Evidence of centromeric histone 3 chaperone involved in DNA damage repair pathway

  1. Prakhar Agarwal
  2. Santanu Kumar Ghosh

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

The centromeric protein-A (CENP-A) is an evolutionary conserved histone H3 variant that marks the identity of the centromeres. Several mechanisms regulate the centromeric deposition of CENP-A as its mislocalization causes erroneous chromosome segregation, leading to aneuploidy-based diseases, including cancers. The most crucial deposition factor is a CENP-A specific chaperone, HJURP (Scm3 in budding yeast), which specifically binds to CENP-A. However, the discovery of HJURP as a DDR (DNA damage repair) protein and evidence of its binding to Holliday junctions in vitro indicate a CENP-A-deposition-independent role of these chaperones. In this study, using budding yeast, we demonstrate that Scm3 is crucial for the DDR pathway as scm3 cells are sensitive to DNA damage. We further observe that the scm3 mutant interacts with the rad52 DDR mutant and is compromised in activating DDR-mediated arrest. We demonstrate that Scm3 associates with the DNA damage sites and undergoes posttranslational modifications upon DNA damage. Overall, from this report and earlier studies on HJURP, we conclude that DDR functions of CENP-A chaperones are conserved across eukaryotes. Thus, the revelation that these chaperones confer genome stability in more than one pathway has clinical significance.

Article activity feed

  1. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    We sincerely thank all three reviewers for their professional and constructive feedback. We appreciate the thorough evaluation of our manuscript and are committed to revising both the manuscript and supplemental materials based on the suggestions. We have carefully considered each comment and have addressed most of them in the initial revised version, which has been transferred. Additionally, we are currently conducting new experiments to provide the requested data to address a few comments. We are confident that these revision experiments will be completed in a couple of months or so, which will significantly enhance the quality of our study.

    Reviewer #1 (Evidence, reproducibility and clarity (Required)):

    In the manuscript by Agarwal and Ghosh, the authors examine yeast Scm3 function in the DNA damage response. They show that Scm3 loss results in DNA damage sensitivity and more Rad52 foci. Importantly, Scm3 is recruited to DSB sites using an HO-cut site and its loss results in an attenuated DNA damage checkpoint as measured by Rad53 phosphorylation. The authors demonstrate convincingly that Scm3, like its human counterpart HJURP, plays a role in the DNA damage response through altered Rad53 activation. However, what its specific role in DNA repair is remains ambiguous.

    Major comments:

    1. It is unusual to see multiple DNA repair foci as those observed in Figure 2B. What is the distribution of cells with 1, 2, 3, 4, or more foci? Are more observed in SCM3-AID cells perhaps suggesting that the DSB ends are not being clustered as would be expected in WT cells exposed to DNA damage?

    Response: As per the reviewer’s comment, we have included a graph (Figure R2) showing the distribution of cells with 1, 2, 3, 4, or more Rad52-GFP foci when they are treated with MMS. There are more cells with 4 or more foci when Scm3 is depleted (SCM3-AID + Auxin) compared to the wild type (SCM3-AID). The average number of Rad52-GFP foci per cell presented in Figure 2B (2.8 in the mutant vs. 1.9 in the wild type) is well in accord with the previous report (Conde and San-Segundo, 2008), where the same was reported as ~2.5 in the cells lacking a methyl transferase Dot1, vs ~ 1.5 in the wild type. More Rad52-GFP foci in MMS-treated cells lacking Scm3 may arise due to the creation of too many damaged sites to be accommodated in 1-2 foci and/or due to the inability of the cells to cluster the DSB ends.

    This result has been incorporated as a new supplementary Figure S4C and new text has been added in the revised manuscript as: “We further quantified the distribution of cells with 1, 2, 3, or >4 Rad52-GFP foci in wild type (SCM3-AID) or Scm3 depleted (*SCM3-AID *+ auxin) cells treated with MMS. Scm3 depleted cells showed a significantly higher number of cells with more than >4 Rad52-GFP foci, suggesting the possibility of the creation of too many damaged sites to be accommodated in 1-2 foci or the inability of such cells to cluster the DSB ends.” in page 7, lines: 237-241.

    2. The peaks with increased Scm3 recruitment by ChIP-seq upon MMS is confusing as MMS does not induce specific damage at genomic locations. Is Scm3 being recruited at other genomic sites that might be more susceptible to DNA damage? Is Scm3 recruited to Pol2 sites for example? Or fragile sites?

    __Response: __We believe that the MMS induced increase in association of Scm3 with the non-centromeric chromatin loci depends on MMS sensitive vulnerable chromosomal sites. We agree with the reviewer that MMS might cause DNA damage at these sites, leading to Scm3 occupancy at those sites. Therefore, we compared the sites of Scm3 occupancy with possible such sites available from the literature that include fragile sites, RNA Pol II binding sites, double strand break hotspots, and coldspots. Based on our analysis, we have included the following lines in the ‘discussion’ section in page 16-17, lines 566-594 as follows:

    “Moreover, an overall increase in the chromatin association of Scm3 in response to MMS also suggests that Scm3 might be recruited to several repair centers or sites that are susceptible to DNA damage, for example, the fragile sites (Figure 3B, C, E, S6). These sites in yeast are DNA regions prone to breakage under replication stress, often corresponding to replication-slow zones (RSZs) (Lemoine et al., 2005). These regions include replication termination (TER) sequences, tRNA genes, long-terminal repeats (LTRs), highly transcribed genes, inverted repeats/palindromes, centromeres, autonomously replicating sequences (ARS), telomeres, and rDNA (Song et al., 2014). Since the helicase Rrm3 is often associated with these fragile regions (Song et al., 2014), we compared Scm3 binding sites with the top 25 Rrm3 binding sites from the literature (Azvolinsky et al., 2009). In untreated cells, Scm3 sites overlapped with three Rrm3 sites on chromosomes X, XII, and XIV. Whereas in MMS treated cells, overlapping was found with four Rrm3 sites, with two (on chromosomes XII and XIV) shared with untreated cells and two new sites were observed on chromosomes II and XII (Table R1). Mapping of the Scm3 sites with the tRNA genes and LTRs revealed that these sites from the untreated cells did not overlap with the LTRs (Raveendranathan et al., 2006). However, the same from the treated cells showed overlap with two LTRs on chromosome XVI. No overlap with tRNA genes was observed in the treated cells (Table R1). We next examined Scm3 occupancy at 71 TERs documented in the literature (Fachinetti et al., 2010). Scm3 was found to bind to 6 TERs in both untreated and MMS-treated cells. Notably, MMS treatment resulted in three new peaks, while three peaks were shared with untreated samples (Table R1). Lastly, we compared Scm3 sites with top 25 RNA Pol II sites obtained from the literature (Azvolinsky et al., 2009). In untreated cells, Scm3 was found at only one of these Pol II sites, whereas after MMS treatment, Scm3 sites overlapped with four such sites (Table R1). We further checked the occupancy of Scm3 at a few DSB hotspots (BUD23, ECM3, and CCT6) and DSB coldspot (YCR093W) as mentioned in the literature (Dash et al., 2024; Nandanan et al., 2021). However, we did not find Scm3 binding to these sites. Overall, in-silico analysis of the binding sites indicates that the non-centromeric enrichment of Scm3 occurs at sites that are amenable to DNA damage.”

    Table R1: The table summarising the occupancy of Scm3 in untreated or MMS treated conditions at the indicated regions

    Region

    Chromosome

    Scm3 occupancy

    Untreated

    MMS treated

    Rrm3 binding sites

    Chr II

    YES

    Chr X

    YES

    Chr XII

    YES

    Chr XII

    YES

    YES

    Chr XIV

    YES

    YES

    LTRs

    Chr XVI

    YES

    Chr XVI

    YES

    tRNA

    Chr XV

    YES

    TERs

    Chr IV

    YES

    Chr V

    YES

    Chr VI

    YES

    YES

    Chr VII

    YES

    Chr X

    YES

    Chr X

    YES

    Chr XIV

    YES

    YES

    Chr XV

    YES

    YES

    Chr XVI

    YES

    Pol II binding sites

    Chr II

    YES

    Chr X

    YES

    Chr XII

    YES

    Chr XII

    YES

    Chr XV

    YES

    The Table R1 has been incorporated as Table S1 in the revised manuscript.

    3. The phosphorylation aspect of Scm3 is intriguing and the authors show that Mec1 is not responsible for mediating its phosphorylation. Tel1 is another kinase that should be examined.

