Replisome-cohesin interactions provided by the Tof1-Csm3 and Mrc1 cohesion establishment factors

The replication checkpoint and sister chromatid cohesion establishment

The replication checkpoint is crucial in moments of replication fork stalling, e.g., at sites of DNA damage or following nucleotide depletion (Pardo et al. 2017). A low level of replication checkpoint activation is detectable even during unperturbed S phase (Forey et al. 2020). It is therefore conceivable that replisome-cohesin encounters lead to temporary checkpoint activation, which might facilitate sister chromatid cohesion establishment. In support of this hypothesis, single molecule observations in Xenopus egg extracts revealed frequent fork stalling upon replisome-cohesin encounters (Kanke et al. 2016).

To determine whether the replication checkpoint contributes to cohesion establishment, we measured sister chromatid cohesion in rad53-11 cells harboring a defective replication checkpoint effector kinase Rad53 (Shimada et al. 2002; Forey et al. 2020). Rad53-11 protein is sparsely expressed and fails to become phosphorylated in response to exposure to the replication inhibitor hydroxyurea (HU) (Fig. S1A). To assess the state of sister chromatid cohesion, we synchronized rad53-11 cells in G1 using α-factor, followed by release to progression through S phase and arrest in mitosis using the spindle poison nocodazole. Cells expressed a tet repressor-GFP fusion protein that marks a tet operator array integrated at the URA3 locus, close to the centromere of chromosome 5. Premature splitting of the resulting GFP dot in a mitotic arrest indicates defective sister chromatid cohesion (Fig. 1A). While cells lacking Mrc1 (mrc1Δ) displayed a substantial cohesion defect, compared to wild-type control cells, rad53-11 cells showed a much smaller increase in GFP dot separation (Fig. 1A). As a complementary approach to assess successful sister chromatid cohesion establishment, we monitored Smc3 acetylation levels during S phase by immunoblotting. As previously seen (Borges et al. 2013), Smc3 acetylation was reduced in mrc1Δ cells compared to wild type, but instead appeared slightly elevated in rad53-11 cells (Fig. 1B). These observations reveal largely successful sister chromatid cohesion establishment in the absence of a functional replication checkpoint.

Fig. 1figure 1

The replication checkpoint and cohesion establishment. A Representative example of cells imaged in the GFP dot assay. One GFP dot (bottom left cell) shows intact cohesion, while separated GFP dots (top right cell) indicate a cohesion defect. One hundred to 200 cells were scored for each indicated genotype. Three independent repeats of the experiment were performed. The means and the individual values are shown, revealing a significant cohesion defect in mrc1Δ cells (unpaired t test p < 0.0001) as well as a smaller yet significant defect in rad53-11 cells (p = 0.018). B Immunoblotting of synchronized cells from A at the indicated time points to analyze Smc3 acetylation and Rad53. Tubulin served as a loading control. The ac-Smc3/tubulin ratios were normalized to the highest ratio observed in wild-type cells. Means and individual values from three independent repeat experiments are shown. Acetylation levels were significantly lower in mrc1Δ cells and higher in rad53-11 (two-way ANOVA tests, p = 0.048 and p = 0.0081, respectively). C Schematic of the Mrc1 mutants analyzed, as well as result from the GFP dot assay to assess sister chromatid cohesion. Compared to increased GFP dot separation in mrc1Δ cells vs. wild type (unpaired t test, p = 0.0002), mrc1AQ-myc and mrc1-DC-myc cells showed a smaller but also significant cohesion defect (p = 0.0064 and p = 0.048, respectively). D Smc3 acetylation was assessed by immunoblotting; Mrc1 was detected via its myc epitope tag. Acetylation differences, assessed using a two-way ANVOA test, remained insignificant at p < 0.05. An antibody background band is marked with an asterisk

The Rad53 checkpoint kinase, as well as the main upstream sensor of the replication checkpoint, Mec1, are essential proteins for their role in regulating dNTP synthesis. They can be deleted in the absence of a cellular inhibitor of dNTP synthesis, Sml1 (Zhao et al. 1998). rad53Δ sml1Δ and mec1Δ sml1Δ cells showed a small increase in the frequency of GFP dot separation when compared to wild-type cells, well below the cohesion defects seen in mrc1Δ cells (Fig. S1B).

