We reasoned that mitotic chromosome size may relate to nuclear size and/or content due to chromatin factors associated with the DNA during interphase. Alternatively, mitotic chromosomes could scale with spindle size through mechanisms operating in mitosis. To distinguish between these possibilities, we performed a time course of whole-embryo immunofluorescence through the late blastula stages of X. laevis development and measured the dimensions of cells, spindles, and metaphase plates (Figure 1A and B, Figure 1—figure supplement 1). Although previous work showed that spindle lengths reach a plateau in cells larger than ~200 μm in diameter (Figure 1—figure supplement 2A; Wühr et al., 2008), measurement of spindle volumes by confocal microscopy revealed that spindles continue to scale at larger cell sizes (Figure 1C), consistent with observations that spindle width correlates more robustly with cell volume than spindle length in cultured cells (Figure 1—figure supplement 2B; Kletter et al., 2021). The combined volume of all mitotic chromosomes (the metaphase plate) also scaled continuously with cell size (Figure 1D), similar to published work describing nuclear scaling (Figure 1—figure supplement 3; Jevtić and Levy, 2015). To assess whether the metaphase plate scaled more with nuclear size or with mitotic spindle size, we binned the data by cell size and plotted average volumes of the different subcellular structures. We found that total mitotic chromosome volume scaled in size remarkably similarly with both spindles and nuclei (Figure 1—figure supplement 4). However, the decrease in nuclear volume during early development was far greater than for mitotic structures: nuclei decreased in size approximately 10-fold over early cleavage divisions, while metaphase plate and spindle volumes decreased by approximately 3-fold and 2-fold, respectively (Figure 1E and F). This comparison suggests that the fold-increase in chromosome compaction as a cell transitions from interphase to metaphase diminishes from 8-fold in early blastula embryos to 1.5-fold in late blastula stages (Figure 1—figure supplement 5). Overall, these results demonstrate that the metaphase plate scales continuously with cell size in the early embryo, suggesting that mitotic chromosomes share scaling features with both nuclei and mitotic spindles.

Figure 1 with 5 supplements see all

Download asset Open asset

Figure 1—source data 1

This file contains all of the source data for Figure 1 and related supplemental figures.

https://cdn.elifesciences.org/articles/84360/elife-84360-fig1-data1-v2.zip

Download elife-84360-fig1-data1-v2.zip

To examine how morphologies of individual mitotic chromosomes change during development and how their size relates to metaphase plate volume, we prepared mitotic cell extracts from stage 3 (4-cell) or stage 8 (~4000-cell) embryos (Wilbur and Heald, 2013) and centrifuged single endogenous mitotic chromosomes onto coverslips for size measurements. We found that median chromosome lengths decreased approximately 2-fold between stage 3 and stage 8 (Figure 2A and B), while chromosome widths increased only by ~1.2-fold (Figure 2—figure supplement 1). These changes were consistent with the magnitude of metaphase plate scaling during this period estimated by whole-embryo immunofluorescence (approximately 2-3-fold, Figure 1D) and demonstrates that shortening of the long axis is the predominant metric underlying the change in mitotic chromosome size during early embryogenesis. We also observed that the median length of endogenous stage 3 mitotic chromosomes was not statistically different from that of replicated sperm chromosomes formed in egg extracts (Figure 2B), suggesting that replicated sperm chromosomes formed in egg extracts may serve as a proxy for mitotic chromosome size during the earliest cell divisions.

Figure 2 with 1 supplement see all

Download asset Open asset

Figure 2—source data 1

This file contains all of the source data for Figure 2 and the related supplemental figure.