    __Response: __We thank the reviewer for the suggestion. We are in the process of examining the role of Tel1 kinase on Scm3 phosphorylation. The results from the experiment will be incorporated in the manuscript.

    Minor comment: 1. It is hard to see what MMS resistance the authors state is observed in Mif2-depleted cells in Figure S3. Perhaps this could be better explained or the claim removed.

    __Response: __We agree with the comment and have removed the claim from the manuscript.

    2. Protein levels of Scm3 or any of the other factors depleted with AID were never assessed.

    __Response: __We have assessed the protein level of Scm3 and a control protein, tubulin using western blotting as per the reviewer’s suggestion (Figure R3). We did not observe any significant change in the protein levels in SCM3-HA or SCM3-HA-AID cells, suggesting that the AID tagging of Scm3 per se did not make the cells non-functional and the protein was degraded as expected upon addition of auxin. Moreover, the SCM3-AID cells were used previously to examine the effect of Scm3 on kinetochore assembly (Lang et al., 2018).

    This result has been incorporated as Figure S2C, and new text has been added in the revised manuscript as: “The depletion of Scm3 was verified by observing a higher percentage of G2/M arrested cells and by western blot analysis verifying degradation of Scm3-AID after auxin treatment (Figure S2B, C).” in page 5, lines: 150-152.

    Reviewer #2 (Evidence, reproducibility and clarity (Required)):

    This manuscript describes a study implicating the Scm3 protein from budding yeast in the DNA damage response (DDR). Scm3 is a chaperone protein, whose main role is considered to be the loading of CENP-A(Cse4) at centromeres to facilitate chromosome segregation. However, the human ortholog of Scm3, HJURP, is known to have a role in DDR and in this study the authors provide evidence that Scm3 is also involved in the DDR in yeast. For the most part, the results presented support the conclusions made.

    Main Points

    1. Figure 1 Could depletion of Scm3 arrest cells in late G2/M and it is this delay that causes increased sensitivity to DNA damaging agents? A control with nocodazole or other means - that also arrests cells at this point - might provide a nice control for this. Perhaps the other kinetochore mutants, used therein, achieve this control - but cell cycle phase would need to be assessed.

    __Response: __ We thank the reviewer for pointing out to a probable effect of the cell cycle stage on the observed MMS sensitivity. In fact, we were also concerned that the observed DNA damage sensitivity in Scm3 depleted cells might be due to G2/M arrest. To rule out this possibility, we monitored Rad52-GFP foci as a marker for DNA damage in the wild type and Scm3 depleted cells both arrested at G2/M using nocodazole (Figure S4). While Scm3 depleted condition exhibited >20% Rad52-GFP positive cells, less than 10% wild type cells showed the same in the absence of any DNA damaging agents (Figure S4E, no MMS, 60 mins). Upon challenging these cells with MMS in the presence of nocodazole, Scm3 depleted condition exhibited over 40% Rad52-GFP positive cells, whereas less than 20% wild-type cells harboured Rad52-GFP. This significant increase in Rad52-GFP positive cells when Scm3 is depleted clearly indicates that the observed MMS sensitivity in these cells is due to the absence of Scm3 rather than due to an effect of a cell cycle stage. Furthermore, we have also used Cdc20 depleted G2/M arrested cells as a wild type control to test the activation of the DNA damage checkpoint by Rad53 phosphorylation. These cells showed robust Rad53 activation in response to MMS, in contrast to poor activation in Scm3 depleted cells (Figure 6), suggesting that G2/M arrest is not the reason for the DNA damage sensitivity observed in the latter cells.

    However, as per the reviewer's suggestion, we examined the MMS sensitivity of the wild type cells arrested at G2/M by nocodazole. As expected, these cells did not show increased sensitivity which further confirms that the DNA damage sensitivity observed in the scm3 mutant is not due to G2/M arrest (Figure R4B). This result has been incorporated within Figure S3, replacing the earlier Figure S3.

    To include this result, we have included new text, and revised the result section in page 5-6, lines 160-181 as follows:

    “The increased sensitivity of scm3-depleted cells to DNA-damaging agents could be due to the weakening of the kinetochores as Scm3-mediated deposition of Cse4 promotes kinetochore assembly or due to the delay in cell cycle, as Scm3 depleted cells arrest in late G2/M phase (Camahort et al., 2007; Cho and Harrison, 2011). If either of these holds true, perturbation of the kinetochore by degradation of other kinetochore proteins or wild type cells arrested at metaphase must show a similar sensitivity to MMS. In budding yeast, Ndc10 is recruited to the centromeres upstream of Scm3 (Lang et al., 2018), whereas the centromeric localization of Mif2, another essential inner kinetochore protein, depends on Scm3 and Cse4 (Xiao et al., 2017). We constructed *NDC10-AID *and *MIF2-AID *strains and used them for our assay to represent the proteins independent or dependent on Scm3 for centromeric localization, respectively. We also included one non-essential kinetochore protein, Ctf19, a protein of the COMA complex, to remove any possible mis-judgement in distinguishing cell-growth-arrest phenotype occurring due to drug-sensitivity vs. auxin-mediated degradation of essential proteins. The COMA complex is directly recruited to the centromeres through interaction with the N terminal tail of Cse4, hence dependent on Scm3 (Chen et al., 2000; Fischböck-Halwachs et al., 2019). Mid-log phase cells were harvested and spotted on the indicated plates, however, we did not observe any increased sensitivity of such cells to MMS (Figure S3). Further, wild type cells, when challenged in the presence of nocodazole and MMS, also did not show any increased sensitivity to MMS. Therefore, the increased sensitivity to MMS in scm3 mutant but not in other kinetochore mutant or metaphase arrested cells indicates that Scm3 possesses an additional function in genome stability besides its role in kinetochore assembly.”

    Further we have also revised the discussion section to include the observed results in page 15, lines 502-510 as follows:

    “However, since the primary function of Scm3 is to promote kinetochore formation by depositing Cse4 at the centromeres, it is important to address if the observed sensitivity is due to perturbation in kinetochores or due to cell cycle delay imposed in the absence of Scm3. Therefore, we similarly partially depleted two essential kinetochore proteins, Ndc10 and Mif2, and deleted one non-essential kinetochore protein, Ctf19, in separate cells and also challenged wild type cells to metaphase block but failed to detect any increased sensitivity to DNA damage stress (Figure S3). These results indicate that the drug sensitivity phenotype of Scm3 depleted cells is not due to weakly formed kinetochores or cell cycle delay.”

    2. Mutants of the HR pathway in yeast (e.g. rad52∆ with mre11∆ for example) are typically epistatic. The observation that Scm3 depletion is not epistatic with rad52∆ (Figure 1C) suggests the Scm3 acts via another pathway than the classic Rad52 HR pathway. This should be pointed out and discussed.

    __Response: __We have now included the discussion “In yeast, although HR is the preferred repair pathway, in the case of perturbed HR, an alternate pathway named non-homologous end joining (NHEJ) can occur. The absence of epistatic interaction between SCM3 and RAD52 (Figure 1C) suggests that Scm3 may function in ways other than the Rad52-mediated classical HR pathway. In this context, it would be interesting to test how Scm3 might interact with the key proteins of the NHEJ pathway, such as Ku70/Ku80 and Lig4 (Gao et al., 2016). It is possible that Scm3 may promote a certain chromatin architecture facilitating the DSB ends to stay together to be accessible for NHEJ-mediated end joining.” in page 16, lines 541-548.

    3. Figure 2 should include auxin treatment of RAD52-GFP cells (without the Scm3 degron) to show that the auxin treatment alone does not increase Rad52 foci.

    __Response: __ We performed the suggested experiment and did not observe any significant increase in Rad52-GFP positive cells when treated the cells with auxin+DMSO as compared to only DMSO (Figure R5).

    This result has been incorporated as a new supplementary Figure S4A,B and new text has been added in the revised manuscript as “To rule out the possibility that auxin treatment alone can cause increased Rad52-GFP foci formation, we challenged the wild type (RAD52-GFP) cells with auxin or DMSO and counted the number of cells with Rad52-GFP foci. We did not observe any increase in Rad52-GFP positive cells when treated with auxin+DMSO as compared to only DMSO (Figure S4A, B).” in page 7, lines: 233-236.

    4. Line 246-247 For the data presented, it seems to me possible that Scm3 depleted cells may indicate a defective DDR pathway (as stated) or may indicate defects in DNA replication or an increase in some other form of DNA damage?