As an additional way to probe a possible replication checkpoint contribution to sister chromatid cohesion, we utilized two checkpoint defective Mrc1 alleles (Fig. 1C). mrc1AQ lacks 17 SQ and TQ motifs that are recognized by the Mec1 upstream kinase to mediate the checkpoint signal (Osborn and Elledge 2003). mrc1-ΔC in turn is missing the Mrc1 C-terminus that harbors a cluster of Rad53 kinase recognition sites (McClure and Diffley 2021). Previous studies reported conflicting observations with mrc1AQ cells, reporting wild-type levels of sister chromatid cohesion (Xu et al. 2004) or cohesion defects similar to mrc1Δ cells (Tsai et al. 2015). We observed an intermediate frequency of GFP dot separation in both mrc1AQ and mrc1-ΔC cells, above wild type but below what is observed in mrc1Δ cells (Fig. 1C). Meanwhile, Smc3 acetylation in mrc1AQ and mrc1-ΔC cells remained at wild-type levels, unlike the acetylation defect seen in mrc1Δ cells (Fig. 1D). It is possible that small sister chromatid cohesion defects, revealed in the GFP dot assay, arose because the mrc1AQ and mrc1-ΔC alleles also compromise Mrc1 function outside of the replication checkpoint. While the exact relationship between the replication checkpoint and cohesion establishment merits further exploration, taken together, our results suggest that cohesion establishment is achieved largely independently of the replication checkpoint.

Replisome speed and cohesion establishment

Mrc1 and Tof1/Csm3 ensure that the replisome progresses at full speed (Tourriere et al. 2005; Yeeles et al. 2017), raising the question of whether successful cohesion establishment requires replication fork encounters at a certain pace. Perhaps cohesion establishment reactions occur at a rate that must be coordinated with that of fork progression. We began to investigate this possibility by manipulating replisome speed independently of Tof1-Csm3 and Mrc1, by altering the dNTP pools available to DNA polymerases (Fig. 2A).

Fig. 2figure 2

Replisome speed and cohesion establishment. A Schematic illustrating modulated DNA polymerase speeds. FACS analysis of DNA content of cells of the indicated genotypes following G1 release into medium containing nocodazole (and HU as indicated). The wild-type profile is overlaid onto the others as a dark grey outline. B Sister chromatid cohesion in the indicated strains, as judged by the GFP dot assay. Mean and individual values of three independent experiments are shown. Unpaired t tests revealed a significant cohesion defect in mrc1Δ cells, compared to wild type (p = 0.0005), but no significant defect following 20 mM HU treatment, and no significant difference between mrcl1Δ and sml1Δ mrc1Δ cells. C Immunoblotting at the indicated time points to analyze Smc3 acetylation and Sml1. Tubulin served as a loading control. The ac-Smc3/tubulin ratios were normalized to the highest ratio observed in wild-type cells. Means and individual values from three independent repeat experiments are shown. Two-way ANOVA tests revealed significantly increased Smc3 acetylation following 20 mM HU treatment (p = 0.029), while other effect sizes did not reach significance at p < 0.05. D As B, using strains of the indicated genotypes. Unpaired t tests confirmed significant cohesion defects in tof1Δ and csm3Δ cells, compared to wild type (p = 0.0001 and p = 0.0002, respectively), but no significant differences due to sml1 deletion. E As C, but only samples taken at 120 min were analyzed. Unpaired t tests confirmed significant acetylation defects in tof1Δ and csm3Δ cells, compared to wild type (p = 0.0055 and p = 0.0007, respectively), but no significant differences due to sml1 deletion

