DNA loop extrusion by condensins and decatenation by DNA topoisomerase II (topo II) are thought to drive mitotic chromosome compaction and individualization. Here, we reveal that the linker histone H1.8 antagonizes condensins and topo II to shape mitotic chromosome organization. In vitro chromatin reconstitution experiments demonstrate that H1.8 inhibits binding of condensins and topo II to nucleosome arrays. Accordingly, H1.8 depletion in Xenopus egg extracts increased condensins and topo II levels on mitotic chromatin. Chromosome morphology and Hi-C analyses suggest that H1.8 depletion makes chromosomes thinner and longer through shortening the average loop size and reducing the DNA amount in each layer of mitotic loops. Furthermore, excess loading of condensins and topo II to chromosomes by H1.8 depletion causes hyper-chromosome individualization and dispersion. We propose that condensins and topo II are essential for chromosome individualization, but their functions are tuned by the linker histone to keep chromosomes together until anaphase.
chromosome compactionmitosislinker histonenucleosomeHi-CchromatinResearch organismXenopushttp://dx.doi.org/10.13039/100000002National Institutes of HealthR35 GM132111FunabikiHironorihttp://dx.doi.org/10.13039/100000002National Institutes of HealthR01 HG003143DekkerJobhttp://dx.doi.org/10.13039/501100000289Cancer Research UKCR-UK C47547/A21536VanniniAlessandrohttp://dx.doi.org/10.13039/100004440Wellcome Trust200818/Z/16/ZVanniniAlessandrohttp://dx.doi.org/10.13039/100000011Howard Hughes Medical InstituteInvestigator ProgramDekkerJobThe funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.Author impact statementLinker histone H1.8 shapes mitotic chromosomes by tuning the number and size of condensin-dependent DNA loops and suppressing condensin and DNA topoisomerase II-dependent individualization.Introduction
Genomic DNA in eukaryotes is compacted by orders of magnitude over its linear length. The extent and mode of the packaging change between interphase and mitosis to support the cell cycle-dependent functions of the DNA. While loosely packed DNA allows efficient decoding of genetic information in interphase, mitotic compaction of DNA enables efficient distribution of genetic information to daughter cells. In addition, chromosome individualization during mitosis ensures that all duplicated chromosomes are independently moved by microtubules and are equally distributed to daughter cells. Despite these common functional requirements of mitotic chromosomes and evolutionary conservation of major known regulators of mitotic chromosome structures, the size and shape of chromosomes vary among species and developmental stages. For example, rod-like individual mitotic chromosomes can be readily visualized in fission yeast Schizosaccharomyces pombe, but not in budding yeast Saccharomyces cerevisiae (Guacci et al., 1994; Umesono et al., 1983). During early embryogenesis in Xenopus and Caenorhabditis elegans, mitotic chromosome lengths become shorter (Ladouceur et al., 2015; Micheli et al., 1993). The mechanistic basis of mitotic chromosome shape regulation remains largely speculative (Heald and Gibeaux, 2018).
Classical experiments on mitotic chromosomes indicated that DNA is organized into loops around a central protein scaffold (Earnshaw and Laemmli, 1983; Paulson and Laemmli, 1977). Two major chromosome-associated ATPases play pivotal roles in mitotic chromatid formation: DNA topoisomerase II (topo II) and the structural maintenance of chromosomes (SMC) family complex, condensin (Kinoshita and Hirano, 2017; Kschonsak and Haering, 2015). Data obtained with chromosome conformation capture assays are consistent with the model that mitotic chromosomes are arranged in a series of loops organized by condensins (Naumova et al., 2013; Gibcus et al., 2018; Elbatsh et al., 2019). Vertebrate cells express two forms of the condensin complex, condensin I and condensin II (Ono et al., 2003; Hirano et al., 1997). Condensin I is loaded onto chromatin exclusively during mitosis, whereas condensin II retains access to chromosomes throughout the cell cycle (Hirota et al., 2004; Ono et al., 2004; Walther et al., 2018). Condensin II plays a role in maintaining chromosome territories in Drosophila nuclei (Rosin et al., 2018; Bauer et al., 2012) and drives sister chromatid decatenation by topo II (Nagasaka et al., 2016). It has been proposed that condensin II acts first in prophase to anchor large outer DNA loops, which are further branched into shorter inner DNA loops by condensin I (Gibcus et al., 2018). This proposal is consistent with their localization as determined by super-resolution microscopy (Walther et al., 2018). In chicken DT40 cells, condensin II drives the helical positioning of loops around a centrally located axis, thus controlling the organization of long distance interactions (6–20 Mb), whereas condensin I appears to control shorter distance interactions (Gibcus et al., 2018). This organization of the condensin I and II loops is also consistent with their roles in maintaining lateral and axial compaction, respectively (Green et al., 2012; Samejima et al., 2012; Bakhrebah et al., 2015). In Xenopus egg extracts, in the presence of wildtype condensin I levels, condensin II depletion does not appear to change mitotic chromosome length, suggesting a reduced role for condensin II on these chromosomes (Shintomi and Hirano, 2011).
The prevailing model suggests that mitotic chromatin loops are formed by the dynamic loop extrusion activity of condensins (Riggs, 1990; Nasmyth, 2001; Alipour and Marko, 2012), although the molecular details of the process remain unclear (Banigan and Mirny, 2020; Cutts and Vannini, 2020; Datta et al., 2020). Single-molecule experiments using purified recombinant yeast and human condensin complexes demonstrated ATP-dependent motor activity and loop extrusion by yeast and human condensins (Terakawa et al., 2017; Ganji et al., 2018; Kong et al., 2020). Condensin-dependent loop extrusion in a more physiological Xenopus extract system has also been shown (Golfier et al., 2020). In silico experiments further suggest that a minimal combination of loop extruders (like condensin) and strand passage activity (such as topo II) can generate well-resolved rod-like sister chromatids from entangled, interphase-like DNA fibers (Goloborodko et al., 2016a). However, it remains unclear if loop extrusion can proceed on chromatin since condensins prefer to bind nucleosome-free DNA (Kong et al., 2020; Zierhut et al., 2014; Shintomi et al., 2017; Toselli-Mollereau et al., 2016; Piazza et al., 2014). Human and yeast condensin complexes are capable of loop extrusion through sparsely arranged nucleosomes in vitro (Kong et al., 2020; Pradhan et al., 2021), but mitotic chromatin adopts a more compact fiber structure (Grigoryev et al., 2016; Ou et al., 2017). Furthermore, large protein complexes such as RNA polymerases are able to limit loop extrusion by SMC protein complexes, such as bacterial condensins and eukaryotic cohesins (Brandão et al., 2019; Hsieh et al., 2020; Krietenstein et al., 2020). Nucleosomes also restrict the diffusion of cohesin in vitro (Stigler et al., 2016). Therefore, the effect of higher-order chromatin fiber structure and other mitotic chromatin proteins on processive loop extrusion by condensin remains unknown.
Interphase nuclei are segmented to chromosome territories, each of which contain highly entangled sister chromatids after replication (Cremer and Cremer, 2010; Sundin and Varshavsky, 1981; Farcas et al., 2011). Replicated pairs of chromatids are linked by cohesin during interphase, and cohesin removal during prophase (except at centromeres and other limited protected loci) promotes resolution of sister chromatids, together with actions of condensin II and topo II (Nagasaka et al., 2016; Gandhi et al., 2006). Different chromosomes are also largely unentangled in interphase HeLa cells (Goundaroulis et al., 2020; Tavares-Cadete et al., 2020), and Ki-67 localization on chromosome peripheries may act as a steric and electrostatic barrier to prevent interchromosomal entanglement during mitosis (Cuylen et al., 2016). However, some interchromosomal linkages were observed in metaphase (Potapova et al., 2019; Marko, 2008). Although it has been shown that active transcription at rDNA results in mitotic interchromosomal links (Potapova et al., 2019), it remains unclear if mitotic chromosome compaction is coupled to resolution of these interchromosomal linkages.
One of the most abundant chromatin proteins beside core histones is the linker histone, which binds to the dyad of the nucleosome and tethers the two linker DNAs emanating from the nucleosome (Bednar et al., 2017; Arimura et al., 2020; Zhou et al., 2015; Zhou et al., 2021). In reconstitution experiments, linker histones cluster oligo-nucleosomes (White et al., 2016; Li et al., 2016) and promote liquid-liquid phase separation (Gibson et al., 2019; Shakya et al., 2020). In vivo, linker histones are also enriched in highly compact chromatin (Izzo et al., 2013; Th’ng et al., 2005; Parseghian et al., 2001). While core histones are evolutionarily highly conserved in eukaryotes, linker histones are much more diversified (Izzo et al., 2008). In the human and mouse genome, 11 H1 paralogs are found, among which some combination of six variants (H1.0–H1.5) is widely expressed in somatic cells (Hergeth and Schneider, 2015). In vertebrate oocytes and early embryos, H1.8 (also known as H1OO, H1foo, H1M, and B4) is the major linker histone variant (Dworkin-Rastl et al., 1994; Wühr et al., 2014). Immunodepletion of H1.8 from Xenopus egg extracts made mitotic chromosomes thinner and longer, causing defective chromosome segregation in anaphase (Maresca et al., 2005). However, the mechanism by which the linker histone affects large-scale chromosome length changes remains unknown.
Here we demonstrate that the linker histone H1.8 suppresses enrichment of condensins and topo II on mitotic chromosomes. In a reconstitution system with purified components, H1.8 inhibits binding of topo II and condensins to nucleosome arrays. Through a combination of chromosome morphological analysis and Hi-C, we show that H1.8 reduces chromosome length by limiting condensin I loading on chromosomes, while H1.8 limits chromosome individualization by antagonizing both condensins and topo II. This study establishes a mechanism by which the linker histone tunes the compaction and topology of mitotic chromosomes.
ResultsLinker histone H1.8 limits enrichment of condensins and topo II on mitotic chromatin
Depletion of linker histone H1.8 in Xenopus egg extracts makes chromosomes thinner and elongated (see Figure 3C, Maresca et al., 2005). We asked if this phenotype may reflect the potential role of H1.8 in regulating condensins and TOP2A (the dominant topo II isoform in Xenopus egg extracts, Wühr et al., 2014), which are essential for mitotic chromosome compaction in Xenopus egg extracts (Hirano and Mitchison, 1994; Adachi et al., 1991; Cuvier and Hirano, 2003). In silico simulation analysis suggests that increasing the number of loop extruders (such as condensin I) on DNA, beyond a minimum threshold, makes chromosomes longer and thinner (Goloborodko et al., 2016a). Reducing condensin I levels on chromatin also made chromosomes shorter experimentally (Shintomi and Hirano, 2011; Elbatsh et al., 2019; Fitz-James et al., 2020). Although it has been reported that H1.8 depletion does not affect chromosomal enrichment of major chromatin proteins (Maresca et al., 2005), we therefore attempted to quantify chromatin-bound levels of condensins and TOP2A.
To investigate whether H1.8 regulates chromatin levels of condensins and topo II, we prepared mitotic chromosomes in Xenopus laevis egg extracts depleted of H1.8. Demembranated X. laevis sperm nuclei were added to either mock (ΔIgG) or H1.8-depleted (ΔH1) extracts from eggs arrested at meiotic metaphase II by cytostatic factor (CSF extracts; Figure 1A and B). Calcium was added to cycle the extract into interphase and induce functional nuclear formation, in which chromosomes were replicated. The corresponding depleted CSF extract was then added to generate metaphase chromosomes (Shamu and Murray, 1992). To eliminate the microtubule-dependent change in chromosome morphology, which may affect quantitative analyses of chromatin proteins, spindle assembly was inhibited using nocodazole. Chromosomes were fixed and the levels of TOP2A and condensin I subunit CAP-G were measured by immunofluorescence (Figure 1C). Depletion of H1.8 increased the levels of both CAP-G and TOP2A on mitotic chromosomes, while adding back recombinant H1.8 rescued the phenotype (Figure 1D). Identical results were obtained when immunofluorescence signal normalization was done by the minor groove DNA-binding dye Hoechst 33342 or by fluorescent dUTP that was incorporated during replication (Figure 1—figure supplement 1A–C). The dUTP quantitation also confirmed that H1.8 depletion did not affect DNA replication (Figure 1—figure supplement 1B), as reported previously (Dasso et al., 1994). The apparent increased signal intensities of CAP-G and TOP2A on chromatin in ∆H1 extracts were not the general consequence of elongated chromosome morphology as H1.8 depletion did not affect a panel of other chromatin proteins, regardless of their binding preference to nucleosomes or nucleosome-free DNA (Zierhut et al., 2014; Figure 1—figure supplement 1D). Enhanced chromosome binding of condensin I and TOP2A in ∆H1.8 extracts was biochemically confirmed by quantifying their levels on purified metaphase chromosomes by western blotting and by mass spectrometry (Figure 1E, Figure 1—figure supplement 1E). Mass spectrometry data also confirmed that H1.8 depletion did not result in a global enrichment of all chromatin-bound proteins. Condensin II levels on chromosomes also showed similar increases in ΔH1 extracts by immunofluorescence and western blots on purified chromosomes, although peptides of condensin II-specific subunits were not detected by mass spectrometry (Figure 1F–H, Figure 1—figure supplement 1E, Figure 1—source data 2).
Linker histone H1.8 suppresses enrichment of condensins and TOP2A on mitotic chromatin.
(A) Experimental scheme to generate replicated chromosomes in Xenopus egg extracts. (B) Western blots of total extracts showing depletion of H1.8 from Xenopus egg extracts and rescue with recombinant H1.8 (rH1.8). (C) Representative images of DNA (Hoechst 33342), CAP-G (condensin I), and TOP2A immunofluorescence on chromosomes in metaphase extracts treated with nocodazole in the indicated conditions. Chromosomes in each nucleus remain clustered in the presence of nocodazole. Bar, 10 µm. (D) Quantification of CAP-G (condensin I) and TOP2A immunofluorescence signals normalized to the DNA (Hoechst) signal for the indicated conditions. Each gray or magenta dot represents the average signal intensity of a single chromosome cluster (from one nucleus). Each black dot represents the median signal intensity from a single experiment. Bars represent mean and SEM of the medians of three independent experiments. (E) Western blots of mitotic chromatin purified from mock (∆IgG) and H1.8-depleted (∆H1) extracts (top) and quantification of band intensities normalized to H3 and H2B (below). Mean and SEM/range from three/two experiments respectively. (F) Representative images of CAP-G2 (condensin II) immunofluorescence on chromosomes in metaphase extracts with nocodazole in the indicated conditions. Bar, 10 µm. (G) Quantification of the CAP-G2 (condensin II) normalized to the DNA (Hoechst) signal for the indicated conditions. Each gray or purple dot represents the average signal intensity of a single chromosome cluster (from one nucleus). Each black dot represents the median signal intensity from a single experiment. Bars represent mean and range of the median of two independent experiments. (H) Western blots of mitotic chromatin purified from mock (∆IgG) and H1.8-depleted (∆H1) extracts. The p-values shown in (D) and (E) were calculated by an unpaired Student’s t-test of the aggregate medians of three independent experiments, after confirming the statistical significance for each experimental dataset by a two-tailed Mann–Whitney U-test. The p-values shown in (G) were calculated on total data from two independent experiments using a two-tailed Mann–Whitney U-test. The number of nuclei imaged in (D) and (G) in each condition for each experiment is indicated in the figure.
Source data for all the figures in <xref ref-type="fig" rid="fig1">Figure 1</xref> and its figure supplement.
Mass spectrometry data for chromatin purified from ΔIgG and ΔH1 metaphase sperm chromosomes.
H1.8 depletion does not lead to global accumulation of DNA-binding proteins.
(A) Experimental scheme to incorporate Cy3-labeled nucleotides to use normalization of immunofluorescence signals on chromosomes. (B) Quantification of Cy3-dUTP signals normalized to Hoechst 33342 signals, showing uniform normalization across two coverslips used for quantification of condensin I and TOP2A. The result also indicates no detectable defect in DNA replication in ∆H1 extracts. (C) CAP-G (condensin I) and TOP2A immunofluorescence signal levels on chromosomes normalized with Hoechst and incorporated Cy3-dUTP. Normalization using Hoechst 33342 and Cy3-dUTP/Cy5-dUTP signals was shown to be consistent across many experiments. (D) Quantification of immunofluorescence signal levels of two proteins that prefer nucleosomes (RCC1 and Dasra A) and two that prefer to bind nucleosome-free DNA (Dppa2 and Xkid) (Zierhut et al., 2014). In (B–D), distribution of signal intensity per chromosome cluster (dots), and median and SEM from one experiment are shown. This analysis was performed twice, and the similar results were obtained. (E) Protein abundance on metaphase sperm chromosomes purified from mock and H1.8-depleted extracts. Subunits from the structural maintenance of chromosomes (SMC) family complexes are labeled in pink. Core histones and their variants are in blue and linker histone variants are shown in red. The lines for 0.25×, 0.5×, 1, 2×, and 4× abundance on ΔH1 compared to ΔIgG chromosomes are shown. The mass spectrometry analysis was performed once. The number of nuclei imaged in each condition for each experiment in (B–D) is indicated above the figure.
Linker histone H1.8 reduces binding of condensins and TOP2A to nucleosome arrays
Since linker histone depletion does not change the nucleosome spacing in Xenopus egg extracts (Ohsumi et al., 1993), we hypothesized that H1.8 depletion results in an increase in linker DNA that becomes accessible to condensins and TOP2A. To test this possibility, we reconstituted nucleosome arrays with purified histones and asked if H1.8 interferes with binding of recombinant human condensins and X. laevis TOP2A (Kong et al., 2020; Ryu et al., 2010; Figure 2A). The purity and the intact nature of the human condensin complexes was confirmed using mass photometry (Figure 2—figure supplement 1; Verschueren, 1985; Young et al., 2018; Sonn-Segev et al., 2020). As previously shown with human condensin I purified from HeLa cells (Kimura et al., 2001), our recombinant human condensin I, but not the ATP-binding defective Q-loop mutant (Hassler et al., 2019; Hopfner et al., 2000; Löwe et al., 2001; Kong et al., 2020), was able to rescue mitotic chromosome morphology defects caused by condensin I depletion in Xenopus extracts (Figure 2—figure supplement 2A), demonstrating that the recombinant human condensin I can also functionally replace Xenopus condensin I. The recombinant X. laevis TOP2A used was also able to rescue chromatid formation in ΔTOP2A extracts and was able to perform ATP-dependent kinetoplast decatenation in vitro (Figure 2—figure supplement 2B and C). The nucleosome array was composed of 19 tandem repeats of 147 bp Widom 601 nucleosome positioning sequence and 53 bp linker DNA (Lowary and Widom, 1998), where full occupancy of the array by a nucleosome core particle (NCP) and H1.8 to each repeat unit was confirmed by native polyacrylamide gel electrophoresis (PAGE) (Figure 2—figure supplement 3A).
Linker histone inhibits binding of condensins and TOP2A to nucleosome arrays.
(A) Experimental scheme for testing the effect of recombinant H1.8 (rH1.8) on binding of purified condensins and TOP2A to arrays of nucleosomes assembled on the Widom 601 nucleosome positioning sequence. (B) Coomassie staining of SDS-PAGE gels, showing input (top) and nucleosome array-bound fraction (middle) of condensin I, rH1.8, and core histones. The rightmost lanes represent the streptavidin beads-only negative control. Buffer contains 2.5 mM MgCl2, 5 mM ATP, and indicated concentrations of NaCl. The band intensities of condensin I subunits were normalized to the histone bands and the binding at 50 mM NaCl for nucleosome arrays without H1.8. Mean and SEM of three independent experiments are shown (bottom). (C) Same as (B), except that nucleosome array binding of condensin II is shown. Mean and SEM (wildtype)/range (Q-loop mutant) of three (wildtype) or two (Q-loop mutant) independent experiments are shown. (D) Same as (B), except that nucleosome array binding of TOP2A in buffer containing 1 mM MgCl2 is shown. Mean and range of two independent experiments are shown.
Source data for all the figures in <xref ref-type="fig" rid="fig2">Figure 2</xref> and its figure supplements.
Mass photometry of condensin complexes.
Recombinant condensin complexes were diluted to 50 nM in buffer (10 mM HEPES pH 8, 2.5 mM MgCl2, 1 mM DTT, 5 mM ATP) supplemented with the indicated sodium chloride concentration and the indicated temperatures for 30 min subjected to mass photometry to determine whether the complexes remain intact in the buffer conditions used for the binding assays. Histograms of the particle count at the indicated molecular mass. The molecular mass was calibrated using bovine serum albumin and urease. The black lines are the Gaussian fits to the peaks. The peak of the Gaussian fit and the percentage of particles in the peak are indicated above each peak. Theoretical molecular weights of condensin I and condensin II complexes are 638 kDa and 658 kDa, respectively.
Recombinant human condensin I and <italic>X. laevis</italic> TOP2A are functional<italic>.</italic>
(A) Representative Hoechst (DNA) and CENP-A immunofluorescence images of chromosomes in indicated metaphase egg extracts after dilution, which disperses individualized chromosomes (left). Extracts depleted of condensin I (ΔCAP-G) were complemented with recombinant human condensin I and condensin I Q-loop mutant protein complexes. Total egg extract western blots of depletion and the rescue conditions (right). (B) Recombinant X. laevis TOP2A is functional in Xenopus egg extracts. Sperm nuclei were added to undepleted, TOP2A-depelted CSF extracts (∆TOP2A), or ∆TOP2A extracts with purified recombinant TOP2A. Representative Hoechst (DNA) images of sperm are shown. In undepleted or ∆TOP2A extracts with supplemented TOP2A, proper sperm remodeling led to mitotic chromosomes formation. In ∆TOP2A extracts, sperm remodeling failed and sperm nuclei remained compact crescent-like shape. (C) Recombinant X. laevis TOP2A possesses decatenating activity in vitro. Agarose gel of kinetoplast decatenation assay. Recombinant TOP2A promoted ATP-dependent decatenation of kinetoplast DNA. Bar, 20 µm.
Condensin binding is inhibited by magnesium.
(A) Native PAGE gel analysis of nucleosome array beads loaded with or without H1.8 after digestion of the array with AvaI, which released monomers of nucleosome positioning sequence. A complete shift of monomer bands by H1.8 addition indicates the saturated occupancy of nucleosome and H1.8. (B) Purified recombinant human condensin I was incubated with nucleosome array beads in buffer containing 2.5 mM MgCl2 with and without 5 mM ATP. Coomassie staining of input and bead fractions is shown. (C) Purified human condensin I was incubated with nucleosome array beads in buffer containing 1 or 2.5 mM MgCl2 with and without 1 mM ATP. Coomassie staining of input and bead fractions is shown. (D) Purified human condensin I Q-loop mutant was incubated with nucleosome array beads in buffer containing 2.5 mM MgCl2 with the addition of 5 mM ATP or 5 mM EDTA. Coomassie staining of input and bead fractions is shown. (E) Alexa647-labeled 196 bp mononucleosomes were incubated with indicated concentrations of condensin I in the presence of 2.5 mM MgCl2 and with or without 5 mM ATP and electrophoresed on a 5% native PAGE (above). Bands at the well represent the nucleosome-condensin complex. Absence of signals at the well in the absence of condensin I indicates that mononucleosomes do not form large aggregates in the tested experimental conditions. Quantification of the condensin-bound fraction of the nucleosomes showing no increase in binding affinity due to ATP (Figure 2—source data 1).
H1.8 inhibits condensin binding to mononucleosomes.
Alexa647-labeled 196 bp mononucleosomes with or without H1.8 were incubated with indicated concentrations of condensin I and electrophoresed on a 5% native PAGE. Alexa647-labeled DNAs are shown. Condensin I binding curves with mononucleosomes with and without H1.8, showing the large increase in binding constant in the presence of H1.8. The data plotted is the mean and SEM of three independent experiments The binding constants derived from fitting the aggregate data to a sigmoidal binding curve are shown (Figure 2—source data 1).
Condensin I binds weakly to mononucleosomes with linker DNA (Kong et al., 2020). We observed similar weak binding of condensin I to nucleosome arrays, and this binding was stimulated by ATP (Figure 2—figure supplement 3B). Since this enhancement was not seen when magnesium concentration was higher than the ATP concentration (Figure 2—figure supplement 3C), we hypothesized that ATP might have enhanced condensin binding to the nucleosome arrays by chelating magnesium, which is known to induce chromatin compaction (Finch and Klug, 1976; Eltsov et al., 2008). Indeed, both EDTA and ATP, which chelate magnesium, increased binding of the Q-loop mutant of condensin I to the nucleosome array (Figure 2—figure supplement 3D), suggesting that condensin binding to the nucleosome array is sensitive to high magnesium concentration and not due to ATP binding of condensin I. In contrast, ATP did not stimulate condensin I binding to mononucleosomes (Figure 2—figure supplement 3E), indicating that excess magnesium may limit condensin binding through compaction of the nucleosome array.