    __Response: __We agree with the reviewer’s comment that the depletion of Scm3 can cause replication error or other form of DNA damage in addition to the defect in DDR pathway. To include this, we have modified the sentence as “Taken together, Scm3 depleted cells exhibit more Rad52 foci, indicating a compromised DDR pathway in these cells. Although, defects in DNA replication or creation of other DNA lesions producing more foci also cannot be ruled out.” in page 8, lines 255-257.

    5. In Figure 1 and throughout, please describe in the figure legends how error bars and p values are derived, and the number of experiments involved.

    __Response: __We have now verified all the figure legends and described how error bars and p values are derived and have mentioned the number of experiments involved.

    Minor points Line 35 replace 'cell survival' with 'cell division' - non-dividing cells can survive fine without chromosome segregation. See also line 62.

    __Response: __We have now changed ‘cell survival’ with ‘cell division’ in lines 35 and 62.

    Line 52 and throughout, I suggest replacing CenH3 with CENP-A or Cse4. The term CenH3 is confusing since regional centromeres contain both CENP-A nucleosomes and H3 nucleosomes - the latter of which can also be called CenH3 nucleosomes.

    __Response: __We have replaced CenH3 with CENP-A or Cse4 at the appropriate locations.

    Lines 69-79 specific references are needed for the sentences starting "HJURP was so named...", "In addition,...", "As a corollary,..." and "Finally,..." The final sentence of this paragraph, starting "Perhaps due to..." is unclear.

    __Response: __We have included the reference as mentioned by the reviewer. Also, we have changed the last line as “Notably, HJURP has been visualized to be diffusely present throughout the nucleus (Dunleavy et al., 2009; Kato et al., 2007), which may be due to its global chromatin binding and involvement in DDR.” in page 3, lines 77-79.

    Line 96 "gross chromatin" is unclear; also line 476.

    __Response: __We have changed gross chromatin to “bulk of the chromatin.” and incorporated it into the main text.

    Line 103 "dimerize"

    __Response: __We have replaced ‘dimerizes’ with ‘dimerize’

    Line 109 "most" and "highly" don't work together - perhaps better to say "the functions appear conserved from humans to yeast".

    __Response: __We have changed the wording as the reviewer suggested.

    Line 175 "grown" to "phase", see also line 223.

    __Response: __We have changed the wording as the reviewer suggested.

    Line 293 delete "besides"

    __Response: __We have deleted the word ‘besides’.

    Figure 5 - panels C & D, please make x axis labels clearer - they are directly underneath the 2kb ChIP. They should include a horizontal bar to indicate that all 5 ChIP experiments are included in each time point.

    __Response: __We have now included a horizontal bar in both Figure 5 and the corresponding supplementary Figure S8, to better represent the ChIP experiments. We thank the reviewer for pointing this out.

    Reviewer #3 (Evidence, reproducibility and clarity (Required)):

    Summary This manuscript studies the role of the Cse4 histone chaperone Scm3 in the S. cerevisiae DNA damage response. The authors show that decreased Scm3 levels exhibit genetic interactions with mutations in the Rad52 gene and sensitivity to MMS. They go on to show that some Scm3 co-localizes with sites of DNA damage using spreads and ChIP-seq techniques and that decreased levels of Scm3 have a reduced DNA damage checkpoint response. The Scm3 protein is also phosphorylated in response to DNA damage. Taken together, the authors propose a model whereby DNA damage recruits the Scm3 protein and Scm3 then helps mediate the checkpoint response. Overall, the data make a case that Scm3 has a relationship to the DNA damage checkpoint but the authors should be careful not to over-conclude that it has a precise role in checkpoint activation based on the data.

    Major Comments

    1. The western blots in the paper are not always entirely convincing. In addition, they are not described in enough detail to understand if a membrane was cut or if multiple gels were run. For example, the tubulin loading in Figure 6D is interrupted toward the end of the blot and the bands in Figure 7D go in different directions for the two blots for the MMS treated cells. In figure 6B, there are no detectable phospho-forms of Rad53 detected on the upper blot for the WT and scm3 lanes even though quantification is given on the right. It would be good to present better examples of the westerns or at least better describe what the reader is visualizing so the quantification and conclusions can be understood. How were the blots quantified? How were the westerns run and processed?

    Response: We have now included a separate paragraph in materials and methods regarding the gel run, processing and quantification of the western blots in the revised manuscript for better understanding of the readers:

    “To detect Scm3-6HA, Rad53, and g-H2A, the total proteins isolated from the appropriate cells were run on 12%, 8%, and 15% SDS gels, respectively. The proteins were transferred to the membranes, which were cut to detect the above proteins and the control protein tubulin separately. For the quantification of the bands on the western blots, a region of interest (ROI) was made around the band of interest, and the intensity of the band was calculated using ImageJ. A same ROI from a no-band area of the blot was used to calculate the background intensity. The background intensity was subtracted from the band intensity. The same process was done for the tubulin bands. The intensity of the target bands (Scm3-6HA, Rad53, and g-H2A) was divided by control tubulin band intensity to get the normalized values for the target bands, which were plotted using GraphPad Prism 9.0 (Version 9.4.1) software.” This has been added in page 25, lines: 881-890.

    Furthermore, we will again perform the experiments for a better representation of the western blots in figures 6B, D, and 7D.

    2. The argument that scm3 depletion leads to a defect in DNA damage checkpoint activation is not strongly supported. Monitoring exit from the cell cycle by multibudding is not the most rigorous assay, especially since the image shows one cell with 5 nuclei. The authors should release cells from G1 into auxin and MMS and monitor cell cycle progression at least one other cell cycle marker, such as anaphase onset, DNA replication and/or Pds1 levels.

    Response: As per the reviewer’s suggestion, in order to support our argument that the absence of Scm3 causes a defect in DNA damage checkpoint activation, we will examine if these cells abrogate G2/M arrest and show an early anaphase onset. For this, we will monitor the levels of Pds1, as a marker of anaphase onset, along the cell cycle in wild type and Scm3-depleted cells both deleted for Mad2 to remove any inadvertent effect of spindle assembly checkpoint. The schematics of the experimental workflow is given in Figure R1. Typically, the cells will be released from alpha factor arrest in the absence or presence of auxin (for the depletion of Scm3) and in the absence or presence of MMS. The samples will be harvested at the indicated time points and will be analyzed for:

    1. Western blot: Pds1-Myc (to detect anaphase onset)
    2. Western blot: Rad53 and p-Rad53 (to detect DNA damage activation)
    3. Immunofluorescence: Tubulin (to detect cell cycle stages) The results of the above experiment will be incorporated in the revised manuscript.

    3. The quantification in Figure 3B is not clear. Is it done on a per/nuclei basis? What pools of Scm3 and Ndc10 are being normalized?

    Response: The intensity was calculated as done before (Mittal et al., 2020, Shah et al., 2023). Typically, the intensity was first measured from the total signal of Scm3/Ndc10 from each chromatin mass or spread (DAPI) by making a polygon (ROI) around the Scm3/Ndc10+DAPI signal. The same ROI was dragged to the background area, from where two separate intensities were calculated. The average of the background intensities was then subtracted from the Scm3/Ndc10 intensity obtained from the same spread to get the normalized intensity depicting each dot in the box plot of Figure 3B. At least 30 spreads were quantified in a similar manner.

    We have mentioned this in the materials and methods section under “Microscopic image analysis.” section in page 22, lines 770-777 as follows: “For intensity calculation, a Region of Interest (ROI) was drawn around the Scm3/Ndc10/g-H2A+DAPI signal, and the intensity of Scm3/Ndc10/g-H2A was measured from each chromatin mass or spread (DAPI). An ROI of the same size was put elsewhere in the background area, from where two separate intensities were calculated. The average of the background intensities was then subtracted from the Scm3/Ndc10/g-H2A intensity obtained from the same spread to get the normalized intensity depicting each dot in the box plot of the respective figures as mentioned previously (Mittal et al., 2020; Shah et al., 2023).”

    Minor Comments 1. The model is elegant but there are chromatin pools (beyond the kinetochore pool) of Scm3 that do not contain Rad52 and/or gamma-H2X and vice versa. It would be helpful if the authors could speculate on how to reconcile these different pools. It might be premature to suggest such a detailed model at this point since the function of Scm3 in the checkpoint is still very unclear so I would encourage the authors to make a less detailed model.

