Altered cohesin dynamics and H3K9 modifications contribute to mitotic defects in the cbf11Δ lipid metabolism mutant

Mitosis is an elaborate process critical for cell reproduction, and the fidelity of mitosis is crucial for the faithful distribution of genetic information into the daughter cells. Mitotic defects can result in aneuploidy, as frequently seen in cancer cells, or in catastrophic mitosis and cell death. As such, mitosis is tightly controlled and regulated at many different levels (Bähler, 2005; Cullati and Gould, 2019; Grallert et al., 2015). The fission yeast Schizosaccharomyces pombe has been instrumental in identifying the factors and processes involved in mitosis, and in defining their effects on mitotic fidelity (Hayles and Nurse, 2018; Hayles et al., 2013). S. pombe features regional, repetitive and heterochromatinized centromeres resembling the centromeres of human cells. The special chromatin structure at the centromeres is important for proper kinetochore establishment and its attachment to the mitotic spindle (Roy et al., 2013; Tong et al., 2019). In post-replication chromosomes, sister chromatids are held together by cohesin rings, which in fission yeast are removed, predominantly at the onset of anaphase, to allow chromosome segregation (Peters et al., 2008). Unlike what occurs in human cells, S. pombe has a fully closed mitosis, during which the nuclear membrane does not break down and the spindle forms and elongates within the confines of the nuclear envelope (Zhang and Oliferenko, 2013).

Intriguingly, chemical inhibitors or genetic manipulations perturbing lipid metabolism have also been shown to affect mitotic fidelity in S. pombe. Inhibition of the fatty acid synthase or downregulation of acetyl coenzyme A carboxylase (Cut6 in S. pombe) result in a form of catastrophic mitosis called the ‘cut’ (for ‘cell untimely torn’) phenotype. In this lethal event, the cell nucleus is bisected by the cytokinetic septum before mitosis is completed (Makarova et al., 2016; Saitoh et al., 1996; Takemoto et al., 2016; Zach and Převorovský, 2018). Recent studies have highlighted the importance of rapid nuclear envelope expansion during anaphase for successful closed mitosis. The surface area of the nucleus enlarges by ∼30% in anaphase, and this process requires an adequate supply of membrane phospholipids. Failure to expand the nucleus is associated with mitotic spindle bending and breaking, chromosome segregation defects, and/or the cut phenotype. Consequently, it has been postulated that insufficient membrane phospholipid supply is the key factor underlying the mitotic defects observed in cells with perturbed lipid metabolism (Makarova et al., 2016; Takemoto et al., 2016).

Cbf11 is a CSL family transcription factor that directly regulates several lipid metabolism genes in S. pombe. Cells lacking Cbf11 show defects in cell cycle progression, including the cut phenotype. The cbf11Δ cells also show decreased content of storage lipid droplets, which would be consistent with a decreased capacity for membrane phospholipid production and, therefore, being prone to catastrophic mitosis (Převorovský et al., 2015, 2016). However, the deletion of cbf11 has been shown to have intriguing genetic interactions with mutations in cohesin loaders, a chromatin modification complex and kinetochore-related factors (Chen et al., 2012; Guo et al., 2014; Rallis et al., 2017). These findings suggest that there might be additional factors contributing to the occurrence of mitotic defects in cells with perturbed lipid metabolism. In this study, we set out to characterize in detail the mitotic progression in cbf11Δ cells and to explore the previously identified genetic interactions mentioned above. We report altered cohesin occupancy and histone H3K9 modifications at the centromeres, accompanied by an increased propensity for chromosome loss in these cells. Furthermore, we identify cohesin dynamics and chromatin structure as factors likely contributing to mitotic fidelity in the cbf11Δ lipid metabolism mutant.

