Abstract
The evolutionarily conserved protein complex, condensin, is central to chromosome dynamics, including mitotic chromosome condensation and segregation. Genome-wide localization of condensin is correlated with transcriptional activity; however, the significance of condensin accumulation in transcribed regions remains unclear. Here, we demonstrate that condensin relieves the obstructive effect of mitotic transcription on sister chromatid separation in fission yeast, Schizosaccharomyces pombe. Time-lapse visualization of sister chromatid DNA separation revealed that mutant condensin causes delayed segregation specifically at mitotically transcribed, condensin-bound gene locus, ecm33+. Contrarily, the delay was abolished by transcriptional shut-off of the actively transcribed gene. We also showed that delayed separation at a heat shock-inducible gene locus, ssa1+, in condensin mutants was significantly alleviated by deletion of the gene. Since condensin has ability to remove ssDNA-binding proteins and RNA from unwound ssDNAs or DNA-RNA hybrids in vitro, we propose a model that condensin-mediated removal of mitotic transcripts from chromosomal DNA is the primary mechanism of sister chromatid separation.
Introduction
In proliferative cells, mitotic chromosome condensation and subsequent segregation of replicated chromosomes into daughter cells are essential for faithful inheritance of genomic DNA. ‘Condensin’ has been named as an essential factor for mitotic chromosome condensation, originally demonstrated in mitotic frog egg extracts (Hirano and Mitchison, 1994; Hirano et al., 1997). Today, condensin is known as a key chromosome organizer for many types of chromosome regulation in interphase and mitosis (Hirano, 2016). In fission yeast, Schizosaccharomyces pombe, condensin temperature-sensitive (ts) mutants have been identified by detecting the failure of mitotic chromosome condensation in the same period of condensin discovery in the frog extracts(Hirano and Mitchison, 1994; Saka et al., 1994). Two condensin ts mutants, cut3-477 and cut14-208, were originally isolated, and gene cloning identified two SMC (structural maintenance of chromosomes) family proteins, SMC4/Cut3 and SMC2/Cut14, as the responsible gene products (Hirano, 2006; Saka et al., 1994; Uhlmann, 2016). SMC subunits consist of two globular head domains at the N- and C-termini, which are connected by coiled-coil regions, interrupted by a central hinge region. The condensin Cut3-Cut14 dimer binds to three non-SMC condensin subunits (Cnd2/Barren and HEAT repeat-containing Cnd1 and Cnd3) forming the heteropentameric holo-complex (Sutani et al., 1999; Yoshimura et al., 2002).
Mapping of condensin binding sites at the whole chromosome level proposed a tight correlation between condensin association and transcription. Besides condensin enrichment at centromeres and RNA polymerase I (RNAP I)-transcribed ribosomal DNA (rDNA) repeats (Nakazawa et al., 2008; Wang et al., 2005), genomewide ChIP (Chromatin immunoprecipitation) analyses revealed that condensin associates with RNA polymerase III (RNAP III)-transcribed genes, such as tRNA genes, as well as retrotransposons in budding and fission yeasts (D’Ambrosio et al., 2008; Iwasaki et al., 2010; Tanaka et al., 2012). Fission yeast condensin also accumulates at genes highly transcribed by RNA polymerase II (RNAP II) in interphase and mitosis (Nakazawa et al., 2015; Sutani et al., 2015; Toselli-Mollereau et al., 2016). TATA box-binding protein, TBP, and mitotic transcription factors reportedly recruit condensin to RNAP II-transcribed, transcriptionally activated genes (Iwasaki et al., 2015; Kim et al., 2016). Thus, condensin binding at chromosomal DNA accompanies active transcription by RNAPs I, II, and III. Preferential binding to active genes has also been proposed for higher eukaryotic condensin I and II (Dowen et al., 2013; Kim et al., 2013; Kranz et al., 2013; Sutani et al., 2015), and even for prokaryotic condensin (Gruber and Errington, 2009), indicating that transcription-driven condensin accumulation is critical to understand the molecular action of condensin at mitotic chromosomes. However, physiological significance of condensin binding at actively transcribed DNA regions is still enigmatic.
In this study, we performed detailed, time-lapse, live cell analysis of sister chromatid DNA separation during fission yeast mitosis, and demonstrated that condensin ensures sister chromatid separation at mitotically upregulated gene and heat shock-inducible gene loci. Our results provide insight into how mitotic transcription is reconciled with chromosome segregation.