    Response: By showing the green hallow, we have depicted the nuclear pool of Scm3, and we have not shown that the pool contains DDR proteins viz., Rad52 or g-H2A. Rather, we have shown the recruitment of these proteins at the DNA damage sites. Since the focus of this manuscript is on the non-centromeric functions of Scm3, we have not shown the kinetochore pool of Scm3. Although the model is a detailed one, the contribution from this work has been mentioned legitimately at every stage so that the readers can judge the merit of this work. We believe that a detailed model would provide a better perspective to the readers to correlate the revealed as well as yet-to-reveal functions of Scm3 in a spatiotemporal manner with the other players of the DDR pathway. Therefore, we prefer to keep the model in a detailed form.

    2. The chip-seq data is not publicly accessible. There is no reference to the data being available to review.

    __Response: __The data will be uploaded to the public domain.

    3. Line 103: not clear what "both the proteins dimerize" means...probably should be "both proteins dimerize"

    __Response: __We have changed the wording to “both proteins dimerize”.

    4. The argument that Ndc10 does not have a growth defect on MMS is a weak conclusion given that almost no control cells grow on auxin in the absence of MMS.

    __Response: __We have now repeated the spotting assay with a lesser concentration of auxin and replaced Figure S3 with a new Figure S3 (Figure R4) to better represent and conclude that the loss of Ndc10 does not cause MMS sensitivity.

  2. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    Summary

    This manuscript studies the role of the Cse4 histone chaperone Scm3 in the S. cerevisiae DNA damage response. The authors show that decreased Scm3 levels exhibit genetic interactions with mutations in the Rad52 gene and sensitivity to MMS. They go on to show that some Scm3 co-localizes with sites of DNA damage using spreads and ChIP-seq techniques and that decreased levels of Scm3 have a reduced DNA damage checkpoint response. The Scm3 protein is also phosphorylated in response to DNA damage. Taken together, the authors propose a model whereby DNA damage recruits the Scm3 protein and Scm3 then helps mediate the checkpoint response. Overall, the data make a case that Scm3 has a relationship to the DNA damage checkpoint but the authors should be careful not to over-conclude that it has a precise role in checkpoint activation based on the data.

    Major Comments

    1. The western blots in the paper are not always entirely convincing. In addition, they are not described in enough detail to understand if a membrane was cut or if multiple gels were run. For example, the tubulin loading in Figure 6D is interrupted toward the end of the blot and the bands in Figure 7D go in different directions for the two blots for the MMS treated cells. In figure 6B, there are no detectable phospho-forms of Rad53 detected on the upper blot for the WT and scm3 lanes even though quantification is given on the right. It would be good to present better examples of the westerns or at least better describe what the reader is visualizing so the quantification and conclusions can be understood. How were the blots quantified? How were the westerns run and processed?

    2. The argument that scm3 depletion leads to a defect in DNA damage checkpoint activation is not strongly supported. Monitoring exit from the cell cycle by multibudding is not the most rigorous assay, especially since the image shows one cell with 5 nuclei. The authors should release cells from G1 into auxin and MMS and monitor cell cycle progression at least one other cell cycle marker, such as anaphase onset, DNA replication and/or Pds1 levels.

    3. The quantification in Figure 3B is not clear. Is it done on a per/nuclei basis? What pools of Scm3 and Ndc10 are being normalized?

    Minor Comments

    1. The model is elegant but there are chromatin pools (beyond the kinetochore pool) of Scm3 that do not contain Rad52 and/or gamma-H2X and vice versa. It would be helpful if the authors could speculate on how to reconcile these different pools. It might be premature to suggest such a detailed model at this point since the function of Scm3 in the checkpoint is still very unclear so I would encourage the authors to make a less detailed model.

    2. The chip-seq data is not publicly accessible. There is no reference to the data being available to review.

    3. Line 103: not clear what "both the proteins dimerize" means...probably should be "both proteins dimerize"

    4. The argument that Ndc10 does not have a growth defect on MMS is a weak conclusion given that almost no control cells grow on auxin in the absence of MMS.

    Significance

    Significance

    This is the first examination of the role of Scm3 in the DNA damage response in S. cerevisiae. My expertise is in the chromatin and segregation fields, but I believe this work will be of interest to the DNA damage field as well. While the homologs of Scm3 are known to have a role in DNA damage, it was unclear if this was conserved in budding yeast. The data in this manuscript are consistent with findings in other organisms but the precise role of the chaperone in the DNA damage response is still unclear.

  3. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    This manuscript describes a study implicating the Scm3 protein from budding yeast in the DNA damage response (DDR). Scm3 is a chaperone protein, whose main role is considered to be the loading of CENP-A(Cse4) at centromeres to facilitate chromosome segregation. However, the human ortholog of Scm3, HJURP, is known to have a role in DDR and in this study the authors provide evidence that Scm3 is also involved in the DDR in yeast. For the most part, the results presented support the conclusions made.

    Main Points:

    1. Figure 1 Could depletion of Scm3 arrest cells in late G2/M and it is this delay that causes increased sensitivity to DNA damaging agents? A control with nocodazole or other means - that also arrests cells at this point - might provide a nice control for this. Perhaps the other kinetochore mutants, used therein, achieve this control - but cell cycle phase would need to be assessed.

    2. Mutants of the HR pathway in yeast (e.g. rad52∆ with mre11∆ for example) are typically epistatic. The observation that Scm3 depletion is not epistatic with rad52∆ (Figure 1C) suggests the Scm3 acts via another pathway than the classic Rad52 HR pathway. This should be pointed out and discussed.

    3. Figure 2 should include auxin treatment of RAD52-GFP cells (without the Scm3 degron) to show that the auxin treatment alone does not increase Rad52 foci.

    4. Line 246-247 For the data presented, it seems to me possible that Scm3 depleted cells may indicate a defective DDR pathway (as stated) or may indicate defects in DNA replication or an increase in some other form of DNA damage?

    5. In Figure 1 and throughout, please describe in the figure legends how error bars and p values are derived, and the number of experiments involved.

    Minor points:

    1. Line 35 replace 'cell survival' with 'cell division' - non-dividing cells can survive fine without chromosome segregation. See also line 62.

    2. Line 52 and throughout, I suggest replacing CenH3 with CENP-A or Cse4. The term CenH3 is confusing since regional centromeres contain both CENP-A nucleosomes and H3 nucleosomes - the latter of which can also be called CenH3 nucleosomes.

    3. Lines 69-79 specific references are needed for the sentences starting "HJURP was so named...", "In addition,...", "As a corollary,..." and "Finally,..." The final sentence of this paragraph, starting "Perhaps due to..." is unclear.

    4. Line 96 "gross chromatin" is unclear; also line 476.

    5. Line 103 "dimerize"

    6. Line 109 "most" and "highly" don't work together - perhaps better to say "the functions appear conserved from humans to yeast".

    7. Line 175 "grown" to "phase", see also line 223.

    8. Line 293 delete "besides"

    9. Figure 5 - panels C & D, please make x axis labels clearer - they are directly underneath the 2kb ChIP. They should include a horizontal bar to indicate that all 5 ChIP experiments are included in each time point.

    Significance

    This is a nice complement to the human work on HJURP and provides convincing evidence that Scm3 can be used to model the function of HJURP. Since yeast is such a tractable model, this work provides a route to study the role of this chaperone in DNA damage repair, which may also be true for human HJURP. The work itself is perhaps not too surprising, but is a solid advance in our understanding of the role of Scm3.

    My own expertise is in yeast DNA repair and chromosome segregation.

  4. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    In the manuscript by Agarwal and Ghosh, the authors examine yeast Scm3 function in the DNA damage response. They show that Scm3 loss results in DNA damage sensitivity and more Rad52 foci. Importantly, Scm3 is recruited to DSB sites using an HO-cut site and its loss results in an attenuated DNA damage checkpoint as measured by Rad53 phosphorylation. The authors demonstrate convincingly that Scm3, like its human counterpart HJURP, plays a role in the DNA damage response through altered Rad53 activation. However, what its specific role in DNA repair is remains ambiguous.

    Major comment:

    1. It is unusual to see multiple DNA repair foci as those observed in Figure 2B. What is the distribution of cells with 1, 2, 3, 4, or more foci? Are more observed in SCM3-AID cells perhaps suggesting that the DSB ends are not being clustered as would be expected in WT cells exposed to DNA damage?

    2. The peaks with increased Scm3 recruitment by ChIP-seq upon MMS is confusing as MMS does not induce specific damage at genomic locations. Is Scm3 being recruited at other genomic sites that might be more susceptible to DNA damage? Is Scm3 recruited to Pol2 sites for example? Or fragile sites?

