Deficiency of Polη in Saccharomyces cerevisiae reveals the impact of transcription on damage-induced cohesion.

The structural maintenance of chromosome (SMC) complex cohesin mediates sister chromatid cohesion established during replication, and damage-induced cohesion formed in response to DSBs post-replication. The translesion synthesis polymerase Polη is required for damage-induced cohesion through a hitherto unknown mechanism. Since Polη is functionally associated with transcription, and transcription triggers de novo cohesion in Schizosaccharomyces pombe, we hypothesized that transcription facilitates damage-induced cohesion in Saccharomyces cerevisiae. Here, we show dysregulated transcriptional profiles in the Polη null mutant (rad30Δ), where genes involved in chromatin assembly and positive transcription regulation were downregulated. In addition, chromatin association of RNA polymerase II was reduced at promoters and coding regions in rad30Δ compared to WT cells, while occupancy of the H2A.Z variant (Htz1) at promoters was increased in rad30Δ cells. Perturbing histone exchange at promoters inactivated damage-induced cohesion, similarly to deletion of the RAD30 gene. Conversely, altering regulation of transcription elongation suppressed the deficient damage-induced cohesion in rad30Δ cells. Furthermore, transcription inhibition negatively affected formation of damage-induced cohesion. These results indicate that the transcriptional deregulation of the Polη null mutant is connected with its reduced capacity to establish damage-induced cohesion. This also suggests a linkage between regulation of transcription and formation of damage-induced cohesion after replication.

Introduction

Dynamic disassembly and reassembly of nucleosomes—the building blocks of chromatin—facilitates processes such as replication and transcription. During the course of chromatin assembly, the canonical histones are exchanged with histone variants or post-translationally modified histones. This affects the physical and chemical properties of nucleosomes, as well as chromatin accessibility. Replication-independent nucleosome assembly, or so-called histone exchange, aids and regulates RNA polymerase II (RNAPII) passage through the nucleosomes during transcription initiation and elongation [1]. This is accomplished through histone chaperones, in concert with histone modifying enzymes and chromatin remodelers [2].

Transcription is not only the instrument for gene expression, but is also connected to cohesin localization on chromosomes. Cohesin is one of the structural maintenance of chromosomes (SMC) protein complexes, with the core formed by Smc1, Smc3 and the kleisin Scc1. Cohesin dynamically associates with chromosomes at intergenic regions of convergent genes, possibly as a result of active transcription [3,4]. Cohesin and its chromatin loader Scc2 have been implicated in gene regulation [5-7] and also in spatial organization of chromosomes into topologically associated domains (TADs) through DNA loop extrusion [8-12].

In addition to the roles described above, the canonical role of cohesin is to mediate sister chromatid cohesion. Cohesin is recruited to chromatin by the cohesin loading complex Scc2-Scc4 from late G1 phase in S. cerevisiae [13], and continuously through the cell cycle [14,15]. During S-phase, cohesin becomes cohesive through acetylation of Smc3 by the acetyltransferase Eco1 [16-18]. The established sister chromatid cohesion is then maintained until anaphase [19], ensuring faithful chromosome segregation.

At the end of S phase, Eco1 is targeted for degradation. However, induction of double strand breaks (DSBs) post-replication (G2/M) is sufficient to stabilize Eco1 [20,21]. Presence of active Eco1 then allows generation of damage-induced cohesion in G2/M, which is established close to the break, and also genome wide on undamaged chromosomes [22-24]. We previously showed that Polymerase eta (Polη), one of the three translesion synthesis (TLS) polymerases in S. cerevisiae, is specifically required for genome wide damage-induced cohesion [25].

Polη (encoded by the RAD30 gene) is well characterized for bypassing bulky lesions induced by ultraviolet irradiation [26], yet emerging evidence suggest that Polη also exhibits TLS-independent functions [27]. Polη is the only TLS polymerase required for damage-induced cohesion [25], independently of its polymerase activity, but dependent on Polη-S14 phosphorylation; potentially mediated by the cyclin dependent kinase, Cdc28 [28]. However, the underlying role of Polη in damage-induced cohesion remains unclear. Thus, absence of Polη does not affect break-proximal damage-induced cohesion or DSB repair. Lack of Polη also does not perturb Eco1 stabilization, cohesin chromatin association or Smc3 acetylation after induction of DSBs in G2/M [25].

Based on the following two observations, we hypothesized that active transcription facilitates damage-induced cohesion genome wide. First, Polη is enriched at actively transcribed regions, and required for expression of several active genes in S. cerevisiae [29]. Second, activated transcription leads to establishment of local de novo cohesion in S. pombe [30]. In other words, it is possible that transcription is deregulated in the Polη null mutant, and that this subsequently affects formation of damage-induced cohesion. Here, we showed that chromatin association of RNAPII is reduced in the absence of Polη, or if Polη-S14-phosphorylation is abolished. In addition, the transcriptional program in the Polη null mutant (rad30Δ) is altered both before and after DSB induction, with expression of genes involved in chromatin assembly and positive transcription regulation being downregulated compared to WT cells. Perturbing histone exchange at promoter regions by a HIR1 or HTZ1 deletion negatively affects damage-induced cohesion, in a similar fashion as in rad30Δ cells, while deletion of the transcription elongation regulator SET2 suppresses the lack of damage-induced cohesion in the rad30Δ mutant. Importantly, the potential linkage between transcription and formation of damage-induced cohesion was further supported by the fact that inhibiting transcription negatively affects its formation. Taken together, our results suggest that the transcription deregulation in the Polη null mutant is relevant to its deficient damage-induced cohesion. This provides new insight into formation of damage-induced cohesion post-replication, of importance for future investigations.

Results

Chromatin association of RNAPII is reduced in the Polη null and Polη-S14A mutants

To test if active transcription is correlated with generation of damage-induced cohesion, we initially assessed sensitivity of the damage-induced cohesion deficient rad30Δ and Polη-S14A cells to transcription elongation inhibitors. Viability of both mutants decreased when exposed to actinomycin D (Fig 1A). In addition, consistent with a previous report [29], rad30Δ cells were sensitive to mycophenolic acid (MPA). This was also true for the Polη-S14A point mutant (Fig 1A). Sensitivity of both mutants to MPA was reversed by supplementing the media with guanine (Fig 1A), verifying that it was due to depletion of the guanylic nucleotide pool [31].

Fig 1

Chromatin association of RNAPII is reduced in the Polη null and Polη-S14A mutants.

(A) Spot assay to monitor sensitivity of the rad30Δ and Polη-S14A mutants to the transcription elongation inhibitors, actinomycin D and mycophenolic acid (MPA). Tenfold serial dilutions of indicated mid-log phase cells on controls (-Ura plate ± guanine), and drug-containing plates, after 3 days incubation at room temperature. (B) ChIP-qPCR analyses to determine chromatin association of Rpb1 in indicated strains, on selected actively transcribed genes in G2/M arrested WT cells. Error bars indicate the mean ± STDEV of two independent experiments. Statistical differences compared to the WT cells at indicated position were evaluated by One-way ANOVA, Tukey post hoc test. The respective p values (<0.05) for each mutant relative to WT are (a) 0.000, (b) 0.004, (c) 0.010, (d) 0.026, (e) 0.039, (f) 0.044, (g) 0.034, (h) 0.000, (i) 0.000, (j) 0.010, (k) 0.011, (l) 0.026, (m) 0.017, (n) 0.003, (o) 0.004, (p) 0.047, (q) 0.047. p,promoter; m, mid; e, end of gene body. n1 and n2, low-binding controls. (C-D) Western blot analysis of Rpb1 stability. G2/M arrested cells from indicated strains, with or without one-hour P-HO break induction, were pelleted and resuspended in media containing cycloheximide (CHX) to monitor Rpb1 protein levels without further protein synthesis. Cdc11 was used as loading control. M, protein marker.

Sensitivity to elongation inhibitors might be due to reduced transcriptional capacity. We therefore monitored chromatin association of Rpb1, the largest subunit of RNAPII, in these mutants. Binding of Rpb1 at promoters and coding regions of selected active genes was reduced in both rad30Δ and Polη-S14A mutants compared to WT cells (Fig 1B). The reduced chromatin association was accompanied by an increased level of total Rpb1 (Figs 1C and S1A). Furthermore, Rpb1 stability in the rad30Δ and Polη-S14A mutants was not affected, regardless of DSB induction (Figs 1C, 1D and S1A and S1B). Here and throughout the study the DSBs were induced at the MAT locus on chromosome III (P-HO) for one-hour, unless otherwise stated. These results together suggest that Polη may facilitate chromatin association of RNAPII for proper transcription initiation and elongation, likely through phosphorylation of Polη-S14 but independently of DNA damage.

Transcription is perturbed in rad30Δ mutants

To further pinpoint a potential connection between transcription and formation of damage-induced cohesion, we focused on the rad30Δ mutant for the following investigations. To begin with, we analyzed gene expression of G2/M arrested WT and rad30Δ cells, before and after one-hour break induction, by RNA-sequencing analysis (RNA-seq). Prior to RNA-seq, G2/M arrest and break induction were confirmed (S2A and S2B Fig). Principal component analysis (PCA) showed that the individual data sets were distributed as distinct clusters (S2C Fig). Differences in gene expression patterns between WT and rad30Δ cells were readily observed before break induction, with 395 genes upregulated and 439 genes downregulated in the G2/M arrested rad30Δ mutant (Fig 2A). In response to DSB induction, the WT cells showed 473 genes up- and 519 genes down-regulated (Fig 2B), whereas there were 360 genes up- and 230 genes down-regulated in the rad30Δ mutant (Fig 2C and S1 Data). While the differentially expressed genes in WT and rad30Δ cells after break induction significantly overlapped (S2D Fig) and trended in the same direction, the up- and down-regulation after DSBs was of greater magnitude in the WT cells (Fig 2D and 2E). This implies that the response to break induction in the rad30Δ cells is similar, but relatively attenuated in comparison to the response in WT cells. Furthermore, we noted that short genes were preferentially upregulated compared to long genes in WT cells after DSB induction (Fig 2F), similar to the reported gene length dependent changes of expression after UV exposure [32,33]. In contrast, differential expression after DSBs is independent of gene length in the rad30Δ mutant (Fig 2F), further indicating a difference between WT and rad30Δ cells in their transcriptional responses. From these results we conclude that RAD30 deletion leads to transcription deregulation, both in unperturbed G2/M phase and in response to break induction.