HU inhibits ribonucleotide reductase, the enzyme complex that synthesizes dNTPs. While high HU concentrations (100–200 mM) block replication progression, lower concentrations lead to gradual replisome slowdown. Wild-type cells supplemented with 20 mM HU show replication rates similar to those observed in mrc1Δ cells without HU (both ca. 1100 bp/min, compared to a wild-type replication speed of ca. 2100 bp/min) (Theulot et al. 2022). We therefore examined sister chromatid cohesion in wild-type cultures, supplemented with 20 mM HU during their progression through S phase. Flow cytometric analysis of DNA content at 5-min intervals confirmed slower DNA synthesis, comparable to that in mrc1Δ cells (Fig. 2A). At later times, mrc1Δ cells completed S phase more efficiently than wild-type cells exposed to 20 mM HU, which could be due to activation of late replication origins in mrc1Δ but not wild-type cells (Koren et al. 2010).

Unlike mrc1Δ cells, wild-type cells treated with 20 mM HU did not show increased GFP dot separation—they had therefore successfully established sister chromatid cohesion (Fig. 2B). Measuring Smc3 acetylation levels in cells treated with 20 mM HU revealed no reduction, compared to that seen in mrc1Δ cells (Fig. 2C). On the contrary, acetylation levels were slightly higher than in wild-type control cells, an observation that we will return to below. Together, these observations suggest that a slowdown of replication fork progression is not by itself a cause of sister chromatid cohesion defects.

As an alternative way to probe the relationship between fork speed and sister chromatid cohesion establishment, we aimed to restore fork progression in mrc1Δ cells to wild-type rates. Removing the ribonucleotide reductase inhibitor Sml1 results in augmented cellular dNTP pools and an increased fork progression rate (Poli et al. 2012; Theulot et al. 2022). Faster replication fork progression in sml1Δ single mutant cells, compared to wild type, did not cause sister chromatid cohesion defects (Fig. 2A–C), strengthening the impression that cohesion establishment occurs independently of a specific fork speed. Sml1 deletion in mrc1Δ cells returned DNA synthesis rates in the double mutant cells close to those observed in wild-type controls (Fig. 2A). However, sister chromatid cohesion or Smc3 acetylation did not improve (Fig. 2B, C). This suggests that the cohesion defect in mrc1Δ cells did not arise due to slow replication fork progression.

Tof1-Csm3 act in a cohesion establishment pathway parallel to Mrc1 (Xu et al. 2007). We therefore also explored the effect of restoring replication speed to tof1Δ and csm3Δ cells. However, neither sister chromatid cohesion nor Smc3 acetylation were improved in tof1Δ or csm3Δ cells by removing Sml1 (Fig. 2D, E). Replication speed in the absence of Tof1 or Csm3 decreases more modestly when compared to mrc1Δ cells, akin to what is observed in wild-type cells treated with 5–10 mM HU (Yeeles et al. 2017; Theulot et al. 2022). However, treating wild-type cells with 5 mM or 10 mM HU also did not result in noticeable sister chromatid cohesion or Smc3 acetylation defects (Fig. S2A). Together, we conclude that successful sister chromatid cohesion establishment is possible at a range of DNA polymerase speeds.

HU augments Smc3 acetylation, independent of cohesion establishment

We had noticed that 20 mM HU treatment led to increased Smc3 acetylation in wild-type cells (Fig. 2C). Titrating HU concentrations between 5 and 20 mM (Fig. S2B) revealed a dose-dependent increase in Smc3 acetylation. While HU-mediated replication slowdown did not interfere with cohesion establishment, we wondered whether, on the contrary, increased Smc3 acetylation due to HU treatment could improve cohesion establishment. We therefore supplemented cultures of tof1Δ, csm3Δ, and mrc1Δ cells with 20 mM HU and monitored Smc3 acetylation and sister chromatid cohesion establishment. Indeed, HU exposure restored Smc3 acetylation in tof1Δ, csm3Δ, and mrc1Δ cells to levels seen in wild-type cells. However, increased Smc3 acetylation was not accompanied by improved sister chromatid cohesion (Fig. S2C). This reveals an Smc3 acetylation reaction, in response to HU treatment, that is independent of successful sister chromatid cohesion establishment. The nature of this reaction, as well as its relationship with Smc3 acetylation during undisturbed replication fork progression, remains to be investigated.