    3. The phosphorylation aspect of Scm3 is intriguing and the authors show that Mec1 is not responsible for mediating its phosphorylation. Tel1 is another kinase that should be examined.

    Minor comment:

    1. It is hard to see what MMS resistance the authors state is observed in Mif2-depleted cells in Figure S3. Perhaps this could be better explained or the claim removed.

    2. Protein levels of Scm3 or any of the other factors depleted with AID were never assessed.

    Significance

    As mentioned above, a clear link for Scm3 in DNA damage repair has now been established in this work but its function in this process is descriptive.

  5. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    We sincerely thank all three reviewers for their professional and constructive feedback. We appreciate the thorough evaluation of our manuscript and are committed to revising both the manuscript and supplemental materials based on the suggestions. We have carefully considered each comment and have addressed most of them in the initial revised version, which has been transferred. Additionally, we are currently conducting new experiments to provide the requested data to address a few comments. We are confident that these revision experiments will be completed in a couple of months or so, which will significantly enhance the quality of our study.

    Reviewer #1 (Evidence, reproducibility and clarity (Required)):

    In the manuscript by Agarwal and Ghosh, the authors examine yeast Scm3 function in the DNA damage response. They show that Scm3 loss results in DNA damage sensitivity and more Rad52 foci. Importantly, Scm3 is recruited to DSB sites using an HO-cut site and its loss results in an attenuated DNA damage checkpoint as measured by Rad53 phosphorylation. The authors demonstrate convincingly that Scm3, like its human counterpart HJURP, plays a role in the DNA damage response through altered Rad53 activation. However, what its specific role in DNA repair is remains ambiguous.

    Major comments:

    1. It is unusual to see multiple DNA repair foci as those observed in Figure 2B. What is the distribution of cells with 1, 2, 3, 4, or more foci? Are more observed in SCM3-AID cells perhaps suggesting that the DSB ends are not being clustered as would be expected in WT cells exposed to DNA damage?

    Response: As per the reviewer’s comment, we have included a graph (Figure R2) showing the distribution of cells with 1, 2, 3, 4, or more Rad52-GFP foci when they are treated with MMS. There are more cells with 4 or more foci when Scm3 is depleted (SCM3-AID + Auxin) compared to the wild type (SCM3-AID). The average number of Rad52-GFP foci per cell presented in Figure 2B (2.8 in the mutant vs. 1.9 in the wild type) is well in accord with the previous report (Conde and San-Segundo, 2008), where the same was reported as ~2.5 in the cells lacking a methyl transferase Dot1, vs ~ 1.5 in the wild type. More Rad52-GFP foci in MMS-treated cells lacking Scm3 may arise due to the creation of too many damaged sites to be accommodated in 1-2 foci and/or due to the inability of the cells to cluster the DSB ends.

    This result has been incorporated as a new supplementary Figure S4C and new text has been added in the revised manuscript as: “We further quantified the distribution of cells with 1, 2, 3, or >4 Rad52-GFP foci in wild type (SCM3-AID) or Scm3 depleted (*SCM3-AID *+ auxin) cells treated with MMS. Scm3 depleted cells showed a significantly higher number of cells with more than >4 Rad52-GFP foci, suggesting the possibility of the creation of too many damaged sites to be accommodated in 1-2 foci or the inability of such cells to cluster the DSB ends.” in page 7, lines: 237-241.

    2. The peaks with increased Scm3 recruitment by ChIP-seq upon MMS is confusing as MMS does not induce specific damage at genomic locations. Is Scm3 being recruited at other genomic sites that might be more susceptible to DNA damage? Is Scm3 recruited to Pol2 sites for example? Or fragile sites?

    __Response: __We believe that the MMS induced increase in association of Scm3 with the non-centromeric chromatin loci depends on MMS sensitive vulnerable chromosomal sites. We agree with the reviewer that MMS might cause DNA damage at these sites, leading to Scm3 occupancy at those sites. Therefore, we compared the sites of Scm3 occupancy with possible such sites available from the literature that include fragile sites, RNA Pol II binding sites, double strand break hotspots, and coldspots. Based on our analysis, we have included the following lines in the ‘discussion’ section in page 16-17, lines 566-594 as follows:

    “Moreover, an overall increase in the chromatin association of Scm3 in response to MMS also suggests that Scm3 might be recruited to several repair centers or sites that are susceptible to DNA damage, for example, the fragile sites (Figure 3B, C, E, S6). These sites in yeast are DNA regions prone to breakage under replication stress, often corresponding to replication-slow zones (RSZs) (Lemoine et al., 2005). These regions include replication termination (TER) sequences, tRNA genes, long-terminal repeats (LTRs), highly transcribed genes, inverted repeats/palindromes, centromeres, autonomously replicating sequences (ARS), telomeres, and rDNA (Song et al., 2014). Since the helicase Rrm3 is often associated with these fragile regions (Song et al., 2014), we compared Scm3 binding sites with the top 25 Rrm3 binding sites from the literature (Azvolinsky et al., 2009). In untreated cells, Scm3 sites overlapped with three Rrm3 sites on chromosomes X, XII, and XIV. Whereas in MMS treated cells, overlapping was found with four Rrm3 sites, with two (on chromosomes XII and XIV) shared with untreated cells and two new sites were observed on chromosomes II and XII (Table R1). Mapping of the Scm3 sites with the tRNA genes and LTRs revealed that these sites from the untreated cells did not overlap with the LTRs (Raveendranathan et al., 2006). However, the same from the treated cells showed overlap with two LTRs on chromosome XVI. No overlap with tRNA genes was observed in the treated cells (Table R1). We next examined Scm3 occupancy at 71 TERs documented in the literature (Fachinetti et al., 2010). Scm3 was found to bind to 6 TERs in both untreated and MMS-treated cells. Notably, MMS treatment resulted in three new peaks, while three peaks were shared with untreated samples (Table R1). Lastly, we compared Scm3 sites with top 25 RNA Pol II sites obtained from the literature (Azvolinsky et al., 2009). In untreated cells, Scm3 was found at only one of these Pol II sites, whereas after MMS treatment, Scm3 sites overlapped with four such sites (Table R1). We further checked the occupancy of Scm3 at a few DSB hotspots (BUD23, ECM3, and CCT6) and DSB coldspot (YCR093W) as mentioned in the literature (Dash et al., 2024; Nandanan et al., 2021). However, we did not find Scm3 binding to these sites. Overall, in-silico analysis of the binding sites indicates that the non-centromeric enrichment of Scm3 occurs at sites that are amenable to DNA damage.”

    Table R1: The table summarising the occupancy of Scm3 in untreated or MMS treated conditions at the indicated regions

    Region

    Chromosome

    Scm3 occupancy

    Untreated

    MMS treated

    Rrm3 binding sites

    Chr II

    YES

    Chr X

    YES

    Chr XII

    YES

    Chr XII

    YES

    YES

    Chr XIV

    YES

    YES

    LTRs

    Chr XVI

    YES

    Chr XVI

    YES

    tRNA

    Chr XV

    YES

    TERs

    Chr IV

    YES

    Chr V

    YES

    Chr VI

    YES

    YES

    Chr VII

    YES

    Chr X

    YES

    Chr X

    YES

    Chr XIV

    YES

    YES

    Chr XV

    YES

    YES

    Chr XVI

    YES

    Pol II binding sites

    Chr II

    YES

    Chr X

    YES

    Chr XII

    YES

    Chr XII

    YES

    Chr XV

    YES

    The Table R1 has been incorporated as Table S1 in the revised manuscript.

    3. The phosphorylation aspect of Scm3 is intriguing and the authors show that Mec1 is not responsible for mediating its phosphorylation. Tel1 is another kinase that should be examined.

    __Response: __We thank the reviewer for the suggestion. We are in the process of examining the role of Tel1 kinase on Scm3 phosphorylation. The results from the experiment will be incorporated in the manuscript.

    Minor comment: 1. It is hard to see what MMS resistance the authors state is observed in Mif2-depleted cells in Figure S3. Perhaps this could be better explained or the claim removed.

    __Response: __We agree with the comment and have removed the claim from the manuscript.

    2. Protein levels of Scm3 or any of the other factors depleted with AID were never assessed.

    __Response: __We have assessed the protein level of Scm3 and a control protein, tubulin using western blotting as per the reviewer’s suggestion (Figure R3). We did not observe any significant change in the protein levels in SCM3-HA or SCM3-HA-AID cells, suggesting that the AID tagging of Scm3 per se did not make the cells non-functional and the protein was degraded as expected upon addition of auxin. Moreover, the SCM3-AID cells were used previously to examine the effect of Scm3 on kinetochore assembly (Lang et al., 2018).