https://cdn.elifesciences.org/articles/84360/elife-84360-fig2-data1-v2.zip

Download elife-84360-fig2-data1-v2.zip

Previously, it was shown that mixing mitotic extracts prepared from early and late blastula stage embryos resulted in spindles of intermediate size due to equilibration of cytoplasmic spindle scaling factors (Wilbur and Heald, 2013). Likewise, combining interphase extracts at different ratios from two Xenopus species with different sized nuclei produced a graded effect on nuclear size (Levy and Heald, 2010). To test whether cytoplasmic factors similarly modulate mitotic chromosome size, we combined metaphase-arrested egg extracts in a 1:1 ratio with stage 8 mitotic embryo extracts containing endogenous mitotic chromosomes (Figure 2B). However, we observed no increase in chromosome length, indicating that mitotic chromosome scaling factors are not exchangeable in the cytoplasm during metaphase. To test whether cytoplasmic extract could alter mitotic chromosome size if added before the onset of chromosome condensation, we filtered stage 8 extracts to remove endogenous chromosomes, then added sperm nuclei either directly or following replication in interphase egg extracts (Figure 2C). In both cases, sperm chromosomes were at least 2-fold longer than the endogenous stage 8 chromosomes (Figure 2D), suggesting that stage 8 metaphase cytoplasm is sufficient to remodel sperm nuclei into mitotic chromosomes, but unable to recapitulate embryo chromosome size. Thus, mitotic chromosome size is predominantly set by factors loaded during interphase that are not readily exchangeable during metaphase, making mitotic chromosome scaling fundamentally distinct from nuclear or spindle size scaling.

The results above indicated that mitotic chromosome size is largely determined by nuclear rather than cytoplasmic factors. We next confirmed previous results that G2-arrested nuclei isolated from blastula-stage embryos and added to metaphase egg extracts produced mitotic chromosomes approximately 2-fold shorter than replicated sperm chromosomes formed in the same extract (Figure 3A and B; Kieserman and Heald, 2011). This finding further supports the idea that mitotic chromosome size is determined prior to entry into metaphase, likely by chromatin factors loaded during interphase. A difference in chromosome size was also recapitulated in extracts depleted of membranes through ultracentrifugation, which are incapable of forming spindles but competent for mitotic chromosome assembly, indicating that spindle formation is not required for mitotic chromosome scaling (Figure 3—figure supplement 1). However, the magnitude decrease in chromosome size was diminished (2-fold in crude egg extracts compared to 1.4-fold in clarified extracts), suggesting that the spindle and/or membrane-associated factors may contribute to the decrease in mitotic chromosome size observed in blastula-stage embryos.

Figure 3 with 6 supplements see all

Download asset Open asset

Figure 3—source data 1

This file contains all of the source data for Figure 3 and related supplemental figures.

https://cdn.elifesciences.org/articles/84360/elife-84360-fig3-data1-v2.zip

Download elife-84360-fig3-data1-v2.zip

Previous work in C. elegans suggested that mitotic chromosome size correlates with intranuclear density and nuclear size (Hara et al., 2013). We observed that embryo nuclei were roughly 2-fold larger in diameter than interphase sperm nuclei (Figure 3—figure supplement 2A), consistent with the doubling of genome size due to the presence of both paternal and maternal genomes in embryo nuclei, suggesting that intranuclear DNA density is comparable between the two sources of nuclei. Yet, mitotic chromosomes formed by adding stage 8 embryo nuclei into egg extracts were 2-fold shorter than those formed from replicated sperm nuclei (Figure 3B), while mitotic spindles formed with either source of nuclei were barely distinguishable in size (Figure 3—figure supplement 2B). These data suggest that neither nuclear size, intranuclear DNA density during interphase, nor spindle size during metaphase determines mitotic chromosome size and further supports the idea that scaling of mitotic spindles, nuclei, and mitotic chromosomes are not necessarily coordinated.

Interestingly, although mitotic chromosome scaling could be recapitulated by adding embryo nuclei to metaphase-arrested egg extracts, chromosome morphologies were distinct. The separation of sister chromatid arms resulting in X-shaped mitotic chromosomes in both stage 3 and stage 8 mitotic embryo extracts (Figure 2A) was not observed when stage 8 embryo nuclei were added to egg extracts as chromosome arms remained tightly associated along their lengths (Figure 3C). Taken together, these results indicate that the factors determining mitotic chromosome size remain associated with chromatin in G2-arrested embryo nuclei even when introduced into metaphase egg extracts, while other factors that determine mitotic chromosome morphology do not.