DNA helicase speed and cohesion establishment

Above, we altered replisome progression by modulating DNA polymerase speed via cellular dNTP pools. In comparison, Tof1-Csm3 and Mrc1 associate with the CMG helicase, where they are thought to affect replisome progression by regulating the speed of DNA unwinding (Eickhoff et al. 2019; Baretić et al. 2020; McClure and Diffley 2021). The effects on replisomes of altering helicase or polymerase speed might differ. For example, slowing down polymerases while DNA unwinding continues at a constant rate might result in increased availability of single-stranded DNA (ssDNA). In contrast, slower DNA unwinding at a constant rate of DNA synthesis might reduce the amount of accessible ssDNA. As ssDNA is a substrate for cohesion establishment (Murayama et al. 2018), we sought a way to emulate the Tof1-Csm3 and Mrc1 effects more accurately by modulating the rate of DNA unwinding.

DNA unwinding is controlled by the replication checkpoint. Rad53 phosphorylates Mrc1 on a number of C-terminal phosphorylation sites to slow down replication fork progression in response to checkpoint activation (McClure and Diffley 2021) (Fig. 3A). An MRC1-8D allele in which 8 of these phosphorylation sites were changed to phosphorylation-mimicking aspartates imposes constitutive replication fork slowdown. We therefore analyzed the proficiency of MRC1-8D cells to establish sister chromatid cohesion. Despite slower DNA unwinding, we did not observe any cohesion establishment defect in MRC1-8D cells, as monitored by GFP dot separation and Smc3 acetylation (Fig. 3B, C). This suggests that sister chromatid cohesion establishment is proficient even at slower helicase progression rates.

Fig. 3figure 3

Helicase speed and cohesion establishment. A Schematic depicting the MRC1 8D allele and its effect on the speed of DNA unwinding. FACS analysis of DNA content as cells of the indicated genotypes were released from G1 block for synchronous progression through S phase into nocodazole-imposed mitotic arrest. The wild-type profile is overlaid onto the others as a dark grey outline. B Sister chromatid cohesion in the same experiment was assessed by the GFP dot assay at the 120-min time point. Means and individual values of three independent repeat experiments are shown. Unpaired t tests confirmed a significant cohesion defect in mrc1Δ cells compared to wild type (p = 0.0012), but no significant difference between MRC1-FLAG and MRC1 8D-FLAG cells. C Smc3 acetylation during the same experiment was analyzed by immunoblotting. Tubulin served as a loading control. The ac-Smc3/tubulin ratios were normalized to the highest ratio observed in wild-type cells. Means and individual values from three independent repeat experiments are shown. Two-way ANOVA tests showed that acetylation differences remained insignificant at p < 0.05

As a complementary approach to address whether the availability of ssDNA is limiting the efficiency of sister chromatid cohesion establishment in tof1Δ, csm3Δ, and mrc1Δ cells, we utilized a mutation in the large subunit of the single-stranded DNA binding protein rfa1(G77E), which shows reduced affinity for single-stranded DNA. This allele improved the cohesion defect seen in the absence of Ctf18-RFC, a cohesion establishment factor that loads PCNA to serve as a potential adaptor for cohesin loading (Murayama et al. 2018; Liu et al. 2020; Minamino et al. 2023). Unlike in the case of Ctf18-RFC, the cohesion defects in the absence of Tof1-Csm3 or Mrc1 remained unaltered by the rfa1(G77E) allele (Fig. S3A). This observation supports the idea that Tof1-Csm3 and Mrc1 act in sister chromatid cohesion establishment independently of affecting ssDNA accessibility.