    This result has been incorporated as Figure S2C, and new text has been added in the revised manuscript as: “The depletion of Scm3 was verified by observing a higher percentage of G2/M arrested cells and by western blot analysis verifying degradation of Scm3-AID after auxin treatment (Figure S2B, C).” in page 5, lines: 150-152.

    Reviewer #2 (Evidence, reproducibility and clarity (Required)):

    This manuscript describes a study implicating the Scm3 protein from budding yeast in the DNA damage response (DDR). Scm3 is a chaperone protein, whose main role is considered to be the loading of CENP-A(Cse4) at centromeres to facilitate chromosome segregation. However, the human ortholog of Scm3, HJURP, is known to have a role in DDR and in this study the authors provide evidence that Scm3 is also involved in the DDR in yeast. For the most part, the results presented support the conclusions made.

    Main Points

    1. Figure 1 Could depletion of Scm3 arrest cells in late G2/M and it is this delay that causes increased sensitivity to DNA damaging agents? A control with nocodazole or other means - that also arrests cells at this point - might provide a nice control for this. Perhaps the other kinetochore mutants, used therein, achieve this control - but cell cycle phase would need to be assessed.

    __Response: __ We thank the reviewer for pointing out to a probable effect of the cell cycle stage on the observed MMS sensitivity. In fact, we were also concerned that the observed DNA damage sensitivity in Scm3 depleted cells might be due to G2/M arrest. To rule out this possibility, we monitored Rad52-GFP foci as a marker for DNA damage in the wild type and Scm3 depleted cells both arrested at G2/M using nocodazole (Figure S4). While Scm3 depleted condition exhibited >20% Rad52-GFP positive cells, less than 10% wild type cells showed the same in the absence of any DNA damaging agents (Figure S4E, no MMS, 60 mins). Upon challenging these cells with MMS in the presence of nocodazole, Scm3 depleted condition exhibited over 40% Rad52-GFP positive cells, whereas less than 20% wild-type cells harboured Rad52-GFP. This significant increase in Rad52-GFP positive cells when Scm3 is depleted clearly indicates that the observed MMS sensitivity in these cells is due to the absence of Scm3 rather than due to an effect of a cell cycle stage. Furthermore, we have also used Cdc20 depleted G2/M arrested cells as a wild type control to test the activation of the DNA damage checkpoint by Rad53 phosphorylation. These cells showed robust Rad53 activation in response to MMS, in contrast to poor activation in Scm3 depleted cells (Figure 6), suggesting that G2/M arrest is not the reason for the DNA damage sensitivity observed in the latter cells.

    However, as per the reviewer's suggestion, we examined the MMS sensitivity of the wild type cells arrested at G2/M by nocodazole. As expected, these cells did not show increased sensitivity which further confirms that the DNA damage sensitivity observed in the scm3 mutant is not due to G2/M arrest (Figure R4B). This result has been incorporated within Figure S3, replacing the earlier Figure S3.

    To include this result, we have included new text, and revised the result section in page 5-6, lines 160-181 as follows:

    “The increased sensitivity of scm3-depleted cells to DNA-damaging agents could be due to the weakening of the kinetochores as Scm3-mediated deposition of Cse4 promotes kinetochore assembly or due to the delay in cell cycle, as Scm3 depleted cells arrest in late G2/M phase (Camahort et al., 2007; Cho and Harrison, 2011). If either of these holds true, perturbation of the kinetochore by degradation of other kinetochore proteins or wild type cells arrested at metaphase must show a similar sensitivity to MMS. In budding yeast, Ndc10 is recruited to the centromeres upstream of Scm3 (Lang et al., 2018), whereas the centromeric localization of Mif2, another essential inner kinetochore protein, depends on Scm3 and Cse4 (Xiao et al., 2017). We constructed *NDC10-AID *and *MIF2-AID *strains and used them for our assay to represent the proteins independent or dependent on Scm3 for centromeric localization, respectively. We also included one non-essential kinetochore protein, Ctf19, a protein of the COMA complex, to remove any possible mis-judgement in distinguishing cell-growth-arrest phenotype occurring due to drug-sensitivity vs. auxin-mediated degradation of essential proteins. The COMA complex is directly recruited to the centromeres through interaction with the N terminal tail of Cse4, hence dependent on Scm3 (Chen et al., 2000; Fischböck-Halwachs et al., 2019). Mid-log phase cells were harvested and spotted on the indicated plates, however, we did not observe any increased sensitivity of such cells to MMS (Figure S3). Further, wild type cells, when challenged in the presence of nocodazole and MMS, also did not show any increased sensitivity to MMS. Therefore, the increased sensitivity to MMS in scm3 mutant but not in other kinetochore mutant or metaphase arrested cells indicates that Scm3 possesses an additional function in genome stability besides its role in kinetochore assembly.”

    Further we have also revised the discussion section to include the observed results in page 15, lines 502-510 as follows:

    “However, since the primary function of Scm3 is to promote kinetochore formation by depositing Cse4 at the centromeres, it is important to address if the observed sensitivity is due to perturbation in kinetochores or due to cell cycle delay imposed in the absence of Scm3. Therefore, we similarly partially depleted two essential kinetochore proteins, Ndc10 and Mif2, and deleted one non-essential kinetochore protein, Ctf19, in separate cells and also challenged wild type cells to metaphase block but failed to detect any increased sensitivity to DNA damage stress (Figure S3). These results indicate that the drug sensitivity phenotype of Scm3 depleted cells is not due to weakly formed kinetochores or cell cycle delay.”

    2. Mutants of the HR pathway in yeast (e.g. rad52∆ with mre11∆ for example) are typically epistatic. The observation that Scm3 depletion is not epistatic with rad52∆ (Figure 1C) suggests the Scm3 acts via another pathway than the classic Rad52 HR pathway. This should be pointed out and discussed.

    __Response: __We have now included the discussion “In yeast, although HR is the preferred repair pathway, in the case of perturbed HR, an alternate pathway named non-homologous end joining (NHEJ) can occur. The absence of epistatic interaction between SCM3 and RAD52 (Figure 1C) suggests that Scm3 may function in ways other than the Rad52-mediated classical HR pathway. In this context, it would be interesting to test how Scm3 might interact with the key proteins of the NHEJ pathway, such as Ku70/Ku80 and Lig4 (Gao et al., 2016). It is possible that Scm3 may promote a certain chromatin architecture facilitating the DSB ends to stay together to be accessible for NHEJ-mediated end joining.” in page 16, lines 541-548.

    3. Figure 2 should include auxin treatment of RAD52-GFP cells (without the Scm3 degron) to show that the auxin treatment alone does not increase Rad52 foci.

    __Response: __ We performed the suggested experiment and did not observe any significant increase in Rad52-GFP positive cells when treated the cells with auxin+DMSO as compared to only DMSO (Figure R5).

    This result has been incorporated as a new supplementary Figure S4A,B and new text has been added in the revised manuscript as “To rule out the possibility that auxin treatment alone can cause increased Rad52-GFP foci formation, we challenged the wild type (RAD52-GFP) cells with auxin or DMSO and counted the number of cells with Rad52-GFP foci. We did not observe any increase in Rad52-GFP positive cells when treated with auxin+DMSO as compared to only DMSO (Figure S4A, B).” in page 7, lines: 233-236.

    4. Line 246-247 For the data presented, it seems to me possible that Scm3 depleted cells may indicate a defective DDR pathway (as stated) or may indicate defects in DNA replication or an increase in some other form of DNA damage?

    __Response: __We agree with the reviewer’s comment that the depletion of Scm3 can cause replication error or other form of DNA damage in addition to the defect in DDR pathway. To include this, we have modified the sentence as “Taken together, Scm3 depleted cells exhibit more Rad52 foci, indicating a compromised DDR pathway in these cells. Although, defects in DNA replication or creation of other DNA lesions producing more foci also cannot be ruled out.” in page 8, lines 255-257.

    5. In Figure 1 and throughout, please describe in the figure legends how error bars and p values are derived, and the number of experiments involved.

    __Response: __We have now verified all the figure legends and described how error bars and p values are derived and have mentioned the number of experiments involved.

    Minor points Line 35 replace 'cell survival' with 'cell division' - non-dividing cells can survive fine without chromosome segregation. See also line 62.