Fig 2

Transcription is perturbed in rad30Δ mutants.

(A-C) Volcano plots showing differentially expressed genes between WT and rad30Δ cells, before and after DSBs, determined by RNA-seq. Each dot represents one gene. Red and blue dots represent up- and down-regulated genes respectively. Numbers of differentially expressed genes (padj < 0.05) are indicated. Black dots indicate genes without significant changes in expression. padj, adjusted p value. (D-E) Comparisons between expression level of genes significantly up (D) or downregulated (E) in the WT+DSB relative to the G2/M arrested WT cells, and expression of the same set of genes in the rad30Δ mutant, based on RNA-seq analysis. Significant differences compared to the WT cells were evaluated by paired t-test. (F) Plot of fold change moving median, sorted by length (300 genes/window) to monitor the trend of gene expression after DSBs in relation to gene length, comparing WT and rad30Δ cells. Fold change values were based on the changes of gene expression in WT and rad30Δ cells after DSBs, determined by RNA-seq.

Polη is more frequently associated with closed-, FN- and TATA-containing promoters

As an attempt to better understand the possible role of Polη during transcription, we used published datasets to analyze if the deregulated genes in rad30Δ cells were associated with specific types of promoters, in a similar manner as reported [34]. These datasets classify genes according to type of promoter: (i) open/closed promoters, either with or without a nucleosome free region [35], (ii) promoters with fragile/stable nucleosome (FN/SN), defined by sensitivity of the -1 nucleosome to MNase digestion [36], and (iii) the canonical TATA-containing or TFIID dominated promoters [37,38]. Notably, a significant number of downregulated genes in G2/M arrested rad30Δ cells were classified under the group of closed promoters (Table 1). In addition, the up- and down-regulated genes in G2/M arrested rad30Δ cells were dominated by TATA-containing promoters (obs/exp>1). These data imply that Polη more frequently associates with promoters in closed configuration and TATA-containing promoters, primed for transcriptional activation in G2/M phase. Interestingly, this prediction was supported by a Polη-ChIP-sequencing analysis. By monitoring genome-wide distribution of Polη during G2/M phase, we found that Polη was enriched 100 bp upstream of transcription start sites (TSSs) and downstream of transcription end sites (TESs) but not at gene bodies (S3A Fig). Furthermore, Polη more frequently associated with closed, FN and TATA-containing promoters (Fig 3A–3C); rather than the open, SN and TFIID-dominated promoters. To better understand if the transcriptional deregulation seen in rad30Δ cells was a direct or indirect effect, we set out to compare gene expression in the rad30Δ mutant with that of a ‘Polη-degron’ strain. The ‘Polη-degron’ strain harbors a combined auxin-inducible degron (AID) and Tet-off system [39,40], which allowed us to temporally deplete Polη during G2/M by addition of auxin and doxycycline (S3B Fig). We initially selected six Polη-bound or unbound promoters according to the Polη-ChIP-sequencing analysis (Fig 3D), and tested if expression of the corresponding genes was affected in rad30Δ and Polη-depleted cells. Expression of the Polη-bound genes (RIM4, PUT1, ECM29) was as expected reduced in rad30Δ cells (Fig 3E), although Polη-binding at the ECM29 promoter was less pronounced (Fig 3D). This was on the contrary not the case if depleting Polη specifically during G2/M phase (Figs 3F and S3B). Since expression of the DDR48 and PUT1 genes was upregulated in Polη-depleted cells, as compared to the untreated control (Fig 3F), we examined expression of five additional Polη-associated genes (S3C and S3D Fig). Expression of these genes, however, showed no difference between auxin/doxycycline-treated and untreated cells (S3D Fig). Thus, Polη appears to play an indirect role during transcription, and the transcriptional deregulation observed in rad30Δ cells is likely accumulated through multiple cell cycles under persistent absence of Polη.

Table 1

Association of differentially expressed genes with promoter type in G2/M arrested rad30Δ cells.

rad30Δ G2 vs. WT G2	upregulated (395)				downregulated (439)
	overlap	obs/exp	p values		Overlap	obs/exp	p values
closed promoter (1596)	118	1.1	0.046		146	1.3	3.309e-04	*
open promoter (3504)	228	1.0	0.459		237	0.9	0.077
FN promoter (1953)a	139	1.1	0.086		156	1.1	0.054
SN promoter (3066)b	206	1.0	0.223		245	1.1	0.008
TATA-containing (1090)	96	1.4	6.726e-04	*	132	1.7	5.069e-11	*
TFIID-dominated (5130)	299	0.9	7.636e-06	*	326	0.9	4.377e-08	*

Number of genes in each group is indicated in parentheses. The numbers in bold indicate that the overlap is higher than expected, observation/expectation (obs/exp)>1. Asterisks indicate significant overlap (p<0.001), evaluated as described in materials and methods. aFN: fragile nucleosome, bSN: stable nucleosome.

Fig 3

Polη is more frequently associated with closed-, FN- and TATA-containing promoters.

(A) Metagenome plot showing accumulation of Polη at closed or open promoters, from 365 bp upstream to 50 bp downstream of the transcription start site (TSS) in G2/M phase. The samples were first normalized to their respective input and then the values were scaled to the maximum value of the plot. (B-C) As in (A), except plotting accumulation of Polη at 500 bp upstream of TSS, to compare its relative enrichment at SN or FN promoter in (B); at TATA or TFIID-dominated promoters in (C). (D) Representative Integrative Genomics Viewer (IGV) tracks of Polη-ChIP-seq at selected promoters, with all Y-axes in the same scale. The samples were normalized to their respective input and the library size. (E) Expression of selected genes in G2/M arrested WT and rad30Δ cells, measured by RT-qPCR. Error bars indicate the mean ± STDEV of two independent experiments. Statistical differences between WT and rad30Δ cells were evaluated by two-tailed t-test. (F) Expression of selected genes with or without depletion of Polη during G2/M, measured by RT-qPCR. Polη was temporally depleted by addition of auxin and doxycycline; the mock control was denoted as ‘WT’. The differences of ΔCt values between samples before and after addition of drugs were calculated as ΔΔCt, presented as 2-ΔΔCt in the graph. The same calculations were applied to the mock control. Error bars indicate the mean ± STDEV of three independent experiments. Statistical differences between WT and Polη-depleted cells were evaluated by two tailed t-test.

Genes involved in chromatin assembly and positive transcription regulation pathways are downregulated in the absence of Polη

To gain mechanistic insight into the diverse transcriptional responses detected in WT and rad30Δ cells, differential gene expression between WT and rad30Δ cells (before and after DSBs) were analyzed by Gene Set Enrichment Analysis (GSEA), followed by generation of enriched pathway maps with Cytoscape as shown in Fig 4. The gene sets under each annotated group are listed in S2 and S3 Data. During G2/M arrest, genes that belong to biological pathways such as chromatin assembly and positive transcription regulation were downregulated in rad30Δ compared to WT cells (Fig 4A). Consistent with downregulation of genes involved in the chromatin assembly pathway, we observed that the global nucleosome occupancy of rad30Δ cells was moderately increased compared to WT cells (S4A Fig). Although this may raise a concern about cohesin binding in rad30Δ cells, as nucleosome-free regions at promoters are required for cohesin loading [6,41], we previously noted that absence of Polη does not result in apparent differences in overall cohesin binding [25]. However, by revisiting our published Scc1 ChIP-sequencing dataset (GSE42655) and performing genome-wide meta-analysis, we found that association of cohesin around TSS was increased in rad30Δ compared to WT cells (S4B and S4C Fig). Notably, this increased binding was not found around TES (S4D and S4E Fig), and was independent of DSB induction (S4B–S4E Fig). This could reflect that cohesin bound at TSS becomes less dynamic when transcription is dysregulated, as in rad30Δ cells.

Fig 4

Genes involved in chromatin assembly and positive transcription regulation pathways are downregulated in the absence of Polη.

(A) Relatively enriched pathways in G2/M arrested WT and rad30Δ cells, plotted with Cytoscape after gene set enrichment analysis (GSEA). The GSEA was performed with gene lists ranked by log10 p value (multiplied by the sign of the fold change) of each gene. The number of genes in each gene set is proportional to the circle size. Lines connect gene sets with similarity greater than 0.7. All gene sets have FDR < 0.05. (B) Gene set enrichment analysis after DSB induction, plotted with Cytoscape to depict the difference between WT and rad30Δ cells in up- or down-regulation of indicated pathways after DSBs. Gene expression of WT and rad30Δ cells after DSBs was compared to that of respective G2/M arrested cells. GSEA was performed as in (A). The lines indicate the same as in (A). All gene sets have FDR < 0.05 and a normalized enrichment score > 2 for at least one of the WT or rad30Δ cells.

When comparing gene expression after break induction, the pathways illustrated in Fig 4B were clearly differentially regulated between WT and rad30Δ cells. The nucleotide metabolism and amino acid metabolism pathways in WT cells, for instance, were upregulated to less extent compared to rad30Δ cells. This further indicates deregulation of gene expression in the rad30Δ mutant. Considering the fact that DNA damage response (DDR) proteins contribute to formation of damage-induced cohesion [22,24], we looked into the DDR pathway after DSB induction. Despite that some genes belonging to the cellular response to DNA damage stimulus pathway (GO: 6974) were upregulated in WT cells after DSB induction, this pathway was overall not significantly enriched. In addition, activation of the DNA damage checkpoint, as indicated by phosphorylation of Rad53, was only observed during the recovery period after DSB induction in both WT and rad30Δ cells (S4F and S4G Fig), with no difference in cell cycle progression between populations (S4H Fig). These results indicate that the lack of damage-induced cohesion in rad30Δ cells is not due to a possible difference in activation of the DNA damage checkpoint. Furthermore, in response to DSBs, expression of the acetyltransferase ECO1 was not enhanced in either WT or rad30Δ cells (S4I Fig). Altogether, this made it plausible to investigate the potential connection between transcription and damage-induced cohesion, and for this we focused on two of the upregulated gene sets in WT cells before DSB induction—chromatin assembly and positive transcription regulation.