Robust recapitulation of chromosome scaling in metaphase-arrested egg extracts enabled molecular-level analysis of potential scaling factors, which was not technically feasible in embryo extracts that cannot transit the cell cycle in vitro. We examined three proteins known to influence chromosome size and morphology in Xenopus, condensin I (the predominant condensin in Xenopus eggs and embryos), topoisomerase II (topo II), and the maternal linker histone, termed H1.8 (Maresca et al., 2005; Nielsen et al., 2020; Shintomi and Hirano, 2011). After performing immunostaining of short embryo chromosomes or long sperm chromosomes formed in the same egg extracts, the abundance of each factor was calculated by normalizing immunofluorescence signal to DNA dye intensity (Figure 3C, see ‘Materials and methods’). We found that short embryo chromosomes contained less condensin I and topo II, but more histone H1.8 relative to long replicated sperm chromosomes (Figure 3D, Figure 3—figure supplement 3). We also examined condensin II, which is 5-fold less abundant than condensin I in Xenopus egg extract and plays a minor role in setting mitotic chromosome size (Ono et al., 2003; Shintomi and Hirano, 2011), and similarly observed lower levels on embryo chromosomes relative to sperm chromosomes (Figure 3—figure supplement 4). These results are consistent with a study showing that depletion of histone H1.8 from egg extracts lengthens mitotic chromosomes (Maresca et al., 2005) and more recent work showing that histone H1.8 inhibits binding of condensin I and topo II to mitotic chromosomes (Choppakatla et al., 2021). Therefore, differential recruitment of condensins, topo II, or histone H1.8 may contribute to mitotic chromosome scaling during embryogenesis.

Our previous work showed that short embryo chromosomes could be reset to lengths observed in replicated sperm samples by cycling the mitotic chromosomes through an additional interphase in egg extracts (Figure 3—figure supplement 5A and B; Kieserman and Heald, 2011). To test whether the abundance of candidate scaling factors was affected, we performed immunofluorescence on mitotic embryo chromosomes formed during the first or second metaphase and found that the abundance of all three factors increased in the second metaphase (Figure 3—figure supplement 5C–F, Figure 3—figure supplement 6). Of the three factors, condensin I levels increased the most (2-fold), returning to levels found on replicated sperm chromosomes (Figure 3—figure supplement 5C and D, Figure 3—figure supplement 6A). Our observation that histone H1.8 levels increased slightly after the second metaphase suggests that condensin I recruitment is not necessarily regulated by H1.8, and that condensin I can override the DNA compaction activity of the linker histone to lengthen embryo chromosomes.

Condensins shape mitotic chromosomes through their ability to form and extrude loops from the central axis (Ganji et al., 2018; Goloborodko et al., 2016). In silico models of loop extrusion activity suggested that tuning the abundance of condensin could dramatically alter DNA loop architecture and thus chromosome dimensions (Goloborodko et al., 2016). However, these models have not been tested under physiological conditions that relate to chromosome size changes in vivo.

To assess how DNA loop size and arrangement are altered in the context of mitotic chromosome scaling, we performed Hi-C on long sperm chromosomes and short embryo chromosomes formed in egg extracts. Hi-C contact maps indicated that short embryo chromosomes had increased longer-range genomic contacts along their entire length, as evidenced by thickening of the diagonal (Figure 4A). To quantify this effect, we plotted the decay of contact frequencies (P) as a function of genomic distance in base pairs (s) (Figure 4B). The shape of P(s) was similar to that observed in earlier work on mitotic chromosomes from human, chicken, and Xenopus (Choppakatla et al., 2021; Gibcus et al., 2018; Naumova et al., 2013), and for rod-shaped dinoflagellate chromosomes (Nand et al., 2021), indicating the same overall organization of mitotic chromosomes across diverse species. Changes in the slope of P(s) have been informative for modeling the underlying organization of DNA into layers of loops (Gibcus et al., 2018) and are more easily visualized by plotting the derivative of P(s) (Figure 4C). Based on this previous work, the amount of DNA contained within a layer is estimated by the genomic distance at which the derivative is at its minimum, which was 10⁶ bp for sperm chromosomes compared to ~10⁷ bp for embryo chromosomes (Figure 4C, gray bar). Within a layer, DNA loop size can be estimated from where the derivative peaks at smaller genomic distances, which was between 10⁴–10⁵ bp for sperm chromosomes vs. 10⁵–10⁶ bp for embryo chromosomes (Figure 4C, orange bar). Combined with our immunofluorescence results from Figure 3, these data are consistent with a model in which mitotic chromosomes scale smaller during development through decreased recruitment of condensin I, resulting in larger DNA loops and more DNA per layer, thus accommodating more DNA on a shorter chromosome axis (Figure 4D).

