Cohesin Causes Replicative DNA Damage by Trapping DNA Topological Stress

Abstract

DNA topological stress inhibits DNA replication fork (RF) progression and contributes to DNA replication stress. In Saccharomyces cerevisiae, we demonstrate that centromeric DNA and the rDNA array are especially vulnerable to DNA topological stress during replication. The activity of the SMC complexes cohesin and condensin are linked to both the generation and repair of DNA topological-stress-linked damage in these regions. At cohesin-enriched centromeres, cohesin activity causes the accumulation of DNA damage, RF rotation, and pre-catenation, confirming that cohesin-dependent DNA topological stress impacts on normal replication progression. In contrast, at the rDNA, cohesin and condensin activity inhibit the repair of damage caused by DNA topological stress. We propose that, as well as generally acting to ensure faithful genetic inheritance, SMCs can disrupt genome stability by trapping DNA topological stress.

Keywords: DNA damage; DNA replication stress; DNA topology; SMC; cohesin; condensin; topoisomerases.

Graphical abstract

Figure 1

Depletion of Top2 during S Phase Causes H2AS129P Enrichment at Centromeres and over the rDNA Repeats (A) Experimental setup of ChIP-seq experiments, indicating how the post-replication cell populations used for the ChIP-seq experiments were prepared. A representative FACS analysis of DNA content of each of the indicated stages of the experiment is shown. (B) The relative enrichment of H2AS129P over H2A ChIP around centromeres in cells is shown either with wild-type expression of Top2 in parental cells (green) or depleted of Top2 (blue) in top2-td cells, both released into the cell cycle under the restrictive conditions. Graph shown is generated from a pile up of the profiles of all centromeres and is an average of two repeats. (C) The relative enrichment of H2AS129P over H2A ChIP across the rDNA repeats in cells either with wild-type expression of Top2 (green) or depleted of Top2 (blue) in top2-td cells, both released into the cell cycle under the restrictive conditions. Graph shown is an average of two repeats. (D) The relative enrichment of H2AS129P over H2A ChIP around centromeres in cells is shown either with wild-type expression of Top2 in parental cells (turquoise) or in top2-4 cells (purple), both released into the cell cycle under the restrictive conditions. Graph shown is generated from a pile up of the profiles of all centromeres and is an average of two repeats. (E) The relative enrichment of H2AS129P over H2A ChIP across the rDNA repeats in cells either with wild-type expression of Top2 in parental cells (turquoise) or in top2-4 cells (purple), both released into the cell cycle under the restrictive conditions. Graph shown is an average of two repeats. See also Figures S1, S2, and S7.

Figure 2

Depletion of Top2 during S Phase Causes H2AS129P Enrichment at Centromeres and over the rDNA Repeats in a Replication-Fork-Dependent Manner (A) The relative enrichment of H2AS129P over H2A ChIP around all centromeres in cdc45-td-depleted cells either with WT expression of Top2 in cdc45-td (brown) or depleted for Top2 and Cdc45 (pink) in cdc45-td top2-td cells, both released into the cell cycle under the restrictive conditions. Data for WT (green) and top2-td cells (blue) from Figure 1 are shown for comparison. Graph shown is generated from a pile up of the profiles of all centromeres and is an average of two repeats. (B) The relative enrichment of H2AS129P over H2A ChIP across the rDNA repeats in cdc45-td-depleted cells either with WT expression of Top2 in cdc45-td (brown) or depleted for Top2 and Cdc45 (pink) in cdc45-td top2-td cells, both released into the cell cycle under the restrictive conditions. Data for WT (green) and top2-td cells (blue) from Figure 1 are shown for comparison. Graph shown is an average of two repeats. (C) FACS analysis of DNA content for one repeat of each of the indicated stages of the experiment. Second repeat is shown in Figure S7.

Figure 3

Depletion of Top2 during S Phase Causes H2AS129P Enrichment at Centromeres and over the rDNA Repeats in a Manner that Causes Accumulation of ssDNA (A) The relative enrichment of RPA1-ChIP over input around all centromeres in cells either with WT expression of Top2 in parental cells (green) or depleted of Top2 (blue) in top2-td cells, both released into the cell cycle under the restrictive conditions. Graph shown is generated from a pile up of the profiles of all centromeres and is an average of three repeats. (B) The relative enrichment of RPA1-ChIP over input across the rDNA repeats in cells either with WT expression of Top2 (green) or depleted of Top2 (blue), both released into the cell cycle under the restrictive conditions. Graph shown is an average of three repeats. (C) The relative enrichment of RPA1-ChIP over input in WT cells released into 200 mM HU under the restrictive conditions for 60 min around all origins (blue), late origins (black), and early origins (red). Origin data were used from Soriano et al. (2014). Graph shown represents one experiment. See also Figure S7.