To monitor the dynamics and outcomes of mitosis in the cbf11Δ mutant, we performed live-cell time-lapse microscopy using strains with fluorescently tagged histone H3 (Hht2–GFP) and α-tubulin (mCherry–Atb2) (Syrovatkina and Tran, 2015) to visualize the chromosomes and mitotic spindle, respectively. Using wild-type (WT) and cbf11Δ cells, we detected three different mitotic outcomes: (1) normal successful mitosis concluded by the formation of two daughter cells with equally segregated chromosomes, (2) catastrophic mitosis in the form of the cut phenotype with unequal chromosome distribution in the daughter cells, and (3) nuclear displacement, which results in the formation of a viable diploid daughter cell and a non-viable anucleate daughter cell (Fig. 1A,B). During the 336 min of the experiment, 51% of WT cells underwent mitosis, with only ∼6% (∼3% of total cells) of these showing aberrant mitotic outcomes (n=263 total cells). In contrast, during the same period, only 13% of cbf11Δ cells underwent mitosis, but ∼31% of those (∼4% of total cells) showed aberrant mitotic outcomes (n=333 total cells) (Fig. 1C). The lower percentage of mitotic cells in the observed cbf11Δ population might be linked to the slow-growth phenotype of this mutant. Interestingly, the nuclear displacement could explain the previously reported continual emergence of non-sporulating diploid subpopulations in cbf11Δ cultures (Převorovský et al., 2009). A similar, actin-dependent, nuclear displacement and evasion of catastrophic mitosis coupled with diploidization was recently documented in several other fission yeast cut mutants (Yukawa et al., 2021) and upon manipulation of regulators and enzymes responsible for the production of membrane phospholipid precursors (Foo et al., 2023).

To better understand the nature of mitotic defects in cbf11Δ cells, we performed a more detailed analysis of our live-cell microscopy dataset from Fig. 1. We randomly selected 10 cells from the categories ‘WT normal’, ‘cbf11Δ normal’ and ‘cbf11Δ abnormal’, and measured the timing and dynamics of nuclear division, chromosome segregation, spindle elongation, and cytokinesis (Fig. 2). WT cells displayed rapid and relatively uniform mitotic progression, with spindle formation starting ∼12–18 min before the onset of anaphase (Fig. 2A) and the median mitotic duration being 26 min (Fig. 2G). In contrast, the timing of mitotic progression was markedly perturbed in the cbf11Δ mutant (median mitotic duration=41 min). In these cells, the interval between spindle formation and anaphase onset was more variable and typically longer than in WT (∼10–30 min; Fig. 2B,H), suggesting possible problems with the initial attachment of chromosomes to spindle microtubules. Furthermore, in mutant cells that completed mitosis successfully, the spindle elongation was slower (Fig. 2C–E), and total mitotic duration was longer and more variable compared to WT (Fig. 2G).

In the case of cbf11Δ mitoses with aberrant outcomes, the spindle elongation rate was even more impaired (Fig. 2D,E), and the maximum spindle length reached before its disassembly was markedly shorter than in either WT or during successful cbf11Δ mitoses (Fig. 2F). Notably, the nuclear displacement phenotype was associated with mid-anaphase spindle bending and/or detachment of one of the daughter chromosome masses from the spindle, and subsequent spindle disassembly and merger of the two daughter chromosome masses into one diploid nucleus (Figs 1B and 2B). These observations are similar to those from previous reports of spindle bending and breaking in fission yeast cells with chemically inhibited fatty acid synthesis (Takemoto et al., 2016; Foo et al., 2023). Our data are therefore compatible with the hypothesis that fatty acid synthesis mutants are prone to mitotic defects as a result of insufficient supply of membrane phospholipids needed for the nuclear envelope expansion during anaphase (Makarova et al., 2016; Takemoto et al., 2016). However, as we show below, other factors also contribute to the mitotic defects in cbf11Δ cells.

The increased duration of the pre-anaphase mitotic period in cbf11Δ cells (Fig. 2B,H) prompted us to investigate the role of the spindle assembly checkpoint (SAC) in mitotic fidelity of the cbf11Δ mutant. Through APC/C inactivation, the SAC inhibits premature sister chromatid separation and mitotic exit until all chromosomes are properly attached to the spindle microtubules (Musacchio, 2015). Theoretically, SAC activation could provide cbf11Δ cells with more time to deal with any mitotic problems they might encounter. Under such a scenario, genetic perturbation of the SAC mechanism would be expected to lead to a higher incidence of unresolved mitotic defects. On the other hand, a failure to silence the SAC after chromosomes have been successfully attached to the spindle might result in prolonged metaphase arrest and perturbed timing of downstream mitotic and cytokinetic events. Therefore, under this alternative scenario, SAC abolition might actually suppress the mitotic defects of the cbf11Δ strain. Indeed, it was shown that SAC abolition can improve the viability of a cut mutant (Elmore et al., 2014). However, when we inactivated SAC in the cbf11Δ background by deleting either bub1 or mad2, we did not observe any significant changes in the incidence of catastrophic mitosis compared to what was seen for cbf11Δ single mutants (Fig. 3A,B). These results suggest that SAC activity only plays a minor role (if any) in the mitotic defects observed in cbf11Δ cells, or that the defects are not caused by problems with kinetochore-microtubule attachment.