Figure 4

Download asset Open asset

Figure 4—source data 1

This file contains all of the source data for Figure 4 and related supplemental figures.

https://cdn.elifesciences.org/articles/84360/elife-84360-fig4-data1-v2.zip

Download elife-84360-fig4-data1-v2.zip

We next investigated possible mechanisms that could decrease the abundance of condensin I on mitotic chromosomes as they scale smaller during development. Characteristic features of cleavage divisions during early embryogenesis include a lack of cell growth and minimal gene expression, which results in exponentially increasing copies of the genome within the same total volume of cytoplasm. The increase in N/C ratio, defined here as the number of nuclei per volume of cytoplasm, titrates a finite maternal pool of DNA binding factors so that they are distributed to more and more genome copies with each subsequent cell cycle, thus lowering their abundance per genome. This effect is thought to underlie activation of zygotic transcription at the mid-blastula transition (Amodeo et al., 2015; Collart et al., 2013), and titration of the histone chaperone Npm2 was shown to play a role in nuclear scaling (Chen et al., 2019).

To test whether N/C ratio could play a role in mitotic chromosome scaling, we tested two different concentrations of sperm nuclei corresponding to either early (~78 sperm nuclei/μL) or late (~1250 nuclei/μL) blastula stage embryos (Figure 5A). After allowing the nuclei to replicate in interphase egg extracts, we added back metaphase-arrested egg extracts, isolated mitotic chromosomes for length measurements and performed immunofluorescence for condensin I, topo II, and H1.8. We found that mimicking increased N/C ratio by adding a higher concentration of sperm nuclei into egg extract decreased mitotic chromosome length ~1.3-fold (Figure 5B), consistent with the decrease in metaphase plate size observed in vivo at the stage of development corresponding to the N/C ratios tested (stages 6–7, Figure 1D). This size change was accompanied by an ~1.6-fold decrease in condensin I abundance on mitotic chromosomes (Figure 5C), with less significant changes for histone H1.8 and topo II (Figure 5—figure supplements 1 and 2). Interestingly, we found that the increased N/C ratio mimicked in this experiment did not significantly affect nuclear size and increased spindle size (Figure 5D and E), suggesting that the N/C ratio is only capable of scaling mitotic chromosome size, but not these other subcellular structures. Previous work using lipid droplets to encapsulate spindles and nuclei in different sized compartments showed that both structures scale to compartment volume through a limiting-component mechanism (Chen et al., 2019; Good et al., 2013; Leech et al., 2022). However, the range of N/C ratios (defined again as genome copies per volume cytoplasm) tested here was significantly lower (approximately 10-fold) than in those studies, suggesting that maternal components do not become size-limiting for spindles and nuclei until later developmental stages.

Figure 5 with 4 supplements see all

Download asset Open asset

Figure 5—source data 1

This file contains all of the source data for Figure 5 and related supplemental figures.

https://cdn.elifesciences.org/articles/84360/elife-84360-fig5-data1-v2.zip

Download elife-84360-fig5-data1-v2.zip

Our previous data strongly suggested that the major molecular determinants of mitotic chromosome size are loaded during interphase. We tested this model further by varying the concentration of stage 8 embryo nuclei added to metaphase egg extracts and observed no difference in mitotic chromosome length (Figure 5—figure supplement 3A). This observation is consistent with the idea that titration of maternal factors had already occurred in the embryo by the time nuclei were collected at G2. Also, when sperm chromosomes were added directly to metaphase egg extracts, thus skipping the initial round of replication, there was again no effect of N/C ratio on chromosome length (Figure 5—figure supplement 3B). Thus, the factors that set mitotic chromosome length as a function of N/C ratio are loaded during interphase, not metaphase.