    __Response: __We have now changed ‘cell survival’ with ‘cell division’ in lines 35 and 62.

    Line 52 and throughout, I suggest replacing CenH3 with CENP-A or Cse4. The term CenH3 is confusing since regional centromeres contain both CENP-A nucleosomes and H3 nucleosomes - the latter of which can also be called CenH3 nucleosomes.

    __Response: __We have replaced CenH3 with CENP-A or Cse4 at the appropriate locations.

    Lines 69-79 specific references are needed for the sentences starting "HJURP was so named...", "In addition,...", "As a corollary,..." and "Finally,..." The final sentence of this paragraph, starting "Perhaps due to..." is unclear.

    __Response: __We have included the reference as mentioned by the reviewer. Also, we have changed the last line as “Notably, HJURP has been visualized to be diffusely present throughout the nucleus (Dunleavy et al., 2009; Kato et al., 2007), which may be due to its global chromatin binding and involvement in DDR.” in page 3, lines 77-79.

    Line 96 "gross chromatin" is unclear; also line 476.

    __Response: __We have changed gross chromatin to “bulk of the chromatin.” and incorporated it into the main text.

    Line 103 "dimerize"

    __Response: __We have replaced ‘dimerizes’ with ‘dimerize’

    Line 109 "most" and "highly" don't work together - perhaps better to say "the functions appear conserved from humans to yeast".

    __Response: __We have changed the wording as the reviewer suggested.

    Line 175 "grown" to "phase", see also line 223.

    __Response: __We have changed the wording as the reviewer suggested.

    Line 293 delete "besides"

    __Response: __We have deleted the word ‘besides’.

    Figure 5 - panels C & D, please make x axis labels clearer - they are directly underneath the 2kb ChIP. They should include a horizontal bar to indicate that all 5 ChIP experiments are included in each time point.

    __Response: __We have now included a horizontal bar in both Figure 5 and the corresponding supplementary Figure S8, to better represent the ChIP experiments. We thank the reviewer for pointing this out.

    Reviewer #3 (Evidence, reproducibility and clarity (Required)):

    Summary This manuscript studies the role of the Cse4 histone chaperone Scm3 in the S. cerevisiae DNA damage response. The authors show that decreased Scm3 levels exhibit genetic interactions with mutations in the Rad52 gene and sensitivity to MMS. They go on to show that some Scm3 co-localizes with sites of DNA damage using spreads and ChIP-seq techniques and that decreased levels of Scm3 have a reduced DNA damage checkpoint response. The Scm3 protein is also phosphorylated in response to DNA damage. Taken together, the authors propose a model whereby DNA damage recruits the Scm3 protein and Scm3 then helps mediate the checkpoint response. Overall, the data make a case that Scm3 has a relationship to the DNA damage checkpoint but the authors should be careful not to over-conclude that it has a precise role in checkpoint activation based on the data.

    Major Comments

    1. The western blots in the paper are not always entirely convincing. In addition, they are not described in enough detail to understand if a membrane was cut or if multiple gels were run. For example, the tubulin loading in Figure 6D is interrupted toward the end of the blot and the bands in Figure 7D go in different directions for the two blots for the MMS treated cells. In figure 6B, there are no detectable phospho-forms of Rad53 detected on the upper blot for the WT and scm3 lanes even though quantification is given on the right. It would be good to present better examples of the westerns or at least better describe what the reader is visualizing so the quantification and conclusions can be understood. How were the blots quantified? How were the westerns run and processed?

    Response: We have now included a separate paragraph in materials and methods regarding the gel run, processing and quantification of the western blots in the revised manuscript for better understanding of the readers:

    “To detect Scm3-6HA, Rad53, and g-H2A, the total proteins isolated from the appropriate cells were run on 12%, 8%, and 15% SDS gels, respectively. The proteins were transferred to the membranes, which were cut to detect the above proteins and the control protein tubulin separately. For the quantification of the bands on the western blots, a region of interest (ROI) was made around the band of interest, and the intensity of the band was calculated using ImageJ. A same ROI from a no-band area of the blot was used to calculate the background intensity. The background intensity was subtracted from the band intensity. The same process was done for the tubulin bands. The intensity of the target bands (Scm3-6HA, Rad53, and g-H2A) was divided by control tubulin band intensity to get the normalized values for the target bands, which were plotted using GraphPad Prism 9.0 (Version 9.4.1) software.” This has been added in page 25, lines: 881-890.

    Furthermore, we will again perform the experiments for a better representation of the western blots in figures 6B, D, and 7D.

    2. The argument that scm3 depletion leads to a defect in DNA damage checkpoint activation is not strongly supported. Monitoring exit from the cell cycle by multibudding is not the most rigorous assay, especially since the image shows one cell with 5 nuclei. The authors should release cells from G1 into auxin and MMS and monitor cell cycle progression at least one other cell cycle marker, such as anaphase onset, DNA replication and/or Pds1 levels.

    Response: As per the reviewer’s suggestion, in order to support our argument that the absence of Scm3 causes a defect in DNA damage checkpoint activation, we will examine if these cells abrogate G2/M arrest and show an early anaphase onset. For this, we will monitor the levels of Pds1, as a marker of anaphase onset, along the cell cycle in wild type and Scm3-depleted cells both deleted for Mad2 to remove any inadvertent effect of spindle assembly checkpoint. The schematics of the experimental workflow is given in Figure R1. Typically, the cells will be released from alpha factor arrest in the absence or presence of auxin (for the depletion of Scm3) and in the absence or presence of MMS. The samples will be harvested at the indicated time points and will be analyzed for:

    1. Western blot: Pds1-Myc (to detect anaphase onset)
    2. Western blot: Rad53 and p-Rad53 (to detect DNA damage activation)
    3. Immunofluorescence: Tubulin (to detect cell cycle stages) The results of the above experiment will be incorporated in the revised manuscript.

    3. The quantification in Figure 3B is not clear. Is it done on a per/nuclei basis? What pools of Scm3 and Ndc10 are being normalized?

    Response: The intensity was calculated as done before (Mittal et al., 2020, Shah et al., 2023). Typically, the intensity was first measured from the total signal of Scm3/Ndc10 from each chromatin mass or spread (DAPI) by making a polygon (ROI) around the Scm3/Ndc10+DAPI signal. The same ROI was dragged to the background area, from where two separate intensities were calculated. The average of the background intensities was then subtracted from the Scm3/Ndc10 intensity obtained from the same spread to get the normalized intensity depicting each dot in the box plot of Figure 3B. At least 30 spreads were quantified in a similar manner.

    We have mentioned this in the materials and methods section under “Microscopic image analysis.” section in page 22, lines 770-777 as follows: “For intensity calculation, a Region of Interest (ROI) was drawn around the Scm3/Ndc10/g-H2A+DAPI signal, and the intensity of Scm3/Ndc10/g-H2A was measured from each chromatin mass or spread (DAPI). An ROI of the same size was put elsewhere in the background area, from where two separate intensities were calculated. The average of the background intensities was then subtracted from the Scm3/Ndc10/g-H2A intensity obtained from the same spread to get the normalized intensity depicting each dot in the box plot of the respective figures as mentioned previously (Mittal et al., 2020; Shah et al., 2023).”

    Minor Comments 1. The model is elegant but there are chromatin pools (beyond the kinetochore pool) of Scm3 that do not contain Rad52 and/or gamma-H2X and vice versa. It would be helpful if the authors could speculate on how to reconcile these different pools. It might be premature to suggest such a detailed model at this point since the function of Scm3 in the checkpoint is still very unclear so I would encourage the authors to make a less detailed model.

    Response: By showing the green hallow, we have depicted the nuclear pool of Scm3, and we have not shown that the pool contains DDR proteins viz., Rad52 or g-H2A. Rather, we have shown the recruitment of these proteins at the DNA damage sites. Since the focus of this manuscript is on the non-centromeric functions of Scm3, we have not shown the kinetochore pool of Scm3. Although the model is a detailed one, the contribution from this work has been mentioned legitimately at every stage so that the readers can judge the merit of this work. We believe that a detailed model would provide a better perspective to the readers to correlate the revealed as well as yet-to-reveal functions of Scm3 in a spatiotemporal manner with the other players of the DDR pathway. Therefore, we prefer to keep the model in a detailed form.