Results and Discussion
Sister chromatid DNA separation delays at mitotically transcribed gene locus in condensin mutant
We have previously shown that S. pombe condensin preferentially accumulates at mitotically transcribed genes and heat shock-inducible genes during mitosis (Nakazawa et al., 2015). To understand how condensin contributes to segregate transcriptionally activated regions in mitosis, we examined separation timing of the locus that binds to condensin. To visualise the locus, 7.8kb lacO arrays were chromosomally integrated within 2 kb of the mitotically up-regulated ecm33+ gene locus (Fig. 1A, left) in wild-type and temperature-sensitive (ts) cut14-208 mutant cells. LacI-GFP was expressed in these cells, and lacO arrays that bound to LacI-GFP were visualized (Nabeshima et al., 1998). lacO arrays were also integrated along the long arm of chromosome 1, within 2 kb of four other genes (sod2+, SPAC4G8.03c, sck1+, and nsk1+), which were not transcriptionally activated during mitosis. The level of bound condensin was negligible at the four genes (Nakazawa et al., 2015). To visualise the spindle pole body (SPB; equivalent to a centrosome), we used Sad1-mCherry, which is located at the SPBs (Fig. 1A, right) (Hagan and Yanagida, 1995). To monitor the time-course of lacO segregation, cultures grown at 26°C were shifted to 36°C (the restrictive temperature for the cut14-208 mutation in the coiled-coil region, S861P) (Sutani and Yanagida, 1997), and >40 mitotic cells were analyzed by taking movies for each lacO/LacI-GFP locus in individual wild-type and cut14-208 cells.
Sister GFP loci (green dots) were separated smoothly in wild-type cells, when the SPB distance (between magenta SPB dots) reached 3-4 µm (Fig. 1B, arrows). Sister loci were considered separated when GFP signals were fully split (> 0.5 µm). SPB distances were then determined at the time point in which the two GFP-dots separated. Measurements from a number of movies (~50 cells at each locus, total 635 movies, statistical data is shown in Supplemental Table 1) revealed an inverse relationship between the distance of the loci from the telomere and the mean SPB distance for GFP signal splitting (median bars in Fig. 1C). The telomere-proximal sod2+ locus took longer to split than the centromere-proximal nsk1+ locus.
Representative time-lapse images for cut14-208 condensin mutant cells are shown in Fig. 1D. The SPB distance at the time of sister-locus separation for sod2+, SPAC4G8.03c, sck1+, and nsk1+ was slightly longer (4-5 µm) than in WT (3-4 µm). On the other hand, the SPB distance for splitting of the mitotically transcribed ecm33+ locus was strikingly longer. Splitting did not occur until the SPB distance reached a mean of 6.5 µm in mutant cells. Thus, the mitotically highly transcribed ecm33+ gene locus showed an anomalous segregation delay in condensin mutant cells, relative to the 4 control loci (Fig. 1E).
Transcriptional shut-off abolishes the delay of sister chromatid DNA separation at actively transcribed genes in condensin mutant
To examine whether transcription at ecm33+ affects separation timing at this locus, a promoter-less mutant, Δpro ecm33, was constructed. ecm33+ expression was greatly diminished (~20%) in Δpro ecm33 compared to that in WT cells, as measured by reverse transcription-PCR (Supplemental. Fig. S1). In Δpro ecm33 cells, the delay of separation at the ecm33+ locus was essentially abolished (the mean SPB distance became ~5µm; Fig. 1E), suggesting an opposing effect of mitotic transcription on sister chromatid DNA separation.
We have also visualized an hsp gene locus, ssa1+, by integrating lacO arrays 6 kb upstream of the gene (Fig. 2A). In WT cells, separation timing of the ssa1+-proximal lacO DNAs was not significantly changed in the deletion mutant of ssa1+ (Δssa1) after heat shock at 36°C (Fig. 2B, Supplemental Table 2). Contrarily, separation timing at that locus became significantly earlier by deleting the ssa1+ gene in cut14-208 (Fig. 2C). We thus concluded that in the condensin mutant, transcription per se caused the delay in sister chromatid DNA splitting at both mitotically-activated and heat-shock responsible gene loci during mitosis.
Condensin-mediated elimination of RNA transcripts may facilitate mitotic chromosome segregation
The results indicate that condensin negates the inhibitory effect of mitotic transcription on chromosome segregation at actively transcribed gene loci. If we consider in vitro activity of the condensin SMC heterodimer (Cut3-Cut14), it promotes the DNA renaturation reaction which winds up of complementary ssDNA to dsDNA (Sutani and Yanagida, 1997). In addition, a ssDNA-binding protein, Ssb1/RPA, bound to ssDNA, and also RNA molecules hybridized to ssDNA were removed by condensin SMC through the DNA renaturation reaction (Akai et al., 2011). This activity is strikingly reduced in condensin ts mutants, cut14-208 (coiled-coil mutation, S861P) or cut14-y1 (hinge mutation, L543S), implying that RNA transcripts bound at actively transcribed regions interfere with chromosome segregation. Transcriptional attenuation using transcription inhibitors or mutation in the transcription mediator protein suppresses chromosome segregation defects in condensin mutants (Supplemental Fig. S2) (Sutani et al., 2015). Taken together, we now propose that condensin-mediated elimination of RNA transcripts on transcribed DNA regions facilitates mitotic chromosome segregation. Actually, many genes are actively transcribed even during mitosis. Our results provide a novel insight into mitotic chromosome segregation, which may be accomplished by removing obstructive RNA and/or proteins on chromosomes, in addition to forming condensed mitotic chromosomes. Further analysis is definitively required to understand the inhibitory effect of bound RNA on chromosome segregation during mitosis.
