrested in early S phase using hydroxyurea for 4 hours 30 minutes, cells were washed 3X, and then resuspended in media containing nocodazole. 1 hour 15 minutes later, the cultures were split to either 30° C (active Eco1p) or 34° C (inactive Eco1p). 45 minutes later, cultures were again split and the HO endonuclease and either Mcd1p or mcd1p-K84Q, K210Q were induced for 1 hour by addition of galactose to the media. Cultures were then shifted to 37.5° C for 45 minutes to inactivate S phase cohesion generated with mcd1-1, and cells were prepared for microscopy. mcd1-1p is active (a) at 34° C and inactive 37.5° C (i). eco1-203p is active (a) at 30° C and inactive at temperatures at or above 34° C (i). (D) DSB-induced cohesion assay in JH4153 (galactose-inducible mcd1-K84Q, K210Q^6HA, eco1-203, mcd1-1) and the control strain EU3326 (galactose-inducible Mcd1^6HA, eco1-203, mcd1-1), which was used to generate the third column (3a). N=3. Error bars indicate standard deviation.

(A) A model of where the Smc3p and Mcd1p mutations fall in the cohesin complex, based off the structure of the C terminus of Mcd1p and the Smc1p head domain (B) smc3-based genome-wide DSB-induced cohesion assay in JH4181 (smc3-42, galactose-inducible Smc3) and JH4192 (smc3-42, galactose-inducible Smc3 and HO endonuclease), which was used to generate the fourth panel. N=3. The assay was preformed in a similar fashion as the mcd1-based cohesion assay with a few exceptions. Cells establish cohesion during S phase using thermo-sensitive smc3-42 (referred to as S phase cohesin). These cells are synchronously released into a nocodazole-induced G2/M arrest from an S phase arrest using Hydroxyurea. Then, wild-type Smc3p is expressed and the HO endonuclease is induced to create a single DSB at the MAT locus. To test whether the newly loaded cohesin containing Smc3p has become cohesive, the S phase cohesin is specifically inactivated by a temperature shift to 37.5° C. Cohesion of sister chromatids of the unbroken Chr IV is visualized by LacI-GFP fusion proteins bound to a tandem array of Lac operators (LacO) integrated at LYS4. The presence of a single GFP spot in the cell indicates cohesion between two sister chromatids whereas two GFP spots indicate loss of cohesion. (C) smc3-based genome-wide DSB-induced cohesion assay in JH5218 (galactose-inducible Smc3, wpl1Δ) and JH5219 (galactose-inducible Smc3 and HO endonuclease, wpl1Δ). N=3. (D) Smc3-based genome-wide DSB-induced cohesion assay in JH5234 (galactose-inducible Smc3, mcd1-K84R, K210R) and JH5236 (galactose-inducible Smc3, mcd1-K84R, K210R, wpl1Δ). N=3. Error bars indicate standard deviation.

Cohesin (yellow circles) is loaded onto chromosomes in G1/S by the Scc2p/Scc4p protein complex. Concomitant with DNA replication, Eco1p-dependent acetylation of Smc3p (blue stars) promotes cohesion generation by antagonizing Wpl1p. Once the cell passes into G2/M, cohesin loading onto chromatin continues but cohesion generation is inhibited, because the relative activity of Eco1p is not sufficient to antagonize Wpl1p. Upon a DSB, cohesin continues to load to CAR sites throughout the genome, and cohesin is also additionally recruited to the break site through Mec1p dependent phosphorylation of H2AX. Chk1p dependent phosphorylation of Mcd1p promotes Eco1p-dependent acetylation of Mcd1p (red circles), which in turn combats Wpl1p to allow cohesion generation in G2/M.