The dynamics of sister chromatid cohesion are crucial for the successful execution of mitosis. Interestingly, the deletion of cbf11 was found to have negative genetic interactions with mutations in the Mis4 cohesin loading factor (adherin; mis4-242) and the Eso1 cohesin N-acetyltransferase (eso1-G799D), which are both important for the establishment of cohesion (Chen et al., 2012). Therefore, we tested whether cohesin plays any role in the mitotic defects of cbf11Δ cells. We have previously shown that cohesin (and condensin) expression is not regulated by Cbf11 (Převorovský et al., 2015). Now we first quantified cohesin (Psm1) localization to the known major cohesin-associated regions (CARs) on chromosome II (Bhardwaj et al., 2016) in asynchronous WT and cbf11Δ cells using chromatin immunoprecipitation (ChIP)-qPCR (Fig. 4D). Although cohesin occupancy along the chromosome arms was similar in both strains, we found significantly higher cohesin occupancy at the centromeric dh and dg repeats. Notably, centromeres are the regions where sister chromatin cohesion is abolished last during mitosis (Peters et al., 2008). Given that cbf11Δ cells show altered cell cycle and pre-anaphase mitotic duration compared to WT cells (Fig. 2), the observed difference in cohesin occupancy might merely reflect these changes in the timing of cell cycle progression. Alternatively, altered cohesin dynamics could play a role in the cbf11Δ mitotic defects. To test this hypothesis, we created double mutants of cbf11Δ and factors involved in cohesin loading and unloading (pds5Δ and wpl1Δ) (Murayama and Uhlmann, 2015; Tanaka et al., 2001) and determined their rates of catastrophic mitosis (Fig. 3A,B). Notably, the deletion of wpl1 resulted in significant suppression of the mitotic defects compared to in the cbf11Δ single mutant, and a noticeable but statistically unsignificant decrease was also observed upon deletion of pds5. It is, therefore, possible that altered timing of cohesin loading and/or removal and the ensuing changes in sister chromatid cohesion contribute to the incidence of catastrophic mitosis in cells lacking Cbf11.

We also used our panel of cbf11Δ double mutants to test whether suppression of mitotic defects correlated with suppression of slow growth (Fig. S1) and abnormal cell morphology, such as non-cylindrical or branched cells (Převorovský et al., 2009) (Fig. 3A). Although we did not observe any improvements in growth rate or cell morphology in the cbf11Δ double mutants with mad2Δ, bub1Δ or pds5Δ, we found that the cbf11Δ wpl1Δ double mutant also showed suppression of aberrant cell morphology, but not of growth defects. Therefore, these three types of defects (mitotic defects, slow growth and abnormal cell morphology) of cbf11Δ cells are at least partially independent of each other and can be separated genetically.

A previous high-throughput screen suggested that cbf11Δ cells are sensitive to the microtubule poison thiabendazole (TBZ) (Han et al., 2010). Furthermore, other studies showed that the loss of cbf11 displays intriguing genetic interactions with mutations in chromatin- and kinetochore-related factors. Namely, cbf11Δ interacts negatively with the deletion of sgf73, a subunit of the SAGA histone acetyltransferase complex, which shows decreased histone acetylation at H3K9 and H3K16 (Guo et al., 2014). Also, cbf11Δ interacts negatively with the deletion of the gsk3 kinase gene (Rallis et al., 2017). Notably, Gsk3 is important for localization of the Mis12 protein to the kinetochore and, subsequently, proper chromosome segregation during mitosis (Goshima, 2003). This prompted us to analyze the centromeric chromatin in cbf11Δ cells in more detail.