Deleting HIR1 leads to partially deficient damage-induced cohesion

To assess if transcriptional activity is related to generation of damage-induced cohesion, we utilized a genetic approach by testing mutants which in theory should mimic or reverse the transcriptional deregulation in rad30Δ cells. One of the interesting candidates was Hir1 (a component of the HIR complex) that is known to be involved in chromatin assembly. The HIR complex and the histone chaperone Asf1 mediate histone H3 exchange with post-translationally modified H3, independently of replication [42,43]. The exchange mainly takes place at promoters and correlates with active transcription. However, basal H3 exchange also occurs to poise inactive promoters for optimal transcription [44,45]. Therefore, the relevance between transcriptional activation and formation of damage-induced cohesion could be investigated through the hir1Δ mutant.

To monitor damage-induced cohesion, DSBs and ectopic P-SMC1-MYC expression were induced by addition of galactose to G2/M arrested cells. Due to the smc1-259 ts background, cohesion established during replication was inactivated by raising the temperature. Damage-induced cohesion generated with the ectopic Smc1-Myc was examined with an integrated TetO/TetR-GFP array on Chr. V (illustrated in S5A Fig). G2/M arrest, break induction and protein expression of the ectopic Smc1-Myc were confirmed for all experiments, with examples shown in S5B–S5D Fig. Interestingly, formation of damage-induced cohesion was partially deficient in the hir1Δ mutant, while the hir1Δrad30Δ double resembled the rad30Δ single mutant, although with slower sister separation (Fig 5A). This indicated that Hir1 and Polη are both required for efficient damage-induced cohesion; possibly acting in the same pathway.

Fig 5

Deleting HIR1 leads to partially deficient damage-induced cohesion.

(A) Damage-induced cohesion assays of the hir1Δ single and hir1Δrad30Δ double mutants after P-HO induction, performed as illustrated in S5A Fig. Means ± STDEV from two independent experiments are shown. Two-hundred cells were counted for each time point, in each experiment. (B-C) Damage-induced cohesion assays of the hhf1-hht1Δ and hht2-hhf2Δ mutants after P-HO induction, performed as in (A). Means ± STDEV from two independent experiments are shown. At least two-hundred cells were counted for each time point, in each experiment.

By using the hir1Δ mutant, we determined if the HIR/Asf1-dependent histone exchange affected formation of damage-induced cohesion. However, the observed deficiency of the hir1Δ cells might be due to de-repression of histone genes, as the HIR complex also negatively regulates expression of histone genes [46,47]. If so, reducing the histone gene dosage should be beneficial for the rad30Δ mutant in generation of damage-induced cohesion. Yet, deletion of any H3-H4 coding gene pair (HHT1-HHF1 and HHT2-HHF2) did not affect formation of damage-induced cohesion, neither on their own nor in rad30Δ cells (Fig 5B and 5C). This indicates that the partial deficiency of the hir1Δ mutant is not due to altered histone gene dosage, and points to a need for histone exchange during transcription for formation of damage-induced cohesion.

Perturbing histone exchange at promoters negatively affects formation of damage-induced cohesion

To further investigate the effect of pertubing histone exchange on formation of damage-induced cohesion, we tested HTZ1 deleted cells. Htz1, the histone variant of H2A, is preferentially incorporated at basal/repressed promoters. It is however susceptible to be evicted from the nucleosome, and that in turn promotes its exchange for H2A. This facilitates transcriptional activation [48,49], and relieves the +1 nucleosome barrier to RNAPII [50,51]. Since the htz1Δ mutant does not respond to P-HO induction [52], γ-irradiation was utilized as source of DSB induction (see materials and methods). Similar to the hir1Δ mutant (S6A and S6B Fig), the htz1Δ mutant showed impaired damage-induced cohesion (Fig 6A). In contrast to a previous report [53], we did not observe a cohesion maintenance defect due to HTZ1 deletion (S6C Fig).

Fig 6

Perturbing histone exchange at promoters negatively affects formation of damage-induced cohesion.

(A) Damage-induced cohesion assay of the htz1Δ mutant after γ-irradiation, performed according to the procedure described in the materials and methods. Means ± STDEV from two independent experiments are shown. For each experiment, at least two-hundred cells were counted for each time point. (B-C) ChIP-qPCR analyses to determine Htz1 occupancy at promoters of selected genes, before (B) and after DSB induction (C) in G2/M arrested WT and rad30Δ cells. SPF1, RAD23 and HAT2 are located at the left arm of chromosome V, where damage-induced cohesion was monitored. Error bars indicate the mean ± STDEV of three independent experiments for (B) and two independent experiments for (C). Statistical differences compared to the WT cells were evaluated by t-test. n, low-binding control. (D) Western blot analysis of the total Htz1 protein level in WT and rad30Δ cells, before and after DSB induction during G2/M phase. Cdc11 was used as loading control. M, protein marker.

Since Htz1 is required for formation of damage-induced cohesion, we investigated if there was a difference in Htz1 occupancy at promoters between WT and rad30Δ cells. For this, we focused on three of the genes analyzed for Rpb1 binding in Fig 1B and three additional genes around the URA3 on Chr. V, where we monitored damage-induced cohesion. We selected genes with TATA-less promoters for analyses because Htz1 is relatively enriched at these promoters [48,49]. Interestingly, Htz1 occupancy at some of the selected promoters was increased in rad30Δ compared to WT cells, particularly after DSB induction in G2/M (Fig 6B and 6C). Despite this difference, the total protein level of Htz1 was similar between WT and rad30Δ cells (Fig 6D). This indicates that the Htz1/H2A exchange at certain promoters was reduced in the absence of Polη, especially in response to DSBs. These results were in line with hir1Δ and htz1Δ cells being deficient in damage-induced cohesion (Figs 5A, 6A and S6B), and suggest that perturbing histone exchange at promoters negatively affects formation of damage-induced cohesion.

Transcriptional deregulation leads to deficient damage-induced cohesion

In addition to the hir1Δ and htz1Δ mutants, we used a set2Δ mutant to test if transcriptional regulation is correlated with generation of damage-induced cohesion. Set2 mediates co-transcriptional H3K36 methylation (H3K36me1/2/3). This promotes restoration of chromatin to the pretranscribed hypoacetylation state and represses histone exchange at coding regions after transcription elongation [54-56]. Presence of Set2 at promoters also suppresses transcription initiation of certain basal repressed genes [57-59]. Interestingly, a set2Δ mutant was reported to suppress sensitivity of certain transcriptional elongation factor mutants to 6-azauracil [59], a mechanistic analog of MPA [60,61]. As we showed that rad30Δ cells are sensitive to transcription elongation inhibitors (Fig 1A), we tested if deletion of SET2 would also rescue rad30Δ cells from this sensitivity. The set2Δ mutant was insensitivite to MPA or actinomycin D, and masked sensitivity of rad30Δ cells especially to actinomycin D (Fig 7A). Through this genetic interaction, we then tested if deletion of SET2 would also suppress the deficiency of rad30Δ cells in damage-induced cohesion. While the set2Δ mutant resembled the WT cells in formation of damage-induced cohesion, deletion of SET2 remarkably suppressed the lack of damage-induced cohesion in the rad30Δ mutant (Fig 7B). In addition, since removing SET2 has been shown to cause an increased RNAPII association towards the 3’-end of actively transcribed genes [62], we monitored chromatin association of Rpb1 in the set2Δrad30Δ mutant. Absence of Set2 in G2/M arrested rad30Δ cells to some extent compensated for the reduced Rpb1 binding in rad30Δ cells (Fig 7C–7E). This trend was however not observed after DSB induction (S7A–S7C Fig). Considering that the differentially expressed genes in WT and rad30Δ cells after DSB induction significantly overlapped (S2D Fig), these data together suggest that general transcription regulation during G2/M phase influences formation of damage-induced cohesion.

Fig 7

Transcriptional deregulation leads to deficient damage-induced cohesion.

(A) Spot assay to monitor the effect of SET2 deletion on the rad30Δ mutant sensitivity to the transcription elongation inhibitors, actinomycin D and mycophenolic acid (MPA). Tenfold serial dilutions of indicated mid-log phase cells on control (-Ura plate ± guanine) and drug-containing plates, after 3 days incubation. (B) Damage-induced cohesion assay of the set2Δ mutant after P-HO induction, performed as depicted in S5A Fig. Means ± STDEV from two independent experiments are shown. At least two-hundred cells were counted for each time point, in each experiment. (C-E) ChIP-qPCR analyses to determine chromatin association of Rpb1 at promoters and 3’-ends of selected genes, in indicated G2/M arrested cells. Except MSC1 and NPL4, the rest of the selected genes are located at the left arm of chromosome V, where damage-induced cohesion was monitored. Error bars indicate the mean ± STDEV of two independent experiments. n, low-binding control (n2 in Fig 1B). (F) Damage-induced cohesion assay of the Rpb1-anchor away strain. Gamma-irradiation (IR) was used as the source of DSBs. The assay was performed according to the procedure described in materials and methods. Rapamycin (RAP) was added to deplete Rpb1 from the nucleus. Means ± STDEV from three independent experiments are shown. At least two-hundred cells were counted for each time point, in each experiment.

To further examine this idea, we inhibited transcription and monitored formation of damage-induced cohesion after γ-irradiation. Transcription inhibition was achieved through the anchor-away system, which uses rapamycin to induce heterodimerization of the anchor (Rpl13A-FKBP12) and the FRB-tagged target, in our case Rpb1 [63], thereby excluding Rpb1 from the nucleus (S7D Fig). To avoid toxicity of rapamycin and its effect on transcription, the rapamycin binding protein Fpr1 and the rapamycin target Tor1 were either deleted or mutated in our ‘Rpb1-anchor away’ strain ([63-65] and S1 Table). Anchoring Rpb1 away from the nucleus by 1-hour rapamycin treatment caused approximately two-fold reduction in expression of selected genes (S7E Fig), without triggering early DNA damage response (S7F Fig), or compromising the protein level of the P-driven ectopic Smc1-Myc (S7G Fig). In line with previous experiments, without addition of rapamycin, damage-induced cohesion was formed after exposure to γ-irradiation (Fig 7F). However, transcription inhibition induced by rapamycin indeed negatively affected formation of damage-induced cohesion. Altogether, this supports the idea that transcription deregulation, as a consequence of persistent absence of Polη, is connected to deficienct damage-induced cohesion.