Finally, we examined whether the titration effect of chromosome factors observed in vitro corresponded to changes in their abundance during development in vivo by measuring fluorescence intensity levels of condensin I, H3, and histone H1.8 in embryo nuclei isolated at different stages. As predicted, although protein concentrations in embryos did not change over the course of the early cleavage divisions (Figure 5—figure supplement 4A), the levels of all factors found in interphase nuclei decreased as genome copy number increased (Figure 5—figure supplement 4B). Together these observations confirm that maternal factors loaded onto newly synthesized copies of the genome during interphase are titratable and demonstrate that a higher density of nuclei (increased N/C ratio) is sufficient to shorten mitotic chromosomes, likely by decreasing levels of condensin I.

A developmental cue central to the scaling of nuclei and spindles is decreasing cell size. We previously identified importin α as a factor that coordinately scales nuclei and spindles to cell size by regulating nuclear import of lamin proteins and the activity of a microtubule destabilizing protein, respectively (Brownlee and Heald, 2019; Levy and Heald, 2010; Wilbur and Heald, 2013). Palmitoylation of importin α drives a fraction of the total protein to the cell membrane, where it can no longer bind to and inhibit its nuclear localization sequence (NLS)-containing cargo that regulates spindle size (Figure 6A). As cell size decreases during early cleavage divisions, cell surface area/volume (SA/V) increases, causing a larger fraction of importin α to associate with the plasma membrane, thus releasing more scaling factors into the cytoplasm (Figure 6A; Brownlee and Heald, 2019). To test whether importin α also plays a role in mitotic chromosome scaling, we treated egg extracts with palmostatin, an inhibitor of the major depalmitoylation enzyme APT1 to increase the pool of palmitoylated importin α, thus mimicking smaller cells with higher cell SA/V (Figure 6B; Dekker et al., 2010). We found that mitotic chromosome size decreased with palmostatin treatment and was fully rescued by the addition of purified recombinant importin α that cannot be palmitoylatated (NP importin α), but not by addition of wild-type (WT) importin α (Figure 6B and C). Immunofluorescence of DMSO- or palmostatin-treated chromosomes revealed increased abundance of histone H1.8 on short palmostatin-chromosomes, whereas the levels of condensin I and topo II were unchanged (Figure 6D, Figure 6—figure supplement 1). These data are consistent with our observation that short embryo chromosomes contain more histone H1.8 than long sperm chromosomes (Figure 3D) and suggest a model in which importin α partitioning scales mitotic chromosomes to cell size by increasing availability of histone H1.8 in the cytoplasm as cell SA/V increases. Future work will be required to test whether or not H1.8 is a bona fide cargo of importin α or whether a more complicated mechanism is at play.

Figure 6 with 1 supplement see all

Download asset Open asset

Figure 6—source data 1

This file contains all of the source data for Figure 6 and related supplemental figures.

https://cdn.elifesciences.org/articles/84360/elife-84360-fig6-data1-v2.zip

Download elife-84360-fig6-data1-v2.zip

Our observation that condensin I levels were not altered upon palmostatin treatment suggested that scaling of mitotic chromosomes to cell size acts independently of the N/C ratio pathway. To determine when in the cell cycle importin α partitioning was affecting chromosome size, we added palmostatin to extracts either before interphase or following entry into metaphase (Figure 6E). Interestingly, we found that palmostatin treatment during metaphase was sufficient to shrink mitotic chromosomes, while palmostatin treatment during interphase had no effect on chromosome length (Figure 6F). Together, these results demonstrate that partitioning of importin α during metaphase scales mitotic chromosomes to spindle and cell size in a condensin I-independent pathway.