We first validated the previous report of cbf11Δ cells being sensitive to TBZ. When plated on YES medium containing TBZ, the cbf11Δ mutant indeed showed strong sensitivity to the drug (Fig. S2A), in line with the lowered robustness of spindle functioning in these cells (Figs 1B and 2B–F). The fission yeast centromeres consist of a central core, where histone H3 is replaced with the CENP-A histone variant, and inner and outer repeats that are heterochromatinized, hypoacetylated and feature methylation of H3K9 (Kniola et al., 2001; Pidoux and Allshire, 2004). Importantly, perturbations of the centromeric heterochromatin lead to mitotic defects (Allshire et al., 1995; Roche et al., 2016). To assess the status of centromeric and pericentromeric chromatin in cbf11Δ cells, we performed ChIP-seq analysis of H3K9 acetylation (H3K9ac; mediated by SAGA) and dimethylation (H3K9me2). Compared to WT, the cbf11Δ mutant exerted an altered pattern of H3K9me2 occupancy with large centromeric regions being exceedingly hypermethylated. Strikingly, we found very similar H3K9me2 perturbations at the centromeres of a hypomorphic cut6 mutant (Pcut6MUT has an ∼50% decrease in cut6 mRNA levels; Převorovský et al., 2016) (Fig. 4A; Fig. S3). Cut6 is the acetyl coenzyme A carboxylase, the rate-limiting enzyme of fatty acid synthesis, and a regulatory target of Cbf11.

On the other hand, the H3K9ac pattern showed mild alterations in the pericentromeric regions and outer centromeric repeats of the cbf11Δ mutant (Fig. 4A; Fig. S2B). To test whether the altered distributions of histone marks had any impact on gene expression in cbf11Δ cells, we measured transcript levels of genes flanking centromere I (per1 and sdh1) and of the centromeric dh and dg repeats (note that these repeats are present as multiple copies in all three fission yeast centromeres and individual repeats cannot be distinguished in our RT-qPCR assay). Indeed, we found that mRNA levels of per1 and sdh1 were ∼1.5× higher and ∼2× lower, respectively, in cbf11Δ cells compared to WT. Furthermore, the dh repeats showed significantly lower expression in the cbf11Δ mutant (Fig. 4B), possibly due to the hypermethylation of these loci. Finally, we tested whether the fidelity of chromosome segregation during mitosis was negatively affected in cbf11Δ cells. To this end, we analyzed the rate of loss of a non-essential minichromosome (Ch16) derived from chromosome III (Tinline-Purvis et al., 2009). We found that cbf11Δ cells had a ∼9-fold higher rate of Ch16 loss per generation compared to WT (Fig. 4C). Collectively, these results suggest that the organization and function of (peri)centromeric chromatin is perturbed in cells lacking Cbf11 and that this might contribute to their problems with chromosome segregation.

The requirement for fatty acid synthesis for the closed mitosis of S. pombe was noted decades ago (Saitoh et al., 1996). However, it has only recently been suggested that this requirement is connected to adequate membrane phospholipid supply for the rapid nuclear envelope expansion during the anaphase of closed mitosis (Makarova et al., 2016; Takemoto et al., 2016; Zach and Převorovský, 2018). When fatty acid synthesis is inhibited with cerulenin, the mitotic spindle buckles or even breaks, resulting in failed mitosis. Intriguingly, in the related fission yeast Schizosaccharomyces japonicus, which features semi-open mitosis, the nuclear membrane does not expand during anaphase and spindle buckling occurs as normal part of the mitotic process, followed by brief nuclear membrane rupture and establishment of two daughter nuclei (Yam et al., 2011). We have previously shown that the Cbf11 transcription factor regulates lipid metabolism genes, and its absence leads to increased occurrence of catastrophic mitosis (the cut phenotype) (Převorovský et al., 2009, 2016). However, given that Cbf11 regulates other genes and processes apart from lipid metabolism (Převorovský et al., 2015), we set out to characterize in detail the mitotic defects of the cbf11Δ mutant in order to assess whether other factors might also contribute to the mitotic defects observed in this mutant. Although our results are compatible with fatty acid shortage contributing to mitotic defects, we identified several additional factors that affect mitotic fidelity in cbf11Δ cells and, importantly, are not directly related to lipid metabolism. Specifically, these factors are related to (peri)centromeric chromatin structure, gene expression and sister chromatid cohesion dynamics.