As a final approach to modulating the rate of replication fork progression, we utilized a yeast strain lacking the catalytic domain of DNA polymerase ε (polε-Δcat). In these cells, polymerase delta takes over leading strand synthesis, albeit at markedly reduced DNA unwinding and synthesis rates (Kesti et al. 1999; Yeeles et al. 2017) (Fig. S3B). polε-Δcat cells showed a marked increase of GFP dot separation, as well as slightly reduced Smc3 acetylation. This result can be interpreted in two ways. It could be that sister chromatid cohesion establishment is compromised at very low replication speeds in the polε-Δcat strain, which are substantially lower than in tof1Δ, csm3Δ, or mrc1Δ cells. Alternatively, the cohesion defect in polε-Δcat cells could stem from a role of DNA polymerase ε that is independent of its role in DNA replication. In support of the latter possibility, a small C-terminal deletion in DNA polymerase ε elicits a cohesion defect without apparent effect on DNA replication (Edwards et al. 2003). The role of DNA polymerase ε in sister chromatid cohesion establishment merits further exploration.

Tof1’s role in topoisomerase I and FACT recruitment

In addition to impacting on DNA replication, Tof1 is known to recruit auxiliary proteins to the replisome. A predicted unstructured C-terminal Tof1 extension contains the interaction sites for both topoisomerase I and FACT (Shyian et al. 2020; Westhorpe et al. 2020; Safaric et al. 2022) (Fig. 4A). It is conceivable that topological stress at replication forks, which accumulates in the absence of topoisomerase I, impairs cohesion establishment. Alternatively, histone clearance and redeposition, facilitated by FACT, might have to be coordinated with sister chromatid cohesion establishment.

Fig. 4figure 4

Tof1 and auxiliary replisome components. A Schematic depicting the Tof1 topoisomerase I and FACT interaction motifs, as well as the tof1-ΔC truncation. Sister chromatid cohesion was assessed in cells of the indicated genotypes 120 min after release from G1 block into nocodazole-imposed mitotic arrest. Means and individual values of three independent repeat experiments are shown. Unpaired t tests confirmed a significant cohesion defect in tof1Δ cells compared to wild type (p = 0.0001), but no significant difference between TOF1-HA and tof1-ΔC-HA cells, or between top1Δ and the wild-type control. B Immunoblotting of samples from the same experiment at the indicated times to analyze Smc3 acetylation and Tof1. Tubulin served as a loading control. The ac-Smc3/tubulin ratios were normalized to the highest ratio observed in wild-type cells. Means and individual values from three independent repeat experiments are shown. Two-way ANOVA tests confirmed a significant Smc3 acetylation defect in tof1Δ cells compared to wild type (p = 0.0001), but no significant difference between TOF1-HA and tof1-ΔC-HA cells. Acetylation was significantly increased in top1Δ cells (p = 0.0021)

To determine whether topoisomerase I and/or FACT recruitment to the replisome play a role in cohesion establishment, we made use of a C-terminal Tof1 truncation mutant (tof1-ΔC, lacking residues 982–1238) that no longer recruits topoisomerase I and FACT (Westhorpe et al. 2020; Safaric et al. 2022). We then assessed sister chromatid cohesion and Smc3 acetylation. The frequency of GFP dot separation, as well as Smc3 acetylation levels, was indistinguishable between wild-type and tof1-ΔC cells, while cells lacking Tof1 altogether (tof1Δ) displayed the expected cohesion defect signature (Fig. 4A, B). These observations suggest that topoisomerase I or FACT recruitment to the replisome is dispensable for successful sister chromatid cohesion establishment.