In the buffer condition where excess ATP was present over magnesium, preloading of H1.8 to the nucleosome array reduced binding of both wild-type and the ATP-binding-deficient Q-loop mutant of condensin I at physiological salt concentrations (50–150 mM NaCl) (Figure 2B). H1.8 also suppressed binding of condensin II and its Q-loop mutant to the nucleosome array (Figure 2C), though, as expected (Kong et al., 2020), condensin II showed higher affinity to the nucleosome array than condensin I. We noticed that the subunit stoichiometry of condensin I and condensin II was altered in the bead fraction from that in the input. Since the condensin complexes were intact at the assay conditions (Figure 2—figure supplement 1), we suspect that non-SMC subunits of condensins were less stable on the nucleosome array than SMC subunits, while H1.8 reduced binding of all condensin subunits to the nucleosome array. The reduced binding of condensins in the presence of H1.8 can be explained by either direct competition between H1.8 and condensins for the linker DNA or by H1.8-mediated formation of a higher-order structure of the nucleosome array (Song et al., 2014; Rudnizky et al., 2021). The direct competition model can be tested by using mononucleosomes with linker DNAs instead of the nucleosome array since H1.8 binding to mononucleosomes does not promote higher-order structure or aggregation (White et al., 2016). Supporting the direct competition model, H1.8 reduced binding of condensin I to mononucleosomes (Figure 2—figure supplement 4).
We also examined if H1.8 interferes with binding of the recombinant Xenopus TOP2A to nucleosome arrays. As compared to condensins, TOP2A showed more stable binding to the nucleosome array, and H1.8 had no effect on TOP2A binding at low salt concentrations. However, H1.8 did reduce nucleosome array binding of recombinant Xenopus TOP2A at 120 mM NaCl (Figure 2D). Altogether, these data demonstrate that preloaded H1.8 on nucleosomes can directly interfere with binding of condensins and TOP2A to chromatin.
Chromosome elongation by H1.8 depletion is due to increased chromatin-bound condensin I
If the increased amount of condensin I on chromatin is responsible for the chromosome elongation phenotype observed in ∆H1 extracts, reducing condensin I levels should reverse this phenotype. To measure lengths of mitotic chromosomes formed in Xenopus egg extracts, we assembled replicated metaphase chromosomes in extracts depleted with mock IgG, H1.8, CAP-G (condensin I), or CAP-D3 (condensin II) antibodies. These extracts were then diluted to disperse individualized chromosomes (Figure 3A and B; Funabiki and Murray, 2000). As reported previously (Maresca et al., 2005), average chromosome length increased by ~50% upon H1.8 depletion (Figure 3C and D). Supporting our hypothesis, when condensin I was co-depleted (ΔH1ΔCAP-G), chromosomes became even shorter than chromosomes in mock-depleted extracts (ΔIgG) (Figure 3C and D). In contrast, condensin II co-depletion (ΔH1ΔCAP-D3) did not change the chromosome length (Figure 3D). Chromosome length in condensin II-depleted extracts (ΔCAP-D3) was also indistinguishable from mock-depleted chromosomes (Figure 3—figure supplement 1A), consistent with the negligible effect of condensin II depletion in the presence of normal levels of condensin I in Xenopus egg extracts (Shintomi and Hirano, 2011).
Chromosome elongation by H1.8 depletion is due to enhanced condensin I loading on chromatin.
(A) Schematic of extract dilution to disperse individualized chromosomes. (B) Western blots of total egg extracts showing depletions of indicated proteins. (C) Representative images of mitotic chromosomes after dilution of indicated extracts. Bar, 10 μm. (D) Quantification of the chromosome length. Data distribution of the length of individual chromosomes from three independent experiments (green, purple, gray) is shown. Each black dot represents the median length of chromosomes from a single experiment. Bar represents mean and SEM of three independent experiments. (E) Quantification of CAP-G levels normalized to DNA signal (Cy3-dUTP) in the indicated conditions by immunofluorescence. Each dot represents the mean of CAP-G intensity normalized to DNA intensity of a single chromosome cluster (from one nucleus). The data plotted is median ± 95% CI. (F) Chromosome lengths in the indicated condition. Each dot represents length of a single chromosome. Bars represent median ± 95% CI. The p-values in (D) compare the median chromosome lengths in each condition and were calculated using an unpaired Student’s t-test, and those in (E, F) compare the median values in a single experiment and were calculated using a two-tailed Mann–Whitney U-test. The number of nuclei (E) or chromosomes (D, F) imaged in each condition for each experiment is indicated above the figure.
Source data for all the figures in <xref ref-type="fig" rid="fig3">Figure 3</xref> and its figure supplements.
Condensin I loading determines chromosome length.
(A) Length of mitotic chromosomes in the indicated conditions after extract dilution showing no effect of CAP-D3 (condensin II) depletion in both mock and H1.8 depletion background. Each dot represents length of a single chromosome. Bars represent median ± 95% CI. The length of >50 chromosomes was measured in each condition for every experiment. (B) Total egg extract blots showing the partial depletions of condensin in the indicated conditions. The p-value in (A) was calculated by a two-tailed Mann–Whitney U-test. The number of chromosomes imaged in each condition for each experiment in (A) is indicated above the figure (Figure 3—source data 1). (C, D) Biological replicate of the partial condensin I depletion experiment in Figure 3E and F.
TOP2A overloading is not responsible for chromosome elongation.
(A) Sperm nuclei were replicated in undepleted interphase extracts supplemented with 25 nM Cy3-dUTP. The nuclei were then cycled back by mixing with different ratios of ΔTOP2A extracts. Immunofluorescence data showing chromatin-bound TOP2A normalized to Cy3-dUTP signal in each condition. Each dot represents the normalized mean intensity of a single nucleus. (B) Chromosome lengths in the indicated conditions. Each dot represents the length of a single chromosome. (C) Sperm nuclei were replicated in either mock (ΔIgG) and H1.8-depleted (ΔH1) extracts and cycled back using 3 volumes of the corresponding extracts (ΔIgG/ΔH1) or 3 volumes of ΔIgGΔTOP2A/ΔH1ΔTOP2A extracts correspondingly. Immunofluorescence data showing the chromatin-bound TOP2A in the indicated conditions. Each dot represents the normalized mean intensity of a single nucleus. (D) Immunofluorescence data showing chromatin-bound CAP-G (condensin I) and chromosome lengths in the indicated conditions. Each dot represents one nucleus (D) or one chromosome (E). Data plotted for all plots are medians and corresponding 95% CI. The number of nuclei or chromosomes is indicated above the plots.
If H1.8 regulates chromosome length by a mechanism independent of its role in condensin I inhibition, we expect that the chromosome lengths in ∆CAP-G extracts and ∆H1∆CAP-G extracts would be different. Unfortunately, since severe defects in chromosome individualization in ∆CAP-G extracts prevented us from measuring chromosome lengths, this comparison was not possible (see Figure 6B). To circumvent this issue, we asked if similar condensin I levels on both ΔIgG and ΔH1 chromatin would result in similar chromosome lengths. Since condensin I binding to chromosomes is suppressed by H1.8, to load condensin I to chromosomes in ∆H1.8 extracts at the levels seen in control (∆IgG) extracts, we assembled chromosomes in extracts where condensin I was partially depleted (Figure 3—figure supplement 1B). When mitotic chromosomes were assembled in ∆H1.8 extracts that contain 40% (of ΔIgG) condensin I (CAP-G), condensin I level on these chromosomes was equivalent to that on chromosomes in control (∆IgG) extracts (Figure 3E, Figure 3—figure supplement 1C). Consistent with the hypothesis that H1.8 does not contribute to chromosome length regulation independently of condensin I, average chromosome length was essentially identical between these two conditions (ΔH1 40% CAP-G and ∆IgG; Figure 3F, Figure 3—figure supplement 1D).
As H1.8 depletion also results in the accumulation of TOP2A (Figure 1), we investigated whether the increased TOP2A also plays a role in the chromosome elongation in ΔH1 extracts. As TOP2A activity is essential for sperm decondensation (Adachi et al., 1991; Shintomi et al., 2015), we examined the chromosome length upon partial depletion of TOP2A (Figure 3—figure supplement 2A). Similar to previous reports from other systems (Farr et al., 2014; Samejima et al., 2012; Nielsen et al., 2020), mitotic chromosomes prepared in extracts with reduced TOP2A activity show slightly increased chromosome length (Figure 3—figure supplement 2B). Consistent with these observations, partial depletion of TOP2A in H1.8-depleted extracts also led to a further increase and not decrease in chromosome length (Figure 3—figure supplement 2C–E), suggesting that chromosome elongation in H1.8-depleted extracts was not caused by the increased level TOP2A on chromatin. In summary, these results demonstrate that H1.8 controls mitotic chromosome lengths primarily through limiting the chromosome binding of condensin I.
Experimental and simulation studies have shown that mitotic chromosome length and width are sensitive to the number of condensin I molecules on a chromosome, through affecting the loop size (Gibcus et al., 2018; Goloborodko et al., 2016a; Goloborodko et al., 2016b; Fitz-James et al., 2020). Low levels of condensin I are predicted to result in fewer and larger loops, while higher levels of condensin I will lead to a larger number of smaller loops. Therefore, if linker histone H1.8 decreases chromosome length through limiting condensin I association with chromatin, we expect that the average size of mitotic condensin loops decreases upon H1.8 depletion. As an alternative way to quantitatively assess the effect of H1.8 and condensins on mitotic chromosome organization, we used the chromosome conformation capture assay Hi-C (Lieberman-Aiden et al., 2009).
Hi-C contact probability maps were generated from replicated metaphase X. laevis sperm chromosomes 60 min after cycling back into mitosis in the presence of nocodazole. As seen in mitotic chromosomes in somatic cells (Naumova et al., 2013; Gibcus et al., 2018) and in early stages of mouse development (Du et al., 2017), all the Hi-C contact maps showed no checkerboard pattern commonly associated with interphase chromosome compartments and lacked any sign of topologically associating domains (TADs) (Szabo et al., 2019; Figure 4A; data available at GEO under accession no. GSE164434). Hi-C interaction maps are characterized by the decay in contact probability, P, as a function of the genomic distance, s. To derive quantitative information about the polymer structure of the mitotic chromosomes, we plotted the genome-wide average P(s) (Figure 4B). P(s) plots were consistent among two biological replicates (Figure 4—figure supplement 1A) and among different chromosomes (Figure 4—figure supplement 1B). The interaction decay profile of the control (ΔIgG) chromosomes was also qualitatively similar to mitotic chromosomes in both DT40 and human cells (Naumova et al., 2013; Gibcus et al., 2018; Elbatsh et al., 2019). Condensin I depletion (∆CAP-G) caused a major change in the Hi-C map (Figure 4A), reflecting its severe morphological defects in mitotic chromosomes in Xenopus egg extracts (Hirano et al., 1997). Unlike in DT40 cells, where depletion of condensin I and condensin II reduces interactions at shorter (<6 Mb) and longer (>6 Mb) distances, respectively (Gibcus et al., 2018), condensin I depletion affected interactions at longer distances (~10 Mb), whereas condensin II depletion (ΔCAP-D3) did not cause recognizable changes in interactions at both long and short distances (Figure 4—figure supplement 1C). A second diagonal band, which is indicative of a strong helical organization in the chromosome axis (Gibcus et al., 2018), was not seen in Xenopus egg extracts (Figure 4A, Figure 4—figure supplement 1C), likely reflecting the minor contribution of condensin II in this system or perhaps due to a prolonged arrest in mitosis (Gibcus et al., 2018). In H1.8-depleted extracts (∆H1), a dramatic decrease in interactions at long genomic distances (1–10 Mb) was observed (Figure 4A and B), as expected from the thinner chromosomes (Figure 3).
Effects of H1 and/or condensin I and II depletion on mitotic genome folding.
(A) Hi-C maps of metaphase X. laevis chromosome 3S, binned to 250 kb, in the indicated condition. (B) Genome-wide average contact probability decay curves for the indicated conditions showing the changes in longer distance interactions. (C) Derivative plots of the average contact probability decay curves for dispersed chromosomes from mock (ΔIgG) and H1.8-depleted extracts (ΔH1) showing the change in estimated loop size. The solid and dotted lines are from two independent biological replicates. (D) Derivative plots of the genome-wide average contact probability decay curves in the indicated conditions. The dotted lines indicate the layer sizes for each plotted condition. (E) Estimates of layer size from derivatives of genome-wide probability decay curves upon depletion of H1.8 and CAP-G (condensin I) or CAP-D3 (condensin II). The mean and range of two biological replicates are shown.
Source data for all the figures in <xref ref-type="fig" rid="fig4">Figure 4</xref> and its figure supplements.
Effects of H1, condensin I, and condensin II depletion on mitotic genome folding.
(A) Contact probability decay curves of mitotic chromosomes in mock (ΔIgG) and H1-depleted (ΔH1) extracts from two different experiments showing the replicability of the Hi-C features. The solid and dashed lines are from biological replicates. (B) Hi-C maps (left) binned to 500 kb and contact probability curves (right) of single chromosomes showing the uniform effects of H1.8 depletion genome wide. (C) Genome average contact probability decay curves upon H1.8, CAP-G, and CAP-D3 depletions. (D) Genome average contact probability derivative curves of metaphase chromosomes in undiluted mock (ΔIgG) and H1.8-depleted (ΔH1) extracts. The layer size estimates from the valley in the derivative plots are indicated. The dotted lines are a biological replicate. The data for this figure is submitted as Figure 4—source data 1.
H3-H4 depletion leads to even smaller loop sizes.
(A) Western blotting showing the depletion of H3-H4 in total egg extract using the H4K12ac antibody (Zierhut et al., 2014). (B) Quantification of condensin I (CAP-G) immunofluorescence levels on chromosomes normalized to DNA in ∆H1 and ∆H3-H4 extracts. Each dot represents the average signal intensity of a single chromosome cluster (from one nucleus). Data plotted is median and 95% CI > 40 nuclei were quantified for each condition plotted. The statistical significance between ΔH1 and ΔH3-H4 was analyzed by a two-tailed Mann–Whitney U-test. (C) Hi-C probability decay derivative plots estimating the loop sizes in ∆H1 and ∆H3-H4 extracts. The dashed lines indicate the loop sizes in the corresponding conditions and the dotted lines indicate the layer size. Two biological replicates for Hi-C on chromosomes from ΔH3-H4 extracts were performed and similar results were obtained from both experiments. The data for this figure is submitted as Figure 4—source data 1, and the number of nuclei imaged in (B) is indicated above the figure.
P(s) plots are useful to determine the underlying polymer structure of the chromosome such as average loop size (Gibcus et al., 2018). Specifically, the first derivative (slope) of the P(s) plots can reveal both the average loop size and the amount of DNA per layer of the rod-shaped mitotic chromosome (layer size) (Gassler et al., 2017; Abramo et al., 2019; Gibcus et al., 2018; Figure 4C). Average loop size can be estimated from the peak value in the derivative plot (Gassler et al., 2017; Patel et al., 2019). As the derivative plot in our data showed a peak in the 10 kb to 1 Mb range, we estimated that this peak location reflects the average loop size. Unlike mitotic chromosomes in DT40 or HeLa cell lines (Gibcus et al., 2018; Abramo et al., 2019), metaphase chromosomes from Xenopus extracts had a flattened peak (Figure 4—figure supplement 1D), possibly due to a larger variation in loop sizes. Although it is difficult to estimate the exact loop size from these plots, chromosomes from ΔH1 extracts showed a reproducible shift of the peak towards smaller genomic distances (Figure 4—figure supplement 1D). This is consistent with a decrease in loop size due to increased condensin I accumulation on chromatin upon H1.8 depletion. Derivative plots generated from Hi-C maps of dispersed chromosomes in diluted extracts (Figure 3A) showed a similar shift in the peak towards smaller genomic distances upon H1.8 depletion (Figure 4C). The derivative plots from these dispersed chromosomes showed a better-defined peak, allowing a coarse loop size estimate of ~140 kb in control (ΔIgG) extracts and ~110 kb in ΔH1 extracts using the previously reported derivative peak heuristic (Gassler et al., 2017). These coarse loop size estimates are comparable to the estimated size of condensin I-driven loops in DT40 and HeLa cells (Gibcus et al., 2018; Naumova et al., 2013).
P(s) plots for mitotic chromosomes display three regimes that are typical for relatively stiff rod-shaped conformation (Naumova et al., 2013; Gibcus et al., 2018): for small genomic distance, the contact probability is dominated by interactions between pairs of loci located within loops (up to 100–200 kb). For loci separated by up to a few megabases, the contact probability decays slowly with genomic distance. These interactions mostly reflect contacts between loci located in different loops that are relatively closely packed as a radial layer of loops around the central axis (intralayer regime). The third regime is characterized by a steep decay in contact probability at several megabases. This represents pairs of loci separated by a relatively large distance along the axis of the rod-shaped chromosome so that they very rarely interact (interlayer regime). The size of these layers can be obtained from the derivative of P(s), where the steep drop in contact probability curve is marked by sharp drop in the derivative value. For control chromosomes, the average amount of DNA in each layer of loops (layer size) was approximately 5 Mb, given the steep decay in contact probability observed for loci separated by more than 5 Mb for both replicates (Figure 4B and D). For H1.8-depleted chromosomes, we observed a smaller layer size of around 3.5 Mb (Figure 4B and D). This 1.5-fold reduction of DNA content of each layer would explain the observed 1.5-fold increase in chromosome length (Figure 3D). Further, we can derive the number of loops per layer by dividing the layer size by the loop size in the corresponding condition. In both control (ΔIgG) and H1.8-depleted extracts (ΔH1), the number of loops per layer was ~40. This indicates that the change in the layer size upon H1.8 depletion is a result of decreased loop size, while the number of loops per layer is not affected.
The layer size in H1.8/condensin I co-depleted extracts (ΔH1ΔCAP-G) was larger than those in control extracts (ΔIgG) (Figure 4D and E). Further supporting the idea that the layer size anticorrelates with the chromosome length, chromosomes in ΔH1ΔCAP-G extracts were indeed shorter than those in control ∆IgG extracts (Figure 3D and F). Condensin II co-depletion (ΔH1ΔCAP-D3) did not affect the layer size (3.5 Mb) and also the chromosome length (Figures 4D, E, 3D). Condensin II depletion alone (ΔCAP-D3) also did not affect the layer size, consistent with the observed lack of change in the chromosome length (Figure 4D and E, Figure 3—figure supplement 1A). Taken together, these data support the hypothesis that H1.8 limits the condensin I level on chromatin and shortens chromosome lengths, allowing each condensin I to form a longer loop, and consequentially tuning the amount of DNA present in each layer. The data also suggest that unlike in chicken DT 40 cells condensin I and not condensin II plays the dominant role in the organization of loop layers (Gibcus et al., 2018).
Since condensin I binding to DNA is also limited by nucleosomes (Zierhut et al., 2014; Kong et al., 2020; Shintomi et al., 2017), we expected that loss of nucleosomes would similarly reduce the loop and layer sizes. To test this, we depleted H3-H4 tetramers using an antibody to acetylated lysine 12 of histone H4 (H4K12ac; Zierhut et al., 2014) and generated metaphase chromosomes for Hi-C (Figure 4—figure supplement 2A). Since X. laevis sperm contains preloaded paternal H3-H4 (Shechter et al., 2009), the number of nucleosomes in our metaphase chromosomes was expected to be reduced by at most 50%. As nucleosomes occupy a large majority of the genomic DNA (Lee et al., 2007; Chereji et al., 2019), even a partial histone depletion (ΔH3-H4) increased chromatin-bound condensin I beyond that of H1.8 depletion (ΔH1) (Figure 4—figure supplement 2B). Consequentially, the layer size in ∆H3-H4 extracts became much smaller (around 500 kb) than in ∆H1 extracts (Figure 4—figure supplement 2C). Assuming that the number of loops per layer is similar in these chromosomes (~40), the loop size estimate is 12 kb, which is much shorter than in H1.8 depletion. Altogether these results suggest that global occupancy of nucleosomes and linker histones can affect DNA loop size and chromosome through controlling the number of condensin molecules on the chromatin fiber.
Condensins and topo II act in concert to generate mitotic chromosomes from decondensed interphase nuclei (Cuvier and Hirano, 2003). Both experimental observations and in silico experiments suggest that condensin can drive decatenation of sister chromatids (Goloborodko et al., 2016a; Nagasaka et al., 2016; Marko, 2009) even during metaphase arrest (Piskadlo et al., 2017). In addition, it has been suggested that condensin-mediated chromosome compaction also promotes chromosome individualization (Brahmachari and Marko, 2019; Sun et al., 2018), though it remains to be established if different linear chromosomes (non-sisters) are catenated with each other even after completion of mitotic compaction since Ki-67 on chromosome surfaces may act as a barrier to prevent interchromosomal DNA interaction during mitosis (Cuylen et al., 2016).
To assess the role of H1.8, condensins, and topo II in chromosome individualization, we used two different assays (Figure 5A). The first was to measure the three-dimensional surface area of metaphase chromosome clusters. In the presence of the microtubule depolymerizing drug nocodazole, metaphase chromosomes derived from each nucleus clustered to form tight ball-like structures (Figure 5B). This clustering can be quantified by the large reduction in the surface area of metaphase chromosome clusters over control nuclei (Figure 5C, Figure 5—figure supplement 1A). This clustering suggests that, in control extracts, chromosomes are predisposed to accumulate interchromosomal contacts (Figure 5A). To verify if these interchromosomal contacts are topological in nature, we physically dispersed the clustered chromosomes by extract dilution (Figure 5A; Funabiki and Murray, 2000). Since this dilution procedure involves mechanical dispersal by hydrodynamic forces, we expect that chromosomes would remain clustered after dilution only in the presence of strong unresolved interchromosomal contacts. Indeed, when ICRF-193, the drug that inhibits topo II-dependent catenation/decatenation without leaving double-strand DNA breaks (Tanabe et al., 1991), was added to egg extracts at the mitotic entry, chromosome individualization was completely blocked, forming large chromosome clusters after extract dilution (Figure 5—figure supplement 1B–D, ICRF-50 min). Even when ICRF-193 was added to metaphase egg extracts after completion of metaphase chromatid formation, but 2 min before extract dilution (Figure 5—figure supplement 1B, bottom), efficiency of chromosome individualization decreased (Figure 5—figure supplement 1C and D, ICRF-2 min). These results suggest that substantial interchromosomal topological catenations remain unresolved in metaphase and that these can be resolved by TOP2A during mechanical dispersion (via spindle or dilution). To quantify the clustering of unresolved chromosomes, we stained the coverslips for CENP-A, the centromere-specific histone H3 variant, which marks a centromere locus per chromosome (Edwards and Murray, 2005). While each fully individualized chromosome showed one CENP-A focus (CENP-A doublet, representing centromeres of a paired sister chromatids, is counted as one focus), multiple CENP-A foci were observed in a cluster of chromosomes. Since even fully individualized chromosomes stochastically interacted during chromosome dispersion, we redefined the chromosome cluster here for the chromosome mass containing four or more CENP-A foci (Figure 5D and E, Figure 5—figure supplement 1E). Although chromosomes from each nucleus tightly compact into a ball-like mass in nocodazole treated metaphase extracts (Figure 5B), they individualized normally by dispersion upon extract dilution (Figure 5D and E, Figure 5—figure supplement 1E). This suggests that the interchromosomal contacts accumulated in the absence of spindles can be resolved by mechanical dispersion. TOP2A inhibition by treatment with ICRF-193 however almost completely blocked chromosome individualization in nocodazole-treated extracts (Figure 5D and E, Figure 5—figure supplement 1E), confirming the accumulation of interchromosomal catenations in clustered chromosomes.
Substantial interchromosomal links remain in metaphase chromosomes.
(A) Schematic showing the dispersal protocol and the different stages for the two chromosome individualization measurements. (B) Representative images of metaphase chromosomes in extracts containing Alexa Fluor 647-labeled tubulin (blue) with and without 10 µg/ml nocodazole. DNA was visualized using Hoechst 33342 (magenta). Scale bar is 20 µm. (C) Quantification of the three-dimensional surface area normalized to the total DNA intensity in (B). Each dot represents a single chromosomal mass. The number of masses quantified in each condition is indicated. (D) Representative images of the dispersed chromosomal masses in the indicated conditions. Scale bar is 20 µm. (E) Percent frequency of individualized chromosomes (chromosomes in DNA masses with <4 CENP-A foci) in the indicated conditions. Data from two biological replicates is shown. The number of masses quantified in each condition is indicated.