First, we found that it takes the cbf11Δ cells longer to finish all pre-anaphase mitotic processes, suggesting that problems are present even before the insufficient supply of membrane phospholipids can take an effect on the closed mitosis (Fig. 2A,B). This indicates that proper microtubule attachment to kinetochores might be compromised and takes longer to achieve in cbf11Δ cells, possibly triggering the SAC. Intriguingly, SAC inactivation has been shown to suppress the temperature sensitivity of the cut9-665 APC/C mutant, which is also prone to catastrophic mitosis (Elmore et al., 2014). However, SAC inactivation had very little effect on mitotic fidelity in the cbf11Δ mutant (Fig. 3A,B), so either the defects lie in other pre-anaphase step(s), or the SAC is unable to sense and/or respond to kinetochore attachment defects in this particular case.

Second, the mitotic spindle dynamics is compromised in cbf11Δ cells, as cells undergoing aberrant mitosis showed lower spindle elongation rate (Fig. 2D,E), and spindle bending and/or breaking was also observed, together with chromosome detachment from the spindle during anaphase (Figs 1B and 2B). Furthermore, cbf11Δ cells are sensitive to the microtubule poison TBZ (Fig. S2A and Han et al., 2010) and are prone to minichromosome loss (Fig. 4C). These observations once again suggest that kinetochore function might be compromised when Cbf11 is missing. Nevertheless, it is also conceivable that the slower spindle elongation and spindle bending could be caused by the physical forces exerted by a nuclear envelope that expands too slowly, and hence not in synchronization with the growing spindle (Takemoto et al., 2016; Foo et al., 2023).

Third, the deletion of cbf11 was shown to exert negative genetic interactions with mutations in two essential factors involved in cohesin loading, Eso1 (eso1-G799D) and Mis4 (mis4-242) (Chen et al., 2012), suggesting that cohesin loading and sister chromatid cohesion might be compromised in cbf11Δ cells. However, we did not detect any decrease in cohesin occupancy along chromosome arms, and we actually found increased cohesin occupancy at the centromeric dh and dg repeats in this mutant (Fig. 4D). Moreover, deletion of the cohesin loader and unloader Wpl1 (the ortholog of mammalian WAPL) partially suppressed the mitotic defects of cbf11Δ cells (Fig. 3B). Notably, deletion of wpl1 can also rescue the otherwise lethal deletion of eso1 (Feytout et al., 2011), suggesting the requirement for a fine balance between cohesin loading and removal from DNA during the cell cycle. Collectively, these results indicate that altered cohesin dynamics, in general, contribute to the mitotic defects of cbf11Δ cells.

Finally, the above-mentioned centromere-related defects of cbf11Δ cells are accompanied by altered chromatin marks at and around the centromeres (Fig. 4A; Figs S2A, S3), together with altered gene expression from these regions (Fig. 4B). Specifically, H3K9 tends to be exceedingly hypermethylated at the dh and dg centromeric repeats (which also show increased cohesin occupancy), and these repeats have lower expression in cbf11Δ cells compared to WT. Importantly, a mutation in the SAGA histone acetyltransferase complex, which leads to decreased H3K9 acetylation (Nugent et al., 2010) and thus presumably increased H3K9 methylation (Alper et al., 2013; Nakayama et al., 2001), shows a negative genetic interaction with cbf11Δ. This suggests that such a combination of two hypermethylation-promoting mutations is deleterious for the cells, highlighting chromatin structure as yet another important contributor to the low mitotic fidelity of cbf11Δ cells. Although centromeric heterochromatin (H3K9me2) is required for proper chromosome segregation (Ekwall et al., 1996), heterochromatin is a dynamic structure that undergoes tightly regulated re-establishment in every cell cycle (Chen et al., 2008), and it is conceivable that exceedingly high levels of H3K9 methylation, as observed in cbf11Δ cells, might disrupt the cyclic behavior of centromeric chromatin structure and function. It is also important to note that heterochromatin, kinetochore function, cohesin occupancy and gene expression are all interconnected and actually interdependent (Bernard et al., 2001; Folco et al., 2019, 5; Grewal and Jia, 2007; Gullerova and Proudfoot, 2008; Nonaka et al., 2002; Volpe et al., 2002). Therefore, our findings implicating all these structures and processes in the mitotic defects of cbf11Δ cells might potentially represent different facets of a common underlying root cause.