    2. The chip-seq data is not publicly accessible. There is no reference to the data being available to review.

    __Response: __The data will be uploaded to the public domain.

    3. Line 103: not clear what "both the proteins dimerize" means...probably should be "both proteins dimerize"

    __Response: __We have changed the wording to “both proteins dimerize”.

    4. The argument that Ndc10 does not have a growth defect on MMS is a weak conclusion given that almost no control cells grow on auxin in the absence of MMS.

    __Response: __We have now repeated the spotting assay with a lesser concentration of auxin and replaced Figure S3 with a new Figure S3 (Figure R4) to better represent and conclude that the loss of Ndc10 does not cause MMS sensitivity.

  6. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    Summary

    This manuscript studies the role of the Cse4 histone chaperone Scm3 in the S. cerevisiae DNA damage response. The authors show that decreased Scm3 levels exhibit genetic interactions with mutations in the Rad52 gene and sensitivity to MMS. They go on to show that some Scm3 co-localizes with sites of DNA damage using spreads and ChIP-seq techniques and that decreased levels of Scm3 have a reduced DNA damage checkpoint response. The Scm3 protein is also phosphorylated in response to DNA damage. Taken together, the authors propose a model whereby DNA damage recruits the Scm3 protein and Scm3 then helps mediate the checkpoint response. Overall, the data make a case that Scm3 has a relationship to the DNA damage checkpoint but the authors should be careful not to over-conclude that it has a precise role in checkpoint activation based on the data.

    Major Comments

    1. The western blots in the paper are not always entirely convincing. In addition, they are not described in enough detail to understand if a membrane was cut or if multiple gels were run. For example, the tubulin loading in Figure 6D is interrupted toward the end of the blot and the bands in Figure 7D go in different directions for the two blots for the MMS treated cells. In figure 6B, there are no detectable phospho-forms of Rad53 detected on the upper blot for the WT and scm3 lanes even though quantification is given on the right. It would be good to present better examples of the westerns or at least better describe what the reader is visualizing so the quantification and conclusions can be understood. How were the blots quantified? How were the westerns run and processed?

    2. The argument that scm3 depletion leads to a defect in DNA damage checkpoint activation is not strongly supported. Monitoring exit from the cell cycle by multibudding is not the most rigorous assay, especially since the image shows one cell with 5 nuclei. The authors should release cells from G1 into auxin and MMS and monitor cell cycle progression at least one other cell cycle marker, such as anaphase onset, DNA replication and/or Pds1 levels.

    3. The quantification in Figure 3B is not clear. Is it done on a per/nuclei basis? What pools of Scm3 and Ndc10 are being normalized?

    Minor Comments

    1. The model is elegant but there are chromatin pools (beyond the kinetochore pool) of Scm3 that do not contain Rad52 and/or gamma-H2X and vice versa. It would be helpful if the authors could speculate on how to reconcile these different pools. It might be premature to suggest such a detailed model at this point since the function of Scm3 in the checkpoint is still very unclear so I would encourage the authors to make a less detailed model.

    2. The chip-seq data is not publicly accessible. There is no reference to the data being available to review.

    3. Line 103: not clear what "both the proteins dimerize" means...probably should be "both proteins dimerize"

    4. The argument that Ndc10 does not have a growth defect on MMS is a weak conclusion given that almost no control cells grow on auxin in the absence of MMS.

    Significance

    Significance

    This is the first examination of the role of Scm3 in the DNA damage response in S. cerevisiae. My expertise is in the chromatin and segregation fields, but I believe this work will be of interest to the DNA damage field as well. While the homologs of Scm3 are known to have a role in DNA damage, it was unclear if this was conserved in budding yeast. The data in this manuscript are consistent with findings in other organisms but the precise role of the chaperone in the DNA damage response is still unclear.

  7. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    This manuscript describes a study implicating the Scm3 protein from budding yeast in the DNA damage response (DDR). Scm3 is a chaperone protein, whose main role is considered to be the loading of CENP-A(Cse4) at centromeres to facilitate chromosome segregation. However, the human ortholog of Scm3, HJURP, is known to have a role in DDR and in this study the authors provide evidence that Scm3 is also involved in the DDR in yeast. For the most part, the results presented support the conclusions made.

    Main Points:

    1. Figure 1 Could depletion of Scm3 arrest cells in late G2/M and it is this delay that causes increased sensitivity to DNA damaging agents? A control with nocodazole or other means - that also arrests cells at this point - might provide a nice control for this. Perhaps the other kinetochore mutants, used therein, achieve this control - but cell cycle phase would need to be assessed.

    2. Mutants of the HR pathway in yeast (e.g. rad52∆ with mre11∆ for example) are typically epistatic. The observation that Scm3 depletion is not epistatic with rad52∆ (Figure 1C) suggests the Scm3 acts via another pathway than the classic Rad52 HR pathway. This should be pointed out and discussed.

    3. Figure 2 should include auxin treatment of RAD52-GFP cells (without the Scm3 degron) to show that the auxin treatment alone does not increase Rad52 foci.

    4. Line 246-247 For the data presented, it seems to me possible that Scm3 depleted cells may indicate a defective DDR pathway (as stated) or may indicate defects in DNA replication or an increase in some other form of DNA damage?

    5. In Figure 1 and throughout, please describe in the figure legends how error bars and p values are derived, and the number of experiments involved.

    Minor points:

    1. Line 35 replace 'cell survival' with 'cell division' - non-dividing cells can survive fine without chromosome segregation. See also line 62.

    2. Line 52 and throughout, I suggest replacing CenH3 with CENP-A or Cse4. The term CenH3 is confusing since regional centromeres contain both CENP-A nucleosomes and H3 nucleosomes - the latter of which can also be called CenH3 nucleosomes.

    3. Lines 69-79 specific references are needed for the sentences starting "HJURP was so named...", "In addition,...", "As a corollary,..." and "Finally,..." The final sentence of this paragraph, starting "Perhaps due to..." is unclear.

    4. Line 96 "gross chromatin" is unclear; also line 476.

    5. Line 103 "dimerize"

    6. Line 109 "most" and "highly" don't work together - perhaps better to say "the functions appear conserved from humans to yeast".

    7. Line 175 "grown" to "phase", see also line 223.

    8. Line 293 delete "besides"

    9. Figure 5 - panels C & D, please make x axis labels clearer - they are directly underneath the 2kb ChIP. They should include a horizontal bar to indicate that all 5 ChIP experiments are included in each time point.

    Significance

    This is a nice complement to the human work on HJURP and provides convincing evidence that Scm3 can be used to model the function of HJURP. Since yeast is such a tractable model, this work provides a route to study the role of this chaperone in DNA damage repair, which may also be true for human HJURP. The work itself is perhaps not too surprising, but is a solid advance in our understanding of the role of Scm3.

    My own expertise is in yeast DNA repair and chromosome segregation.

  8. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    In the manuscript by Agarwal and Ghosh, the authors examine yeast Scm3 function in the DNA damage response. They show that Scm3 loss results in DNA damage sensitivity and more Rad52 foci. Importantly, Scm3 is recruited to DSB sites using an HO-cut site and its loss results in an attenuated DNA damage checkpoint as measured by Rad53 phosphorylation. The authors demonstrate convincingly that Scm3, like its human counterpart HJURP, plays a role in the DNA damage response through altered Rad53 activation. However, what its specific role in DNA repair is remains ambiguous.

    Major comment:

    1. It is unusual to see multiple DNA repair foci as those observed in Figure 2B. What is the distribution of cells with 1, 2, 3, 4, or more foci? Are more observed in SCM3-AID cells perhaps suggesting that the DSB ends are not being clustered as would be expected in WT cells exposed to DNA damage?

    2. The peaks with increased Scm3 recruitment by ChIP-seq upon MMS is confusing as MMS does not induce specific damage at genomic locations. Is Scm3 being recruited at other genomic sites that might be more susceptible to DNA damage? Is Scm3 recruited to Pol2 sites for example? Or fragile sites?

    3. The phosphorylation aspect of Scm3 is intriguing and the authors show that Mec1 is not responsible for mediating its phosphorylation. Tel1 is another kinase that should be examined.

    Minor comment:

    1. It is hard to see what MMS resistance the authors state is observed in Mif2-depleted cells in Figure S3. Perhaps this could be better explained or the claim removed.

    2. Protein levels of Scm3 or any of the other factors depleted with AID were never assessed.

    Significance

    As mentioned above, a clear link for Scm3 in DNA damage repair has now been established in this work but its function in this process is descriptive.

  9. Version published to 10.1101/2024.07.26.605363v2 on bioRxiv
  10. Version published to 10.1101/2024.07.26.605363v1 on bioRxiv