Figure 4

Inactivation of Cohesin Suppresses H2AS129P Enrichment around Centromeres, although Inactivation of Cohesin or Condensin Suppresses H2AS129P Enrichment across the rDNA Repeats (A) The relative enrichment of H2AS129P over H2A ChIP around centromeres is shown for scc1-73 top2-td (brown), released into the cell cycle under the restrictive conditions. Data for WT parental cells (green) and top2-td cells (blue) from Figure 1 are shown for comparison. Graph shown is generated from a pile up of the profiles of all centromeres and is an average of two repeats. (B) The relative enrichment of H2AS129P over H2A ChIP over the rDNA repeats is shown for scc1-73 top2-td (brown), released into the cell cycle under the restrictive conditions. Data for WT parental cells (green) and top2-td cells (blue) from Figure 1 are shown for comparison. Graph shown is an average of two repeats. (C) The relative enrichment of H2AS129P over to H2A ChIP across centromeres is shown for smc2-td K38I cells (where Smc2 protein is depleted and an enzymatically inactive form of Smc2 smc2K38I is concurrently expressed; dark gray) and smc2td K38I top2-td cells where both smc2 and Top2 are depleted and smc2K38I expressed (orange), both released into the cell cycle under the restrictive conditions. Data for WT parental cells (green) and top2-td cells (blue) from Figure 1 are shown for comparison. Graph shown is generated from a pile up of the profiles of all centromeres and is an average of two repeats. (D) The relative enrichment of H2AS129P over H2A ChIP over the rDNA repeats is shown for smc2-td K38I cells (where Smc2 protein is depleted and an enzymatically inactive form of Smc2, smc2K38I, is concurrently expressed; dark gray) and smc2-td K38I top2-td cells where both smc2 and Top2 are depleted and smc2K38I expressed (orange), both released into the cell cycle under the restrictive conditions. Data for WT parental cells (green) and top2-td cells (blue) from Figure 1 are shown for comparison. Graph shown is an average of two repeats. See also Figures S3 and S7.

Figure 5

Quantitative ChIP-Seq of H2AS129P around Centromeres and rDNA Shows that Cohesin and Condensin Activity Affects Rate of Repair across the rDNA (A) FACS analysis of one repeat for wild type, top2-td, scc1-73 top2-td, or smc2-td K38I top2-td at 25°C exponential followed by restrictive conditions in a G1 arrest and released under restrictive conditions for both 80 min and 120 min. Second repeat is shown in Figure S7. (B) The relative enrichment of H2AS129P over H2A ChIP normalized to S. pombe spike in (see STAR Methods) around centromeres is shown for wild type (green), top2-td (blue), scc1-73 top2-td (brown), or smc2-td K38I top2-td (orange) all released into the cell cycle under the restrictive conditions for 80 min. Graph shown is generated from a pile up of the profiles of all centromeres and is an average of two repeats. (C) The relative enrichment of H2AS129P over H2A ChIP normalized to S. pombe spike in (see STAR Methods) around centromeres is shown for wild type (green), top2-td (blue), scc1-73 top2-td (brown), or smc2-td K38I top2-td (orange) all released into the cell cycle under the restrictive conditions for 120 min. Graph shown is generated from a pile up of the profiles of all centromeres and is an average of two repeats. (D) The relative enrichment of H2AS129P over H2A ChIP normalized to S. pombe spike in (see STAR Methods) across the rDNA repeats is shown for wild type (green), top2-td (blue), scc1-73 top2-td (brown), or smc2-td K38I top2-td (orange) all released into the cell cycle under the restrictive conditions for 80 min. Graph shown is an average of two repeats. (E) The relative enrichment of H2AS129P over H2A ChIP normalized to S. pombe spike in (see STAR Methods) across the rDNA repeats is shown for wild type (green), top2-td (blue), scc1-73 top2-td (brown), or smc2-td K38I top2-td (orange) all released into the cell cycle under the restrictive conditions for 120 min. Graph shown is an average of two repeats. See also Figure S4.

Figure 6

Inactivation of Eco1 Partially Suppresses H2AS129P Enrichment around Centromeres (A) The relative enrichment of H2AS129P over H2A ChIP around all centromeres in eco1-1 cells either with WT expression of Top2 in eco1-1 (black) or depleted of Top2 in eco1-1 top2-td (violet) cells, both released into the cell cycle under the restrictive conditions. Data for WT (green) and top2-td cells (blue) from Figure 1 are shown for comparison. Graph shown is generated from a pile up of the profiles of all centromeres and is an average of two repeats. (B) The relative enrichment of H2AS129P over H2A ChIP across the rDNA repeats in eco1-1 cells either with WT expression of Top2 in eco1-1 (black) or depleted of Top2 in eco1-1 top2-td (violet) cells, all released into the cell cycle under the restrictive conditions. Data for WT (green) and top2-td cells (blue) from Figure 1 are shown for comparison. Graph shown is an average of two repeats. See also Figure S7.

Figure 7

Cohesin Activity Increases Fork Rotation during DNA Replication of Plasmids with Active Transcription Units The frequency of fork rotation on the different plasmid replicons was examined by gel electrophoresis and Southern blotting as described in STAR Methods. Cells containing the top2-4 allele and a CEN plasmid (8 kb pRS315), a CEN plasmid and 3× tRNA genes (3×tRNA pRS316), and a CEN plasmid with two active genes in a converging orientation (SEC53 FMP32 pRS315) were assessed for DNA catenation following one round of DNA replication in the absence of Top2 activity and with or without active cohesin (WT with active cohesin, scc1-73 inactive cohesin). 3×tRNA pRS316 WT sample was taken from Schalbetter et al. (2015). The CEN plasmid with two active genes in a converging orientation, SEC53 FMP32 pRS315, was also modified to inactivate the centromere by a double point mutation (Jehn et al., 1991), SEC53 FMP32 cenmutpRS315, and assessed for DNA catenation following one round of DNA replication in the absence of Top2 activity. Representative autoradiograms of the SEC53 FMP32 pRS315 plasmid with and without active cohesin and the SEC53 FMP32 cenmutpRS315 plasmid are shown in Figure S6. The relative intensity of catenanes generated post-replication was quantified and the population median of the catenanes calculated for each of the conditions. The median of each experiment is plotted on the boxplot along with the boxes representing the middle two quartiles of the distributions of the dataset. p values are derived from paired t tests; a star indicates a significant difference between two conditions (p < 0.05). See also Figures S5 and S6.