Another important question is whether the chromatin-related cause(s) of mitotic defects are specific to the cbf11Δ mutant, or whether they apply more generally to other mutants with perturbed lipid metabolism (Zach and Převorovský, 2018). Here, we show that H3K9 dimethylation is also perturbed in a hypomorphic cut6 fatty acid synthesis mutant (Fig. 4A; Fig. S3), which is predisposed to catastrophic mitosis (Převorovský et al., 2016). Although proper (centromeric) chromatin structure and dynamics might simply become more critical in cells compromised by insufficient membrane phospholipid supply, other possible explanations have recently emerged. Fatty acid synthesis is a major consumer of acetyl-CoA, which also serves as a substrate for histone acetylation. Remarkably, metabolic manipulations leading to altered acetyl-CoA levels have been shown to trigger changes in histone modifications and expression of specific genes in the budding yeast and mammals (Galdieri and Vancura, 2012; McDonnell et al., 2016; Takahashi et al., 2006; Wellen et al., 2009). Moreover, specific changes in chromatin structure and gene expression can be brought about by highly localized alterations in acetyl-CoA availability in the nucleus (Mews et al., 2017). It is thus possible that, in lipid metabolism mutants, the chromatin landscape is altered in such a way that, for example, a subset of genes becomes overexpressed (as in the case of per1 in cbf11Δ cells; Fig. 4B) whereas centromeric heterochromatin is exceedingly hypermethylated, and this imbalance leads to defective mitosis. Intriguingly, we have recently found that mutations in lipid metabolism genes, such as cbf11 and cut6, lead to specific changes in histone modifications, and to derepression of a subset of stress-response genes and increased resistance to oxidative stress. Notably, the whole process is dependent on specific histone acetyl transferases (Princová et al., 2023). In summary, although the exact mechanism of how lipid metabolism is linked to centromeric chromatin function and fidelity of closed mitosis needs to be addressed by future studies, we have demonstrated several novel factors, not directly related to lipid metabolism, that affect mitotic fidelity in cells with perturbed lipid homeostasis. Moreover, given that closed mitosis is typical for many clinically and economically important fungal species, our findings could inform research in the fields of drug design and pest control.

Standard methods and media were used for the cultivation of Schizosaccharomyces pombe strains (Sabatinos and Forsburg, 2010). Unless indicated otherwise, S. pombe strains were grown at 32°C in YES with the appropriate supplements or stressors [yeast extract powder, Formedium; D-(+)-glucose, Merck; SP supplements, Formedium; thiabendazole, Sigma-Aldrich], as required.

The strains used in the study are listed in Table S1 and are available from the authors upon request. Deletion of cbf11 was carried out using the pMP91 targeting plasmid based on pCloneNAT1 as described (Gregan et al., 2006). All other strains used in this study were constructed by standard genetic crosses.

S. pombe cells were pre-cultured to exponential phase in 180 μl of YES in 96-well culture plates at 32°C in the VarioSkan Flash plate reader (Thermo Fisher Scientific). 5 μl of the preculture was used to inoculate 175 μl of fresh YES in 96-well culture plates and incubated for ∼48 h in the VarioSkan Flash plate reader at 32°C with background shaking (1080 rpm, rotation diameter 1 mm, a measuring time of 100 ms and 5 nm bandwidth). Optical densities were measured at 10 min intervals at λ=600 nm. Culture doubling times were calculated in R using the ‘growthrates’ package (https://github.com/tpetzoldt/growthrates).

WT and cbf11Δ cells harboring the Ch16-RMGAH non-essential minichromosome (Hylton et al., 2020; Tinline-Purvis et al., 2009) were grown on YES plates with low adenine content (10 mg/l). Under such conditions, cells that have lost the minichromosome and thus are adenine auxotrophs (ade6-M210) turn dark red. Sectored colonies (at least half-red) lost the minichromosome in the very first cell division after plating and are used to calculate the per-division chromosome loss rate. Fully red colonies lost the minichromosome prior to plating and are excluded from the analysis (Javerzat, 1996).