Previous work in C. elegans showed that mitotic chromosome size correlated positively with nuclear size and negatively with intranuclear density (Hara et al., 2013; Ladouceur et al., 2015). However, in egg extracts, mitotic chromosome size does not necessarily correlate with either spindle or nuclear size. G2-arrested stage 8 embryo nuclei, which contain both the maternal and paternal genomes, are almost 2-fold larger than replicated sperm nuclei (Figure 3—figure supplement 2A), consistent with a doubling of genome size. Yet when added to metaphase egg extracts, they form mitotic chromosomes that are 2-fold shorter than replicated sperm chromosomes (Figure 3B), demonstrating that in this system, mitotic chromosome size does not scale to either intranuclear density or nuclear size. Although mitotic chromosomes scaled continuously with spindle size in vivo (Figure 1), this correlation was abolished in vitro (Figure 3—figure supplement 2B). Furthermore, when N/C ratio was varied in egg extracts, mitotic chromosomes shrank 1.3-fold while spindle size increased almost 2-fold and nuclear size remained constant (Figure 5). Together, these data suggest that although spindles, nuclei, and mitotic chromosomes appear to scale coordinately with one another in vivo, the mechanisms regulating their size changes are distinct.

Whereas mitotic spindle size and nuclear size are set by factors operating during metaphase and interphase, respectively, we found that both phases of the cell cycle contribute to scaling of mitotic chromosomes. Experiments performed in cytoplasmic extracts showed that mitotic chromosome size is determined by factors already present in interphase nuclei and cannot be reset by during metaphase (Figures 2 and 3). In contrast, importin α partitioning between the cytoplasm and plasma membrane scaled mitotic chromosomes to spindle and cell size specifically during metaphase. Consistent with this temporal separation of developmental cues, we found that condensin I acts as a scaling factor only in the N/C ratio pathway and not the importin α pathway (Figure 5C vs. Figure 6D). Together, these data suggest that the NLS-containing cargo that scale mitotic chromosomes, unlike cargos that scale nuclei and spindles, cannot freely exchange with other factors in the egg cytoplasm to lengthen short embryo chromosomes. Since these conclusions are drawn from results obtained in vitro using reconstituted chromosomes, it remains to be determined how these two pathways would interplay in vivo. We speculate that the N/C ratio pathway that operates during interphase sets mitotic chromosome size within a certain range, but final mitotic chromosome size is set by importin α partitioning during metaphase, ensuring that metaphase plate size scales with spindle size. This idea is consistent with an early model that proposed mitotic chromosome size must be coordinated with spindle size in order to prevent chromosome mis-segregation (Schubert and Oud, 1997). Now that we have a more complete understanding of the pathways that regulate both spindle and mitotic chromosome size, it will be interesting to test this model in vivo. Finally, Xenopus embryos divide asymmetrically, with larger cells on the vegetal side. This raises the possibility that even within the same embryo the importin α and N/C ratio pathways could combine to have different effects on subcellular scaling in vivo.

Based on in silico models of condensin I loop extrusion activity, it was predicted that changing condensin I occupancy on mitotic chromosomes would cause major changes in DNA loop size and chromosome dimensions (Goloborodko et al., 2016). In egg extracts, depleting condensin I by 5-fold decreased chromosome size by almost 2-fold (Shintomi and Hirano, 2011). However, it was unclear whether such changes occur in vivo to scale mitotic chromosomes. Our results suggest that within the physiologically relevant range of condensin I concentrations present during early embryogenesis, less condensin I correlates with both increased loop size and increased length-wise compaction. Another recent study showed that immunodepletion of H1.8 from Xenopus egg extracts decreased condensin I occupancy on sperm mitotic chromosomes, reducing their length (Choppakatla et al., 2021). In contrast, we found that an increase in condensin I could act independently of linker histone H1.8 to lengthen embryo chromosomes formed in egg extracts (Figure 3—figure supplement 5), suggesting that condensin I is the major scaling factor for mitotic chromosomes. The discrepancy between this study and the previous one suggests that the interplay between condensin I and H1.8 on mitotic chromosomes is more nuanced in vivo where total protein concentrations are constant during early cleavage divisions (Figure 5—figure supplement 4). For example, other factors could act upstream of H1.8 to regulate condensin I loading. Our importin α data suggest that chromosome size can also be modulated by H1.8 alone, without any changes in condensin I abundance (Figure 6D), suggesting that condensin I is not the only factor determining mitotic chromosome size in vivo. Together, our results reveal a more complex interplay among the factors that shape mitotic chromosomes than previously anticipated, some of which could also be cell-cycle dependent.