Source data for <xref ref-type="fig" rid="fig5">Figure 5</xref> and its figure supplement.
Regulation of chromosome individualization by topo II, condensins, and H1.8 in <italic>Xenopus</italic> egg extracts.
(A) Biological replicate of Figure 5C. (B) Schematic of ICRF-193 addition to check for requirement of topo II activity in individualizing chromosomes. Interphase nuclei were first formed in ∆IgG or ∆H1 extracts, topo II inhibitor ICRF-193 (50 µM) was added to egg extracts, either together with corresponding depleted egg extracts and incubated for 50 min (ICRF-50 min), or 48 min after adding the depleted extracts, followed by 2 min incubation with ICRF-193 (ICRF-2 min). (C) Metaphase extracts processed as (A) were diluted to disperse individualized chromosomes. Representative Hoechst images of chromosomes are shown. Bar, 20 µm. (D) Quantification of the Hoechst-stained area of chromosomes in (B). Each dot represents the area of a single chromosome or a chromosome cluster. Large values indicate the extent of chromosome clusters. The box shows the 10th–90th percentile limits of the sample values. The number of DNA masses imaged in each condition is indicated above the figure. (E) Histogram of the number of chromosomes in clusters containing the indicated number of CENP-A foci in the indicated conditions. The number of chromosomes imaged is the same as in Figure 5E. The data for this figure is submitted as Figure 5—source data 1.
Since condensins promote sister chromatid decatenation through topo II (Baxter et al., 2011; Goloborodko et al., 2016a), we then asked if H1.8-mediated suppression of condensins limits chromosome individualization and resolution of interchromosomal linkages. If so, H1.8 depletion may reduce the minimum required level of condensin activities to support chromosome individualization. In metaphase mock-depleted (ΔIgG) and H1.8-depleted (∆H1) extracts with replicated chromosomes, the extract dilution procedure (Figure 3A) resulted mostly in physically separated single chromosomes but also a small number of clumped chromosomes (Figure 6B, Figure 6—figure supplement 1C). To quantify chromosome individualization, we measured chromosome clustering using CENP-A foci (Figure 6—figure supplement 1), and also independently by chromosome morphology-based classification (Figure 6—figure supplement 1C). Condensin I depletion (ΔCAP-G) resulted in defective chromosome individualization, suggesting that condensin I activity drives resolution of interchromosomal entanglements, and that condensin II is not sufficient to resolve these interchromosomal links in this background. Strikingly, co-depletion of H1.8 and condensin I (ΔH1ΔCAP-G) effectively rescued chromosome individualization without detectable CAP-G on chromatin (Figure 6A–D, Figure 6—figure supplement 1D). This apparent bypass of condensin I requirement in chromosome individualization required condensin II as chromosome individualization failed in the triple-depleted extracts (ΔH1ΔCAP-G∆CAP-D3; Figure 6B,C). The increased condensin II in ΔH1ΔCAP-G extracts may replace the function of condensin I by condensin II in the absence of H1.8 (Figure 6E). These results demonstrate that H1.8-mediated suppression of condensin enrichment on chromatin limits chromosome individualization during mitosis.
H1.8 suppresses condensin to limit chromosome individualization.
(A) Western blots of total egg extracts showing depletion levels in extract of condensin I and condensin II using the CAP-G and CAP-D3 antibodies, respectively. * represents non-specific band. (B) Representative images of chromosomes after extract dilution, which disperses individualized chromosomes. DNA and centromere-associated CENP-A immunofluorescence are shown. Bar, 20 µm. (C) Percent frequency of individualized chromosomes (chromosomes in DNA masses with <4 CENP-A foci) in the indicated conditions. A large majority of DNA masses with no CENP-A foci are derived from ∆CAP-D3 extracts, where CENP-A loading is compromised (Bernad et al., 2011). DNA masses and CENP-A foci were identified using Otsu’s thresholding algorithm and CENP-A foci in a binarized DNA mask were counted. The numbers of DNA masses counted in each condition were as follows: ΔIgG (502, 643), ΔH1 (1279, 839), ΔCAP-G (447, 170), ΔCAP-D3 (937), ΔH1ΔCAP-G (1565, 301), ΔH1ΔCAP-D3 (1536, 300), ΔH1ΔCAP-GΔCAP-D3 (300, 156). (D) Quantification of CAP-G (condensin I) immunofluorescence normalized to the DNA signal for the indicated conditions. Each gray or orange dot represents the average signal intensity of a single chromosome cluster (from one nucleus). Each black dot represents the median signal intensity from a single experiment. Bars represent mean and range of the medians of two independent experiments. (E) Quantification of CAP-G2 (condensin II) immunofluorescence intensity, normalized to the DNA signal for the indicated conditions. Each gray or magenta dot represents the average signal intensity of a single chromosome cluster (from one nucleus). Each black dot represents the median signal intensity from a single experiment. Bars represent mean and range of the medians of two independent experiments. The p-values in (D) and (E) were calculated by an unpaired Student’s t-test and a two-tailed Mann–Whitney U-test respectively. The number of nuclei imaged in each condition in (D) and (E) in each experiment is indicated above the figures.
Source data for all the figures in <xref ref-type="fig" rid="fig6">Figure 6</xref> and its figure supplement.
Additional data for chromosome clustering.
(A) Distribution of CENP-A foci per DNA mass in the experiment shown in Figure 6A–C. Clusters of unresolved chromosomes are represented by higher numbers (>3) of CENP-A foci per DNA mass indicates clusters. No detectable CENP-A focus in ∆CAP-D3 (∆condensin II) extracts is due to low CENP-A signal, reflecting the reported role of condensin II in CENP-A loading (Bernad et al., 2011). The number noted in each sample label is the number of chromosomes/clusters counted in each condition. (B) Percent frequency of single chromosomes (chromosomes in DNA masses with <2 CENP-A foci) in the indicated conditions showing that the cutoff for number of CENP-A foci does not change the relative individualization measured. The number of DNA masses imaged in each condition is the same as in Figure 6C. (C) Categories of chromosome clusters observed upon performing chromosome individualization assay. (D) Percent frequency of DNA clusters categorized as unindividualized nuclei as in (C) in the indicated conditions. Each chromosome mass is classified to one of the three categories based on the morphology. ‘Clumpy chromosomes’ represents a mass containing a few clearly recognizable chromosomes, while ‘unindividualized’ represents a mass with unresolved multiple chromosomes. Mean and SEM from three independent experiments. Each dot represents the percentage of unindividualized chromosome clusters in the indicated condition in an independent biological replicate. The number of DNA masses counted in each condition is as follows: ΔIgG (50, 128, 201), ΔH1 (66, 135, 447), ΔCAP-G (13, 30, 50), ΔH1ΔCAP-G (58, 201, 194).
H1.8 prevents chromosomes from hyper-individualization by suppressing condensins and topo II
Since TOP2A-mediated decatenation played a role in resolving interchromosomal links, we next examined the functional significance of H1.8-mediated suppression of TOP2A enrichment on mitotic chromosomes (Figure 1C–E), asking if H1.8 depletion could reduce the required TOP2A level in extracts for chromosome individualization. Complete loss of TOP2A inhibits decompaction of sperm nuclei, a process associated with replacement of protamines with histones (Adachi et al., 1991), so we addressed this question using extracts partially depleted of TOP2A (Figure 7A). We first generated nuclei with replicated chromosomes in extracts containing the normal level of TOP2A, and then the extracts were diluted with ∆TOP2A extracts to reduce the total TOP2A level to 25% (Figure 7—figure supplement 1A). Under this condition, the level of chromosome-associated TOP2A also reduced to 25% (Figure 7B). Upon reduction of TOP2A to 25%, the frequency of unindividualized chromosome clusters increased about threefold in ΔIgG background (Figure 7C). However, in the absence of H1.8, extracts with 25% levels of TOP2A were still able to support maximum level of chromosome individualization as chromosome-associated levels of TOP2A became equivalent to untreated control extracts (Figure 7C). These data suggest that chromosome individualization is sensitive to TOP2A levels on chromatin, and that H1.8 suppression of TOP2A plays a role in suppressing chromosome individualization.
H1.8 suppresses hyper-individualization through condensins and topo II.
(A) Schematic of partial TOP2A depletion to test sensitivity of chromosome individualization to TOP2A levels. (B) Quantification of chromosome-associated TOP2A upon partial TOP2A depletion. Each dot represents the mean of TOP2A intensity normalized to DNA intensity of a single chromosome cluster (from one nucleus). The data plotted is median ± 95% CI. (C) Percent frequency of DNA clusters categorized as unindividualized nuclei (Figure 6—figure supplement 1C) upon partial TOP2A depletion. Mean and SEM from three independent experiments. Each dot represents the percentage of unindividualized chromosome clusters in the indicated condition in an independent biological replicate. (D) Representative images of nuclei using DNA (Cy5-dUTP) showing the clustering phenotype in the indicated conditions. Scale bar, 10 µm. (E) Quantification of the three-dimensional surface area of the chromosome clusters in (D) normalized to DNA (Cy5-dUTP) signal. Each gray, magenta, green, or orange dot represents the normalized surface area of single nucleus or chromosome cluster, and each black square represents the median surface area of a single experiment. Data plotted is mean and SEM of four independent experiments. (F) Quantification of TOP2A immunofluorescence intensity normalized to the DNA signal for the indicated conditions. Each gray or magenta dot represents the average signal intensity of a single chromosome cluster (from one nucleus). Each gray, magenta, orange, or gray dot represents the median signal intensity from a single experiment. Mean and SEM of the median of four independent experiments are also shown. (G) Representative images of nuclei using DNA (Cy3-dUTP) showing the clustering phenotype in the indicated conditions. Scale bar, 10 µm. (H) Quantification of the three-dimensional surface area of the chromosome clusters in (G) normalized to DNA (Cy3-dUTP) signal. Each black open circle represents the normalized surface area of single nucleus or chromosome cluster. Data plotted is median and 95% CI. The p-values in (C) and (F) were calculated by an unpaired Student’s t-test, and the p-values in (H) were calculated by a two-tailed Mann–Whitney U-test. The number of nuclei imaged in each condition for each experiment in (E), (F), and (H) is indicated above the figure.
Source data for all the figures in <xref ref-type="fig" rid="fig7">Figure 7</xref> and its figure supplement.
H1.8 suppresses over-individualization through condensins and topo II.
(A) Western blots of total egg extracts showing partial depletion of TOP2A. Left three lanes: dilution series of total mock-depleted extracts (∆IgG) for signal quantitation. Middle ∆IgG lane: extracts with 25% level of TOP2A. right two ∆H1 lanes; extracts depleted of H1.8 with either 25 or 100% levels of TOP2A. (B) Quantification of TOP2A signal normalized to the DNA signal in the indicated conditions. Each gray, orange, or violet dot represents the mean normalized TOP2A level in a single nucleus. The black squares represent the median of the TOP2A values of all the nuclei in a single experiment. Data plotted is mean and SEM. The p-value in (C) was calculated using an unpaired Student’s t-test. The number of nuclei imaged in (C) in each condition is indicated above the figure, and the data for this figure is submitted as Figure 7—source data 1.
These data suggest that increased condensins and topo II levels on chromosomes in H1.8-depleted extracts reduce the number of links between chromosomes. To assess the consequence of these reduced interchromosomal links, we measured the chromosome clustering in nocodazole-treated extracts. As shown earlier, chromosomes of each nucleus in nocodazole-treated control extracts cluster into tight balls (Figure 5B,C). In contrast, individual chromosomes were more readily distinguished by DNA staining and spread to larger area in ΔH1 extracts, suggesting that the H1.8-mediated suppression of chromosome individualization is responsible for the chromosome clustering when spindle assembly is compromised (Figure 7D,E). Although condensin I depletion (ΔCAP-G) led to failed chromosome individualization (Figure 6B,C), these chromosomes do not show increased clustering over those seen in control (ΔIgG) extracts. This is consistent with the notion that chromosomes may already be maximally clustered in control extracts. However, H1.8 condensin I co-depleted extracts (ΔH1ΔCAP-G) showed reduced clustering compared to ∆IgG extracts, despite the shorter average chromosome length in ∆H1∆CAP-G extracts (Figure 3D,F), suggesting that chromosome spreading upon H1.8 depletion is not primarily driven by chromosome elongation but by chromosome hyper-individualization. Since TOP2A is enriched on chromosomes upon H1.8 depletion in a condensin-independent manner (Figure 7F, Figure 7—figure supplement 1B), we then asked if this increased TOP2A plays a role in this increased chromosome spreading. Partial TOP2A depletion had no effect on clustering in control (ΔIgG) extracts since these chromosomes were already tightly packed together (Figure 7G,H–25% TOP2A). TOP2A depletion in H1.8-depleted extracts also did not reduce the chromosome spreading even though TOP2A levels on these chromosomes became comparable to that of the control (ΔIgG-100%; Figure 7G,H), perhaps because increased condensin loading can compensate for reduced TOP2A activity. However, partial TOP2A depletion reduced chromosome spreading in H1.8/condensin I co-depleted (ΔH1ΔCAP-G) extracts, suggesting that suppression of both condensin I and topo II by H1.8 keeps egg extract chromosomes together during mitosis even in the absence of spindle microtubules.
Discussion
It has been thought that the linker histone H1 promotes local chromatin compaction through stabilizing linker DNA and facilitating nucleosome-nucleosome interaction (Song et al., 2014; White et al., 2016; Li et al., 2016). Here we demonstrated that H1.8 also plays a major role in regulation of long-range DNA interaction through controlling chromatin loading of condensins. Although condensins and topo II are activated in mitosis to drive chromosome segregation (Kimura et al., 1998; Hirano and Mitchison, 1991), we showed that their functionalities are antagonized by H1.8 to tune chromosome length and prevent hyper-individualization, which increases surface area (Figure 8, Table 1). Thus, the linker histone H1, which can promote local chromatin compaction through promoting nucleosome-nucleosome interaction, can also suppress chromosome individualization and long-range DNA folding.
A graphical model of how H1.8 controls mitotic chromosome length.
In the absence of H1.8, more condensins and topo II bind to more DNA loops of shorter length, resulting in longer and more individualized chromosomes (top). H1.8 limits chromatin levels of condensins and topo II to generate longer and thus fewer DNA loops, resulting in shorter and less individualized chromosomes (bottom).
Summary table of chromatin levels of condensins, topo II, and chromosome phenotypes.
Condensin I
Condensin II
TOP2A
Chromosome length
Chromosome individualization
∆H1
2.5×
2×
3.5×
1.5×
++
∆CAP-G
0.1×
1×
1×
NA
Defective
∆CAP-D3
1×
0.4×
1×
1×
Unchanged
∆H1∆CAP-G
0.2×
2×
3.5×
0.5×
+
∆H1∆CAP-D3
2.5×
0.9×
3.5×
1.5×
++
∆H1∆CAP-G ∆CAP-D3
0.2×
0.9×
3.5×
NA
Defective
Nucleosomes reduce binding of condensins to DNA in vitro (Kong et al., 2020), in vivo (Toselli-Mollereau et al., 2016; Piazza et al., 2014), and in Xenopus egg extracts (Zierhut et al., 2014; Shintomi et al., 2017). Now we showed that the linker histone H1.8 limits binding of condensin I and II to chromatin both in vitro and in Xenopus egg extracts. H1.8 suppressed condensin binding on both mononucleosomes and nucleosome arrays, suggesting that H1.8 is able to compete out condensins for the same linker DNA targets (Rudnizky et al., 2021), though the capacity of linker histones to promote higher-order structures or phase separation may also limit the access of condensin (Song et al., 2014; Gibson et al., 2019). Since condensin I subunits are most abundant chromatin proteins whose levels were enhanced by H1.8 depletion (Figure 1—figure supplement 1E, Figure 1—source data 2), and chromosome lengths can be dictated by the amount of condensins in a manner independently of H1.8 (Figure 3F), we propose that regulating linker histone stoichiometry could serve as a rheostat to control chromosome length through tuning the condensin level on chromatin (Figure 8). Although such a linker histone-mediated chromosome length shortening seems odd in the large oocyte cells, mitotic chromosome length is constrained not only by the cell size but also by the spindle size (Schubert and Oud, 1997). As spindle size in Xenopus embryos does not scale with the cell size during the early divisions (Wühr et al., 2008), lack of chromosome clearance from the spindle midzone due to elongated anaphase chromosomes may result in chromosome breakage by the cytokinesis (Maresca et al., 2005; Janssen et al., 2011). By keeping the spindle and chromosome length short, variations in duration for chromosome segregation would be reduced to better synchronize cell division. Since the larger cell size correlates with reduced mitotic checkpoint strength (Galli and Morgan, 2016; Minshull et al., 1994), limiting chromosome length might be important for their timely and synchronous segregation during the rapid early embryonic cell divisions.
Since the average loop size is an aggregate result of loop extrusion rate, processivity, and number of loop extruders (Goloborodko et al., 2016b; Alipour and Marko, 2012), it is unclear from our data whether H1.8 affects the loop extrusion kinetics of a single condensin molecule. However, our observation is consistent with the in silico simulation showing that increasing the number of loop extruders makes chromosomes thinner and longer by reducing average loop size (Goloborodko et al., 2016a; Goloborodko et al., 2016b). Similar DNA loop shortening accompanied with increased condensin I loading was also reported on integrated fission yeast genome DNA segments in mouse and human chromosomes (Fitz-James et al., 2020). Reduced condensin loading due to mutations in condensin I or expression of phosphomimetic H3 mutants in human cells also led to reduced chromosome length (Elbatsh et al., 2019). Condensin binding is similarly suppressed by nucleosomes, but loop extrusion proceeds unhindered through sparsely distributed nucleosomes (Kong et al., 2020). Since the length of chromosomes with similar condensin I levels is unaffected by the presence of H1.8 (Figure 3E,F), H1.8 may increase the loop size by simply reducing the number of condensin molecules on chromatin but may not necessarily inhibit the loop extrusion rate or processivity.
The regulatory mechanisms of TOP2A recruitment to mitotic chromatin are less clear than those of condensins. Similar to condensin, both TOP2A and TOP2B preferentially localize at active and highly transcribed chromatin, indicating a possible preference for nucleosome-free regions (Canela et al., 2017; Thakurela et al., 2013; Yu et al., 2017). We observe that TOP2A levels on chromatin increase upon linker histone depletion in egg extracts and that linker histone inhibits TOP2A binding to nucleosome arrays in vitro. The preference of topo II for binding linker DNA may also explain the observation of well-spaced TOP2B binding peaks around the well-spaced nucleosomes around the CTCF- binding sites (Canela et al., 2017; Canela et al., 2019). A recent report has also demonstrated the competition between TOP2A and H1.8 in the absence of nucleosomes (Shintomi and Hirano, 2021). Unlike condensin I depletion, TOP2A depletion did not rescue chromosome elongation phenotype of H1.8 depletion. Rather, depletion of topo II leads to elongated chromosomes in vertebrate somatic cell lines and in early embryos of C. elegans (Farr et al., 2014; Samejima et al., 2012; Ladouceur et al., 2017; Nielsen et al., 2020) and also in Xenopus egg extracts (Figure 3—figure supplement 2). These data support the proposed structural role for TOP2A in axial compaction of mitotic chromosomes through its C-terminal chromatin-binding domain (Nielsen et al., 2020; Shintomi and Hirano, 2021; Lane et al., 2013).
Our data also suggest that condensins and topo II combine to resolve extensive and persistent interchromosomal topological entanglement during mitotic compaction in Xenopus egg extracts (Figure 5), but these activities are counterbalanced by H1.8. In HeLa cells, somatic variants of H1 are phosphorylated and evicted along the inter-chromatid axis. This partial H1 eviction in prophase seems to be required for complete decatenation along the chromosome arms (Krishnan et al., 2017). This suggests that local enrichment of condensins (particularly condensin II) upon H1 eviction may play a role in sister chromatid decatenation. In contrast, H1.8 binding to nucleosomes is enhanced upon transition from interphase to M phase in Xenopus egg extracts (Arimura et al., 2020). It might be counterintuitive that condensins and topo II are actively antagonized during mitosis by H1.8 to suppress chromosome individualization, which is a prerequisite for chromosome segregation. Since spindle assembly in oocytes relies on chromatin-induced microtubule nucleation (Heald et al., 1996), keeping chromosomes at metaphase plate might be important for maintaining robust bipolar spindle. As we showed that H1.8 is important for chromosome clustering when spindle assembly is inhibited (Figure 7), this microtubule-independent chromosome clustering may be particularly important for the large oocyte and early embryonic cells since it would be difficult to assemble a bipolar spindle onto all the chromosomes once they disperse into the large space of the cytoplasm. Clustering chromosomes through incomplete individualization may also avoid generation of chromosomes that do not associate with the spindle. Such a mechanism would facilitate effective kinetochore attachment during early embryonic cell divisions when the spindle checkpoint cannot be activated by unattached chromosomes (Mara et al., 2019; Gerhart et al., 1984; Hara et al., 1980). Indeed, it was recently shown that paternal and maternal chromosomes cluster at the interface of two pronuclei prior to the first zygotic mitosis after fertilization to facilitate rapid and efficient kinetochore microtubule attachment in human and bovine embryos (Cavazza et al., 2021). Another possible reason for the suppressed individualization is related to the fact that oocyte chromosomes completely lose cohesion from arms at the end of meiosis I, while maintaining sister chromatid cohesion only at the centromeres (Lister et al., 2010). Normally, this centromeric cohesion is critical for supporting the kinetochore tension to establish bipolar attachment. During long natural arrest at meiotic metaphase II, these centromeres undergo cohesion fatigue, where centromeres prematurely separate. However, proper segregation may still be accomplished due to apparent inter-chromatids DNA linkages (Gruhn et al., 2019). Resolution of these DNA linkages may be prevented by H1.8-mediated suppression of condensin and TOP2A. However, our data do not eliminate the possibility that H1.8 plays a condensin and TOP2A-independent role in regulating chromosome individualization. Regulation of Ki-67, which coats the surface of chromosomes, may be a good candidate of H1-mediated regulation (White et al., 2016; Gibson et al., 2019; Cuylen et al., 2016).
Linker histones are a dynamic component of chromatin (Misteli et al., 2000). Linker histone occupancy varies widely (Woodcock et al., 2006) and can be controlled by both the linker histone variant and their post-translational modifications (Th’ng et al., 2005; Christophorou et al., 2014; Hergeth and Schneider, 2015). We expect that changing the H1 stoichiometry on chromatin by changing the amount, subtype, and/or affinity (e.g., through post-translational modifications) of H1 affects the length and individualization of chromosomes through regulating the levels of condensins and topo II. In Xenopus egg extracts, somatic histone H1 fails to rescue chromosome elongation phenotype of H1.8 depletion since highly abundant importin β sequesters the somatic H1 and suppresses its binding to mitotic chromosomes (Freedman and Heald, 2010). It has been shown that the sperm male pronucleus dynamically interacts with microtubules during its dramatic nuclear reorganization upon exposure to the egg cytoplasm (Xue et al., 2013), during which H1.8 suppresses chromosome fragmentation mediated by microtubule-dependent force (Xiao et al., 2012). Thus, it is possible that H1.8 is tuned to provide chromatin with physical strength of bulk chromatin rather than facilitating chromosome individualization during the unique fertilization process. Linker histones also serve important interphase roles through regulation of transcription (Izzo et al., 2008), and the epigenetic landscape of the chromatin determines the linker histone variant on chromatin (Th’ng et al., 2005; Parseghian et al., 2001; Izzo et al., 2013). Since H1.8 competitively inhibits condensin binding in mitosis, it is tempting to speculate that linker histones also inhibit condensin II and cohesin binding in interphase. We suggest that local and global regulation of chromatin structure and function can be regulated by controlling differential expression of linker histone H1 variants and their modifications not just through promoting inter-nucleosomal interactions but also by controlling accessibility of SMC proteins and topo II.