For chromatin immunoprecipitation (ChIP), 200 ml of fission yeast cultures were grown in YES to the density of 0.5×10⁷–1.2×10⁷ cells/ml, fixed with 1% formaldehyde for 30 min, quenched with 10 ml of 2.5 M glycine for 10 min, washed with water and broken with glass beads in lysis buffer [50 mM HEPES, 1 mM EDTA, 150 mM NaCl, 1% Triton X-100, 0.1% sodium deoxycholate, FY protease inhibitors (Serva); pH 7.6] using the FastPrep-24 instrument (MPI Biomedicals). Chromatin extracts were sheared with the Bioruptor sonicator (Diagenode) to yield DNA fragments of ∼200 bp; one-tenth of the total chromatin extract amount was kept for input DNA control. Next, 5 µg of anti-GFP antibody (ab290, Abcam) was added to ∼1 mg of chromatin extract and incubated for 1 h at 4°C. Antibody–chromatin complexes were then captured with Dynabeads^® Protein A (cat. no. 10002D, Life Technologies). The precipitated material was washed twice with lysis 500 buffer (50 mM HEPES, 1 mM EDTA, 500 mM NaCl, 1% Triton X-100, 0.1% sodium deoxycholate; pH 7.6), LiCl/NP-40 buffer (10 mM Tris-HCl, 1 mM EDTA, 250 mM LiCl, 1% Nonidet P-40, 1% sodium deoxycholate; pH 8.0), once in TE buffer (10 mM Tris-HCl, 1 mM EDTA; pH 8.0), and eluted in elution buffer (50 mM Tris-HCl, 10 mM EDTA, 1% SDS; pH 8.0). Crosslinks were reversed overnight at 65°C, samples were treated with 5 μl (0.5 mg/ml) DNase-free RNase for 1 h at 37°C followed by 7 μl (20 mg/ml) proteinase K for 2 h at 55°C (both Thermo Fisher Scientific), and purified twice using phenol-chloroform extraction and sodium acetate precipitation. The enrichment of specific target DNA sequences and controls in the immunoprecipitated material versus the input was determined by qPCR using the HOT FIREPol^® EvaGreen^® qPCR Supermix (Solis BioDyne) and the LightCycler 480 II instrument (Roche). The primers used are listed in Table S2. The strains used for ChIP were MP322 (h- leu1 psm1-GFP:LEU2) and MP745 (h- leu1 psm1-GFP:LEU2 cbf11Δ::NatR).

ChIP was performed analogously as in ChIP-qPCR with the following modifications. Harvested cells were washed in PBS. For each immunoprecipitation, 5 µg of antibody (H3K9ac, Ab4441; H3K9me2, Ab1220, both Abcam) were incubated with the chromatin extract for 1 h at 4°C with rotation. Immunoprecipitated and input DNA samples were purified using phenol-chloroform extraction and sodium acetate/ethanol precipitation. In the second biological replicate, DNA purification on AMPure XP beads (Beckman Coulter, AC63880) was performed following the phenol-chloroform extraction to remove low-molecular mass fragments and RNA. Library construction and sequencing were performed by BGI Tech Solutions (Hong Kong) using the BGISEQ-500 sequencing system (∼24 M of 50 nt single end reads per sample).

The S. pombe reference genome sequence and annotation were obtained from PomBase (release date 2018-09-04; Lock et al., 2019; Wood et al., 2002). Read quality was checked using FastQC version 0.11.8 (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/), and reads were aligned to the S. pombe genome using HISAT2 2.1.0 (Kim et al., 2019) and SAMtools 1.9 (Bonfield et al., 2021; Li et al., 2009). Read coverage tracks (i.e. target protein occupancy) were then computed and normalized to the respective mapped library sizes using deepTools 3.3.1 (Ramírez et al., 2016). Plots of coverage tracks were made using the IGV genome browser (Thorvaldsdóttir et al., 2013). The raw ChIP-seq data are available from the ArrayExpress (https://www.ebi.ac.uk/arrayexpress/) database under the accession numbers E-MTAB-11081 and E-MTAB-11082. The scripts used for ChIP-seq data processing and analysis are available from https://github.com/mprevorovsky/ox-stress_histones.