Discussion

We previously showed that Polη is specifically required for genome wide damage-induced cohesion [25] but its mechanistic role in this process was unclear. The present study was initiated by the observation that Polη-deficient cells displayed altered transcriptional regulation, both in unchallenged G2/M arrested cells and in response to DSBs. Transcription elongation deficiency was corroborated by increased sensitivity of Polη-deficient cells to transcription elongation inhibitors (Fig 1A). It could be argued that the sensitivity to actinomycin D would be a consequence of DNA damage because actinomycin D also inhibits topoisomerases [66], leading to formation of DSBs. However, since the rad30Δ mutant is insensitive to specific topoisomerase inhibitors, such as camptothecin and etoposide [67,68], this was less likely.

To know which pathways were affected in the absence of Polη, gene set enrichment analysis was performed after RNA-seq. We found that mitochondrial related pathways were enhanced in rad30Δ cells, in contrast to downregulation of genes belonging to the chromatin assembly pathway (Fig 4A and 4B and S2 and S3 Data). This is an interesting observation since genes involved in the tricarboxylic acid cycle and oxidative phosphorylation pathways, which are related to mitochondria, were similarly upregulated in mutants with defective chromatin assembly [69].

To test if the lack of damage-induced cohesion in rad30Δ cells would be due to transcriptional dysregulation, we first tested the requirement of HIR/Asf1 mediated histone exchange for damage-induced cohesion, from the perspective of chromatin assembly. By deleting the HIR1 gene, which is sufficient to disrupt the HIR/Asf1 interaction [43], we found that the hir1Δ mutant is partially deficient in damage-induced cohesion (Figs 5A and S6B). The role of the HIR complex in damage-induced cohesion might appear difficult to pinpoint since it is involved in multiple processes. We thus addressed the possible effect of HIR-dependent repression of histone genes [46] on formation of damage-induced cohesion. This possibility was however excluded because no effect of deleting H3-H4 gene pairs (Fig 5B and 5C) was observed in rad30Δ cells. The HIR complex has also been implicated in formation of a functional kinetochore [70] and heterochromatic gene silencing [71]. However, the chromatin assembly complex-1 (CAF-1) is redundant with the HIR complex in these processes. Deletion of Hir1 is thereby not likely to perturb other processes than histone exchange. We therefore suggest a direct role for HIR-dependent histone exchange in damage-induced cohesion.

Functional importance of Polη in transcription was proposed to depend on its polymerase activity [29], while its role in damage-induced cohesion was not [25]. The finding that transcription supports formation of damage-induced cohesion could therefore be seen as conflicting with the polymerase-independent role of Polη. However, we previously showed that the putative Polη-S14 phosphorylation is required for damage-induced cohesion, but not for cell survival after UV irradiation [28], which depends on Polη polymerase activity. In addition, the Polη-S14A mutant exhibits similar elongation inhibitor sensitivity and altered Rpb1 behaviour as the rad30Δ mutant (Fig 1A–1D). This together indicates that the polymerase activity is not the sole requirement for Polη in transcription.

To gain further insight into the role of Polη in transcription, we analyzed the types of promoters that Polη associates with (Table 1 and Fig 3). We found that the differentially expressed genes in G2/M arrested rad30Δ cells, especially the downregulated genes, were relatively enriched for closed and TATA-containing promoters. In line with this, we showed by ChIP-sequencing that Polη preferentially occupies these two types of promoters. The closed promoters that lack a nucleosome free region, are known to regulate stress related genes [72]. This is consistent with the downregulation of stress response (GO:0033554) in G2/M arrested rad30Δ cells (S2 Data, Ungrouped). Similarly, TATA-box containing genes are highly regulated and associated with stress response [37]. This together suggests that Polη could support transcription for proper stress response. Besides of closed and TATA-containing promoters, Polη was also found to be relatively enriched at FN promoters in our Polη-ChIP-sequencing analysis. The FN promoters typically regulate highly expressed or growth related genes [36]. However, the differentially expressed genes in rad30Δ cells did not significantly overlap with genes regulated by this type of promoter (Table 1), showing that their expression was not affected by persistent absence of Polη. Nevertheless, knowing the types of promoters that Polη preferentially associates with should be helpful for identification of its potential interactors during transcription. It would also be interesting to know how these preferences are correlated with formation of damage-induced cohesion in the future.

To further investigate if Polη acts directly or indirectly in transcription, we depleted Polη temporally during G2/M and analyzed expression of selected genes. It turned out that in contrast to rad30Δ cells, expression of most tested genes was not affected (Figs 3E, 3F and S3D). This suggests an indirect role of Polη in transcription, which would only perturb the transcriptional process if it is persistently absent from cells. We speculate that presence of Polη at specific promoters and Polη-S14-phosphorylation contribute to formation of a certain chromatin state, which is primed for proper transcriptional regulation. This is indicated by the reduced chromatin binding of Rpb1 in the Polη null and Polη-S14A mutants (Fig 1B), the preference of Polη for certain promoters (Fig 3A–3C) and the reduced Htz1/H2A exchange in rad30Δ cells (Fig 6C). In contrast, a temporal depletion of Polη during G2/M would likely have less impact on chromatin state. Precisely how the persistent absence of Polη indirectly affects transcription remains to be investigated further.

Through perturbing histone exchange, removing a transcription elongation regulator (illustrated in S8 Fig), and inhibiting transcription by anchoring Rpb1 away from the nucleus (Figs 7F and S7D–S7G), we show that a regulated transcriptional response connected to chromatin assembly, potentially facilitates generation of damage-induced cohesion post-replication. Since establishment of sister chromatid cohesion is proposed to occur simultaneously with replication fork progression [14,73], in concert with replication-coupled nucleosome assembly [74], it is also possible that replication-independent nucleosome assembly (histone exchange) is utilized as an alternative platform for generation of damage-induced cohesion after replication (S8 Fig, WT). In support of this, deregulated transcription and reduced Htz1/H2A exchange in rad30Δ cells negatively affected formation of damage-induced cohesion (S8 Fig, rad30Δ).

Despite the subtle defect in chromosome segregation observed in the rad30Δ mutant [25], the importance of genome wide damage-induced cohesion remains to be determined. It might be relevant to the increased chromosome mobility in response to DSBs, which presumably facilitates the search of sequence homology for recombination [75,76]. Interestingly, chromosome mobility is at the same time constrained by sister chromatid cohesion [77]. Since unbroken chromosomes are known to be less mobile than broken chromosomes [75,76], formation of genome-wide damage-induced cohesion might further limit the movements of undamaged chromosomes, to reduce the chance of unfavorable recombinations.

In summary, we show that transcriptional deregulation driven by persistent absence of Polη leads to deficient damage-induced cohesion. Through a genetic approach, our study provides new insight into a potential linkage between histone exchange and generation of damage-induced cohesion post-replication. Further studies would be needed to understand how chromatin dynamics during transcription facilitate formation of genome wide damage-induced cohesion, and if damage-induced cohesion could restrict movements of undamaged chromosomes.

Materials and methods

Yeast strains and media

All S. cerevisiae yeast strains, listed in S1 Table, were W303 derivatives (ade2-1 trp1-1 can1-100 leu2-3, 112 his3-11, 15 ura3-1 RAD5 GAL psi). To create null mutants, the gene of interest was replaced with an antibiotic resistance marker through lithium acetate based transformation. Some strains were crossed to obtain desired genotypes. Yeast extract peptone (YEP) supplemented with 40 μg/ml adenine was used as yeast media, unless otherwise stated.

Spot assay

Cell culturing and subsequent serial dilutions were performed as described [28]. Each dilution was sequentially spotted on uracil drop-out (-Ura) media, containing actinomycin D, MPA, or solvent only (final 1.2% ethanol in plates). Guanine was supplemented at 0.3 mM final concentration [78]. The plates were kept at room temperature and documented on the third day. Each spot assay was done at least twice.

Protein extraction and western blotting

Whole cell extracts (WCEs) were prepared with glass bead disruption, TCA or a sodium hydroxide based method [79]. To monitor Rpb1 stability, cycloheximide (Sigma) was supplemented in media (final 100 μg/ml), and the protein extracts were prepared with sodium hydroxide based method. Bolt 4–12% Bis-Tris or NuPAGE 3–8% Tris-Acetate gels (Invitrogen) were used for electrophoresis, with Bolt MOPS, Bolt MES or NuPAGE Tris-Acetate SDS running buffer (Invitrogen). Proteins were transferred to nitrocellulose membranes with the Trans-blot Turbo system (Bio-Rad) or the XCell II Blot Module (Invitrogen). Antibody information is listed in the S2 Table. Odyssey Infrared Imaging and BioRad chemiluminescence system were used for antibodies detections. Image Studio Lite software was used for quantitation of protein bands.

Chromatin immunoprecipitation (ChIP)-qPCR for Rpb1 and Htz1

ChIP was in essence performed as described with some modifications [25]. Cells were crosslinked with final 1% formaldehyde for 20 minutes at room temperature, followed by addition of final 125 mM glycine for 5 minutes. The cells were washed three times in 1X cold TBS and mechanically lysed using a 6870 freezer/mill (SPEX, CertiPrep). WCEs were subjected to chromatin shearing by sonication (Bandelin, Sonopuls) for chromatin fragments of 3–500 bp. Anti-Rpb1 and anti-Htz1 antibodies were coupled to protein A and protein G Dynabeads (Invitrogen) respectively for immunoprecipitation at 4°C, overnight. Crosslinking of eluted IP and input samples was reversed, and DNA was purified. DNA analysis was performed by real time qPCR (RT-qPCR) using SYBR Green (Applied Biosystems), according to manufacturer’s guidelines on an ABI Prism 7000 sequence detection system (Applied Biosystems). The genes of interest were selected based on the RNA-seq results. Primers used are listed in S3 Table. Each gene was analyzed with three technical repeats for each individual experiment. Statistical analysis was performed with SPSS statistics software (IBM).

Polη-Myc ChIP

Preparation of WCEs for ChIP of Myc-tagged Polη was performed as described above, and the ChIP as in [29] with the following modifications. Sonicated cell lysates from 70–80 OD units were incubated with anti-MYC, rotating at 4°C overnight. Protein G Dynabeads (Invitrogen) were then added for immunoprecipitation for 3.5 hours at 4°C. After reversing cross-linking at 65°C for 15 hours, the samples were treated with RNAse (50 μg/ml, final concentration) for 1 hour at 37°C, and finally the chromatin was purified (PCR purification kit, Qiagene). DNA analysis was performed by ChIP Sequencing (see below).