Materials and methods
Reagent type (species) or resource
Designation
Source or reference
Identifiers
Additional information
Cell line (Spodoptera frugiperda)
SF9 insect cells
Gibco
11496015
Cell line (Trichoplusia ni)
High Five insect cells
Gibco
B85502
Cell line(Pichia pastoris)
Yeast cells
Invitrogen/Thermo Fisher Scientific Ryu et al., 2010
C18100
GS115
Biological sample (Xenopus laevis)
Xenopus
NASCO
LM00531RRID:XEP_Xla100
Female, adult frogs
Biological sample (Xenopus laevis)
Xenopus
NASCO
LM00715RRID:XEP_Xla100
Male, adult frogs
Antibody
Anti-H3 (rabbit polyclonal)
Abcam
RRID:AB_302613
WB (1 µg/ml)
Antibody
Anti-H2B (rabbit polyclonal)
Abcam
RRID:AB_302612
WB (1 µg/ml)
Antibody
Anti α-tubulin (mouse monoclonal)
Sigma-Aldrich
RRID:AB_477593
WB (1:10000)
Antibody
Anti-H1.8 (rabbit polyclonal)
Jenness et al., 2018
RU1974
WB (1 µg/ml)
Antibody
Anti-TOP2A (rabbit polyclonal)
Ryu et al., 2010
NA
WB (2 µg/ml)IF (1 µg/ml)
Antibody
Anti-CAP-G (rabbit polyclonal)
Zierhut et al., 2014
RU1008
WB (2 µg/ml)IF (2 µg/ml)
Antibody
Alexa 488-anti-CAP-G (rabbit polyclonal)
This study
NA
IF (4 µg/ml),refer to ‘Antibodies’ section in Methods
Antibody
Anti-CAP-D2 (rabbit polyclonal)
Hirano et al., 1997
NA
WB (2 µg/ml)
Antibody
Anti-CAP-G2 (rabbit polyclonal)
Gift from S. Rankin
OMRF195
WB (4 µg/ml)IF (4 µg/ml)
Antibody
Anti-CAP-D3 (rabbit polyclonal)
This study
RU2042
WB (2 µg/ml),refer to ‘Antibody production’ section in Methods
Anti-rabbit Alexa 488 F(ab’)2 fragment (goat polyclonal)
LifeScience Technologies
RRID:AB_142134
IF (1:1000)
Chemical compound, drug
Nocodazole
Sigma-Aldrich
M1404
10 µg/ml
Chemical compound, drug
ICRF-193
Santa Cruz Biotechnology
sc-200889
50/500 µM
Other
Dynabeads-Protein A
Thermo Fisher Scientific
100-08D
250 ng antibody/1 µl beads
Other
Dynabeads-M280 Streptavidin
Thermo Fisher Scientific
11206D
NA
Software, algorithm
MATLAB
MathWorks
R2019A
NA
Peptide, recombinant protein
DpnII
NEB
R0543
Peptide, recombinant protein
DNA polymerase I, large (Klenow) fragment
NEB
M0210S
Peptide, recombinant protein
T4 DNA ligase 1 U/µl
Invitrogen
15224090
Peptide, recombinant protein
T4 DNA polymerase
NEB
M0203L
Peptide, recombinant protein
T4 polynucleotide kinase
NEB
M0201
Peptide, recombinant protein
Biotin-14-dATP
Invitrogen
19524016
Commercial assay or kit
TruSeq Nano DNA Sample Prep Kit
Illumina
20015964
Peptide, recombinant protein
Klenow fragment (3′ → 5′ exo-)
NEB
M0212L
Experimental model and subject details<italic>X. laevis</italic> frogs
Animal husbandry and protocol (20031) approved by institutional animal care and use committee (IACUC) of the Rockefeller University were followed. Mature female pigmented X. laevis frogs (NASCO- LM00535MX) and male frogs (NASCO-LM00715) were maintained in a temperature-controlled room (16–18°C) using a recirculating water system at The Rockefeller Comparative BioScience Center (CBC). Frogs were temporarily moved to a satellite facility for ovulation.
MethodsAntibodies
H3 was detected with ab1791 (Abcam; 1 µg/ml for western blots). H2B was detected with ab1790 (Abcam; 1 µg/ml for western blots). α-tubulin was detected with T9026 (Sigma; 1:10,000 for western blots).
H1.8 was detected using anti-H1M (H1.8) antibody (Jenness et al., 2018, 1 µg/ml for western blots). Anti-TOP2A was a gift from Y. Azuma (Ryu et al., 2010; 1 µg/ml for western blots and IF). Anti-CAPD2 was a gift from T. Hirano (Hirano et al., 1997; 2 µg/ml for western blots). Anti-CAPG2 was a gift from S. Rankin (4 µg/ml for IF, 2 µg/ml for western blots). xCAP-G (Zierhut et al., 2014) and xCAP-D3 custom antibodies were used at 2 µg/ml for western blots and 1 µg/ml for IF (for xCAP-G only). CENP-A was detected using an antibody against N-terminal 50 amino acids of CENP-A (Wynne and Funabiki, 2015) and used at 4 µg/ml for IF. xCAP-G antibody was also conjugated to Alexa488 using the Alexa488-NHS ester (Thermo Fisher Scientific) using the manufacturer’s instructions. The conjugated antibody was purified using the Sephadex G-25 in PD-10 desalting column (Cytiva). The conjugated antibody was dialyzed into PBS + 50% glycerol and stored in aliquots at –80°C after freezing with liquid nitrogen. The labeled antibody was used at 4 µg/ml for IF.
IRDye 680LT goat anti-mouse IgG (H + L), IRDye 680LT goat anti-rabbit IgG (H + L), IRDye 800CW goat anti-mouse IgG (H + L), and IRDye 800CW goat anti-rabbit IgG (H + L) were used at 1:15,000 (LI-COR Biosciences) dilution for western blots. Alexa 488, Alexa 555, and Alexa647 conjugated secondary antibodies (Jackson Immunoresearch) were used for immunofluorescence.
Antibody production
xCAP-D3 C-terminal peptide (CRQRISGKAPLKPSN) was synthesized at The Rockefeller University Proteomics Resource Center. The peptide was then coupled to keyhole limpet hemocyanin protein according to the manufacturer’s protocol (Thermo Fisher Scientific) and used to immunize rabbits (Cocalico Biologicals). Antibody was purified from the immunized rabbit sera using affinity purification against the same peptide coupled to SulfoLink resin (Thermo Fisher Scientific). The antibody was dialyzed into PBS + 50% glycerol and stored with the addition of 0.05% sodium azide.
<italic>Xenopus</italic> egg extracts and immunodepletion
Cytostatic factor (CSF)-arrested X. laevis egg extracts were generated as previously described (Murray, 1991). To generate replicated mitotic chromosomes, 0.3 mM CaCl2 was added to CSF-arrested extracts containing X. laevis sperm to cycle the extracts into interphase at 20°C. 90 min after adding CaCl2, half the volume of fresh CSF extract and 40 nM of the non-degradable cyclin BΔ90 fragment were added to interphase extracts to induce mitotic entry (Holloway et al., 1993; Glotzer et al., 1991). After 60 min of incubation, extracts were processed for morphological and biochemical assessments. For all experiments involving immunofluorescence, 10 nM nocodazole was added along with the cyclin BΔ90.
For immunodepletions of 50–100 µl extracts, antibodies were conjugated to Protein-A coupled Dynabeads (Thermo Fisher Scientific) at 250 μg/ml beads, either at room temperature for 60 min or overnight at 4°C. Mock (IgG) and H1.8 (H1) antibody beads were crosslinked using 4 mM BS3 (Thermo Fisher Scientific) at room temperature for 45 min and quenched using 10 mM Tris-HCl (Sigma). All antibody beads were washed extensively using Sperm Dilution Buffer (SDB; 10 mM HEPES, 1 mM MgCl2, 100 mM KCl, 150 mM sucrose) and separated from the buffer using a magnet before addition of extract. H1.8 depletions (ΔH1) were performed with two 45 min rounds of depletion at 4°C using 2 volumes of antibody-coupled beads for each round. For condensin I and condensin II depletions, 1.5–2 volumes of xCAP-G or xCAP-D3 antibody-coupled beads were used in a single round for depletion for 60 min at 4°C. For double depletion of condensin I and II, a single round of depletion using 1.5 volume each of xCAP-G and xCAP-D3 antibody-coupled beads was performed. For TopoII depletions (ΔTOP2A), a single round of depletion was performed using 1.2 volume of anti-TopoIIα coupled antibody beads for 60 min at 4°C. After the incubations, the beads were separated using a magnet.
Western blots
For total egg extract samples, 1 µl sample was added to 25 μl 1× sample buffer (50 mM Tris-HCl pH 6.8, 2% SDS, 10% glycerol, 2.5% β-mercaptoethanol) and boiled for 10 min. Samples were spun at 8000 rpm for 3 min before gel electrophoresis and overnight transfer at 4°C. Blotting membranes were blocked with 4% powdered skim-milk (Difco). Primary and secondary antibodies were diluted in LI-COR Odyssey blocking buffer-PBS (LI-COR Biotechnology). Western blots were imaged on a LI-COR Odyssey. Quantifications were done using ImageJ.
Hi-CStandard samples
106X. laevis sperm nuclei were added to 150 µl interphase extract and allowed to replicate at 21°C for 90 min. The extracts were cycled back into mitosis by adding 100 µl CSF extract, 40 nM of the non-degradable cyclin BΔ90 and 10 µM nocodazole (Sigma). After 60 min at metaphase, the samples were diluted into 12 ml of fixing solution (80 mM K-PIPES pH 6.8, 1 mM MgCl2, 1 mM EGTA, 30% glycerol, 0.1% Triton X-100, 1% formaldehyde) and incubated at room temperature with rocking for 10 min. The samples were then quenched with 690 µl 2.5 M glycine for 5 min at room temperature. The samples were then placed on ice for 15 min and then centrifuged at 6000 g at 4°C for 20 min. The pellet was then resuspended in 1 ml ice-cold DPBS. The tube was then centrifuged again at 13,000 g for 20 min at 4°C. The buffer was aspirated, and the pellet was frozen in liquid nitrogen and then stored at –80°C.
Dispersed chromosome samples
The metaphase chromosome samples were prepared as above, but nocodazole was omitted. The metaphase extracts were diluted by adding 1.2 ml chromosome dilution buffer (10 mM K-HEPES pH 8, 200 mM KCl, 0.5 mM EGTA, 0.5 mM MgCl2, 250 mM sucrose) and incubated at room temperature for 8 min. 6 ml fixation buffer (5 mM K-HEPES pH 8, 0.1 mM EDTA, 100 mM NaCl, 2 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 0.5% Triton X-100, 20% glycerol, 1% formaldehyde) was added to the tube, mixed by rotation 10 min at room temperature. 420 μl 2.5 M glycine was added to quench the formaldehyde, and the mixture was incubated for 5 min at room temperature. The samples were then placed on ice for 15 min and then centrifuged at 6500 g at 4°C for 20 min. The pellet was then resuspended in 1 ml ice-cold DPBS. The tube was then centrifuged again at 13,000 g for 20 min at 4°C. The buffer was aspirated, and the pellet was frozen in liquid nitrogen and then stored at –80°C.
Two biological replicates were performed for each sample, and they confirmed similar behavior among the replicates.
Library prep and sequencing
Hi-C protocol was performed as previously described (Belaghzal et al., 2017), with the exception that cell disruption by douncing was omitted. Briefly, pellets were digested by DpnII overnight at 37°C prior to biotin fill-in with biotin-14-dATP for 4 hr at 23°C. After ligation at 16°C for 4 hr, crosslinking was reversed by proteinase K at 65°C overnight. Purified ligation products were sonicated with 200 bp average size, followed by 100–350 bp size selection. End repair was performed on size-selected ligation products, prior to purifying biotin tagged DNA fragments with streptavidin beads. A-tailing was done on the purified DNA fragments followed by Illumina Truseq adapter ligation. Hi-C library was finished by PCR amplification and purification to remove PCR primers. Final library was sequenced on Illumina HiSeq 4000 with PE50.
Hi-C data processing
Hi-C fastq files were mapped to the X. laevis 9.2 genome with the distiller-nf pipeline (https://github.com/open2c/distiller-nf, Flyamer, 2021). The reads were aligned with bwa-mem, afterwards duplicate reads were filtered out. These valid pair reads were aggregated in genomic bins of 10, 25, 50, 100, 250, and 500 kb using the cooler format (Abdennur and Mirny, 2020). Cooler files were balanced using Iterative balancing correction (Imakaev et al., 2012), ignoring first two diagonals to avoid artifacts within the first bin such as re-ligation products. Contact heatmaps from balanced cooler files were viewed and exported with Higlass (Kerpedjiev et al., 2018).
Contact probability (<italic>P</italic>(<italic>s</italic>)) and derivatives
Contacts probability was calculated by contact frequency (P) as function of genomic distance (s). Interaction pairs were selected for genomic distance from 1 kb till 100 Mb binned at log-scale. Within each genomic bin, observed numbers of interactions were divided by total possible number of interactions within the bin. Distance decay plots were normalized by total number interactions, and derivative plots were made from corresponding P(s). The derivative plots plotted in Figure 4C and D, Figure 4—figure supplement 1D were drawn using LOESS smoothing.
Immunofluorescence
Immunofluorescence was performed according to previously published protocols (Desai et al., 1998). 10 µl metaphase extracts containing chromosomes were diluted into 2 ml of fixing solution (80 mM K-PIPES pH 6.8, 1 mM MgCl2, 1 mM EGTA, 30% glycerol, 0.1% Triton X-100, 2% formaldehyde) and incubated at room temperature for 7 min. The fixed chromosomes were then laid onto a cushion (80 mM K-PIPES pH 6.8, 1 mM MgCl2, 1 mM EGTA, 50% glycerol) with a coverslip placed at the bottom of the tube and centrifuged at 5000 g for 15 min at 18°C in a swinging bucket rotor. The coverslips were recovered and fixed with methanol (–20°C) for 4 min. The coverslips were then blocked overnight with antibody dilution buffer (50 mM Tris-Cl pH 7.5, 150 mM NaCl, 2% BSA). Primary and secondary antibodies were diluted in antibody dilution buffer and sealed in Prolong Gold AntiFade mounting media (Thermo Fisher Scientific).
For coverslips stained with Alexa488-anti-CAP-G antibody (Figures 1C, D, 3E, 6D, Figure 3—figure supplements 1C and 2D, Figure 4—figure supplement 2B and Figure 6D and E), coverslips stained with primary and secondary antibodies were washed three times with PBS-T (1× PBS + 0.5% Tween-20). Then, they were blocked with 100 µg/ml rabbit IgG or 30 min and were incubated with Alexa488-anti-xCAP-G antibody without any washing steps in between. The coverslips were then washed three times with PBS-T and then sealed in Prolong Gold AntiFade mounting media (Thermo Fisher Scientific).
Chromosome individualization
Chromosomes from each nucleus often remain clustered in metaphase crude egg extracts. To disperse these clustered chromosomes, extracts containing chromosomes were diluted following a method described before with some modifications (Funabiki and Murray, 2000). 40 µl Chromosome Dilution Buffer (10 mM K-HEPES pH 8, 200 mM KCl, 0.5 mM EGTA, 0.5 mM MgCl2, 250 mM sucrose) was added to 10 µl metaphase extract containing chromosomes and incubated at room temperature for 8 min. 200 µl fixation buffer (5 mM K-HEPES pH 8, 0.1 mM EDTA, 100 mM NaCl, 2 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 0.5% Triton X-100, 20% glycerol, 2% formaldehyde) was added to the tube and incubated for 10 min at room temperature. The samples were laid over a cushion (5 mM K-HEPES pH 8, 0.1 mM EDTA, 100 mM NaCl, 2 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 50% glycerol) with a coverslip placed under the cushion and centrifuged at 7000 g for 20 min at 18°C in a swinging bucket rotor. The coverslips were recovered and fixed with ice-cold methanol for 4 min, washed extensively and blocked overnight with antibody dilution buffer (50 mM Tris-Cl pH 7.5, 150 mM NaCl, 2% BSA). CENP-A immunofluorescence was performed on these coverslips for Figures 5D, E, 6B, C, Figure 6—figure supplement 1A and B.
Chromosome purification
One volume of metaphase extracts with ~3000/µl sperm nuclei was diluted into 3 volumes of DB2 (10 mM K-HEPES, 50 mM β-glycerophosphate, 50 mM NaF, 20 mM EGTA, 2 mM EDTA, 0.5 mM spermine, 1 mM phenylmethylsulfonyl fluoride, 200 mM sucrose) and laid over 1 ml cushion (DB2 with 50% sucrose). The tube was centrifuged in a swinging bucket rotor at 10,000 g for 30 min at 4°C. Most of the cushion was aspirated and the pellet was resuspended in the remaining solution and transferred to a fresh tube. The sample was centrifuged again at 13,000 g for 15 min at 4°C. The pellet was then resuspended in 1× sample buffer and boiled for 10 min before being subject to gel electrophoresis.
Image acquisition and analysis
All the quantitative immunofluorescence imaging and some of the spindle imaging was performed on a DeltaVision Image Restoration microscope (Applied Precision), which is a wide-field inverted microscope equipped with a pco.edge sCMOS camera (pco). The immunofluorescence and surface area measurement samples were imaged with z-sections of 200 nm width with a 100× (1.4 NA) objective and were processed with a iterative processive deconvolution algorithm using the SoftWoRx (Applied Precision). The dispersed chromosomes imaged for length measurements and chromosome individualization were imaged in five 1 µm z-sections with a 63× (1.33 NA) silicone oil objective.
More than 20 nuclei imaged for immunofluorescence and three-dimensional surface area quantification for most of the experiments for each condition. The differences in each experiment were analyzed using a two-tailed Mann–Whitney U-test. The data in Figures 1D and G and 6D and E, 7E and F, Figure 7—figure supplement 1B were combined data from multiple experiments, where the data were normalized to the control (ΔIgG) levels, whose medians were normalized to 1 for all conditions (Lord et al., 2020). The aggregate data were then analyzed using an unpaired Student’s t-test.
For all the immunofluorescence quantifications, the maximum intensity single slice was selected, background subtraction was performed, and average intensities were calculated on a mask generated using the DNA signal. The analysis was performed using custom MATLAB (MathWorks) code available at https://github.com/pavancss/PC2021_microscopy, copy archived at swh:1:rev:bda651f09c81eb4233b0bd078e9180166bf6d6ad (Choppakatla, 2021).
For surface area measurements, images were interpolated into stacks of 67 nm width. A surface mask was built in three-dimensional space and surface area and DNA signal was calculated using the regionprops3 MATLAB function. Only large objects (>10,000 pixel3) were analyzed to compare significant fractions of each nucleus. The analysis was done automatically using custom MATLAB (MathWorks) code available at https://github.com/pavancss/PC2021_microscopy.
For the CENP-A foci counting in Figures 5E and 6C, Figure 5—figure supplement 1E, and Figure 6—figure supplement 1A,B, DNA and CENP-A were segmented by Otsu’s thresholding algorithm. Each independent object in a binarized DNA image was treated as a single chromosomal mass and CENP-A foci were counted in each mass. This was done using custom MATLAB code available at https://github.com/pavancss/PC2021_microscopy.
For the categorization of unindividualized chromosomes in Figure 6—figure supplement 1D and Figure 7C, a large area of a coverslip was imaged in panels and all the observed DNA masses were counted and categorized as in Figure 6—figure supplement 1C in an unblinded fashion. The numbers of DNA masses counted in each condition in Figure 7C were: ΔIgG-100% TOP2A (201, 128, 146), ΔIgG-25% TOP2A (155, 177, 144), ΔH1-100% TOP2A (447, 135, 208), and ΔH1-25% TOP2A (232, 187, 171).
For chromosome length measurements, >54 chromosomes were measured in each experiment to ensure that a relatively even sampling of the 18 different chromosomes of each sperm nucleus was possible. Chromosome length measurements were done by manually tracing the chromosomes on a single maximum intensity slice in ImageJ 1.52p.
Mononucleosomes and nucleosome arrays
Nucleosome arrays were prepared as previously noted (Guse et al., 2011; Zierhut et al., 2014). The plasmid pAS696, which contains 19 repeats of the Widom 601 nucleosome position sequence (Lowary and Widom, 1998), was digested with EcoRI, XbaI, HaeII, and DraI. The fragment containing the array was isolated using polyethylene glycol-based precipitation. The ends of the DNA fragment were filled in with dATP, dGTP, dCTP, and Bio-16-dUTP (Chemcyte) using Klenow DNA polymerase (NEB) and purified using Sephadex G-50 Nick columns (Cytiva Biosciences).
Mononucleosomal DNA were prepared by digesting pAS696 using AvaI. The 196 bp fragment was isolated using polyethylene glycol-based precipitation. The ends of the fragment were filled in with dATP, dGTP, dTTP, and Alexa647-aha-dCTP (Thermo Fisher Scientific) using Klenow DNA polymerase (NEB) and purified using Sephadex G-50 Nick columns (Cytiva Biosciences).
For nucleosome deposition, 10 µg of DNA arrays or mononucleosomal DNA was mixed with equimolar amount of X. laevis H3-H4 tetramer and twice equimolar amount of X. laevis H2A-H2B dimers in 1× TE with 2 M NaCl. The mixture was added into in a Slide-A-Lyzer dialysis cassette (Thermo Fisher Scientific) and placed into 500 ml high salt buffer (10 mM Tris-Cl pH 7.5 @ 4°C, 2 M NaCl, 1 mM EDTA, 5 mM β-mercaptoethanol, 0.01% Triton X-100). Salt was reduced in a gradient by pumping in 2 l of low salt buffer (10 mM Tris-Cl pH 7.5 at 4°C, 100 mM NaCl, 1 mM EDTA, 5 mM β-mercaptoethanol, 0.01% Triton X-100) at constant volume at 1 ml/min. The quality of the nucleosome arrays was ascertained by digesting the nucleosome arrays with AvaI overnight in low magnesium buffer (5 mM potassium acetate, 2 mM Tris-acetate, 0.5 mM magnesium acetate, 1 mM DTT, pH 7.9) and electrophoresed in a 5% polyacrylamide gel made in 0.5× TBE (45 mM Tris-borate, 1 mM EDTA). The mononucleosomes were assayed by direct electrophoresis.
Nucleosome-binding assays
Nucleosome arrays were bound to M280 Streptavidin Dynabeads (Thermo Fisher Scientific) in chromatin bead binding buffer (50 mM Tris-Cl pH 8, 150 mM NaCl, 0.25 mM EDTA, 0.05% Triton X-100, 2.5% polyvinylalcohol) by shaking at 1300 rpm for 3.5 hr. To block the Step tagged condensin complexes from binding the unconjugated streptavidin on the beads during the condensin pull downs, the beads were washed once in chromatin binding buffer and then incubated in 1 mM biotin in chromatin-binding buffer by shaking at 1300 rpm for 1 hr. The beads were then washed with chromatin-binding buffer (50 mM Tris-Cl pH 8, 150 mM NaCl, 0.25 mM EDTA, 0.05% Triton X-100) three times, moved to a new tube, washed twice with SDB (10 mM HEPES, 1 mM MgCl2, 100 mM KCl, 150 mM sucrose), and split into two tubes. SDB with 0.0008% poly-glutamic acid (Sigma; Stein and Künzler, 1983) was mixed with 400 nM recombinant xH1.8 (buffer for control) and incubated for 5 min at room temperature. This mixture was incubated with the beads (half with buffer, half with xH1.8) with rotation at 16°C. The beads were then washed 1× with SDB and 1× with binding buffer (10 mM HEPES pH 8, 40 mM NaCl, 2.5 mM MgCl2, 0.5 mM DTT, 0.05% Triton X-100). Beads were washed 2× with binding buffer with indicated assay salt concentration and resuspended in binding buffer with 100 nM recombinant TOP2A, 380 nM human condensin I, condensin I Q loop mutant, or 320 nM condensin II or condensin II Q loop mutant. The beads were rotated at room temperature for 30 min. Total reaction samples were taken, and the beads were washed three times on a magnet in binding buffer and moved to a new tube. The beads were collected on a magnet and resuspended in 1× sample buffer (50 mM Tris-HCl pH 6.8, 2% SDS, 10% glycerol, 2.5% β-mercaptoethanol) and boiled for 5 min. Gel electrophoresis was performed, and the gels were stained with GelCode Blue Stain reagent (Thermo Fisher Scientific).
Condensin gel shift assays
200 nM Alexa647 labeled 196 bp mononucleosomes were mixed with 0.0008% poly-glutamic acid (Sigma; Stein and Künzler, 1983) and half was mixed with 400 nM recombinant xH1.8 in 1× binding buffer (10 mM HEPES pH 8, 50 mM NaCl, 2.5 mM MgCl2, 5 mM ATP, 0.5 mM DTT, 0.05% Triton X-100) and incubated for 30 min at room temperature. 100 nM of the mononucleosomes with or H1.8 were mixed with the indicated concentration of condensin I in 1× binding buffer at 4°C for 30 min and subject to electrophoresis onto a 5% polyacrylamide gel in 0.5× TBE at room temperature. The gels were imaged on a LI-COR Odyssey (LI-COR Biotechnology). The binding curves were fitted using GraphPad Prism 8.4.3 using the sigmoidal binding curve option of the nonlinear curve fitting.