Previously, we found that S. pombe condensin enriches at central centromeric DNAs where kinetochores are assembled with specialized nucleosomes containing a centromere-specific histone H3 variant, CENP-A/Cnp1 (Allshire and Ekwall, 2015; Nakazawa et al., 2008; Takahashi et al., 2000). In addition to the pericentromeric outer repeat regions (otr), the central centromere also generates a non-coding RNA transcribed by RNAP II (Chan and Wong, 2012; Choi et al., 2011; Talbert and Henikoff, 2018). Condensin enrichment at the central centromere appears to be dependent on transcription (Supplemental Fig. S3), implying that condensin acts on centromeric transcripts. In metaphase, replicated sister centromeric DNAs are pre-separated for establishment of bi-oriented kinetochore-spindle attachment (Nabeshima et al., 1998); however, the pre-separation fails in condensin mutant cells, resulting in unequal segregation of centromere-proximal sister DNAs (Nakazawa et al., 2008). Condensin may play a crucial role in separating centromeres by removing non-coding RNA transcripts, as well as in actively-transcribed chromosomal arm regions.
Materials and Methods
Strains, plasmids, and media
The S. pombe haploid wild-type strain 972 h− and its derivative mutant strains, including the temperature-sensitive (ts) cut14-208, cut3-477 and cnd2-1 were used (Aono et al., 2002; Saka et al., 1994). Disruption of the ecm33+ promoter (Takada et al., 2010) was carried out using a two-step PCR-based gene targeting method (Bähler et al., 1998). The ecm33+ gene was cloned into a pSK248 plasmid under the native promoter and expressed in an ecm33+ promoter-less mutant to alleviate the cell morphology defect (Takada et al., 2010).
Labelling of chromosomal loci at ecm33+, sod2+, SPAC4G8.03c, sck1+, nsk1+, and ssa1+ using the lac repressor (LacI)/ lac operator (lacO) system was performed as previously described (Nabeshima et al., 1998), with the following modifications. The pMK19A-KanR plasmid derived from pMK2A was used for lacO insertion at each locus. pMK19A-KanR contains the lacO repeat (7.8 kb) and the kanamycin-resistance gene KanR+ (1.7 kb fragment from pFA6a-KanMX) (Bähler et al., 1998). Target sequences (1.5-2.0 kb) were then cloned into pMK19A-KanR. Resulting plasmids were linearized by cleavage within the target sequences and transformed into a strain expressing the GFP-LacI-NLS protein.
Thiolutin (Sigma T3450) and 1,10-phenanthroline (Sigma 320056) were diluted with dimethyl sulphoxide (1 mg/mL) and ethanol (30 mg/mL), respectively, and used to inhibit RNA polymerases I, II, and III (Lackner et al., 2007). Culture media used for S. pombe were YPD and SPA sporulation medium (Saka et al., 1994). Cells were counted using a haematology analyzer (Sysmex FDA-500).
Live cell analysis
S. pombe cells were cultured at 26°C in EMM2 medium and shifted to 36°C for 1.5 hr. Before observation, exponentially growing cells were transferred to a glass-bottomed dish (IWAKI Glass) coated with 10 mg/mL concanavalin A (Wako). Time-lapse images were recorded with a DeltaVision microscope system (Applied Precision, GE Healthcare UK Ltd). The objective lens was an oil immersion lens (PlanApo 60x, NA 1.4; Olympus). For observations of lacO/LacI-GFP and Sad1-mCherry signals, a set of images from 7 focal planes at 0.2-µm intervals was obtained every 0.5 min. Image projection and deconvolution were performed using an imaging workstation (SoftWoRx; Applied Precision). More than 40 mitotic cells (for Figure. 1) or 25 (for Figure. 2) were observed for each locus. Image J (NIH) was used to measure the SPB distance when lacO/LacI-GFP signals were separated. We defined the separation timing of sister lacO/LacI-GFP signals when the distance between proximal edges of the GFP signals was greater than 0.5 µm.
RNA extraction, reverse transcription PCR, and chromatin immunoprecipitation
Total RNA from S. pombe cells was extracted using the hot-phenol method. One µg of total RNA was reverse-transcribed using PrimeScript RT reagent (TaKaRa) with oligo dT primers. A genomic DNA eraser supplied with reverse transcription reagent was used to remove contaminating genomic DNA in the RNA sample. cDNA solution was diluted 25-fold with RNase-free water (Ambion) and 5 µL were used for PCR. Results were quantified using real-time PCR with SYBR premix Ex Taq II solution (TaKaRa). Chromatin immunoprecipitation (ChIP) was performed as previously described (Nakazawa et al., 2015).
Acknowledgments
We thank Dr. Steven D. Aird for editing the manuscript. NN was supported by a Grant-in-Aid for Young Scientists (B) 25840013 from the Japan Society for the Promotion of Science (JSPS). We are also grateful for the generous support of OIST.