(A) DSB-induced cohesion assay in strains expressing galactose-inducible mcd1p-K84R (EU3399), mcd1p-K210R (JH4132), or mcd1p-K84R, K210R (JH4134). N=3. The cohesion generated in response to a DSB is significant in mcd1-K84 and mcd1-K210 mutants (p=.002, p=< .001 respectively, Student’s T test) (B) Chromsome binding of G2/M-loaded cohesin in strains EU3291 (galactose-inducible Mcd1^6HA) and JH4134 (galactose-inducible mcd1-K84R, K210R^6HA) by chromatin spreads. Cells grown at 30° C were arrested in G2/M using nocodazole. Mcd1p^6HA or mcd1p-K84R, K210R^6HA was then induced for 1 hr 15 minutes by addition of galactose to the media. Cells were prepared for chromatin spreads as described (see experimental procedures) and Mcd1p or mcd1p-K84R, K210R was visualized by indirect immunofluorescence with an antibody against the HA tag. A representative field was shown, with no observable difference in cohesin binding between cells expressing Mcd1p^6HA or mcd1p-K84R, K210R^6HA (C) The binding of cohesin containing Mcd1p or mcd1p-K84R, K210R around MAT on Chr III upon induction of a DSB. Exponentially growing strains JH4144 containing either wild-type Mcd1p or mcd1p-K84R, K210R provided by a TRP1 CEN6 plasmid. Cells were grown at 30° C and arrested in G2/M using nocodazole. An HO break at the MAT locus was induced by the addition of galactose to the culture media for 90 minutes. Cells were fixed and subjected to chromatin immunporecipitation as described using an antibody against Mcd1p (Kindly provided by V. Guacci) N=2. Cohesin is found at CAR196000 at the MAT locus irrespective of the presence of a DSB, and is enriched to the right upon DSB induction. A representative experiment is shown, with error bars representing standard deviation of data obtained from two separate PCR runs, within the same experiment.

DSB-induced cohesion assay in (A) JH4157 (galactose-inducible Mcd1-S83A, K84Q, K210Q^6HA) and the control strain JH2350 (galactose-inducible Mcd1-S83A^6HA), which was used to generate the third column. (B) JH4159 (chk1Δ, galactose-inducible K84Q, K210Q^6HA) and the control strain JH2375 (chk1Δ, galactose-inducible Mcd1^6HA), which was used to generate the third column. (C) G2/M-induced cohesion assay in a strain which over expresses Eco1p and has galactose-inducible mcd1p-K84R, K210R^6HA (JH4141). (D) DSB-induced cohesion assay in JH4149 (galactose-inducible Mcd1-S83D, K84R, K210R^6HA). N=3. Error bars indicate standard deviation.

(A) Viability assay in strains 3084-2A (eco1-203), 3084-2A-pCM87 (eco1-203, pGAL:MCD1^6HA), and 3084-2A-pCM87 K84Q, K210Q(eco1-203, pGAL:mcd1-K84Q,K210Q^6HA). Five fold serial dilutions of the above strains were plated on YEP/2% Lactic Acid/3% Glycerol/2% Galactose plates and put at 23° C, 30° C, and 37° C respectively. (B) JH5256 (Δsmc3:CLONAT, pEU42(SMC3, URA3, CEN/ARS)) was transformed with pEU41 (SMC3, LEU2, CEN/ARS) or its derivative pEU41 K112R,K113R (smc3-K112R, K113R, LEU2, CEN/ARS) along with pVG164 (MCD1, TRP1, CEN/ARS) or its derivative pVG164 K84Q,K21Q (mcd1-K84Q,K210Q, TRP1, CEN/ARS). Five fold serial dilutions of overnight cultures grown in –Leu, -Trp dropout media were spotted on –Leu control plates, or 5-FOA (5-Fluoroorotic Acid) plates and grown at 30° C. An absence of growth on 5-FOA indicates an obligate retention of the pEU42 plasmid bearing wild-type SMC3 (C) Smc3-based genome-wide DSB-induced cohesion assay in JH5207 (galactose-inducible smc3-K112R, K113R). N=3. Cohesin containing smc3p-K112R, K113R generates DSB-induced cohesion, but fails to generate cohesion to the levels established with the S phase cohesin (p<.001) (D) Smc3-based G2/M cohesion assay in JH5201 (galactose-inducible smc3-K112Q, K113Q) and JH5208 (galactose-inducible smc3-K112Q, K113Q and HO endonuclease), which was used to generate the fourth panel. N=3. The cohesion generated in response to a DSB is not significantly different from the cohesion established during S phase (p=.127). Error bars indicate standard deviation.