Protein purificationH1.8
A pET51b vector expressing X. laevis H1.8 with an N-terminal Strep-Tag II and C-terminal 6× Histidine-tag was a gift from Rebecca Heald (UC Berkeley). E. coli Rosetta2 (DE3 pLysS) cells containing expression plasmids were grown in TBG-M9 media (15 g/l tryptone, 7.5 g/l yeast extract, 5 g/l NaCl, 0.15 g/l MgSO4, 1.5 g/l NH4Cl, 3 g/l KH2PO4, 6 g/l Na2HPO4; 0.4% glucose) at 37°C until they reach OD ~0.6 and were supplemented with 1 mM isopropylthio-β-galactoside (IPTG) and grown at 18°C for 14 hr. Cells were collected and resuspend in lysis buffer (1× PBS, 500 mM NaCl, 10% glycerol, 20 mM imidazole, 0.1% Triton X-100, 10 mM β-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 10 µg/ml leupeptin, 10 µg/ml pepstatin, 10 µg/ml chymostatin). All subsequent steps were carried out at 4°C. After 30 min incubation, the cell suspension was sonicated and centrifuged at 45,000 g for 45 min at 4°C. The supernatant was added to Ni-NTA beads (Bio-Rad) and rotated for 60 min. The beads were then washed with Wash Buffer 1 (1× PBS, 20 mM imidazole, 500 mM NaCl, 4 mM β-mercaptoethanol, 10 mM ATP, 2.5 mM MgCl2, cOmplete EDTA-free protease inhibitor cocktail; Roche). The beads were eluted with NTA elution buffer (1× PBS, 400 mM imidazole, 500 mM NaCl). The correct fractions were collected and dialyzed into PBS supplement with 500 mM NaCl, concentrated using Amicon Ultra centrifugal filters (10k cutoff), flash frozen, aliquoted, and stored at –80°C.
TopoIIα
X. laevis TOP2A tagged with calmodulin-binding protein (CBP) was purified from Pichia pastoris yeast as reported (Ryu et al., 2010) with some modifications. P. pastoris integrated with a CBP tagged TOP2A cassette under the influence of an alcohol oxidase (AOX) promoter (a gift from Yoshiaki Azuma) were grown in BMGY media (1% yeast extract, 2% peptone, 100 mM potassium phosphate pH 6, 1.34% yeast nitrogen base, 4 × 10–5 % biotin, 1% glycerol) containing 50 µg/ml G418 (Thermo Fisher Scientific) at 30°C until OD ~ 4.0. The cells were collected by centrifugation and split into BMMY media (1% yeast extract, 2% peptone, 100 mM potassium phosphate pH 6, 1.34% yeast nitrogen base, 4 × 10–5 % biotin, 0.5% methanol) and grown at 22°C for 14 hr. The cells were collected, packed into a syringe, and extruded into liquid nitrogen in the form of noodles. These frozen noodles were lysed using a Retsch PM100 cryomill (Retsch) with continuous liquid nitrogen cooling. The cyromilled cells were then resuspended in Lysis Buffer (150 mM NaCl, 18 mM β-glycerophosphate, 1 mM MgCl2, 40 mM HEPES [pH 7.8], 5% glycerol, 0.1% Triton X-100, 1 mM DTT, cOmplete EDTA-free protease inhibitor tablet) and sonicated on ice. The cells were centrifuged at 35,000 g for 45 min at 4°C. 2 mM CaCl2 was added to the supernatant along with calmodulin-sepharose beads (Strategene) and the mixture was incubated at 4°C for 120 min. The beads were then washed with ATP-Wash Buffer (Lysis Buffer + 5 mM MgCl2, 2 mM CaCl2, 1 mM ATP), Wash Buffer 1 (Lysis Buffer + 2 mM CaCl2), Wash Buffer 2 (300 mM NaCl, 1 mM MgCl2, 2 mM CaCl2, 20 mM HEPES [pH 7.8], 5% glycerol, 1 mM DTT) and then eluted into elution buffer (300 mM NaCl, 1 mM MgCl2, 5 mM EGTA, 20 mM HEPES [pH 7.8], 5% glycerol, 1 mM DTT).
The eluted protein was then passed through a MonoQ anion exchange column (Cytiva) on an AKTA-FPLC (Cytiva) to separate co-purified DNA. The flowthrough was then digested with TEV protease to cleave the CBP tag and then loaded on a HiTrap Heparin HP column (Cytiva) on an AKTA-FPLC and eluted using a salt gradient of 150 mM NaCl to 1 M NaCl. The selected fractions were then loaded on a Superose 6 gel filtration column (Cytiva) and eluted in freezing buffer (250 mM NaCl, 1 mM MgCl2, 20 mM HEPES pH 7.8, 5% glycerol, 1 mM DTT). The protein was then concentrated and frozen in aliquots at –80°C.
Condensins
Human condensin complexes were purified as described previously (Kong et al., 2020). Briefly, the five subunits of human condensin I and II, sub-complexes, and Q-loop mutations and were assembled into biGBac vectors (Weissmann et al., 2016) to create baculovirus for protein expression in HighFive insect cells. Cell were lysed in condensin purification buffer (20 mM HEPES [pH 8], 300 mM KCl, 5 mM MgCl2, 1 mM DTT, 10% glycerol) supplemented with Pierce protease inhibitor EDTA-free tablet (Thermo Scientific) and Benzonase (Sigma). Cleared lysate was loaded on to a StrepTrap HP (GE), washed with condensin purification buffer, and eluted with condensin purification buffer supplemented with 5 mM Desthiobiotin (Sigma). Protein-containing fractions were pooled, diluted twofold with Buffer A (20 mM HEPES [pH 8], 5 mM MgCl2, 5% glycerol, 1 mM DTT), loaded on to HiTrap Heparin HP column (GE), washed with Buffer A with 250 mM NaCl, then eluted with buffer A with 500 mM NaCl. Finally, size-exclusion chromatography was performed using condensin purification buffer and a Superose 6 16/70 or increase 10/300 column (GE).
Mass photometry
All mass photometry data were taken using a Refeyn OneMP mass photometer (Refeyn Ltd). Movies were acquired for 10,000 frames (100 s) using AcquireMP software (version 2.4.0) and analyzed using DiscoverMP software (version 2.4.0, Refeyn Ltd), all with default settings. Proteins were measured by adding 1 µl of stock solution (50 nM) to a 10 µl droplet of filtered buffer (10 mM HEPES pH 8, 2.5 mM MgCl2, 1 mM DTT, 50–300 mM NaCl, 5 mM ATP). Contrast measurements were converted to molecule weights using a standard curve generated with bovine serine albumin (Thermo 23210) and urease (Sigma U7752).
Mass spectrometry
Sperm chromosomes were purified as previously described (Funabiki and Murray, 2000). Briefly, extracts containing 8000/µl sperm nuclei were replicated along with 5 mM biotin-dUTP for 90 min and cycled back into metaphase with the addition of 1 volume of fresh CSF depleted (correspondingly ΔIgG or ΔH1). After 60 min in metaphase, these chromosomes were diluted in 3 volumes of DB (10 mM K-HEPES [pH 7.6], 100 mM KCl, 2 mM EDTA, 0.5 mM EGTA, 0.5 mM spermine, 250 mM sucrose, 1 mM PMSF, and 10 µg/ml each of leupeptin, pepstatin, and chymostatin) and centrifuged through a 60 DB cushion (DB with 60% [w/v] sucrose). The collected chromosome-enriched pellet was then incubated with 15 µl streptavidin-coupled Dynabeads (M280) and rotated at 4°C for 2 hr. The beads were then collected, washed, and boiled in sample buffer before running on a 6% polyacrylamide gel for 10 min. The gel was stained with Commassie blue, and the protein-containing gel fragments were cut out and processed for mass spectrometry. The mass spectrometry was performed at the Rockefeller University Proteomics Resource Center as previously described (Zierhut et al., 2014), but the peptides were queried against the X. laevis database (Wühr et al., 2014) using the MaxQuant software (Max-Planck Institute).
Additional informationCompeting interests
none
Reviewing editor, eLife
Author contributions
Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Supervision, Validation, Visualization, Writing - original draft, Writing - review and editing
Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Visualization, Writing - review and editing
Resources, Writing - review and editing
Resources, Supervision, Writing - review and editing
Data curation, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing - review and editing
Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Validation, Visualization, Writing - original draft, Writing - review and editing
Ethics
This study was performed in strict accordance with the care standards provided by the 8th edition of the Guide for the Care and Use of Laboratory Animals. African clawed frogs, Xenopus laevis, which were maintained and handled according to approved institutional animal care and use committee (IACUC) protocol (20031) of the Rockefeller University, which is an Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC) accredited research facility.
Additional filesData availability
Hi-C sequencing data have been deposited in GEO under an accession code GSE164434. All other data generated or analyzed during this study are included in the manuscript and supporting source data files.
The following dataset was generated:
ChoppakatlaPDekkerBCuttsEEVanniniADekkerJFunabikiH2021Linker histone H1.8 inhibits chromatin-binding of condensins and DNA topoisomerase II to tune chromosome compaction and individualizationNCBI Gene Expression OmnibusGSE164434
Acknowledgements
We thank C Zierhut for providing ΔH3-H4-depleted extracts for Hi-C analysis; S Rankin, T Hirano, Y Azuma for sharing reagents; C Jenness for anti-H1.8 antibodies; JF Martinez and MP Rout for their help with purifying TOP2A; A North, C Rico, and K Cialowicz at Bioimaging Resource Center (BIRC) for help with imaging; C Steckler and H Molina at The Rockefeller Proteomics Core Facility for help with mass spectrometry and analyzing the data; K Mickolajczyk and T Kapoor for help with mass photometry, Y Arimura, R Heald, L Mirny, A Vannini, and C Zierhut and members of the Funabiki lab for helpful discussions. This work was supported by the National Institutes of Health (NIH) grant R35 GM132111 to HF and NIH grant R01 HG003143 to JD. JD is an investigator of the Howard Hughes Medical Institute. AV is supported by a Cancer Research UK Programme Foundation (CR-UK C47547/A21536) and a Wellcome Trust Investigator Award (200818/Z/16/Z). HF is affiliated with Graduate School of Medical Sciences, Weill Cornell Medicine, and Cell Biology Program, the Sloan Kettering Institute. The authors declare that no competing financial interests exist.
ReferencesAbdennurNMirnyLA2020Cooler: Scalable Storage for Hi-C Data and Other Genomically Labeled ArraysBioinformatics3631131610.1093/bioinformatics/btz54031290943AbramoKValtonALVenevSVOzadamHFoxANDekkerJ2019A chromosome folding intermediate at the condensin-to-cohesin transition during telophaseNature Cell Biology211393140210.1038/s41556-019-0406-231685986AdachiYLukeMLaemmliUK1991Chromosome Assembly in Vitro: Topoisomerase II Is Required for CondensationCell6413714810.1016/0092-8674(91)90215-k1846085AlipourEMarkoJF2012Self-Organization of Domain Structures by DNA-Loop-Extruding EnzymesNucleic Acids Research40112021121210.1093/nar/gks92523074191ArimuraYShihRMFroomRFunabikiH2020Nucleosome structural variations in interphase and metaphase chromosomesMolecular Cell10.1016/j.molcel.2021.08.010BakhrebahMZhangTMannJRKalitsisPHudsonDF2015Disruption of a Conserved CAP-D3 Threonine Alters Condensin Loading on Mitotic Chromosomes Leading to Chromosome HypercondensationThe Journal of Biological Chemistry2906156616710.1074/jbc.M114.62710925605712BaniganEJMirnyLA2020Loop Extrusion: Theory Meets Single-Molecule ExperimentsCurrent Opinion in Cell Biology6412413810.1016/j.ceb.2020.04.01132534241BauerCRHartlTABoscoG2012Condensin II Promotes the Formation of Chromosome Territories by Inducing Axial Compaction of Polyploid Interphase ChromosomesPLOS Genetics8e100287310.1371/journal.pgen.100287322956908BaxterJSenNLópez MartínezVMonturus De CarandiniMESchvartzmanJBDiffleyJFXAragónL2011Positive Supercoiling of Mitotic DNA Drives Decatenation by Topoisomerase II in EukaryotesScience3311328133210.1126/science.1201538BednarJGarcia-SaezIBoopathiRCutterARPapaiGReymerASyedSHLoneINTonchevOCrucifixCMenoniHPapinCSkoufiasDAKurumizakaHLaveryRHamicheAHayesJJSchultzPAngelovDPetosaCDimitrovS2017Structure and Dynamics of a 197 bp Nucleosome in Complex with Linker Histone H1Molecular Cell6638439710.1016/j.molcel.2017.04.01228475873BelaghzalHDekkerJGibcusJH2017Hi-c 2.0: An optimized HI-C procedure for high-resolution genome-wide mapping of chromosome conformationMethods123566510.1016/j.ymeth.2017.04.00428435001BernadRSánchezPRiveraTRodríguez-CorsinoMBoyarchukEVassiasIRay-GalletDArnaoutovADassoMAlmouzniGLosadaA2011Xenopus HJURP and CONDENSIN II are required for CENP-A assemblyThe Journal of Cell Biology19256958210.1083/jcb.20100513621321101BrahmachariSMarkoJF2019Chromosome Disentanglement Driven via Optimal Compaction of Loop-Extruded Brush StructuresPNAS116249562496510.1073/pnas.190635511631757850BrandãoHBPaulPvan den BergAARudnerDZWangXMirnyLA2019RNA Polymerases as Moving Barriers to Condensin Loop ExtrusionPNAS116204892049910.1073/pnas.190700911631548377CanelaAMamanYJungSWongNCallenEDayAKieffer-KwonKRPekowskaAZhangHRaoSSPHuangSCMckinnonPJAplanPDPommierYAidenELCasellasRNussenzweigA2017Genome organization drives chromosome fragilityCell17050752110.1016/j.cell.2017.06.03428735753CanelaAMamanYHuangSYNWutzGTangWZagnoli-VieiraGCallenEWongNDayAPetersJMCaldecottKWPommierYNussenzweigA2019Topoisomerase ii-induced chromosome breakage and translocation is determined by chromosome architecture and transcriptional activityMolecular Cell7525226610.1016/j.molcel.2019.04.03031202577CavazzaTTakedaYPolitiAZAushevMAldagPBakerCChoudharyMBucevičiusJLukinavičiusGElderKBlayneyMLucas-HahnANiemannHHerbertMSchuhM2021Parental genome unification is highly error-prone in mammalian embryosCell1842860287710.1016/j.cell.2021.04.01333964210CherejiRVBrysonTDHenikoffS2019Quantitative MNase-Seq Accurately Maps Nucleosome Occupancy LevelsGenome Biology2019810.1186/s13059-019-1815-z31519205ChoppakatlaP2021Pc2021_microscopyswh:1:rev:bda651f09c81eb4233b0bd078e9180166bf6d6adSoftware Heritagehttps://archive.softwareheritage.org/swh:1:dir:3496b1ef4ea3cf985b256ffe117689fcf24e0e8e;origin=https://github.com/pavancss/PC2021_microscopy;visit=swh:1:snp:11e8c455b9e383d56a784e6e94d88e30bb63025d;anchor=swh:1:rev:bda651f09c81eb4233b0bd078e9180166bf6d6adChristophorouMACastelo-BrancoGHalley-StottRPOliveiraCSLoosRRadzisheuskayaAMowenKABertonePSilvaJCRZernicka-GoetzMNielsenMLGurdonJBKouzaridesT2014Citrullination regulates pluripotency and histone h1 binding to chromatinNature50710410810.1038/nature1294224463520CremerTCremerM2010Chromosome TerritoriesCold Spring Harbor Perspectives in Biology2a00388910.1101/cshperspect.a00388920300217CuttsEEVanniniA2020Condensin Complexes: Understanding Loop Extrusion One Conformational Change at a TimeBiochemical Society Transactions482089210010.1042/BST2020024133005926CuvierOHiranoT2003A Role of Topoisomerase II in Linking DNA Replication to Chromosome CondensationThe Journal of Cell Biology16064565510.1083/jcb.20020902312604590CuylenSBlaukopfCPolitiAZMüller-ReichertTNeumannBPoserIEllenbergJHymanAAGerlichDW2016Ki-67 Acts as a Biological Surfactant to Disperse Mitotic ChromosomesNature53530831210.1038/nature1861027362226DassoMDimitrovSWolffeAP1994Nuclear Assembly Is Independent of Linker HistonesPNAS91124771248110.1073/pnas.91.26.124777809061DattaSLecomteLHaeringCH2020Structural Insights into DNA Loop Extrusion by SMC Protein ComplexesCurrent Opinion in Structural Biology6510210910.1016/j.sbi.2020.06.00932674008DesaiAMurrayAWMitchisonTWalczakCE1998Chapter 20 the use of Xenopus egg extracts to study mitotic spindle assembly and function in vitroMethods in Cell Biology6138541210.1016/S0091-679X(08)61991-3DuZZhengHHuangBMaRWuJZhangXHeJXiangYWangQLiYMaJZhangXZhangKWangYZhangMQGaoJDixonJRWangXZengJXieW2017Allelic reprogramming of 3d chromatin architecture during early mammalian developmentNature54723223510.1038/nature2326328703188Dworkin-RastlEKandolfHSmithRC1994The maternal histone h1 variant, h1m (b4 protein), is the predominant h1 histone in Xenopus pregastrula embryosDevelopmental Biology104210.1006/dbio.1994.1042EarnshawWCLaemmliUK1983Architecture of Metaphase Chromosomes and Chromosome ScaffoldsThe Journal of Cell Biology96849310.1083/jcb.96.1.846826654EdwardsNSMurrayAW2005Identification of Xenopus CENP-A and an Associated Centromeric DNA RepeatMolecular Biology of the Cell161800181010.1091/mbc.e04-09-078815673610ElbatshAMOKimEEeftensJMRaaijmakersJAvan der WeideRHGarcía-NietoABravoSGanjiMUit de BosJTeunissenHMedemaRHde WitEHaeringCHDekkerCRowlandBD2019Distinct Roles for Condensin’s Two ATPase Sites in Chromosome CondensationMolecular Cell7672473710.1016/j.molcel.2019.09.02031629658EltsovMMaclellanKMMaeshimaKFrangakisASDubochetJ2008Analysis of cryo-electron microscopy images does not support the existence of 30-nm chromatin fibers in mitotic chromosomes in situPNAS105197321973710.1073/pnas.081005710519064912FarcasA-MUluocakPHelmhartWNasmythK2011Cohesin’s concatenation of sister DNAs maintains their intertwiningMolecular Cell449710710.1016/j.molcel.2011.07.03421981921FarrCJAntoniou-KourouniotiMMimmackMLVolkovAPorterACG2014The α isoform of topoisomerase II is required for hypercompaction of mitotic chromosomes in human cellsNucleic Acids Research424414442610.1093/nar/gku07624476913FinchJTKlugA1976Solenoidal Model for Superstructure in ChromatinPNAS731897190110.1073/pnas.73.6.18971064861Fitz-JamesMHTongPPidouxALOzadamHYangLWhiteSADekkerJAllshireRC2020Large Domains of Heterochromatin Direct the Formation of Short Mitotic Chromosome LoopseLife9e5721210.7554/eLife.5721232915140FlyamerI2021Distiller-nfv0.3.3Githubhttps://github.com/open2c/distiller-nfFreedmanBSHealdR2010Functional Comparison of H1 Histones in Xenopus Reveals Isoform-Specific Regulation by Cdk1 and RanGTPCurrent Biology201048105210.1016/j.cub.2010.04.02520471264FunabikiHMurrayAW2000The Xenopus Chromokinesin Xkid Is Essential for Metaphase Chromosome Alignment and Must Be Degraded to Allow Anaphase Chromosome MovementCell10241142410.1016/S0092-8674(00)00047-710966104GalliMMorganDO2016Cell Size Determines the Strength of the Spindle Assembly Checkpoint during Embryonic DevelopmentDevelopmental Cell3634435210.1016/j.devcel.2016.01.00326859356GandhiRGillespiePJHiranoT2006Human Wapl Is a Cohesin-Binding Protein That Promotes Sister-Chromatid Resolution in Mitotic ProphaseCurrent Biology162406241710.1016/j.cub.2006.10.06117112726GanjiMShaltielIABishtSKimEKalichavaAHaeringCHDekkerC2018Real-Time Imaging of DNA Loop Extrusion by CondensinScience36010210510.1126/science.aar783129472443GasslerJBrandãoHBImakaevMFlyamerIMLadstätterSBickmoreWAPetersJ-MMirnyLATachibanaK2017A mechanism of cohesin-dependent loop extrusion organizes zygotic genome architectureThe EMBO Journal363600361810.15252/embj.20179808329217590GerhartJWuMKirschnerM1984Cell Cycle Dynamics of an M-Phase-Specific Cytoplasmic Factor in Xenopus laevis Oocytes and EggsThe Journal of Cell Biology981247125510.1083/jcb.98.4.12476425302GibcusJHSamejimaKGoloborodkoASamejimaINaumovaNNueblerJKanemakiMTXieLPaulsonJREarnshawWCMirnyLADekkerJ2018A pathway for mitotic chromosome formationScience359637610.1126/science.aao613529348367GibsonBADoolittleLKSchneiderMWGJensenLEGamarraNHenryLGerlichDWReddingSRosenMK2019Organization of Chromatin by Intrinsic and Regulated Phase SeparationCell17947048410.1016/j.cell.2019.08.03731543265GlotzerMMurrayAWKirschnerMW1991Cyclin is degraded by the ubiquitin pathwayNature34913213810.1038/349132a01846030GolfierSQuailTKimuraHBruguésJ2020Cohesin and condensin extrude DNA loops in a cell cycle-dependent mannereLife9e5388510.7554/eLife.5388532396063GoloborodkoAImakaevMVMarkoJFMirnyLA2016aCompaction and Segregation of Sister Chromatids via Active Loop ExtrusioneLife5e1486410.7554/eLife.1486427192037GoloborodkoAMarkoJFMirnyLA2016bChromosome Compaction by Active Loop ExtrusionBiophysical Journal1102162216810.1016/j.bpj.2016.02.04127224481GoundaroulisDLieberman AidenEStasiakA2020Chromatin Is Frequently Unknotted at the Megabase ScaleBiophysical Journal1182268227910.1016/j.bpj.2019.11.00231818464GreenLCKalitsisPChangTMCipeticMKimJHMarshallOTurnbullLWhitchurchCBVagnarelliPSamejimaKEarnshawWCChooKHAHudsonDF2012Contrasting roles of condensin I and condensin II in mitotic chromosome formationJournal of Cell Science1251591160410.1242/jcs.09779022344259GrigoryevSABascomGBuckwalterJMSchubertMBWoodcockCLSchlickT2016Hierarchical Looping of Zigzag Nucleosome Chains in Metaphase ChromosomesPNAS1131238124310.1073/pnas.151828011326787893GruhnJRZielinskaAPShuklaVBlanshardRCapalboACimadomoDNikiforovDChanACHNewnhamLJVogelIScaricaCKrapchevMTaylorDKristensenSGChengJErnstEBjørnAMBColmornLBBlayneyMElderKLissJHartshorneGGrøndahlMLRienziLUbaldiFMcCoyRLukaszukKAndersenCYSchuhMHoffmannER2019Chromosome errors in human eggs shape natural fertility over reproductive life spanScience3651466146910.1126/science.aav732131604276GuacciVHoganEKoshlandD1994Chromosome Condensation and Sister Chromatid Pairing in Budding YeastThe Journal of Cell Biology12551753010.1083/jcb.125.3.5178175878GuseACarrollCWMoreeBFullerCJStraightAF2011In Vitro Centromere and Kinetochore Assembly on Defined Chromatin TemplatesNature47735435810.1038/nature1037921874020HaraKTydemanPKirschnerM1980A Cytoplasmic Clock with the Same Period as the Division Cycle in Xenopus EggsPNAS7746246610.1073/pnas.77.1.4626928638HasslerMShaltielIAKschonsakMSimonBMerkelFThärichenLBaileyHJMacošekJBravoSMetzJHennigJHaeringCH2019Structural basis of an asymmetric condensin atpase cycleMolecular Cell741175118810.1016/j.molcel.2019.03.03731226277HealdRGibeauxR2018Subcellular scaling: does size matter for cell division?Current Opinion in Cell Biology52889510.1016/j.ceb.2018.02.00929501026HergethSPSchneiderR2015The H1 Linker Histones: Multifunctional Proteins beyond the Nucleosomal Core ParticleEMBO Reports161439145310.15252/embr.20154074926474902HiranoTMitchisonTJ1991Cell Cycle Control of Higher-Order Chromatin Assembly around Naked DNA in VitroThe Journal of Cell Biology1151479148910.1083/jcb.115.6.14791661728HiranoTMitchisonTJ1994A Heterodimeric Coiled-Coil Protein Required for Mitotic Chromosome Condensation in VitroCell7944945810.1016/0092-8674(94)90254-27954811HiranoTKobayashiRHiranoM1997Condensins, Chromosome Condensation Protein Complexes Containing