Total RNA extraction and RT-qPCR

For RNA-seq, G2/M arrested cells (about 9 OD600) were harvested before and after 1-hour P-HO break induction. Equal amount of samples were additionally collected at each time-point as genomic DNA (gDNA) controls. The gDNA content of each sample was determined prior to total RNA extraction. Total RNA extracts were prepared with PureLink RNA Mini Kit (Invitrogen), with some modifications of the manufacture’s guidelines. Collected cell pellets were washed once with SE mix (final 1 M sorbitol and 50 mM EDTA), and resuspended with 100 μl zymolyase lysis buffer (SE mix supplemented with final 3 mg/ml 100T zymolyase (Sunrise Science) and 2.5 μl Ribolock (Invitrogen). The suspension was incubated at 30°C for 60 minutes, followed by addition of 200 μl kit-provided RNA lysing buffer, supplemented with Ribolock. The rest of the procedure was performed according to the manufacture’s guidelines. To elute total RNA from columns, the volume of RNase free water for elution was adjusted according to gDNA content of each sample. For each strain, equal volume of the total RNA extract was further purified with DNA-free Kit (Invitrogen).

For RT-qPCR, purified total RNA (300 or 650 ng) was spiked in with 1 ng luciferase control RNA (Promega) prior to cDNA synthesis. Luciferase was then used as the reference gene for data analyses [80], unless otherwise stated. Primers used are listed in S4 Table. Each gene was analyzed with three technical repeats for each individual experiment.

RNA-sequencing and ChIP-sequencing data analyses

Total RNA samples prepared for RNA-seq (triplicates) were subsequently handled by Novogene for mRNA enrichment, library construction (250–300 bp insert cDNA library) and RNA sequencing (Illumina HiSeq X Ten, paired-end, 10 M reads). Quality controls were included for the total RNA samples and during the procedures for RNA-sequencing.

FASTQC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) was used for quality control of the .fastq-files for both RNA- and ChIP-seq. Adapter and poor quality read trimming was performed with cutadapt [81]. The RNA-seq data was mapped with the splice-aware aligner HISAT2 [82]. The Scc1-ChIP-seq data was mapped using Bowtie [83] with the colorspace option enabled, while the Polη-ChIP-seq data was mapped using Bowtie2 [84]. Afterwards the mapped files were sorted using samtools [85]. All three sets of sequencing data were aligned to the yeast genome version SacCer3 downloaded from UCSC genome browser. Duplicates in the mapped.bam-files were removed using MarkDuplicates (http://broadinstitute.github.io/picard) from the Picard toolset.

For the RNA-seq data set, the reads were counted per gene using featureCounts [86]. The count-files were imported into R and further analyzed using edgeR [87,88] for FPKM calculations and DESeq2 [89] for differential expression analysis. Differential expression analysis yielded fold-changes alongside significance for genes, additionally DESeq2 was used to generate principal component analysis plots. Genes with a total read count below 10 across all samples as well as those producing NAs (not available) in any of the comparisons for fold-change calculation were excluded from the analysis. As all four conditions showed a similar within-group variability in the PCA plot, for all fold-change calculations all samples were run together as opposed to subsetting the samples of interest e.g. WT G2 + DSBs vs. WT G2. This allowed for more accurate estimation of the dispersion parameter and in turn calculation of significance for the fold-changes. Also, the moving average of the fold-change was calculated by ordering the genes included in the DESeq2 dataset by length and then calculating the median of a window of 300 genes around these gene. No moving average was calculated for the 75 longest and shortest genes as they did not have an even number of genes on either site for moving average calculation.

For the Scc1-ChIP-seq dataset, cohesin peaks were called using MACS2 [90]. The files generated were then imported into R, where they were annotated using the package ChIPpeakAnno [91] with gene lists downloaded using the biomaRt package [92]. The lists of genes overlapping or with their gene end closest to the peak middle with cohesin peaks were read into ngs.plot [93] for metagenome analysis. After analysis had been performed, the data were replotted using the internal R plotting. Ngs.plot was also used to perform metagenome analysis for the Polη-ChIP-seq dataset at different promoter types. Additionally, the bigCompare command of the deepTools suite was used to generate bigWig files of the Polη-IPs normalized to both their respective inputs and the library size [94]. These bigwig files were then loaded into the Integrative Genomics Viewer (IGV) for visualization [95,96].

Gene set enrichment analysis (GSEA) was performed using the Broad Institute software (http://www.broad.mit.edu/gsea) [97] using S. cerevisiae gene sets from the Xijin Ge lab (http://ge-lab.org/#/data) [98]. The GSEA enrichment map was created using the EnrichmentMap plugin [99] for Cytoscape [100], broadly following a published protocol [101]. Groupings were facilitated by the Cytoscape AutoAnnotate plugin [102]. In the comparison of WT vs. rad30Δ cells, only gene sets enriched with an adjusted p-value of < 0.05 were plotted. In the comparison of both WT and rad30Δ cells ± DSB induction, only gene sets enriched with an adjusted p-value of < 0.05 and a normalized enrichment score (NES) > 2 for either strain were plotted.

Statistical significance of the overlapping genes in the Venn diagrams and Table 1 were calculated using either a normal approximation or the hypergeometric probability formula. The online tool on http://nemates.org/MA/progs/overlap_stats.html was used for evaluation.

Damage-induced cohesion assay and controls

All strains used harbor the smc1 temperature sensitive allele (smc1-259). The experiments with the P-HO allele for DSB induction were performed as described [28], and illustrated in S5A Fig. The assay utilizing γ-irradiation as DSB source is described in S6A Fig. Considering that the htz1Δ mutant is benomyl sensitive [103], the strains used in this assay contain the P-CDC20 and smc1-259 ts alleles. The strains were grown in methionine drop-out media (-Met) to log phase at 23°C. To arrest cells in G2/M phase, expression of CDC20 was repressed by replacing the media to YEP supplemented with Met (final 2 mM) and 0.1% glucose. Galactose (final 2%) was then added for 1.5 hours to induce expression of ectopic Smc1-Myc, driven by the GAL promoter. The cultures were subsequently split into half and resuspended in 1X PBS. One half for γ-irradiation (250 Gy), and another half as non-irradiated control. After 1-hour recovery in YEP media supplemented with galactose and Met, the temperature was raised to 35°C and damage-induced cohesion was monitored for 90 minutes.

For the Rpb1-anchor away strain, damage-induced cohesion assay was performed as illustrated in S6A Fig with the following modifications. The culture was split after 1-hour GAL-induction, half for addition of rapamycin (final 1 μg/ml) and half for addition of DMSO as control. After 1-hour ± rapamycin treatment, the cultures were spun down and resuspended in PBS supplemented with benomyl (PBS/B). The following procedures were as depicted in S6A Fig, except the cells were allowed to recover for 30 minutes in YEP media supplemented with glucose and benomyl after ± γ-irradiation. Noted that after resuspension in PBS/B, the cultures were always supplemented with rapamycin or DMSO when changing media.

Proper G2/M arrest, expression of the ectopic Smc1-Myc and DSBs induction in these assays were confirmed with FACS analysis, western blot, and pulsed-field gel electrophoresis (PFGE) respectively. Efficiency of γ-irradiation was analyzed with Southern blot after PFGE, with a probe for chromosome XVI, as described [104]. Rpb1-in situ staining was performed as described [28], using a specific anti-Rpb1 antibody.

MNase digestion assay

G2/M arrested cells were crosslinked in vivo with formaldehyde (final 0.5%), for 20 minutes at 23°C. To quench the reaction, glycine (final 125 mM) was added in cultures for 10 minutes. The cells were then harvested and stored at -80°C. Prior to MNase digestion, the cells were resuspended in pre-incubation solution (final 20 mM citric acid, 20 mM Na2HPO4, 40 mM EDTA, pH 8.0), with aliquots taken for cell-counting. The final volume of resuspension was subsequently adjusted to have 4.5 x 107 cells/ml. The cells were pre-treated with freshly added 2-mercaptoethanol (2-ME, final 30 mM in pre-incubation buffer) for 10 minutes at 30°C, followed by zymolyase treatment in zymolyase buffer (final 1 M sorbitol, 50 mM Tris-HCl (pH 7.5), 10 mM 2-ME and 1 mg/ml 100T zymolyase) for 30–35 minutes [105]. Converted spheroplasts were washed once with cold zymolyase buffer without 2-ME, resuspended in nystatin buffer (final 50 mM NaCl, 1.5 mM CaCl2, 20 mM Tris-HCl (pH 8.0), 1 M sorbitol, and 100 μg/ml nystatin (Sigma), and then kept on ice temporarily.

The following MNase digestion was performed for each strain individually. Resuspended spheroplasts were sequentially added into the MNase aliquots (ranged from final 0.0125 to 0.1 U/ml, prepared in nystatin buffer), and incubated at 25°C for 15 minutes. Reactions were stopped by adding 1% SDS/12 mM EDTA (final concentration) [106,107]. Subsequently, the spheroplasts were treated with RNase (final 0.02 μg/μl) at 37°C for 45 minutes, followed by proteinase K (final 0.4 μg/μl) at 65°C, overnight. The DNA samples were purified with phenol/chloroform extraction, precipitated with ethanol overnight and then resuspended in 1X TE. The samples (2.5 μg) were analyzed with gel electrophoresis (1.2% TAE agarose gel, at 35 V overnight) [105].

Quantitation of Rpb1 levels.

(A-B) Relative amounts of Rpb1 after addition of water (A, control) or galactose (B) to induce P DSB induction for one-hour, followed by cycloheximide (CHX) chase up to 150 minutes. Western blots from two independent experiments were quantified to compare Rpb1 levels (relative to Cdc11) between the indicated strains.

(TIFF)

Click here for additional data file.

Control experiments for RNA-seq and Venn diagrams of differentially expressed genes in indicated strains after DSB induction.

(A) FACS analysis to confirm benomyl-induced G2/M arrest. 1G, 1-hour GAL-induction (P). (B) PFGE analysis to monitor DSB induction on chromosome III. G2, G2/M arrest; 1G as in (A). (C) PCA demonstrating distribution of independent data sets between groups and clustering of data sets within groups. (D) Venn diagrams showing overlaps of differentially expressed genes in WT and rad30Δ cells after DSBs, based on RNA-seq. The red and blue arrows indicate up- and down-regulated genes respectively. Statistical significance of the overlapping genes was evaluated as described in Materials and Methods, with * p < 0.001.

(TIFF)

Click here for additional data file.

Genome-wide distribution of Polη and additional gene expression analyses for Polη-depleted cells during G2/M.