Topoisomerase I is a non-essential protein in budding yeast, as topoisomerase II can compensate for its fundamental role in relieving topological stress. We were therefore able to use a strain devoid of topoisomerase I (top1Δ) to independently address whether this enzyme contributes to cohesion establishment. We found that topoisomerase I is dispensable for cohesion establishment, as seen in the GFP dot assay (Fig. 4A). We noticed somewhat elevated Smc3 acetylation levels in top1Δ cells, when compared to wild-type cells (Fig. 4B). The reason for this increase, like that previously seen in HU-treated cells, remains to be elucidated. In summary, we conclude that successful cohesion establishment is possible without interactions that the Tof1 C-terminus provides with auxiliary replication factors, including topoisomerase I.

Direct protein interactions between Tof1-Csm3, Mrc1, and cohesin

The known roles of Tof1-Csm3 and Mrc1 at the replisome, which we so far studied, appear unrelated to the establishment of sister chromatid cohesion. We therefore hypothesized that Tof1-Csm3 and Mrc1 perform a previously unknown function in cohesion establishment. Recent structural studies have placed Tof1-Csm3 and the Mrc1 N-terminus at the front of the replisome (Eickhoff et al. 2019; Baretić et al. 2020), a prime location where cohesion establishment factors would physically encounter cohesin as the replication fork approaches. We therefore investigated whether Tof1-Csm3 and Mrc1 engage in protein interactions with cohesin.

Previous studies using nematode and human cell extracts have reported an interaction between their respective Tof1 orthologs, TIM-1 and TIMELESS, and cohesin (Chan et al. 2003; Leman et al. 2010). However, a mass spectrometry-based interaction screen using human cell extracts failed to confirm a TIMELESS-cohesin interaction (Ivanov et al. 2018). To investigate the possibility of direct Tof1-Csm3 or Mrc1 interactions with cohesin, we biochemically purified budding yeast Tof1-Csm3, Mrc1, and cohesin (Yeeles et al. 2017; Minamino et al. 2018). We also included the cohesion establishment factor Chl1, which was reported to interact with cohesin in human and yeast cell extracts (Parish et al. 2006; Samora et al. 2016).

To investigate the possibility of direct protein interactions, we employed an experimental setup in which a purified target protein is immobilized on beads, before these are briefly immersed with candidate binding partners (C. Smith and J. Diffley, personal communication). We immobilized cohesin using an antibody against a Pk epitope tag fused to its Smc1 subunit. These cohesin-covered beads, or antibody-only control beads, were incubated with Tof1-Csm3, Mrc1, or Chl1. Not reported to interact with cohesin, the GINS complex was used as a control. Beads were then washed, bound protein eluted, and analyzed by SDS-polyacrylamide gel electrophoresis followed by Coomassie Blue staining. Tof1-Csm3, Mrc1, as well as Chl1 were all recovered in the bead-bound fraction in a cohesin-dependent manner (Fig. 5A), suggesting that these cohesion establishment factors directly interact with the cohesin complex. The GINS complex, on the other hand, did not interact with cohesin.

Fig. 5figure 5

Protein interactions between cohesion establishment factors and cohesin. A Interaction screen with cohesin as the bait. Cohesin-coated or control beads were incubated with the indicated replisome components. Ten percent of input proteins are loaded next to the bead-bound fractions. Proteins were visualized by Coomassie Blue staining. B Interaction screen with Mrc1 as the bait. Mrc1-coated or control beads were first incubated with or without Tof1-Csm3, before a further incubation with cohesin or the GINS complex and analyzed as above

To confirm the interaction of Tof1-Csm3 and Mrc1 with cohesin, we took a reverse approach and immobilized Mrc1 on beads using an antibody against a Flag epitope tag on Mrc1. Following incubation with cohesin and washes, the bound protein was eluted by incubation with Flag peptide. Mrc1 on beads, but not control beads, interacted with cohesin (Fig. 5B). Addition of both Tof1-Csm3 and cohesin to Mrc1 beads led to the retention of both complexes, consistent with direct interactions among the three components. As a control, GINS was not retained on Mrc1 beads (to detect any possible GINS traces, elution in this case was with SDS).