XCAP-C, XCAP-E and a Xenopus Homolog of the Drosophila Barren ProteinCell8951152110.1016/S0092-8674(00)80233-09160743HirotaTGerlichDWKochBEllenbergJPetersJ-M2004Distinct Functions of Condensin I and II in Mitotic Chromosome AssemblyJournal of Cell Science1176435644510.1242/jcs.0160415572404HollowaySLGlotzerMKingRWMurrayAW1993Anaphase Is Initiated by Proteolysis Rather than by the Inactivation of Maturation-Promoting FactorCell731393140210.1016/0092-8674(93)90364-V8391932HopfnerKPKarcherAShinDSCraigLArthurLMCarneyJPTainerJA2000Structural biology of RAD50 atpase: Atp-driven conformational control in DNA double-strand break repair and the abc-atpase superfamilyCell10178980010.1016/S0092-8674(00)80890-910892749HsiehTHSCattoglioCSlobodyanyukEHansenASRandoOJTjianRDarzacqX2020Resolving the 3D Landscape of Transcription-Linked Mammalian Chromatin FoldingMolecular Cell7853955310.1016/j.molcel.2020.03.00232213323ImakaevMFudenbergGMcCordRPNaumovaNGoloborodkoALajoieBRDekkerJMirnyLA2012Iterative Correction of Hi-C Data Reveals Hallmarks of Chromosome OrganizationNature Methods9999100310.1038/nmeth.214822941365IzzoAKamieniarzKSchneiderR2008The Histone h1 family: Specific members, specific functions?Biological Chemistry38933334310.1515/BC.2008.03718208346IzzoAKamieniarz-GdulaKRamírezFNoureenNKindJMankeTvan SteenselBSchneiderR2013The genomic landscape of the somatic linker histone subtypes h1.1 to h1.5 in human cellsCell Reports32142215410.1016/j.celrep.2013.05.00323746450JanssenAvan der BurgMSzuhaiKKopsGJPLMedemaRH2011Chromosome Segregation Errors as a Cause of DNA Damage and Structural Chromosome AberrationsScience3331895189810.1126/science.121021421960636JennessCGiuntaSMüllerMMKimuraHMuirTWFunabikiH2018HELLS and CDCA7 Comprise a Bipartite Nucleosome Remodeling Complex Defective in ICF SyndromePNAS115E876E88510.1073/pnas.171750911529339483KerpedjievPAbdennurNLekschasFMcCallumCDinklaKStrobeltHLuberJMOuelletteSBAzhirAKumarNHwangJLeeSAlverBHPfisterHMirnyLAParkPJGehlenborgN2018Higlass: Web-based visual exploration and analysis of genome interaction mapsGenome Biology1912510.1186/s13059-018-1486-130143029KimuraKHiranoMKobayashiRHiranoT1998Phosphorylation and Activation of 13S Condensin by Cdc2 in VitroScience (New York, N.Y.)28248749010.1126/science.282.5388.487KimuraKCuvierOHiranoT2001Chromosome Condensation by a Human Condensin Complex in Xenopus Egg ExtractsThe Journal of Biological Chemistry2765417542010.1074/jbc.C00087320011136719KinoshitaKHiranoT2017Dynamic Organization of Mitotic ChromosomesCurrent Opinion in Cell Biology46465310.1016/j.ceb.2017.01.00628214612KongMCuttsEEPanDBeuronFKaliyappanTXueCMorrisEPMusacchioAVanniniAGreeneEC2020Human Condensin I and II Drive Extensive ATP-Dependent Compaction of Nucleosome-Bound DNAMolecular Cell799911410.1016/j.molcel.2020.04.02632445620KrietensteinNAbrahamSVenevSVAbdennurNGibcusJHsiehTHSParsiKMYangLMaehrRMirnyLADekkerJRandoOJ2020Ultrastructural Details of Mammalian Chromosome ArchitectureMolecular Cell7855456510.1016/j.molcel.2020.03.00332213324KrishnanSSmitsAHVermeulenMReinbergD2017Phospho-H1 Decorates the Inter-Chromatid Axis and Is Evicted along with Shugoshin by SET during MitosisMolecular Cell6757959310.1016/j.molcel.2017.07.00828781233KschonsakMHaeringCH2015Shaping Mitotic Chromosomes: From Classical Concepts to Molecular MechanismsBioEssays3775576610.1002/bies.20150002025988527LadouceurA-MDornJFMaddoxPS2015Mitotic Chromosome Length Scales in Response to Both Cell and Nuclear SizeThe Journal of Cell Biology20964565110.1083/jcb.20150209226033258LadouceurAMRanjanRSmithLFaderoTHeppertJGoldsteinBMaddoxASMaddoxPS2017CENP-A and Topoisomerase-II Antagonistically Affect Chromosome LengthThe Journal of Cell Biology2162645265510.1083/jcb.20160808428733327LaneABGiménez-AbiánJFClarkeDJ2013A Novel Chromatin Tether Domain Controls Topoisomerase IIα Dynamics and Mitotic Chromosome FormationThe Journal of Cell Biology20347148610.1083/jcb.20130304524217621LeeWTilloDBrayNMorseRHDavisRWHughesTRNislowC2007A High-Resolution Atlas of Nucleosome Occupancy in YeastNature Genetics391235124410.1038/ng211717873876LiWChenPYuJDongLLiangDFengJYanJWangPYLiQZhangZLiMLiG2016Fact remodels the tetranucleosomal unit of chromatin fibers for gene transcriptionMolecular Cell6412013310.1016/j.molcel.2016.08.02427666592Lieberman-AidenEvan BerkumNLWilliamsLImakaevMRagoczyTTellingAAmitILajoieBRSaboPJDorschnerMOSandstromRBernsteinBBenderMAGroudineMGnirkeAStamatoyannopoulosJMirnyLALanderESDekkerJ2009Comprehensive mapping of long-range interactions reveals folding principles of the human genomeScience32628929310.1126/science.118136919815776ListerLMKouznetsovaAHyslopLAKalleasDPaceSLBarelJCNathanAFlorosVAdelfalkCWatanabeYJessbergerRKirkwoodTBHöögCHerbertM2010Age-related meiotic segregation errors in mammalian oocytes are preceded by depletion of cohesin and sgo2Current Biology201511152110.1016/j.cub.2010.08.02320817533LordSJVelleKBMullinsRDFritz-LaylinLK2020SuperPlots: Communicating Reproducibility and Variability in Cell BiologyThe Journal of Cell Biology219610.1083/jcb.20200106432346721LowaryPTWidomJ1998New DNA Sequence Rules for High Affinity Binding to Histone Octamer and Sequence-Directed Nucleosome PositioningJournal of Molecular Biology276194210.1006/jmbi.1997.14949514715LöweJCordellSCvan den EntF2001Crystal Structure of the SMC Head Domain: An ABC ATPase with 900 Residues Antiparallel Coiled-Coil InsertedJournal of Molecular Biology306253510.1006/jmbi.2000.437911178891MaraLPaimGFitzharrisG2019Cell-Size-Independent Spindle Checkpoint Failure Underlies Chromosome Segregation Error in Mouse Report Cell-Size-Independent Spindle Checkpoint Failure Underlies Chromosome Segregation ErrorCurrent Biology2986587310.1016/j.cub.2018.12.042MarescaTJFreedmanBSHealdR2005Histone H1 Is Essential for Mitotic Chromosome Architecture and Segregation in Xenopus laevis Egg ExtractsThe Journal of Cell Biology16985986910.1083/jcb.20050303115967810MarkoJF2008Micromechanical Studies of Mitotic ChromosomesChromosome Research1646949710.1007/s10577-008-1233-718461485MarkoJF2009Linking Topology of Tethered Polymer Rings with Applications to Chromosome Segregation and Estimation of the Knotting LengthPhysical Review E - Statistical, Nonlinear, and Soft Matter Physics7911610.1103/PhysRevE.79.051905MicheliGLuzzattoARCarrìMTde CapoaAPellicciaF1993Chromosome Length and DNA Loop Size during Early Embryonic Development of Xenopus laevisChromosoma10247848310.1007/BF003571038375216MinshullJSunHTonksNKMurrayAW1994A MAP Kinase-Dependent Spindle Assembly Checkpoint in Xenopus Egg ExtractsCell7947548610.1016/0092-8674(94)90256-97954813MisteliTGunjanAHockRBustinMBrownDT2000Dynamic binding of histone H1 to chromatin in living cellsNature40887788110.1038/3504861011130729MurrayAW1991Chapter 30 Cell Cycle ExtractsMethods in Cell Biology3658160510.1016/S0091-679X(08)60298-8NagasakaKHossainMJRobertiMJEllenbergJHirotaT2016Sister Chromatid Resolution Is an Intrinsic Part of Chromosome Organization in ProphaseNature Cell Biology1869269910.1038/ncb335327136266NasmythKA2001Disseminating the Genome: Joining, Resolving, and Separating Sister Chromatids During Mitosis and MeiosisAnnual Review of Genetics3567374510.1146/annurev.genet.35.102401.09133411700297NaumovaNImakaevMFudenbergGZhanYLajoieBRMirnyLADekkerJ2013Organization of the mitotic chromosomeScience34294895310.1126/science.123608324200812NielsenCFZhangTBarisicMKalitsisPHudsonDF2020Topoisomerase iiα is essential for maintenance of mitotic chromosome structurePNAS117121311214210.1073/pnas.200176011732414923OhsumiKKatagiriCKishimotoT1993Chromosome Condensation in Xenopus Mitotic Extracts without Histone H1Science2622033203510.1126/science.82660998266099OnoTLosadaAHiranoMMyersMPNeuwaldAFHiranoT2003Differential contributions of condensin I and condensin II to mitotic chromosome architecture in vertebrate cellsCell11510912110.1016/S0092-8674(03)00724-414532007OnoTFangYSpectorDLHiranoT2004Spatial and Temporal Regulation of Condensins I and II in Mitotic Chromosome Assembly in Human CellsMolecular Biology of the Cell153296330810.1091/mbc.E04-03-024215146063OuHDPhanSDeerinckTJThorAEllismanMHO’SheaCC2017ChromEMT: Visualizing 3D Chromatin Structure and Compaction in Interphase and Mitotic CellsScience357eaag002510.1126/science.aag002528751582ParseghianMHNewcombRLHamkaloBA2001Distribution of Somatic H1 Subtypes Is Non-Random on Active vs. Inactive Chromatin II: Distribution in Human Adult FibroblastsJournal of Cellular Biochemistry8364365910.1002/jcb.122411746507PatelLKangRRosenbergSCQiuYRaviramRCheeSHuRRenBColeFCorbettKD2019Dynamic Reorganization of the Genome Shapes the Recombination Landscape in Meiotic ProphaseNature Structural & Molecular Biology2616417410.1038/s41594-019-0187-030778236PaulsonJRLaemmliUK1977The Structure of Histone-Depleted Metaphase ChromosomesCell1281782810.1016/0092-8674(77)90280-X922894PiazzaIRutkowskaAOriAWalczakMMetzJPelechanoVBeckMHaeringCH2014Association of Condensin with Chromosomes Depends on DNA Binding by Its HEAT-Repeat SubunitsNature Publishing Group21610.1038/nsmb.2831PiskadloETavaresAOliveiraRA2017Metaphase Chromosome Structure Is Dynamically Maintained by Condensin I-Directed DNA (de)CatenationeLife6e2612010.7554/eLife.2612028477406PotapovaTAUnruhJRYuZRancatiGLiHStampferMRGertonJL2019Superresolution Microscopy Reveals Linkages between Ribosomal DNA on Heterologous ChromosomesThe Journal of Cell Biology2182492251310.1083/jcb.20181016631270138PradhanBBarthRKimEDavidsonIFBauerBLaarTVYangW2021SMC Complexes Can Traverse Physical Roadblocks Bigger than Their Ring SizebioRxiv10.1101/2021.07.15.452501RiggsAD1990DNA Methylation and Late Replication Probably Aid Cell Memory, and Type I DNA Reeling Could Aid Chromosome Folding and Enhancer FunctionPhilosophical Transactions of the Royal Society of London. Series B, Biological Sciences32628529710.1098/rstb.1990.00121968665RosinLFNguyenSCJoyceEF2018Condensin II Drives Large-Scale Folding and Spatial Partitioning of Interphase Chromosomes in Drosophila NucleiPLOS Genetics14e100739310.1371/journal.pgen.100739330001329RudnizkySKhamisHGinosarYGorenEMelamedPKaplanA2021Extended and Dynamic Linker Histone-DNA Interactions Control Chromatosome CompactionMolecular Cell611210.1016/j.molcel.2021.06.006RyuHFurutaMKirkpatrickDGygiSPAzumaY2010PIASy-dependent SUMOylation regulates DNA topoisomerase IIalpha activityThe Journal of Cell Biology19178379410.1083/jcb.20100403321079245SamejimaKSamejimaIVagnarelliPOgawaHVargiuGKellyDAde Lima AlvesFKerrAGreenLCHudsonDFOhtaSCookeCAFarrCJRappsilberJEarnshawWC2012Mitotic chromosomes are compacted laterally by kif4 and condensin and axially by topoisomerase iiαThe Journal of Cell Biology19975577010.1083/jcb.20120215523166350SchubertIOudJL1997There Is an Upper Limit of Chromosome Size for Normal Development of an OrganismCell8851552010.1016/S0092-8674(00)81891-79038342ShakyaAParkSRanaNKingJT2020Liquid-Liquid Phase Separation of Histone Proteins in Cells: Role in Chromatin OrganizationBiophysical Journal11875376410.1016/j.bpj.2019.12.02231952807ShamuCEMurrayAW1992Sister chromatid separation in frog egg extracts requires dna topoisomerase II activity during anaphaseThe Journal of Cell Biology11792193410.1083/jcb.117.5.9211315785ShechterDNicklayJJChittaRKShabanowitzJHuntDFAllisCD2009Analysis of histones in Xenopus laevis I A distinct index of enriched variants and modifications exists in each cell type and is remodeled during developmental transitionsThe Journal of Biological Chemistry2841064107410.1074/jbc.M80727320018957438ShintomiKHiranoT2011The Relative Ratio of Condensin I to II Determines Chromosome ShapesGenes & Development251464146910.1101/gad.206031121715560ShintomiKTakahashiTSHiranoT2015Reconstitution of Mitotic Chromatids with a Minimum Set of Purified FactorsNature Cell Biology171014102310.1038/ncb318726075356ShintomiKInoueFWatanabeHOhsumiKOhsugiMHiranoT2017Mitotic Chromosome Assembly despite Nucleosome Depletion in Xenopus Egg ExtractsScience3561284128710.1126/science.aam970228522692ShintomiKHiranoT2021Guiding Functions of the C-Terminal Domain of Topoisomerase IIα Advance Mitotic Chromosome AssemblyNature Communications12291710.1038/s41467-021-23205-w34006877SongFChenPSunDWangMDongLLiangDXuRMZhuPLiG2014Cryo-em study of the chromatin fiber reveals a double helix twisted by tetranucleosomal unitsScience34437638010.1126/science.125141324763583Sonn-SegevABelacicKBodrugTYoungGVanderLindenRTSchulmanBASchimpfJFriedrichTDipPVSchwartzTUBauerBPetersJMStruweWBBeneschJLPBrownNGHaselbachDKukuraP2020Quantifying the heterogeneity of macromolecular machines by mass photometryNature Communications11177210.1038/s41467-020-15642-w32286308SteinAKünzlerP1983Histone H5 Can Correctly Align Randomly Arranged Nucleosomes in a Defined in Vitro SystemNature30254855010.1038/302548a06835386StiglerJÇamdereGÖKoshlandDEGreeneEC2016Single-Molecule Imaging Reveals a Collapsed Conformational State for DNA-Bound CohesinCell Reports1598899810.1016/j.celrep.2016.04.00327117417SunMBiggsRHornickJMarkoJF2018Condensin Controls Mitotic Chromosome Stiffness and Stability without Forming a Structurally Contiguous ScaffoldChromosome Research2627729510.1007/s10577-018-9584-130143891SundinOVarshavskyA1981Arrest of Segregation Leads to Accumulation of Highly Intertwined Catenated Dimers: Dissection of the Final Stages of SV40 DNA ReplicationCell2565966910.1016/0092-8674(81)90173-26269752SzaboQBantigniesFCavalliG2019Principles of Genome Folding into Topologically Associating DomainsScience Advances5eaaw166810.1126/sciadv.aaw166830989119TanabeKIkegamiYIshidaRAndohT1991Inhibition of Topoisomerase II by Antitumor Agents Bis(2,6-Dioxopiperazine) DerivativesCancer Research5149034908Tavares-CadeteFNorouziDDekkerBLiuYDekkerJ2020Multi-Contact 3C Reveals That the Human Genome during Interphase Is Largely Not EntangledNature Structural & Molecular Biology271105111410.1038/s41594-020-0506-532929283TerakawaTBishtSEeftensJMDekkerCHaeringCHGreeneEC2017The Condensin Complex Is a Mechanochemical Motor That Translocates along DNAScience35867267610.1126/science.aan651628882993ThakurelaSGardingAJungJSchübelerDBurgerLTiwariVK2013Gene Regulation and Priming by Topoisomerase IIα in Embryonic Stem CellsNature Communications4347810.1038/ncomms3478Th’ngJPHSungRYeMHendzelMJ2005H1 Family Histones in the NucleusJournal of Biological Chemistry280278092781410.1074/jbc.m501627200Toselli-MollereauERobelletXFauqueLLemaireSSchiklenkCKleinCHocquetCLegrosPN’GuyenLMouillardLChautardEAuboeufDHaeringCHBernardP2016Nucleosome eviction in mitosis assists condensin loading and chromosome condensationThe EMBO Journal351565158110.15252/embj.20159284927266525UmesonoKHiraokaYTodaTYanagidaM1983Visualization of Chromosomes in Mitotically Arrested Cells of the Fission Yeast Schizosaccharomyces pombeCurrent Genetics712312810.1007/BF0036563724173154VerschuerenH1985Interference Reflection Microscopy in Cell Biology: Methodology and ApplicationsJournal of Cell Science7527930110.1242/jcs.75.1.2793900106WaltherNHossainMJPolitiAZKochBKueblbeckMØdegård-FougnerØLampeMEllenbergJ2018A Quantitative Map of Human Condensins Provides New Insights into Mitotic Chromosome ArchitectureThe Journal of Cell Biology2172309232810.1083/jcb.20180104829632028WeissmannFPetzoldGVanderLindenRHuis In ’t VeldPJBrownNGLampertFWestermannSStarkHSchulmanBAPetersJM2016BiGBac Enables Rapid Gene Assembly for the Expression of Large Multisubunit Protein ComplexesPNAS113E2564E256910.1073/pnas.160493511327114506WhiteAEHiebARLugerK2016A Quantitative Investigation of Linker Histone Interactions with Nucleosomes and ChromatinScientific Reports61912210.1038/srep1912226750377WoodcockCLSkoultchiAIFanY2006Role of Linker Histone in Chromatin Structure and Function: H1 Stoichiometry and Nucleosome Repeat LengthChromosome Research14172510.1007/s10577-005-1024-316506093WührMChenYDumontSGroenACNeedlemanDJSalicAMitchisonTJ2008Evidence for an Upper Limit to Mitotic Spindle LengthCurrent Biology181256126110.1016/j.cub.2008.07.09218718761WührMFreemanRMPreslerMHorbMEPeshkinLGygiSKirschnerMW2014Deep Proteomics of the Xenopus laevis Egg Using an MRNA-Derived Reference DatabaseCurrent Biology241467147510.1016/j.cub.2014.05.04424954049WynneDJFunabikiH2015Kinetochore function is controlled by a phospho-dependent coexpansion of inner and outer componentsThe Journal of Cell Biology21089991610.1083/jcb.20150602026347137XiaoBFreedmanBSMillerKEHealdRMarkoJF2012Histone H1 Compacts DNA under Force and during Chromatin AssemblyMolecular Biology of the Cell234864487110.1091/mbc.E12-07-051823097493XueJZWooEMPostowLChaitBTFunabikiH2013Chromatin-Bound Xenopus Dppa2 Shapes the Nucleus by Locally Inhibiting Microtubule AssemblyDevelopmental Cell27475910.1016/j.devcel.2013.08.00224075807YoungGHundtNColeDFinebergAAndreckaJTylerAOlerinyovaAAnsariAMarklundEGCollierMPChandlerSATkachenkoOAllenJCrispinMBillingtonNTakagiYSellersJREichmannCSelenkoPFreyLRiekRGalpinMRStruweWBBeneschJLPKukuraP2018Quantitative mass imaging of single biological macromoleculesScience36042342710.1126/science.aar583929700264YuXDavenportJWUrtishakKACarilloMLGosaiSJKolarisCPBylJAWRappaportEFOsheroffNGregoryBDFelixCA2017Genome-wide TOP2A DNA cleavage is biased toward translocated and highly transcribed lociGenome Research271238124910.1101/gr.211615.11628385713ZhouBJiangJFengHGhirlandoRXiaoSBaiYZhouB2015Structural Mechanisms of Nucleosome Recognition by Linker Histones Article Structural Mechanisms of Nucleosome Recognition by Linker HistonesMolecular Cell5962863810.1016/j.molcel.2015.06.025ZhouBRFengHKaleSFoxTKhantHde ValNGhirlandoRPanchenkoARBaiY2021Distinct Structures and Dynamics of Chromatosomes with Different Human Linker Histone IsoformsMolecular Cell8116618210.1016/j.molcel.2020.10.03833238161ZierhutCJennessCKimuraHFunabikiH2014Nucleosomal Regulation of Chromatin Composition and Nuclear Assembly Revealed by Histone DepletionNature Structural & Molecular Biology2161762510.1038/nsmb.28452495259310.7554/eLife.68918.sa1Decision letterMarstonAdèle LReviewing EditorUniversity of EdinburghUnited KingdomMarstonAdèle LReviewerUniversity of EdinburghUnited Kingdom
Our editorial process produces two outputs: (i) public reviews designed to be posted alongside the preprint for the benefit of readers; (ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.
Acceptance summary:
This paper takes advantage of the Xenopus egg cell free system, combining classical immunofluorescence assays with chromosome conformation (Hi-C) analyses to elucidate the contribution of linker histone H1 to mitotic chromosome organization. The authors find that linker histone H1.8 limits the association of condensin and topoisomerase II to control both chromosome length and individualization. These effects of histone HI on larger scale chromosome organisation may be important for clustering chromosomes for segregation in cells with a large cytoplasm, such as oocytes.
Decision letter after peer review:
Thank you for submitting your article "Linker histone H1.8 inhibits chromatin-binding of condensins and DNA topoisomerase II to tune chromosome length and individualization" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, including Adèle L Marston as the Reviewing Editor and Reviewer #1, and the evaluation has been overseen by Jessica Tyler as the Senior Editor.
The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.
Essential revisions:
1) The authors conclude from their study that Condensins and topo II "functionalities are antagonized by H1.8 to tune chromosome length and prevent hyper-individualization". This interpretation comes from the removal of histone H1.8 from extracts/chromatin and the authors suggest that this is due to a direct competition of H1.8 with Condensins and topo II. However, they show in Figure 1—figure supplement 2 that the abundance of several chromatin proteins is also increased when depleting histone H1.8. Since one of those proteins is histone H1.3, it is unclear how the author can directly ascribe the increased binding of topo II and Condensins on chromosome to the removal of H1.8. Could it not be that removal of H1.8 leads to enrichment of H1.3 and that the latter histone has topo II/Condensin-loading properties on chromosomes? In fact, any of the proteins whose abundance increases due to the loss of H1.8 could act as a topo II/Condensin-loader on chromosomes, which would also explain the authors observations.
2) The purified Condensin complexes used in the DNA/nucleosome binding experiments have major issues with its subunit stoichiometry. Inspection of band patterns in the Condensin input gels reveals that the purified preparations used in this study fluctuate between nearly stoichiometric (Figure 2—figure supplement 2B), to partly stoichiometric (Figure 2—figure supplement 2A) to severely unbalanced (Figure 2A,B; Figure 2—figure supplement 2C). In all cases, the 8S (CAP-E/C) subcomplex of Condensin is overrepresented relative to the 11S subcomplex (CAP-G/H/D2). This indicates enzyme subunit dissociation during purification. This subunit imbalance is even more dramatic when comparing the DNA-bound (bead) fraction to the inputs in these experiments. The bead fraction is typically pulling down ~10-fold more CAP-E/C than other subunits. This indicate that the DNA/nucleosome binding behavior attributed to the Condensin "complex" is in fact reflective of only the SMC subcomplex (CAP-E/C without CAP-G/H/D2). As a consequence, most or all the DNA/nucleosome array binding experiments need to be repeated with genuine/ stoichiometric Condensin complexes to provide valid conclusions on the DNA binding behavior of this enzyme.