Total RNA was extracted from exponentially growing S. pombe cells using the MasterPure Yeast RNA Purification Kit (Epicentre). The procedure for removal of DNA contamination was as per the specification of the manufacturer but was performed twice to efficiently remove the DNA from the samples. Purified RNA was converted into cDNA using random primers and the RevertAid™ Reverse Transcriptase kit (Fermentas) using the protocol specified by the manufacturer. qPCR was performed using the HOT FIREPol^® EvaGreen^® qPCR Supermix (Solis BioDyne) and the LightCycler 480 II instrument (Roche). For RT-qPCR, act1 (actin) and/or rho1 (Rho GTPase) were used as reference genes. The primers used are listed in Table S2.

For nuclear staining, exponentially growing S. pombe cells were pelleted by centrifugation (1000 g, 3 min, 25°C) and fixed in 70% ethanol. Fixed cells were centrifuged again and rehydrated in deionized H₂O. Cells were stained with 0.5 μg/ml DAPI. Cells were applied on glass slides coated with soybean lectin, covered with a glass cover slip, and imaged using the 60× objective of the Olympus CellR widefield microscope with oil immersion (NA 1.4). The frequency of catastrophic mitosis occurrence was determined by manual evaluation of microscopic images using the counter function of ImageJ software, version 1.52p (Schneider et al., 2012). At least 400 cells from the asynchronous populations were analyzed per sample, and mitotic defects were scored based on nuclear morphology and septum presence/position.

For visualizing live cells, ∼4×10⁶–8×10⁶ exponentially growing S. pombe cells were collected by centrifugation (1000 g, 3 min, 25°C) and resuspended in ∼5 μl of YES. 1 μl of cell suspension was applied on 2% agarose-YES medium solidified on a PDMS spacer (generously provided by Phong Tran, Institut Curie, Paris) and covered with a coverslip. Cells were imaged using the Olympus CellR microscope with automatic z-axis objective movement with a 60× objective fitted with a 12-bit monochromatic Hamamatsu ORCA CCD camera. The images captured consisted of 10–11 Z-stacks of 0.3–0.5 µm sections, 500 ms exposure in red and green channels with a 2 min cycle, and 2×2 binning. Images were then processed using ImageJ 1.52p (Schneider et al., 2012) with maximum intensity projection. A custom ImageJ macro script was used for automating the image processing. Nuclear and spindle dynamics were analyzed using the tools available in ImageJ. The nuclear distance was measured by using Hht2–GFP signals and converting the green channel images into binary, measuring the maximum distance between the Hht2–GFP signals using the plot profile function in ImageJ. Spindle length was quantified by drawing a line along the length of the spindle (using mCherry–Atb2 signals) at each timepoint and measuring the length of the line using ImageJ. Spindle elongation rate was calculated by computing the slope from the exponential region of the spindle length versus time (see Fig. 2). Statistical analyses and generating of figures were performed in R studio and Graphpad Prism.

Please note that the observed frequencies of mitotic defects are not directly comparable between live and fixed cells. Following catastrophic mitosis, the dead cells rapidly lose histone–GFP fluorescence (imaging of live cells), but their DNA can still be visualized with DAPI for a much longer period (imaging of fixed cells), resulting in higher apparent defect frequencies in fixed cells.

Source numeric data for plots are provided in Table S3.

We are very grateful to Phong Tran for his advice and material support for live-cell microscopy. We thank Karl Ekwal, Charalampos Rallis, Tim Humphrey, and Phong Tran for providing strains. The bub1Δ, mad2Δ, pds5Δ, wpl1Δ, and psm1-GFP strains were provided by The Yeast Genetic Resource Center Japan. We also thank Adéla Kracíková for technical assistance and Ondřej Šebesta for assistance with microscopy. Microscopy was performed in the Laboratory of Confocal and Fluorescence Microscopy co-financed by the European Regional Development Fund and the state budget of the Czech Republic, projects no. CZ.1.05/4.1.00/16.0347 and CZ.2.16/3.1.00/21515, and supported by the Czech-BioImaging large RI project LM2018129. Computational resources were supplied by the project ‘e-Infrastruktura CZ’ (e-INFRA LM2018140) provided within the program Projects of Large Research, Development, and Innovations Infrastructures.

The peer review history is available online at https://journals.biologists.com/jcs/lookup/doi/10.1242/jcs.261265.reviewer-comments.pdf.