(A) Metagenome plot showing distribution of Polη, with 100 bp flanking regions upstream and downstream of the gene bodies during G2/M phase. The samples were first normalized to their respective input and then the values were scaled to the maximum value of the plot. (B) Western blot to check depletion of Polη in G2/M arrested cells. Final concentrations of auxin and doxycycline were 6 mM and 20 μg/ml respectively. IAA, auxin; dox, doxycycline; t0, the 0-time point after addition of IAA/dox; t1.5, 90 minutes after treatment. The drug solvents (50% ethanol and water) were added in the ‘-IAA/dox’ mock control. The western blot image, including the protein marker, was cropped to show selected samples. Cdc11 was used as loading control. (C) Representative Integrative Genomics Viewer (IGV) tracks showing the differences in distribution of Polη at selected promoters. The samples were normalized to their respective input and library size. (D) Expression of selected genes with or without depletion of Polη during G2/M, measured by RT-qPCR. Calculations were the same as described in the legend of Fig 3F. Error bars indicate the mean ± STDEV of three independent experiments.

(TIFF)

Click here for additional data file.

The rad30Δ mutant showed increased nucleosome occupancy, but no difference in activation of DNA damage checkpoint and ECO1 gene expression compared to WT cells.

(A) Monitoring nucleosome occupancy based on sensitivity of cells to MNase digestions. The concentrations of MNase were 0, 0.0125, 0.025, 0.05, 0.1 U/ml (final). One representative gel electrophoresis from at least two independent assays performed is shown. The gel images were cropped to show selected samples. M, DNA ladder; Un, undigested; 1x, monomer; 2x, dimer; 3x, trimer; 4x, tetramer. (B) Metagenome plot showing cohesin enrichment ± 1000 bp from the transcription start site (TSS) in WT and rad30Δ cells ± DSB induction in G2/M phase. The samples were first normalized to their respective input and then the values were scaled to the maximum value of the plot. (C) The data from (B) plotted relative to the WT-DSB sample. After normalizing to the input, all samples were also normalized to WT-DSB sample to visualize the changes between the WT and rad30Δ cells. (D) Metagenome plot showing cohesin distribution 1000 bp downstream and 100 bp upstream from the transcription end site (TES) in WT and rad30Δ cells ± DSB induction in G2/M phase. Plotted as in (B). (E) As in (C), except plotting cohesin distribution around the TES according to (D). (F) Monitoring activation of the DNA damage checkpoint (phosphorylation of Rad53) after DSB induction with western blot. Galactose was added into the G2/M arrested cell cultures to induce P break induction for 1- or 1.5-hour, denoted as 1G or 1.5G. Sample collected from G2/M arrested WT cells, treated with phleomycin (final 15 μg/ml) for 1.5 hours was included as positive control (PC). Cdc11 was used as loading control. M, protein marker. (G) Monitoring activation of DNA damage checkpoint during DSB recovery. DSBs were induced for 1- or 1.5-hour, as in (F). The cells were then allowed to recover in YEP media supplemented with glucose and benomyl for another 1.5 hour (1.5 R) at 35°C, to mimic the damage-induced cohesion assay. 1G, 1.5G, PC, M as in (F). Cdc11 was used as loading control. (H) FACS analyses of cell cycle progression in WT and rad30Δ cells, at indicated time points after release into YEP media supplemented with glucose to recover from DSB induction. Samples without DSBs were included as control. B, benomyl; R, recovery. (I) ECO1 gene expression in G2/M arrested WT and rad30Δ cells ± P (left) and ± γ-irradiation (right). The relative gene expression was measured by RT-qPCR. FBA1 was used as a reference gene for the ± P samples. Error bars indicate the mean ± STDEV of two independent experiments.

(TIFF)

Click here for additional data file.

The method and related control experiments for a typical damage-induced cohesion assay.

(A) Damage-induced cohesion assay performed with GAL induced DSBs on chromosome III (P). Strains harboring the temperature sensitive smc1-259 allele are arrested in G2/M by addition of benomyl (‘B’). Galactose is then added for expression of ectopic P (Smc1 WT) and induction of DSBs, for 1-hour. The temperature is then raised to 35°C, restrictive to the smc1-259 allele, for disruption of S-phase cohesion (blue rings). The Tet-O/TetR-GFP system (green dots) is used to monitor damage-induced cohesion (red rings) on chr. V. Chr., chromosome; III, three; V, five. B1 and 2 indicate replacement of media with freshly prepared benomyl. (B) FACS analysis to confirm G2/M arrest during the time course of a typical damage-induced cohesion assay. 3B, 3-hour benomyl arrest. (C) PFGE analysis to detect DSB induction on chromosome III. 1, G2/M arrest; 2, 1-hour GAL-induction (P and P). (D) Western blot to check expression of the GAL promoter driven ectopic Smc1-Myc protein. G2, G2/M arrest; 1G, 1-hour GAL-induction as in (C). Cdc11 was used as loading control. M, protein marker.

(TIFF)

Click here for additional data file.

Damage-induced cohesion assay performed with γ-irradiation and the maintenance of sister chromatid cohesion in htz1Δ cells.

(A) Damage-induced cohesion assay performed with γ-irradiation. Formation of damage-induced cohesion is monitored on chr. V with the same Tet-O/TetR-GFP system, as in S5A Fig, with slight differences in the experimental procedure. Strains with smc1-259 background are arrested in G2/M by addition of benomyl (‘B’), expression of ectopic P (Smc1 WT) is then induced by addition of galactose. The cells are subsequently pelleted, resuspended in 1X PBS supplemented with benomyl. The resuspension is split in one half for irradiation, and half as non-irradiated control. After irradiation, both ± irradiated cells are recovering in YEP media supplemented with galactose and benomyl. Subsequently, the media is changed to YEP containing glucose and benomyl, and the temperature raised to 35°C, to monitor formation of damage-induced cohesion. (B) Damage-induced cohesion assay of the hir1Δ mutant in response to γ-irradiation, performed as depicted in (A). Means ± STDEV from two independent experiments are shown. For each experiment, two-hundred cells were counted for each time point. (C) Sister chromatid cohesion maintenance of the htz1Δ mutant under prolonged G2/M arrest. The cells were initially synchronized in G1 by α-factor in YEP media containing galactose. Expression of P was then shut off by switching the carbon source to glucose (YEPD), which resulted in the subsequent prolonged G2/M arrest as monitored by FACS (left panel). Sister chromatid separation was monitored at the URA3 locus on Chr. V by the TetO/TetR-GFP system. Means ± STDEV from three independent experiments are shown (right panel). A rad61Δ mutant with known high sister separation under prolonged G2/M arrest was included as control. Parts of the results from the same experiments were previously published [28]. Chr., chromosome.

(TIFF)

Click here for additional data file.

Control experiments for the Rpb1-anchor away method.

(A-C) ChIP-qPCR analyses to determine chromatin association of Rpb1 at promoters and 3’-ends of selected genes, in G2/M arrested cells after DSB induction. The same genes as in Fig 7C–7E were analyzed. Error bars indicate the mean ± STDEV of three independent experiments. n, low-binding control (n2 in Fig 1B). (D) Representative in situ immunofluorescence images for samples collected from the damage-induced cohesion assays in Fig 7F. The cells were stained with anti-Rpb1 and then counterstained with DAPI. t0, the time point before splitting the culture for addition of rapamycin (RAP); 1h RAP, 1-hour after ± rapamycin; 2nd step, the secondary antibody alone as control. (E) Fold reduction of selected genes after 1-hour rapamycin treatment, measured by RT-qPCR. The 2-ΔCt values of untreated samples were set as 1. (F) Western blot to monitor early DNA damage response, as indicated by H2AS129-phosphorylation. RAP, rapamycin; R, recovery; IR, γ-irradiation (250 Gy); M, protein marker. Cdc11 was used as loading control. (G) Western blot to check expression of the ectopic Smc1-Myc, driven by the GAL promoter. G2, G2/M arrest; 1G, 1-hour GAL-induction. RAP, M, Cdc11 as in (F).

(TIFF)

Click here for additional data file.

A summary of the main results.

In G2/M arrested WT cells, genes belonging to the positive transcription regulation and chromatin assembly pathways are enriched compared to rad30Δ cells. Reduced chromatin assembly in rad30Δ cells results in less dynamic chromatin, indicated by additional nucleosomes. Deregulated transcription and sensitivity to elongation inhibitors in rad30Δ cells are indicated by thin arrows over the TSS and ORF. Histone exchange between H3 and the post-translationally modified H3 (H3K56Ac) at promoter regions is reduced in the hir1Δ mutant, while histone exchange of H2A.Z for H2A predominantly at the +1 nucleosome is prevented in the htz1Δ mutant, hampering transcriptional regulation. Both mutants were deficient in damage-induced cohesion. In contrast, deletion of SET2 compensated for reduced transcriptional capacity of the rad30Δ mutant, and suppressed the lack of damage-induced cohesion in rad30Δ cells. Taken together, histone exchange during transcription may facilitate formation of damage-induced cohesion. Transcriptional regulation is perturbed in rad30Δ cells, and this appeared to have a consequence on generation of damage-induced cohesion. Cells with a single green dot indicates established damage-induced cohesion while cells with two dots indicates lack of damage-induced cohesion. Since Polη may play an indirect role in transcription, recruitment of Polη to the promoter region is indicated with a dashed double ended arrow. ORF, open reading frame.

(TIFF)

Click here for additional data file.

Differential gene expression analysis.

(XLSX)

Click here for additional data file.

GSEA summary rad30Δ G2 versus WT G2.

(XLSX)

Click here for additional data file.

GSEA summary DSB versus G2.

(XLSX)

Click here for additional data file.

Numerical data of graphs.

(XLSX)

Click here for additional data file.

Summary of statistical analyses.

(XLSX)

Click here for additional data file.

Strains used in this study.

(DOCX)

Click here for additional data file.

Information on used primary antibodies.

(DOCX)

Click here for additional data file.

Primers used in ChIP-qPCR.

(DOCX)

Click here for additional data file.

Primers used in RT-qPCR.

(DOCX)

Click here for additional data file.

13 Apr 2021

Dear Dr Ström,

Thank you very much for submitting your Research Article entitled 'The assisting role of Pol η in transcription facilitates formation of damage-induced cohesion' to PLOS Genetics.