3) Regarding chromosome individualization, the authors show that the percentage of "individualized chromosomes" is the same in mock and H1 depleted extracts (after dilution of the chromosome preparation before fixation). However, the dispersion of the chromosomes (processed for immunofluorescence without dilution) is clearly higher in the latter. What is driving the clustering? Interchromosomal entanglements? If so, shouldn´t the entanglements prevent chromosome individualization upon dilution?
4) Related to point (3), the link between H1, condensin, Top2A and chromosome individualization is not clear. In Figure 6, chromosome individualization is measured by the separation of CENP-A foci, while in Figure 7, the authors examine surface area. These measurements may underlie quite distinct phenomena and so it is unclear what the different measurements in each of the conditions refer to. The authors need to clarify what these differences mean.
5) On the same line, the authors write at the beginning of Discussion: "H1 (…) can also suppress chromosome dispersion that is driven by condensins and topo II that organize the long-range DNA folding". Two issues here:
a) It is not clear which results in the paper actually connect long range DNA folding with the dispersion phenotype.
b) The role of topoII in long range DNA folding is not tested in the paper.
Maybe the contribution of topoII to chromosome length could be tested using the same scheme as in Figure 6G-H and the protocol for observing individualized chromosomes (dilution before fixation)
6) A previous study showed that depletion of condensin I results in complete loss of individualization of replicated chromosomes in Xenopus egg extracts (Shintomi and Hirano, Genes Dev 2011), whereas in the current study the authors see only a partial effect (Figure 5C). Could this mean that depletion is not complete (western bloting in Figure 5A suggests this might be the case)? Incomplete depletion would complicate the interpretation of the rescue experiment. Does the amount of protein in the extract and relate this to the extent of chromosomal condensation observed? The authors need to explain this discrepancy with previously published work.
7) The authors propose that one of histone H1.8 role is to reduce chromosome length (a function achieved by antagonizing topo II/Condensin loading on chromosome). However, histone H1.8 is an embryonic linker histone and cells in early embryos are typically large and less dependent on chromosome size-reduction mechanisms. One would assume that chromosome size should be maximally reduced in small somatic cells and, consequently, shortening of chromosomes should be promoted by somatic linker histone H1 subtypes. In Xenopus laevis, histone H1.8 accumulates during oogenesis, persists until the mid-blastula transition, and is then replaced by somatic H1 subtypes. It seems like H1.8 disappears when its putative chromosome shortening function is most needed. The authors did not explain how their results are consistent with this biological context.
Reviewer #1 (Recommendations for the authors):
1. Is there a threshold amount of condensin I that needs to be associated with chromosomes for "normal" length. Can the authors partially deplete histone H1 and see a correlation between the amount of condensin and chromosome length?
2. Related to 2. If chromosome individualization is affected in H1 depletion, the authors should be able to observe inter-chromosomal interactions in their Hi-C dataset. Can this be quantified for the different conditions?
Reviewer #2 (Recommendations for the authors):
This is a very nice study from a laboratory that has contributed several important papers to the field of mitosis over the years. Experiments were well designed and carefully performed for the most part (with the exception of DNA/nucleosome binding experiments with purified Condensins). The flow of experiments in this manuscript is logical. Overall, I would say this manuscript addresses an important biological question, and this topic has obvious implications for development and genome stability. The manuscript would however benefit from toning down several overinterpretations and would also be improved by providing a more compelling rationale to justify a role for an early embryonic histone in the reduction of chromosome length.
1 – All experiments presented in this study should be performed a minimum of 3 times as this represents a basic requirement for reproducibility in biomedical research. This includes all Hi-C experiments (two replicates are reported) and some nucleosome-binding experiments (Figure 2—figure supplement 3).
2 – There is a missing reference in the following sentence on page 4: Since Condensins prefer to bind nucleosome-free DNA (Kong et al. 2020; Zierhut et al., 2014; Shintomi et al., 2017; Toselli‐Mollereau et al., 2016). This observation was originally made by Piazza et al., 2014, Nat Struct Mol Biol.
Reviewer #3 (Recommendations for the authors):
It could be very useful to present a table in the Discussion with a summary of:
1 – the relative amounts of condensins I and II and topo II on DNA , taken from figures 5D, 5E, and Figure 6_supplement1 C.
2 –the phenotype observed in terms of chromosome length, chromosome individualization, and clustering, under the different conditions tested (control, δ H1, deltaH1deltaG, deltaH1deltaD3 and deltaH1deltaGdeltaD3).
This would help understand the co-dependencies in recruitment of the factors under study and the correlation between the different phenotypes.
[Editors' note: further revisions were suggested prior to acceptance, as described below.]
Thank you for resubmitting your work entitled "Linker histone H1.8 inhibits chromatin-binding of condensins and DNA topoisomerase II to tune chromosome length and individualization" for further consideration by eLife. Your revised article has been reviewed by 3 peer reviewers, including Adèle L Marston as the Reviewing Editor and Reviewer #1, and the evaluation has been overseen by Jessica Tyler as the Senior Editor.
The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:
You will see below that Reviewer #2 has two outstanding issues. Please address these as follows:
1. Please add a sentence to the Results section for the condensin binding assays to qualify that SMC subunits may be retained on the chromatin beads more efficiently than the other condensin subunits. You may also want to speculate why this is the case, as in your rebuttal letter.
2. While three repetitions of every experiment are not insisted on, it is important that the statistical analysis used is appropriate and clearly reported throughout the manuscript. In a number of places in the manuscript, t tests appear to have been carried out where n=2. If/where this is the case, please reanalyse the data using the appropriate statistical test and state this clearly in the figure legend. Where key experiments fail to show significance using these criteria, an additional repeat may be appropriate. For example, the legend of Figure 2 states "The statistical significance of the changes in each immunofluorescence experiment was assessed by a Mann Whitney U-test and the p-values shown in D, E and G are calculated by an unpaired Student's t-test of the aggregate medians (immunofluorescence) or blot intensities of each single experiment." A t-test would not be appropriate for Figure 2D where it appears that two immunofluorescence experiments are presented, if this is indeed what is shown. Please also refer to the comment by reviewer #2.
In addition, please note suggestions for textual changes from Reviewer #1.
Reviewer #1 (Recommendations for the authors):
The authors have addressed all of the essential revisions. In addition they have provided convincing responses to all reviewers comments. There are a few places where the text could be clarified:
Line 185. Please explain the Q loop mutant
Line 195 "Although SMC2 and SMC4 appeared to be better retained on nucleosome beads than the non-SMC subunits"
Line 318 "This is consistent with the shorter chromosomes in DH1DCAP-G1 extracts as compared to control…"
Line 362: Please explain ICRF-193.
Line 368: Suggest: To quantify the clustering, we stained the coverslips for CENP-A and used the CENP-A foci (CENP-A doublet is counted as one focus) to measure the number of chromosomes in each chromosome mass. The fraction of chromosomes was measured in small clusters (<4 CENP-A foci, since a few chromosomes can colocalizae stochastically).
Line 383 "better" can be deleted here since the assay has already been introduced.
Line 421 "showed reduced clustering compared to.."
Line 427 Suggest replacing "background" with "extracts"
Line 428 "TOP2A depletion in H1.8 depleted extracts also did not reduce the chromosome spreading even though TOP2A levels on these chromosomes became comparable to that of the control".
Reviewer #2 (Recommendations for the authors):
The revised manuscript has been improved by the authors but two issues raised in previous reviews have not been addressed in the latest version of the manuscript.
1 – Condensin stoichiometry: Experiments shown in Figure 2 and associated supplemental figures use Condensin complexes with non-stoichiometric subunit composition. The authors recognize this fact in their rebuttal letter, stating that the "stoichiometry of condensin subunits seems to be changed in DNA-bound fraction in DNA-beads pull-down experiments…" The issue is not convincingly resolved by the mass photometry experiment shown in Figure 2—figure supplement 1 because this experiment does not fully recapitulate conditions of the DNA binding reactions.
The authors suggest problems with stoichiometry are not relevant because Condensin complexes used in their experiments were "purified as reported previously (Kong et al., 2020)" and "are similar to those seen in previous reports in both recombinant yeast condensin and immunoprecipitated Xenopus and human condensins (Kong et al., 2020; Ganji et al., 2018; Kimura, Cuvier, and Hirano 2001; Kimura and Hirano 1997)." Examination of Condensin purification gels from Kong et al., shows that the complexes purified by these authors are actually stoichiometric (see Figure 1D in PMID 32445620). The same conclusion can be reached for the yeast complexes purified by the Häring group (see Figure 1C in PMID 28882993). The other two studies cited by the authors (Kimura, Cuvier, and Hirano 2001; Kimura and Hirano 1997) use silver staining to label Condensin subunits, a staining method that most biochemists recognize as non-linear and inappropriate to compare the abundance of different proteins. The contention that subunit imbalances seen in the Condensin complexes used by Choppakatla and colleagues are normal and/or reflect imbalances seen in other purified Condensin complexes cannot be verified in the published literature.
The DNA/nucleosome binding behavior attributed to Condensin is this study is likely reflective of the SMC subcomplex (CAP-E/C without CAP-G/H/D2), not the genuine Condensin 1 enzyme. As a consequence, most or all the DNA/nucleosome array binding experiments need to be repeated with stoichiometric Condensin complexes to provide valid conclusions on the DNA binding behavior of the entire/native enzyme.
2 – Several experiments presented in the revised manuscript have been performed as duplicates, and on one occasion (Figure 3E-F) the results presented come from a single experiment. This reviewer argued previously that all experiments in this study should be performed a minimum of 3 independent times. The authors stated in their rebuttal letter that they "do not see that the level of confidence significantly increases by enforcing every experiment to be done 3 times, statistically speaking." I disagree with this statement. There is ample evidence to show that statistical analyses are less reliable with sample sizes of 2. Many textbooks /reference papers in statistics support this view and, as an example, I would refer the authors to "A biologist's guide to statistical thinking and analysis." The authors are correct, though, in their statement that eLife has no explicit publication policy on the minimal number of independent replicates required for publication.
The insistence of the authors to use two independent datasets/experiments for most figures has statistical implications. For example, on some occasions, the authors use a t-test to compare two groups, which is a parametric test. On other occasions, the authors use a Mann Whitney U test, which is a non-parametric test. Parametric testing assumes that the data are normally distributed. With n=2, it is expected that the data would not pass a normality test, hence a Mann-Whitney U test is likely the only appropriate option for most of the figures presented in the manuscript. Separate from these points, when comparing multiple groups on a graph, a One-Way ANOVA is more appropriate than performing multiple independent t-tests. Overall, the authors need to justify their rationale for the choice of statistical analyses performed throughout the manuscript, especially when using only n=2. In many cases, performing an extra experiment to bring datasets to n=3 would likely alleviate the concerns expressed above.
Reviewer #3 (Recommendations for the authors):
The authors have clarified most of my doubts and answered carefully to the criticisms raised during revision. The new figures and changes in the text have improved the manuscript and have made it more clear and accessible to readers.
10.7554/eLife.68918.sa2Author response
Essential revisions:
1) The authors conclude from their study that Condensins and topo II "functionalities are antagonized by H1.8 to tune chromosome length and prevent hyper-individualization". This interpretation comes from the removal of histone H1.8 from extracts/chromatin and the authors suggest that this is due to a direct competition of H1.8 with Condensins and topo II. However, they show in Figure 1—figure supplement 2 that the abundance of several chromatin proteins is also increased when depleting histone H1.8. Since one of those proteins is histone H1.3, it is unclear how the author can directly ascribe the increased binding of topo II and Condensins on chromosome to the removal of H1.8. Could it not be that removal of H1.8 leads to enrichment of H1.3 and that the latter histone has topo II/Condensin-loading properties on chromosomes? In fact, any of the proteins whose abundance increases due to the loss of H1.8 could act as a topo II/Condensin-loader on chromosomes, which would also explain the authors observations.
Our data show that H1.8 depletion results in increased accumulation of condensins and topo II on mitotic chromatin in Xenopus egg extracts (Figure 1) and that H1.8 can directly inhibit binding of condensins and topo II to the linker DNA between nucleosomes in vitro with purified components (Figure 2). Thus, we believe that a direct competition between H1.8 and condensins and topo II for the linker DNA represents a very likely model to explain the chromatin enrichment of condensins and topo II in egg extracts. Other DNA binding proteins also showed some changes in chromatin binding in H1.8 depleted extracts (Figure 1-figure supplement 1E, figure supplement 2, 3), but condensins and topo II are the most abundant chromatin proteins whose chromatin binding is significantly affected by H1.8 depletion. The amount of H1.3 bound to chromatin upon H1.8 depletion was around two orders of magnitude lower than that of condensins and topo II on chromatin (Figure 1-figure supplement 1E) and was 6-fold lower than the residual undepleted H1.8. Therefore, it would be highly unlikely that H1.3 directly stimulates condensin and topo II recruitment unless it can exhibit an unknown catalytic action. If that is the case, it is clearly out of the scope of this manuscript. Interestingly, we now showed that partial depletion of TOP2A did not rescue chromosome elongation phenotype of H1.8 depletion. We believe that this new result would further highlight the unique role of condensin in chromosome elongation.
2) The purified Condensin complexes used in the DNA/nucleosome binding experiments have major issues with its subunit stoichiometry. Inspection of band patterns in the Condensin input gels reveals that the purified preparations used in this study fluctuate between nearly stoichiometric (Figure 2—figure supplement 2B), to partly stoichiometric (Figure 2—figure supplement 2A) to severely unbalanced (Figure 2A,B; Figure 2—figure supplement 2C). In all cases, the 8S (CAP-E/C) subcomplex of Condensin is overrepresented relative to the 11S subcomplex (CAP-G/H/D2). This indicates enzyme subunit dissociation during purification. This subunit imbalance is even more dramatic when comparing the DNA-bound (bead) fraction to the inputs in these experiments. The bead fraction is typically pulling down ~10-fold more CAP-E/C than other subunits. This indicate that the DNA/nucleosome binding behavior attributed to the Condensin "complex" is in fact reflective of only the SMC subcomplex (CAP-E/C without CAP-G/H/D2). As a consequence, most or all the DNA/nucleosome array binding experiments need to be repeated with genuine/ stoichiometric Condensin complexes to provide valid conclusions on the DNA binding behavior of this enzyme.
The human condensin complexes used in the binding assays in Figure 2 were purified as reported previously (Kong et al., 2020). These complexes eluted as a single peak in a gel filtration column, and the varied intensities of the subunits in the complex are similar to those seen in previous reports in both recombinant yeast condensin and immunoprecipitated Xenopus and human condensins (Kong et al., 2020; Ganji et al., 2018; Kimura, Cuvier, and Hirano 2001; Kimura and Hirano 1997). Apparent difference in band intensities in the input samples is most likely due to the different numbers of Coomassie dye binding sites among different condensin subunits. In order to confirm that the condensin complexes used in the binding assays were intact full-length complexes, we also performed mass photometry in the binding assay conditions and added these as Figure 2—figure supplement 1. We had also shown that the wildtype human condensin I was functional as it was able to rescue chromosome morphology in condensin I depleted egg extracts (Figure 2-figure supplement 2A). Although stoichiometry of condensin subunits seems to be changed in DNA-bound fraction in DNA-beads pull-down experiments, most likely due to selective loss of non-SMC subunits during bead-washing steps, H1.8 clearly reduced DNA-binding of all condensin subunits. Since we observed H1.8 mediated inhibition of condensin binding by both chromatin pulldowns (Figure 2) and using electrophoresis mobility shift assays (EMSA Figure2-figure supplement 4), we believe that this represents the behavior of full length condensin complexes.
3) Regarding chromosome individualization, the authors show that the percentage of "individualized chromosomes" is the same in mock and H1 depleted extracts (after dilution of the chromosome preparation before fixation). However, the dispersion of the chromosomes (processed for immunofluorescence without dilution) is clearly higher in the latter. What is driving the clustering? Interchromosomal entanglements? If so, shouldn´t the entanglements prevent chromosome individualization upon dilution?
4) Related to point (3), the link between H1, condensin, Top2A and chromosome individualization is not clear. In Figure 6, chromosome individualization is measured by the separation of CENP-A foci, while in Figure 7, the authors examine surface area. These measurements may underlie quite distinct phenomena and so it is unclear what the different measurements in each of the conditions refer to. The authors need to clarify what these differences mean.
To better explain our observations and the apparent contradictions in the data, we added an additional figure discussing the two metrics for individualizations (Figure 5). We measured chromosome individualization in two contexts, with mechanical dispersal (Figure 6- Dilution assay) and in nocodazole treated extracts (Figure 7-Surface area).
Two lines of evidence demonstrate that significant interchromosomal links persist in mitosis and the TOP2A activity is needed to resolve these links. The first involves the clustering of chromosomes in the absence of a spindle (Figure 5B, 5C) which we quantify using the surface area of the chromosomal masses. The second is shown using the chromosome dilution assay (Figure 5A). In this assay, entropic forces, generated by dissolving the spindle matrix and reduced molecular crowding, disperse chromosomes and generate single chromosomes. The number of CENP-A foci per chromosome mass was a convenient way to quantitate chromosome clustering/individualization. Although the collapse of spindles due to the addition of nocodazole leads to chromosome clustering due to the accumulation of interchromosomal links (Figure 5A), the mechanical dispersion in the dilution assay can resolve the additional interchromosomal links (Figure 5D, 5E). We also then showed that the resolution of these interchromosomal links requires TOP2A activity during the dilution process (Figure 5D, 5E, figure supplement 1).
Due to the differences in the two assays, chromosome clustering observations by the dilution assay can differentiate between normal individualization in control (ΔIgG) and the hyperindividualization in ΔH1 extracts. However, the accumulation of interchromosomal links in ΔCAP-G extracts can be observed readily (Figure 6B, 6C). The surface area measurements assay measure individualization in nocodazole treated extracts (Figure 5A). Since no mechanical forces separate the chromosomes in this experiment, control extracts already appear to show maximal clustering (Figure 5B, 5C) and no differences between control and ΔCAP-G extracts seem to be observed. The hyper-individualization due to the loss of H1.8 mediated suppression of condensins and TOP2A is however observed as chromosome declustering.
5) On the same line, the authors write at the beginning of Discussion: "H1 (…) can also suppress chromosome dispersion that is driven by condensins and topo II that organize the long-range DNA folding". Two issues here:
a) It is not clear which results in the paper actually connect long range DNA folding with the dispersion phenotype.
b) The role of topoII in long range DNA folding is not tested in the paper.
Maybe the contribution of topoII to chromosome length could be tested using the same scheme as in Figure 6G-H and the protocol for observing individualized chromosomes (dilution before fixation)
(a)We apologize for a grammatical mistake in this sentence, which should have read;
“Thus, the linker histone H1, which can promote local chromatin compaction through promoting nucleosome-nucleosome interaction, can also suppress chromosome dispersion that is driven by condensins and topo II, which organize the long-range DNA folding.”
Instead of “which”, we incorrectly connected the clauses with “that”. We meant to say that H1.8 plays a role in both long-distance genome folding and chromosome dispersion and this happens through its regulation of condensins and topo II. We do not wish to directly connect the long-range folding phenotype to the chromosome individualization phenotype. This is evident from our observations that shortened chromosomes with larger DNA loop layers in ΔH1ΔCAP-G extracts (Figure 3C, D, Figure 4D, E) are more dispersed than chromosomes in control extracts. Thus, although condensins and topo II may play a role in both processes, the processes appear to be regulated differently. To avoid the confusion, the rephrased sentence reads;
“Thus, the linker histone H1, which can promote local chromatin compaction through promoting nucleosome-nucleosome interaction, can also suppress chromosome dispersion and long-range DNA folding.”
(b) The data presented in Figure 3E, 3F show that chromosome length was determined solely by the condensin I levels on the chromatin, since condensin I co-depletion did notaffect the topo II overloading upon H1.8 depletion (Figure 7F). We also added Figure 3-figure supplement 2 where we show that partial TOP2A depletion leads to a small increase in chromosome length and that this does not rescue chromosome elongation in ΔH1 extracts. This is consistent with previous reports of the effect of TOP2A depletion or inhibition in vertebrate cells (Farr et al., 2014; Ladouceur et al., 2017; Nielsen et al., 2020; Samejima et al., 2012).
6) A previous study showed that depletion of condensin I results in complete loss of individualization of replicated chromosomes in Xenopus egg extracts (Shintomi and Hirano, Genes Dev 2011), whereas in the current study the authors see only a partial effect (Figure 5C). Could this mean that depletion is not complete (western bloting in Figure 5A suggests this might be the case)? Incomplete depletion would complicate the interpretation of the rescue experiment. Does the amount of protein in the extract and relate this to the extent of chromosomal condensation observed? The authors need to explain this discrepancy with previously published work.
As the reviewer notes, previous reports suggest that complete condensin I depletion fails to generate well-individualized chromosomes (Shintomi and Hirano 2011). We believe that the phenotype reported in Figure 6B, 6C is consistent with this above data. Since we mechanically separate chromosomes by a dilution step before fixing them, we believe that we can separate a small number of ‘fuzzy’ chromosomes from each nucleus in ΔCAP-G extracts. Since we observe only ~15-20% of the chromosomes are individualized even after mechanical dispersion, we believe that this represents a complete failure of chromosome individualization.
Please note that the interpretation of partial condensin depletion effect would have been complicated only if we had argued that apparent chromosome individualization in ΔH1ΔCAPG extracts is executed by a condensin-independent mechanism. Since additional condensin II co-depletion in triple depletion experiment (ΔH1ΔCAP-GΔCAP-D3) resulted in failed individualization, we instead argue that apparent chromosome individualization in ΔH1ΔCAPG extracts is due to increased chromatin-binding of condensin II (and likely due to residual condensin I). Although we may not eliminate the potential role of other proteins in promoting chromosome individualization in egg extract chromosomes, this possibility would not affect our conclusion that H1 suppresses chromosome individualization through limiting chromatinloading of condensin I and condensin II.
7) The authors propose that one of histone H1.8 role is to reduce chromosome length (a function achieved by antagonizing topo II/Condensin loading on chromosome). However, histone H1.8 is an embryonic linker histone and cells in early embryos are typically large and less dependent on chromosome size-reduction mechanisms. One would assume that chromosome size should be maximally reduced in small somatic cells and, consequently, shortening of chromosomes should be promoted by somatic linker histone H1 subtypes. In Xenopus laevis, histone H1.8 accumulates during oogenesis, persists until the mid-blastula transition, and is then replaced by somatic H1 subtypes. It seems like H1.8 disappears when its putative chromosome shortening function is most needed. The authors did not explain how their results are consistent with this biological context.
The reviewers are correct that one may assume that long chromosomes can be allowed in a large oocyte/egg. This is exactly why our discovery that the oocyte-specific H1.8 actively suppresses the actions of condensin is surprising and has important implications. Mitotic chromosome length is constrained not only by the cell size but also by the spindle size (Schubert and Oud 1997). Importantly, the spindle size in Xenopus embryos does not scale with the cell size during the early divisions (Wuhr et al., 2008). Functional significance of the upper limit of the spindle size remains speculative, but in the discussion, we suggested a potential role for limiting the spindle size to reduce the time of chromosome segregation/mitosis in rapid, synchronized cell divisions in frogs without cell cycle checkpoint controls. The longer the spindle is, the variations in duration for chromosome segregation may increase, making it difficult to maintain the cell division synchrony. Chromosomes prepared in H1.8 depleted extracts (ΔH1) are too long to fit into the mitotic spindles and this may lead to anaphase defects (Maresca, Freedman, and Heald 2005). H1.8 depletion also results in more fragile chromosomes, suggesting that condensin overloading due to loss of H1.8 may result in chromosome fragmentation due to spindle forces (Xiao et al., 2012).
Despite their smaller cell sizes, mitotic durations can be longer in somatic cells with checkpoint controls than in early embryonic cells with larger cell sizes. Indeed, it was suggested that the larger cell size in early embryos correlates with reduced mitotic checkpoint strength (Galli and Morgan 2016). Thus, it rather makes sense that having shorter mitotic chromosome length is more important in larger early embryonic cells that have reduced mitotic checkpoint strength. We have emphasized these points in our revised Discussion.