We also immobilized Tof1-Csm3 on beads, which specifically interacted with cohesin (Fig. S4A). This interaction was unaltered by the inclusion of benzonase during the incubations, excluding the possibility that the interaction was mediated through DNA that might have contaminated our purified proteins.

So far, we performed our interaction incubations in buffer containing 100 mM potassium glutamate, conditions close to a budding yeast physiological environment. To test how stably Tof1-Csm3 interacts with cohesin, we added increasing NaCl concentrations during the binding incubation. The interaction was gradually lost in the presence of more than 250 mM NaCl (Fig. S4B), suggesting that polar or charge interactions contribute. We also performed benzonase treatment and salt titration for the Mrc1-cohesin interaction with similar outcomes (Fig. S4C,D).

Finally, we tested whether complexes formed between Tof1-Csm3, Mrc1 and cohesin are stable enough to be characterized by size exclusion chromatography. Following incubation, individual or combined protein samples were separated on a gel filtration column. Immunoblot analysis showed that cohesin and Tof1-Csm3 (visualized by a CBP tag on Csm3) individually eluted at volumes expected of their respective sizes (Fig. S5). After co-incubation, a faint Csm3 band became detectable in earlier fractions where cohesin elutes, while the bulk of the Csm3 profile remained unchanged, suggesting that the Tof1-Csm3 interaction with cohesin is weak. In a similar experiment, Mrc1 mostly co-eluted with cohesin, suggesting that Mrc1 and cohesin engage in a more durable interaction.

Tof1-Csm3 and Mrc1 deploy multipronged cohesin interactions

To explore the cohesion establishment factor interactions with cohesin, we performed protein crosslinking mass spectrometry (CLMS) to identify interaction sites. In two repeats of the experiment, we included cohesin, Tof1-Csm3 and Chl1, either with or without Mrc1. Cα atom distances spanned by crosslinks inside known structured regions of individual proteins fell within the expected range of the sulfo-SDA crosslinker (Fig. S6). Furthermore, we identified numerous crosslinks between subunits of the cohesin complex, as well as between Tof1 and Csm3, that were consistent with prior structural knowledge (Fig. 6A) (Baretić et al. 2020; Higashi et al. 2020), thereby overall validating the CLMS experiment.

Fig. 6figure 6

Multipronged cohesion establishment factor interactions with the cohesin complex. A Collated circos plot of interprotein crosslinks detected by CLMS in the two samples. Grey lines represent intersubunit crosslinks within the cohesin or Tof1-Csm3 complexes. Linkages between cohesin and Tof1 (green), Mrc1 (red), and Chl1 (purple) are highlighted. B Atomic model of the budding yeast Smc1-Smc3 dimer and Scc1 N-terminus, built using Phyre2 (Kelley et al. 2015) based on a fission yeast cohesin model (Higashi et al. 2020). CLMS linkage sites are highlighted. C Interaction screen with Tof1-Csm3 (left) or Mrc1 (right) on beads, comparing cohesin tetramer, cohesin trimer, and Smc1-Smc3 dimer as binding partners. Ten percent of the input and the bead bound fractions were analyzed by SDS-polyacrylamide gel electrophoresis followed by Coomassie Blue staining

Additionally, we detected crosslinks between the cohesion establishment factors and cohesin (Fig. 6A). Most prevalent were interactions of Tof1 and Mrc1 with Smc1 and Smc3, with an additional link each between Mrc1 and Scc3, as well as Chl1 and Scc1. Figure 6B shows these interactions mapped onto a Smc1-Smc3 structural model, revealing interaction clusters along both SMC coiled coils. To experimentally validate the conclusion that Tof1-Csm3 and Mrc1 preferentially interact with the Smc1-Smc3 dimer within the cohesin complex, we extended our biochemical interaction analysis. We immobilized Tof1-Csm3 or Mrc1 on beads that we then incubated with either purified cohesin tetramer complexes (Smc1-Smc3-Scc1-Scc3), cohesin trimers (

留言 (0)

沒有登入
gif