The manuscript was fully evaluated at the editorial level and by independent peer reviewers. The reviewers appreciated the attention to an important problem, but raised some substantial concerns about the current manuscript. Based on the reviews, we will not be able to accept this version of the manuscript, but we would be willing to review a much-revised version. Reviewers 1 and 3 raise a number of major concerns that would need to be addressed experimentally.  We cannot, of course, promise publication at that time.

Should you decide to revise the manuscript for further consideration here, your revisions should address the specific points made by each reviewer. We will also require a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript.

If you decide to revise the manuscript for further consideration at PLOS Genetics, please aim to resubmit within the next 60 days, unless it will take extra time to address the concerns of the reviewers, in which case we would appreciate an expected resubmission date by email to plosgenetics@plos.org.

If present, accompanying reviewer attachments are included with this email; please notify the journal office if any appear to be missing. They will also be available for download from the link below. You can use this link to log into the system when you are ready to submit a revised version, having first consulted our Submission Checklist.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please be aware that our data availability policy requires that all numerical data underlying graphs or summary statistics are included with the submission, and you will need to provide this upon resubmission if not already present. In addition, we do not permit the inclusion of phrases such as "data not shown" or "unpublished results" in manuscripts. All points should be backed up by data provided with the submission.

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool.  PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

PLOS has incorporated Similarity Check, powered by iThenticate, into its journal-wide submission system in order to screen submitted content for originality before publication. Each PLOS journal undertakes screening on a proportion of submitted articles. You will be contacted if needed following the screening process.

To resubmit, use the link below and 'Revise Submission' in the 'Submissions Needing Revision' folder.

[LINK]

We are sorry that we cannot be more positive about your manuscript at this stage. Please do not hesitate to contact us if you have any concerns or questions.

Yours sincerely,

Lorraine S. Symington

Associate Editor

PLOS Genetics

Gregory P. Copenhaver

Editor-in-Chief

PLOS Genetics

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: Cohesin binds around DSBs and is important for repair. Interestingly, DSBs trigger a genome-wide reinforcement of cohesion. This manuscript seeks to build on previous work implicating RAD30 in the genome-wide process but not DSB-local cohesion. They use an engineered budding yeast strain with a ts mutation in smc1, engineered to make a gal inducible HO break on chromosome 3 and monitor cohesion on chromosome 5. In a few cases gamma irradiation is used instead of an HO break. The authors investigate how RAD30 contributes to transcription as a mechanism by which it generates genome-wide cohesion. The hypothesis put forward is that RAD30 facilitates DI-cohesion via its role in the transcriptional response to DSBs. The data in support of this hypothesis needs to be stronger to warrant publication.

Summary of Figures:

In figure 1 they demonstrate Rad30 mutant strains are sensitive to transcriptional inhibitors and Rpb1 ChIP shows it has reduced localization to several genes despite normal protein levels. In figure 2 they demonstrate that gene expression is disrupted in the RAD30 deletion strain under normal growth conditions and the transcriptional response to a DSB is attenuated. The authors argue that this attenuated transcriptional program could underlie the lack of genome-wide damage-induced cohesion. However, by this logic, any mutation that reduces the transcriptional response to a DSB would have this phenotype-this is a point that should be further probed experimentally. In Figure 3, they show that rad30 mutants have more cohesin specifically at the TSS, with no difference in the TES, and irrespective of DSB induction. The relationship of this information to the main point could be more clearly stated. Figure 4 shows the gene expression signatures, further analyzing the data from Figure 2. In the next section of the manuscript the authors switch to genetic approaches to probe chromatin proteins that influence DI-cohesion, based in part on the gene expression data in the rad30 mutant. In figure 5 they show that deletion of HIR1 partially rescues DI cohesion in the Rad30 mutant background whereas histone deletions have no DI-cohesion defects and do not rescue. In figure 6 they test the effect of deleting the Htz1 histone variant. They cannot use the HO system so they induce breaks with radiation. They show that Htz1 levels are higher by ChIP following a break in the rad30 mutant. In Figure 7 they show Set2 can partially rescue DI-cohesion in the rad30 background. They also show ChIP data for Rpb1 but I can’t figure out what I am supposed to take away from the ChIP experiment. In Figure S7 they attempt to integrate their findings into a working model. I appreciate the effort to show a working model but it needs more work. In summary it seems that Hir1 and Htz1 affect histone turnover and DI-cohesion is defective. Do they mean to imply that Rad30 affects histone turnover directly? Indirectly? Why does altered elongation with set 2 deletion rescue the rad30 phenotype? If DI-cohesion is defective in rad30 due to an attenuated transcriptional response to a DSB, does set2 deletion rescue by rescuing the transcriptional response in rad30?

For me, the major question presented in the introduction, which is whether transcription contributes to genome-wide DI-cohesion, is an interesting one. But the manuscript does not meet the bar to answer this question, or even partially answer it with respect to RAD30 function. Results are not sufficiently integrated with each other, making it difficult to understand how each experiment addresses the overarching question of how transcription contributes to genome-wide DI cohesion. The data provided is not sufficient to support the hypothesis that Rad30 deletion blocks the formation of damage induced cohesion via altering the transcriptional program. Furthermore, its not clear whether the effect is direct or indirect. I have suggested three experiments below that would potentially begin to further examine the hypothesis, but depending on the outcome, more experiments may be warranted. This includes ChIP of Rad30, to be integrated with gene expression data, inhibiting transcription and examining DI-cohesion, and further experimentation to understand how deletion of Set2 rescues Rad30.

Major concerns and suggestions for how to improve the manuscript

1. Are the promoter types that are associated with differentially expressed genes in the rad30 mutant promoters at which Rad30 binds? This information would help argue for a direct role for Rad30 in the transcription of these genes, and would help support the model.

2. It’s a shame thiolutin induced a DNA damage response on its own and prevented them from testing DI-cohesion when transcription was inhibited, as this could help support the model. Could the authors find a different way to do this experiment? What about ActD treatment?

3. It’s not clear to me why the authors decided to examine the specific chromatin mutants chosen (Hir1, Htz1, Set2). Were these genes highly dysregulated in the gene expression data or chosen some other way? To me the manuscript seems to jumps around from one mutant to another without much logic and just when we start to understand how one mutation is acting they move to another. A more in depth analysis would be helpful to arrive as a more mechanistic understanding, perhaps focusing on how deletion of set2 rescues rad30. One experiment that could be done is examine whether the double mutant has a transcriptional profile that is more wild-type relative to the rad30 mutant profile, which would support the model.

Minor concerns

4. Please add relevant statistics to Figure 2 D and E to demonstrate the change from WT.

5. I’m not sure the results regarding short genes in figure 2F are relevant to the overall message.

6. The cohesion defect for RAD30 in part 5A is much larger than in part C and so a partial rescue would be much harder to observe.

7. For Figure 6, the authors should check Htz1 proteins levels by western to help with interpretation of the ChIP. Are the levels staying the same or going up in the rad30 mutant?

8. The authors have used a graphical image of budding yeast cartoon to diagram cohesion but it doesn’t print well in my copy. It needs a black outline.

Reviewer #2: In their manuscript PGENETICS-D-20-00364 Wu and colleagues attempts to define the contribution of the yeast TLS DNA Pol eta to DNA damage-induced cohesion formation, a phenomenon the same group reported in 2013 in Plos Genetics. Here, they conclude that Pol eta affects damage-induced cohesion through its novel role in transcription. First they confirm the role of Pol eta in transcription, previously reported by others, using RNA seq., establishing that a few hundred genes are up or downregulated in the absence of the RAD30 gene coding for Pol eta. The reduced association of RNA PolII at promoters and coding regions in rad30 delta detected by ChIP-qPCR agrees with a transcriptional regulatory role of Pol eta. Intriguingly, cohesin association at promoters is increased in rad30 delta, still damaged-induced cohesion is impaired. They found similar deficiency when deleting HIR1 needed for H3 exchange for transcription activation, and suppression of the rad30 delta transcriptional and cohesin formation defects by deletion of the SET2 histone methyl transferase repressing histone exchange during transcription elongation. Based on these they conclude that deregulated transcription in rad30 delta, which perturbs dynamic nucleosome assembly affects formation of damage-induced cohesion.

The topic of the paper DNA damage induced cohesion formation is highly interesting, and the results represent substantial contribution. The experiments are carefully executed with all the necessary controls, and the paper is clearly written.

I have one concern regarding the experimental repetitions. The authors state that the results indicate the mean ± STDEV of at least two independent experiments, in almost all figure legends. Two experiments are not enough for solid statistics and conclusions, especially when STDEV is high. They should indicate the exact numbers of biological and technical replicates for each experiment, and perform additional repetitions if necessary.

Reviewer #3: This manuscript investigates the potential links between transcription and damage-induced cohesion through a common role of Rad30, the DNA polymerase eta involved in translesion synthesis. The authors first tested the toxicity of transcription inhibitors in rad30D strains and observed increased sensitivity as well as a decrease in the amount chromatin bound PolII. Similar observations were obtained with the DI-cohesion mutant of Rad30, Rad30-S14A. The authors observed deregulation of transcription profiles for Rad30D in the presence and absence of DNA damage hence suggesting a role for Rad30D in general transcription regulation. Moreover, the levels of cohesin binding at TSS in Rad30D cells were increased and this led the authors investigate the link between DI-cohesion and transcription at promoter sites. To this aim they used hir1D and htz1D mutations, which disrupt histone assembly at gene promoters. They found that both mutants decrease the ability of cells to establish DI-cohesion, thus concluding that histone exchange disruption at promoters causes defects in damage-induced cohesion. Finally, the authors show that deletion of histone methyltransferase set2 suppresses the DI-cohesion defects of Rad30. The authors conclude from this data that transcription influences formation of damage-induced cohesion, and that Rad30 is required for damage-induced cohesion because it facilitates transcription.

Overall, this manuscript makes some interesting observations and indeed makes some weak links between Rad30, transcription and DI-cohesion. It has been previously shown that cohesin loading requires nucleosome free regions and therefore changes in transcription or nucleosome assembly/chromatin at promoters (in hir1D and hzt1D) are indeed expected to have an impact on DI-cohesion. The role of Rad30 is however intriguing and although the changes in transcription in Rad30D cells could indeed explain some effects, the links are very vague and often do not offer a clear cut explanation of how Rad30 might be either involved in transcription regulation or facilitate DI-cohesion (by altering transcription), as the authors only offer some genetic interactions. Also, the use of Rad30D mutants might not be the best experimental system. The authors should use Rad30 auxin degron strains where they could remove Rad30 upon DNA damage and avoid the background effect of the deletion.