It is also possible that limiting the chromosome size may be a byproduct of preventing chromosome hyper-chromosome individualization by limiting the number of chromatin-bound condensins and TOP2A. As we pointed out in Discussion, clustering chromosomes during or prior to spindle formation by limiting chromosome individualization may be particularly important in a large oocyte to ensure that a bipolar spindle assembles on the whole set of chromosomes. Pronuclear clustering in human and bovine oocytes has also recently been shown to be important to ensure efficient chromosome segregation (Cavazza et al., 2021). Previous studies have also shown that, when bound to chromatin, somatic linker histones can rescue the chromosome elongation observed in H1.8 depleted extracts (Freedman and Heald 2010). However, exogenously provided somatic linker histones are sequestered by importin β during mitosis and are thus excluded largely from mitotic chromatin and somatic linker histones can only rescue loss of H1.8 upon release from this sequestration (Freedman and Heald 2010). This suggests that H1.8 plays a specialized role in oocyte and early embryo mitoses to restrict the chromosome length and individualization in these cells. Somatic linker histones may also play a similar role in the chromosome shortening observed in post MBT embryos as importin sequestration is weaker in smaller cells (Freedman and Heald 2010).
Reviewer #1 (Recommendations for the authors):
1. Is there a threshold amount of condensin I that needs to be associated with chromosomes for "normal" length. Can the authors partially deplete histone H1 and see a correlation between the amount of condensin and chromosome length?
In Figure 3E, F, by co-depletion of H1.8 and condensin I, we were able to generate a broad range of chromosome-bound condensin I levels. The chromosome lengths are highly correlated with the amount of chromosome-bound condensin I in this range. Since condensin I is essential for chromosome individualization, a process that is needed for chromosome length measurement, it might not be simple to address the question of whether a threshold level of condensin is required for “normal” chromosome length. Based on Figure 3E, F and Hi-C data, we would assume that chromosome length is linearly correlated with the chromosome-bound levels of condensin I when there is enough condensin I to make chromosome individualized.
2. Related to 2. If chromosome individualization is affected in H1 depletion, the authors should be able to observe inter-chromosomal interactions in their Hi-C dataset. Can this be quantified for the different conditions?
This is an interesting question, and we tried to monitor inter-chromosomal interactions by Hi-C. Unfortunately, the data were difficult to interpret. As the ligation step was performed in situ as described previously (Belaghzal, Dekker, and Gibcus 2017), the lack of a cellular membrane in pelleted mitotic nuclei from Xenopus egg extracts likely resulted in increased inter-nuclear trans contacts. Although this data still captures the mitotic chromosome structure through cis contacts, we observe a high incidence of trans contacts that may be artifactual. While it may be possible to reduce internuclear trans contacts by modifying the dilution procedure, this requires substantial efforts and is thus beyond the scope of the current manuscript.
Reviewer #2 (Recommendations for the authors):
This is a very nice study from a laboratory that has contributed several important papers to the field of mitosis over the years. Experiments were well designed and carefully performed for the most part (with the exception of DNA/nucleosome binding experiments with purified Condensins). The flow of experiments in this manuscript is logical. Overall, I would say this manuscript addresses an important biological question, and this topic has obvious implications for development and genome stability. The manuscript would however benefit from toning down several overinterpretations and would also be improved by providing a more compelling rationale to justify a role for an early embryonic histone in the reduction of chromosome length.
Regarding the functional implications of our findings in oocytes and early embrios, please note our response to Essential Point #7, where we explain why it rather makes sense to keep chromosomes short and suppress hyper chromosome individualization prior to anaphase in a large egg/oocyte.
1 – All experiments presented in this study should be performed a minimum of 3 times as this represents a basic requirement for reproducibility in biomedical research. This includes all Hi-C experiments (two replicates are reported) and some nucleosome-binding experiments (Figure 2—figure supplement 3).
We disagree with an idea that all experiments must be repeated a minimum of 3 times in biomedical research before sharing the results with public; it is not an established rule in the cell biology field (or eLife publication policy), as far as we are concerned. We employed a variety of complementary methods (quantitative microscopy, biochemistry and Hi-C) to make conclusive remarks. We do not see that the level of confidence significantly increases by enforcing every experiment to be done 3 times, statistically speaking. Ultimately, reproducibility must be independently tested by other researchers. However, we have repeated the mono-nucleosome-binding experiment (Figure 2—figure supplement 4), and further strengthened our original conclusion.
2 – There is a missing reference in the following sentence on page 4: Since Condensins prefer to bind nucleosome-free DNA (Kong et al. 2020; Zierhut et al. 2014; Shintomi et al. 2017; Toselli‐Mollereau et al. 2016). This observation was originally made by Piazza et al. 2014, Nat Struct Mol Biol.
We thank the reviewer for pointing out this. We will include this reference.
Reviewer #3 (Recommendations for the authors):
It could be very useful to present a table in the Discussion with a summary of:
1 – the relative amounts of condensins I and II and topo II on DNA , taken from figures 5D, 5E, and Figure 6_supplement1 C
2 –the phenotype observed in terms of chromosome length, chromosome individualization, and clustering, under the different conditions tested (control, δ H1, deltaH1deltaG, deltaH1deltaD3 and deltaH1deltaGdeltaD3).
This would help understand the co-dependencies in recruitment of the factors under study and the correlation between the different phenotypes.
Thank you for the great idea. We added this table as a Table 1.
References
Belaghzal, Houda, Job Dekker, and Johan H. Gibcus. 2017. “Hi-C 2. 0 : An Optimized Hi-C Procedure for High-Resolution Genome-Wide Mapping of Chromosome Conformation.” Methods 123: 56–65. https://doi.org/10.1016/j.ymeth.2017.04.004.
Cavazza, Tommaso, Yuko Takeda, Antonio Z Politi, Magomet Aushev, Patrick Aldag, Clara Baker, Meenakshi Choudhary, et al. 2021. “Parental Genome Unification Is Highly Error-Prone in Mammalian Embryos.” Cell, May, 1–18. https://doi.org/10.1016/j.cell.2021.04.013.
Farr, Christine J., Melissa Antoniou-Kourounioti, Michael L. Mimmack, Arsen Volkov, and Andrew C.G. Porter. 2014. “The α Isoform of Topoisomerase II Is Required for Hypercompaction of Mitotic Chromosomes in Human Cells.” Nucleic Acids Research 42 (7): 4414–26. https://doi.org/10.1093/nar/gku076.
Freedman, Benjamin S., and Rebecca Heald. 2010. “Functional Comparison of H1 Histones in Xenopus Reveals Isoform-Specific Regulation by Cdk1 and RanGTP.” Current Biology 20 (11): 1048–52. https://doi.org/10.1016/j.cub.2010.04.025.
Freedman, Benjamin S., Kelly E. Miller, and Rebecca Heald. 2010. “Xenopus Egg Extracts Increase Dynamics of Histone H1 on Sperm Chromatin.” PLoS ONE 5 (9): 1–10. https://doi.org/10.1371/journal.pone.0013111.
Galli, Matilde, and David O. Morgan. 2016. “Cell Size Determines the Strength of the Spindle Assembly Checkpoint during Embryonic Development.” Developmental Cell 36 (3): 344–52. https://doi.org/10.1016/j.devcel.2016.01.003.
Ganji, Mahipal, Indra A. Shaltiel, Shveta Bisht, Eugene Kim, Ana Kalichava, Christian H. Haering, and Cees Dekker. 2018. “Real-Time Imaging of DNA Loop Extrusion by Condensin.” Science 360 (6384): 102–5. https://doi.org/10.1126/science.aar7831.
Kimura, Keiji, Olivier Cuvier, and Tatsuya Hirano. 2001. “Chromosome Condensation by a Human Condensin Complex in Xenopus Egg Extracts.” The Journal of Biological Chemistry 276 (8): 5417–20. https://doi.org/10.1074/jbc.C000873200.
Kimura, Keiji, and Tatsuya Hirano. 1997. “ATP-Dependent Positive Supercoiling of DNA by 13S Condensin: A Biochemical Implication for Chromosome Condensation.” Cell 90 (4): 625–34. https://doi.org/10.1016/S0092-8674(00)80524-3.
Kong, Muwen, Erin E. Cutts, Dongqing Pan, Fabienne Beuron, Thangavelu Kaliyappan, Chaoyou Xue, Edward P. Morris, Andrea Musacchio, Alessandro Vannini, and Eric C. Greene. 2020. “Human Condensin I and II Drive Extensive ATP-Dependent Compaction of Nucleosome-Bound DNA.” Molecular Cell, May, 683540. https://doi.org/10.1016/j.molcel.2020.04.026.
Ladouceur, Anne-Marie, Rajesh Ranjan, Lydia Smith, Tanner Fadero, Jennifer Heppert, Bob Goldstein, Amy Shaub Maddox, and Paul S. Maddox. 2017. “CENP-A and Topoisomerase-II Antagonistically Affect Chromosome Length.” The Journal of Cell Biology 216 (9): jcb.201608084. https://doi.org/10.1083/jcb.201608084.
Maresca, Thomas J., Benjamin S. Freedman, and Rebecca Heald. 2005. “Histone H1 Is Essential for Mitotic Chromosome Architecture and Segregation in Xenopus laevis Egg Extracts.” Journal of Cell Biology 169 (6): 859–69. https://doi.org/10.1083/jcb.200503031.
Nielsen, Christian F., Tao Zhang, Marin Barisic, Paul Kalitsis, and Damien F. Hudson. 2020. “Topoisomerase IIa Is Essential for Maintenance of Mitotic Chromosome Structure.” Proceedings of the National Academy of Sciences of the United States of America 117 (22). https://doi.org/10.1073/pnas.2001760117.
Samejima, Kumiko, Itaru Samejima, Paola Vagnarelli, Hiromi Ogawa, Giulia Vargiu, David A. Kelly, Flavia de Lima Alves, et al. 2012. “Mitotic Chromosomes Are Compacted Laterally by KIF4 and Condensin and Axially by Topoisomerase IIα.” Journal of Cell Biology 199 (5): 755–70. https://doi.org/10.1083/jcb.201202155.
Schubert, I., and J. L. Oud. 1997. “There Is an Upper Limit of Chromosome Size for Normal Development of an Organism.” Cell 88 (4): 515–20. https://doi.org/10.1016/S0092-8674(00)81891-7.
Shintomi, Keishi, and Tatsuya Hirano. 2011. “The Relative Ratio of Condensin I to II Determines Chromosome Shapes.” Genes and Development 25 (14): 1464–69. https://doi.org/10.1101/gad.2060311.
Woodcock, Christopher L., Arthur I. Skoultchi, and Yuhong Fan. 2006. “Role of Linker Histone in Chromatin Structure and Function: H1 Stoichiometry and Nucleosome Repeat Length.” Chromosome Research 14 (1): 17–25. https://doi.org/10.1007/s10577-005-1024-3.
Wuhr, Martin, Yao Chen, Sophie Dumont, Aaron C. Groen, Daniel J. Needleman, Adrian Salic, and Timothy J. Mitchison. 2008. “Evidence for an Upper Limit to Mitotic Spindle Length.” Current Biology 18 (16): 1256–61. https://doi.org/10.1016/j.cub.2008.07.092.
Xiao, B., Benjamin S. Freedman, K. E. Miller, Rebecca Heald, and John F. Marko. 2012. “Histone H1 Compacts DNA under Force and during Chromatin Assembly.” Molecular Biology of the Cell 23 (24): 4864–71. https://doi.org/10.1091/mbc.E12-07-0518.
[Editors' note: further revisions were suggested prior to acceptance, as described below.]
You will see below that Reviewer #2 has two outstanding issues. Please address these as follows:
1. Please add a sentence to the Results section for the condensin binding assays to qualify that SMC subunits may be retained on the chromatin beads more efficiently than the other condensin subunits. You may also want to speculate why this is the case, as in your rebuttal letter.
We revised the corresponding sentences to read:
“We noticed that the subunit stoichiometry of condensin I and condensin II was altered in the bead fraction from that in the input. Since the condensin complexes were intact at the assay conditions (Figure 2—figure supplement 1), we suspect that non-SMC subunits of condensins were less stable on the nucleosome array than SMC subunits, while H1.8 reduced binding of all condensin subunits to the nucleosome array.”
2. While three repetitions of every experiment are not insisted on, it is important that the statistical analysis used is appropriate and clearly reported throughout the manuscript. In a number of places in the manuscript, t tests appear to have been carried out where n=2. If/where this is the case, please reanalyse the data using the appropriate statistical test and state this clearly in the figure legend. Where key experiments fail to show significance using these criteria, an additional repeat may be appropriate. For example, the legend of Figure 2 states "The statistical significance of the changes in each immunofluorescence experiment was assessed by a Mann Whitney U-test and the p-values shown in D, E and G are calculated by an unpaired Student's t-test of the aggregate medians (immunofluorescence) or blot intensities of each single experiment." A t-test would not be appropriate for Figure 2D where it appears that two immunofluorescence experiments are presented, if this is indeed what is shown. Please also refer to the comment by reviewer #2.
We assume that the editor meant Figure 1, not Figure 2.
In general (Figures 1D, 3D, 6D, 7E, 7F), the statistical analysis of superplots with data from multiple experiments was performed as described (Lord et al., 2020). The t-test comparing just the medians of the samples is the most stringent significance test that ignores the confidence generated by the technical replicates in each sample. We also note that the t-test is still valid for n=2, although it reduces statistical power (Naegle, Gough, and Yaffe 2015). This may result in false negatives, where smaller differences between samples are likely to be deemed non-significant. However, since the changes we observed were substantial, we observed significantly low p values with the reduced power of n=2 experiments. However, as the reviewer notes, the t-test with the reduced n value may lead to erroneous conclusions. As suggested, we changed the analysis in those experiments where the experiment was only repeated twice to a non-parametric test (Figure 1G, 6E). We also added an additional experiment for Figure 1D, where we now applied Student’s t-test. For Figure 6B, 6C, although similar experiments were repeated multiple times, the CENP-A metric was only performed twice. To supplement this, we now include additional data performed independently thrice, where chromosome individualization was quantified based on manual chromosome morphological classification (Figure 6—figure supplement 1C).
We also wish to stress that none of the conclusions of the paper rely on the statistical significance of any single measurement. The functional consequences of condensin and TOP2A suppression by H1.8 were shown by several different orthogonal experiments.
In addition, please note suggestions for textual changes from Reviewer #1.
Reviewer #1 (Recommendations for the authors):
The authors have addressed all of the essential revisions. In addition they have provided convincing responses to all reviewers comments. There are a few places where the text could be clarified:
We appreciate these specific suggestions. We have clarified several of the statements and added more context where it was necessary.
Line 185. Please explain the Q loop mutant.
The revised text reads, “the ATP-binding deficient Q-loop mutant of condensin I”. Please note that this was also introduced at line 164.
Line 195 "Although SMC2 and SMC4 appeared to be better retained on nucleosome beads than the non-SMC subunits"
This section now reads, “We noticed that the subunit stoichiometry of condensin I and condensin II was altered in the bead fraction from that in the input. Since the condensin complexes are intact at the assay conditions (Figure 2—figure supplement 1), we suspect that non-SMC subunits of condensins were less stable on the nucleosome array than SMC subunits, while H1.8 reduced binding of all condensin subunits to the nucleosome array.”
Line 318 "This is consistent with the shorter chromosomes in DH1DCAP-G1 extracts as compared to control…"
The revised sentence reads:
“Further supporting the idea that the layer size anticorrelates with the chromosome length, chromosomes in ΔH1ΔCAP-G extracts were indeed shorter than those in control ΔIgG extracts (Figure 3D, 3F).”
Line 362: Please explain ICRF-193.
The revised sentence reads, “Indeed, when ICRF-193, the drug that inhibits topo II-dependent catenation/decatenation without leaving double-strand DNA breaks (Ikegami, Ishida, and Andoh 1991), was added to egg extracts at the mitotic entry, chromosome individualization was completely blocked, forming large chromosome clusters after extract dilution (Figure 5—figure supplement 1B-D, ICRF-50 min).”
Line 368: Suggest: To quantify the clustering, we stained the coverslips for CENP-A and used the CENP-A foci (CENP-A doublet is counted as one focus) to measure the number of chromosomes in each chromosome mass. The fraction of chromosomes was measured in small clusters (<4 CENP-A foci, since a few chromosomes can colocalizae stochastically).
As we realized that we never introduced CENP-A, we revised this section to better explain this method as below:
“To quantify the clustering of unresolved chromosomes, we stained the coverslips for CENP-A, the centromere-specific histone H3 variant, which marks a centromere locus per chromosome (Edwards and Murray 2005). While each fully individualized chromosome showed one CENP-A focus (CENP-A doublet, representing centromeres of a paired sister chromatids, is counted as one focus), multiple CENP-A foci were observed in a cluster of chromosomes. Since even fully individualized chromosomes stochastically interacted during chromosome dispersion, we redefined the chromosome cluster here for the chromosome mass containing four or more CENP-A foci (Figure 5D, 5E, Figure 5—figure supplement 1E). Although chromosomes from each nucleus tightly compact into a ball-like mass in nocodazole treated metaphase extracts (Figure 5B), they individualized normally by dispersion upon extract dilution (Figure 5D, 5E, figure supplement 1E).”
Line 383 "better" can be deleted here since the assay has already been introduced.
We changed the sentence as advised.
Line 421 "showed reduced clustering compared to.."
We changed the sentence as advised.
Line 427 Suggest replacing "background" with "extracts"
We followed the advice.
Line 428 "TOP2A depletion in H1.8 depleted extracts also did not reduce the chromosome spreading even though TOP2A levels on these chromosomes became comparable to that of the control".
We incorporated this suggestion.
Reviewer #2 (Recommendations for the authors):
The revised manuscript has been improved by the authors but two issues raised in previous reviews have not been addressed in the latest version of the manuscript.
We respectfully disagree with the assessment of the reviewer.
1 – Condensin stoichiometry: Experiments shown in Figure 2 and associated supplemental figures use Condensin complexes with non-stoichiometric subunit composition. The authors recognize this fact in their rebuttal letter, stating that the "stoichiometry of condensin subunits seems to be changed in DNA-bound fraction in DNA-beads pull-down experiments…" The issue is not convincingly resolved by the mass photometry experiment shown in Figure 2—figure supplement 1 because this experiment does not fully recapitulate conditions of the DNA binding reactions.
The authors suggest problems with stoichiometry are not relevant because Condensin complexes used in their experiments were "purified as reported previously (Kong et al., 2020)" and "are similar to those seen in previous reports in both recombinant yeast condensin and immunoprecipitated Xenopus and human condensins (Kong et al., 2020; Ganji et al., 2018; Kimura, Cuvier, and Hirano 2001; Kimura and Hirano 1997)." Examination of Condensin purification gels from Kong et al. shows that the complexes purified by these authors are actually stoichiometric (see Figure 1D in PMID 32445620). The same conclusion can be reached for the yeast complexes purified by the Häring group (see Figure 1C in PMID 28882993). The other two studies cited by the authors (Kimura, Cuvier, and Hirano 2001; Kimura and Hirano 1997) use silver staining to label Condensin subunits, a staining method that most biochemists recognize as non-linear and inappropriate to compare the abundance of different proteins. The contention that subunit imbalances seen in the Condensin complexes used by Choppakatla and colleagues are normal and/or reflect imbalances seen in other purified Condensin complexes cannot be verified in the published literature.
The DNA/nucleosome binding behavior attributed to Condensin is this study is likely reflective of the SMC subcomplex (CAP-E/C without CAP-G/H/D2), not the genuine Condensin 1 enzyme. As a consequence, most or all the DNA/nucleosome array binding experiments need to be repeated with stoichiometric Condensin complexes to provide valid conclusions on the DNA binding behavior of the entire/native enzyme.
The mass photometry experiments shown in Figure 2—figure supplement 1 were performed in the assay buffers and at assay conditions (Room temperature incubation). It is not immediately obvious to us why this does not completely recapitulate the experimental conditions in the binding reactions in Figure 2. Since mass photometry is a more direct measurement of the complex stoichiometry than the Coomassie staining on the gels, we are confident that the input condensin complexes are stoichiometric in nature. Please note that the mass photometry analysis was done on the aliquots from the same lot of purified condensin complexes used in all experiments shown in this manuscript. Although the stoichiometry in the condensin complexes bound to the beads appears to be different, we feel that this may be due to a differential elution from the chromatin beads during the washing step. Supporting this possibility, the electrophoretic mobility shift assays (EMSA) performed in solution also recapitulate the H1.8 mediated inhibition of condensin binding (Figure 2—figure supplement 4). Importantly, we verified that the purified condensin I complex can support condensin I-depletion phenotypes in Xenopus egg extracts (Figure 2—figure supplement 2A). This is an ultimate verification that our purified condensin I is fully functional.
2 – Several experiments presented in the revised manuscript have been performed as duplicates, and on one occasion (Figure 3E-F) the results presented come from a single experiment. This reviewer argued previously that all experiments in this study should be performed a minimum of 3 independent times. The authors stated in their rebuttal letter that they "do not see that the level of confidence significantly increases by enforcing every experiment to be done 3 times, statistically speaking." I disagree with this statement. There is ample evidence to show that statistical analyses are less reliable with sample sizes of 2. Many textbooks /reference papers in statistics support this view and, as an example, I would refer the authors to "A biologist's guide to statistical thinking and analysis." The authors are correct, though, in their statement that eLife has no explicit publication policy on the minimal number of independent replicates required for publication.
The insistence of the authors to use two independent datasets/experiments for most figures has statistical implications. For example, on some occasions, the authors use a t-test to compare two groups, which is a parametric test. On other occasions, the authors use a Mann Whitney U test, which is a non-parametric test. Parametric testing assumes that the data are normally distributed. With n=2, it is expected that the data would not pass a normality test, hence a Mann-Whitney U test is likely the only appropriate option for most of the figures presented in the manuscript. Separate from these points, when comparing multiple groups on a graph, a One-Way ANOVA is more appropriate than performing multiple independent t-tests. Overall, the authors need to justify their rationale for the choice of statistical analyses performed throughout the manuscript, especially when using only n=2. In many cases, performing an extra experiment to bring datasets to n=3 would likely alleviate the concerns expressed above.
In general (Figures 1D, 3D, 6D, 7E, 7F), the statistical analysis of superplots with data from multiple experiments was performed as described here (Lord et al., 2020). The t-test comparing just the medians of the samples is the most stringent significance test that ignores the confidence generated by the technical replicates in each sample. We also note that the t-test is still valid for n=2, although it loses statistical power (Naegle, Gough, and Yaffe 2015). This may result in false negatives, where smaller differences between samples are likely to be deemed non-significant. However, since the changes we observed were substantial, we observed significantly low p values with the reduced power of n=2 experiments. However, as the reviewer notes, the t-test with the reduced n value may lead to erroneous conclusions. As suggested, we changed the analysis in those experiments where the experiment was only repeated twice to a non-parametric test (Figure 1G, 6E). We also added an additional experiment for Figure 1D, where we now applied Student’s t-test. For Figure 6B, 6C, although similar experiments were repeated multiple times, the CENP-A metric was only performed twice. To supplement this, we now include additional data performed independently thrice, where chromosome individualization was quantified based on chromosome morphological classification (Figure 6- figure supplement 1C).
Regarding the experiment shown in Figure 3E-F, the result was reproduced in a biological replicate (Figure 3- figure supplement 1C,D), as we had reported in the previous revision.
We also wish to stress that none of the conclusions of the paper rely on the statistical significance of any single measurement. The functional consequences of condensin and TOP2A suppression by H1.8 were shown by several different orthogonal experiments. Collectively, we believe that it is highly unlikely that addition or even subtraction of an experiment would debunk our major conclusions described in this paper. Thus, we stand by the scientific rigor of the manuscript.
References
Edwards, Nathaniel S, and Andrew W Murray. 2005. “Identification of Xenopus CENP-A and an Associated Centromeric DNA Repeat.” Molecular Biology of the Cell 16 (4): 1800–1810. https://doi.org/10.1091/mbc.e04-09-0788.
Ikegami, Yoji, Ryoji Ishida, and Toshiwo Andoh. 1991. “Inhibition of Topoisomerase II by Antitumor Agents Bis(2,6-Dioxopiperazine) Derivatives.” Cancer Research 51 (18): 4903–8.
Lord, Samuel J, Katrina B Velle, R Dyche Mullins, and Lillian K. Fritz-Laylin. 2020. “SuperPlots: Communicating Reproducibility and Variability in Cell Biology.” Journal of Cell Biology 219 (6). https://doi.org/10.1083/jcb.202001064.
Naegle, Kristen, Nancy R Gough, and Michael B Yaffe. 2015. “Criteria for Biological Reproducibility: What Does ‘n ‘ Mean?” Science Signaling 8 (371): fs7–fs7. https://doi.org/10.1126/scisignal.aab1125.