To support their claims it would be important for the authors to illustrate their points using a handful of specific genome sites where they can investigate changes in transcription and cohesin loading (or cohesion establishment) upon damage in a manner dependent on Rad30 (using an auxin degron approach). This would significantly strengthen their study.

**********

Have all data underlying the figures and results presented in the manuscript been provided?

Large-scale datasets should be made available via a public repository as described in the PLOS Genetics data availability policy, and numerical data that underlies graphs or summary statistics should be provided in spreadsheet form as supporting information.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

12 Jul 2021

Submitted filename: 210707 Wu et al Rebuttal.docx

Click here for additional data file.

26 Jul 2021

Dear Dr Ström,

Thank you very much for submitting your Research Article entitled 'Transcriptional deregulation driven by absence of Polη brings negative impact on damage-induced cohesion' to PLOS Genetics.

The manuscript was fully evaluated at the editorial level and by independent peer reviewers. While two of the reviewers were satisfied with the revision, reviewer 1 identified some concerns that we ask you address in a revised manuscript. In particular, it would be important to confirm that addition of rapamycin does not impact DI-cohesion in WT cells. You might also consider the alternative title for the manuscript suggested by reviewer 1.

We therefore ask you to modify the manuscript according to the review recommendations. Your revisions should address the specific points made by each reviewer.

In addition we ask that you:

1) Provide a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript.

2) Upload a Striking Image with a corresponding caption to accompany your manuscript if one is available (either a new image or an existing one from within your manuscript). If this image is judged to be suitable, it may be featured on our website. Images should ideally be high resolution, eye-catching, single panel square images. For examples, please browse our archive. If your image is from someone other than yourself, please ensure that the artist has read and agreed to the terms and conditions of the Creative Commons Attribution License. Note: we cannot publish copyrighted images.

We hope to receive your revised manuscript within the next 30 days. If you anticipate any delay in its return, we would ask you to let us know the expected resubmission date by email to plosgenetics@plos.org.

If present, accompanying reviewer attachments should be included with this email; please notify the journal office if any appear to be missing. They will also be available for download from the link below. You can use this link to log into the system when you are ready to submit a revised version, having first consulted our Submission Checklist.

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Please be aware that our data availability policy requires that all numerical data underlying graphs or summary statistics are included with the submission, and you will need to provide this upon resubmission if not already present. In addition, we do not permit the inclusion of phrases such as "data not shown" or "unpublished results" in manuscripts. All points should be backed up by data provided with the submission.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

PLOS has incorporated Similarity Check, powered by iThenticate, into its journal-wide submission system in order to screen submitted content for originality before publication. Each PLOS journal undertakes screening on a proportion of submitted articles. You will be contacted if needed following the screening process.

To resubmit, you will need to go to the link below and 'Revise Submission' in the 'Submissions Needing Revision' folder.

[LINK]

Please let us know if you have any questions while making these revisions.

Yours sincerely,

Lorraine S. Symington

Associate Editor

PLOS Genetics

Gregory P. Copenhaver

Editor-in-Chief

PLOS Genetics

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: I appreciate that the authors executed and added a number of additional experiments to address reviewers’ concerns. The overarching question is quite interesting. Most of the individual experiments have proper controls, repeats, and quantification. However, I do not think the authors have reached the bar. The third reviewer commented in the first round of reviews that the link between Rad30, transcription, chromatin, and DI-cohesion is vague, and I think this problem persists in the revised version, although it has improved some in terms of knitting together a story. The constitutive genetic deletion approach makes it difficult to ascertain how the mutations are affecting transcription in the short term and long term, and if the effects are direct versus indirect, and how this corresponds with the effect on the loss of damage induced cohesion. All the null backgrounds allow for transcription to persist, so it must be something more specific about the transcriptional process that leads to loss of damage-induced cohesion. The concerns below are not an exhaustive list of weaknesses. Even if these issues could be addressed, the manuscript may still not meet the bar.

1. New evidence is presented suggesting that the effect of Rad30 deletion is probably indirect as the binding does not correlate with transcriptional changes. This correlation between binding and transcription, or lack thereof, would have been much stronger if the experiments were performed genome-wide and not at a handful of genes.

2. It’s not clear to me if loss of Rad30 changes histone turnover or histone acetylation. It seems premature to conclude that Rad30 perturbs transcription through a similar mechanism to either Htz1 or Hir1, based on the data included.

3. I appreciate that the authors added the experiment to induce the removal of Rpb1 to impact transcription. This is potentially a very important experiment. As a control the authors should validate that the addition of rapamycin alone does not impact damage-induced cohesion in a WT background, since rapamycin will impact the transcriptional program even in the absence of the anchors away allele of Rpb1.

4. Rad30 seems to have its effect on genes with a specific group of promoters. How should I think about the effect on specific promoters relative to damage induced cohesion? Does the actual gene expression program have any bearing on DI-cohesion? The focus on the genes up and down regulated in the null background, which is a long term effect, and the promoter types bound, confuses the issue of the effect on DI-cohesion for me. Stripping away extraneous information may help convey the message better.

Minor concerns.

1. The adjusted title is poorly formulated. The authors should state their findings in the positive. A preferable title would be something like “Transcription-related chromatin features facilitate damage-induced cohesion”

2. The writing style in the manuscript is improved from the original but it is still often difficult to follow the experiments and the logic.

3. The reply to reviewers has, at some points, a similar problem with clarity.

4. The quality/size of the figures was so poor in my pdf that I had to download figures one by one, and enlarge them.

5. Given comments 2-4, the reviewer’s work is much more challenging than for an average review assignment. The authors and journal should strive to provide a manuscript in a more easily reviewable state.

Reviewer #2: I accept the modifications.

Reviewer #3: The authors have addressed, or explained, most of my concerns, I find the revised version improved. I am now supportive of publication.

**********

Have all data underlying the figures and results presented in the manuscript been provided?

Large-scale datasets should be made available via a public repository as described in the PLOS Genetics data availability policy, and numerical data that underlies graphs or summary statistics should be provided in spreadsheet form as supporting information.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

4 Aug 2021

Submitted filename: 210803 Wu et al Rebuttal.docx

Click here for additional data file.

5 Aug 2021

Dear Dr Ström,

We are pleased to inform you that your manuscript entitled "Deficiency of Polη in Saccharomyces cerevisiae reveals the impact of transcription on damage-induced cohesion" has been editorially accepted for publication in PLOS Genetics. Congratulations!

Before your submission can be formally accepted and sent to production you will need to complete our formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Please note: the accept date on your published article will reflect the date of this provisional acceptance, but your manuscript will not be scheduled for publication until the required changes have been made.

Once your paper is formally accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you’ve already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosgenetics@plos.org.

In the meantime, please log into Editorial Manager at https://www.editorialmanager.com/pgenetics/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production and billing process. Note that PLOS requires an ORCID iD for all corresponding authors. Therefore, please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field.  This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager.

If you have a press-related query, or would like to know about making your underlying data available (as you will be aware, this is required for publication), please see the end of this email. If your institution or institutions have a press office, please notify them about your upcoming article at this point, to enable them to help maximise its impact. Inform journal staff as soon as possible if you are preparing a press release for your article and need a publication date.

Thank you again for supporting open-access publishing; we are looking forward to publishing your work in PLOS Genetics!

Yours sincerely,

Lorraine S. Symington

Associate Editor

PLOS Genetics

Gregory P. Copenhaver

Editor-in-Chief

PLOS Genetics

www.plosgenetics.org

Twitter: @PLOSGenetics

----------------------------------------------------

Comments from the reviewers (if applicable):

----------------------------------------------------

Data Deposition

If you have submitted a Research Article or Front Matter that has associated data that are not suitable for deposition in a subject-specific public repository (such as GenBank or ArrayExpress), one way to make that data available is to deposit it in the Dryad Digital Repository. As you may recall, we ask all authors to agree to make data available; this is one way to achieve that. A full list of recommended repositories can be found on our website.

The following link will take you to the Dryad record for your article, so you won't have to re‐enter its bibliographic information, and can upload your files directly:

http://datadryad.org/submit?journalID=pgenetics&manu=PGENETICS-D-21-00364R2

More information about depositing data in Dryad is available at http://www.datadryad.org/depositing. If you experience any difficulties in submitting your data, please contact help@datadryad.org for support.

Additionally, please be aware that our data availability policy requires that all numerical data underlying display items are included with the submission, and you will need to provide this before we can formally accept your manuscript, if not already present.

----------------------------------------------------

Press Queries

If you or your institution will be preparing press materials for this manuscript, or if you need to know your paper's publication date for media purposes, please inform the journal staff as soon as possible so that your submission can be scheduled accordingly. Your manuscript will remain under a strict press embargo until the publication date and time. This means an early version of your manuscript will not be published ahead of your final version. PLOS Genetics may also choose to issue a press release for your article. If there's anything the journal should know or you'd like more information, please get in touch via plosgenetics@plos.org.

4 Sep 2021

PGENETICS-D-21-00364R2

Deficiency of Polη in Saccharomyces cerevisiae reveals the impact of transcription on damage-induced cohesion

Dear Dr Ström,

We are pleased to inform you that your manuscript entitled "Deficiency of Polη in Saccharomyces cerevisiae reveals the impact of transcription on damage-induced cohesion  " has been formally accepted for publication in PLOS Genetics! Your manuscript is now with our production department and you will be notified of the publication date in due course.

The corresponding author will soon be receiving a typeset proof for review, to ensure errors have not been introduced during production. Please review the PDF proof of your manuscript carefully, as this is the last chance to correct any errors. Please note that major changes, or those which affect the scientific understanding of the work, will likely cause delays to the publication date of your manuscript.

Soon after your final files are uploaded, unless you have opted out or your manuscript is a front-matter piece, the early version of your manuscript will be published online. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.

Thank you again for supporting PLOS Genetics and open-access publishing. We are looking forward to publishing your work!

With kind regards,

Andrea Szabo

PLOS Genetics

On behalf of:

The PLOS Genetics Team

Carlyle House, Carlyle Road, Cambridge CB4 3DN | United Kingdom

plosgenetics@plos.org | +44 (0) 1223-442823

plosgenetics.org | Twitter: @PLOSGenetics