Previous work in zebrafish and frogs suggested that progressive titration of maternal factors such as histone H3 due to increasing N/C ratio plays an important role in regulating the timing of Zygotic Genome Activation (ZGA) in the embryo (Amodeo et al., 2015; Joseph et al., 2017). Here, we recapitulated this titration effect simply by increasing the concentration of nuclei in egg extracts and found that N/C ratio is sufficient to regulate mitotic chromosome size but not spindle or nuclear size (Figure 5). Together, these results suggest that N/C ratio could be a universal mechanism for regulating chromatin structure across the cell cycle, but for very different functions: transcriptional regulation during interphase and chromosome segregation during metaphase. Two recent studies in Xenopus and zebrafish propose a model in which increasing nuclear import rates during early cleavage divisions triggers ZGA as the genome gains access to key pioneer factors that create open chromatin states permissive for transcription (Nguyen et al., 2022; Shen et al., 2022). Future studies leveraging both the in vitro and in vivo power of the Xenopus system will be invaluable for discovering new mechanistic links between subcellular scaling, chromatin structure, and function during early embryogenesis.

Testes dissected from X. laevis males were gently cleaned with kimwipes and stored at 4°C in 1× MR (100 mM NaCl, 1.8 mM KCl, 2 mM CaCl₂, 1 mM MgCl_2, and 5 mM HEPES-NaOH pH 7.6) for 1 wk. X. laevis females were primed with 100 U of pregnant mare serum gonadotropin (PMSG, National Hormone and Peptide Program, Torrance, CA) at least 48 hr before use and boosted with 500 U of hCG 16 hr before an experiment. Once ovulated, females were gently squeezed along the lower spine to deposit eggs onto Petri dishes coated with 1.5% agarose in 0.1× MMR (100 mM NaCl, 2 mM KCl, 1 mM MgCl₂, 2 mM CaCl₂, 5 mM HEPES-NaOH pH 7.6, 0.1 mM EDTA). 1/4-1/3 of a X. laevis testes was added to 1 mL of mqH₂O in a 1.5 mL tube and homogenized using scissors and a plastic pestle for at least 30 s. The entire solution of sperm was added to a dish of eggs and swirled gently to mix. After a 10 min incubation, dishes were flooded with 0.1× MMR and incubated for an additional 10 min. Jelly coats were removed with a 3% L-cysteine solution in mqH₂O-NaOH, pH 7.8. After extensive washing (at least five times) with 0.1× MMR, embryos were incubated at 23°C. After the first cleavage division, fertilized embryos were sorted and placed in fresh dishes containing 0.1× MMR coated with 1.5% agarose in 0.1× MMR.

Successfully dividing embryos were fixed in MAD (two parts methanol, two parts acetone, one part DMSO) at 5 min intervals during each developmental stage when mitosis was likely to be occurring. After 1–3 hr of fixation, embryos were transferred to fresh MAD before storing at –20°C for up to 3 mo. Embryos were then gradually rehydrated into 0.5× SSC (75 mM NaCl, 7.5 mM sodium citrate, pH 7.0), bleached in 2% H₂O₂, 5% formamide and 0.5× SSC under direct light or 2–3 hr. Bleached embryos were blocked in 1× PBS, 0.1% Triton X-100, 2 mg/mL BSA, 10% goat serum, 5% DMSO for 16–24 hr at 4°C. Primary antibodies were incubated at 2–10 μg/mL at 4°C for 60 hr, washed in PBT (1× PBS, 0.1% Triton X-100, 2 mg/mL BSA) for 30 hr. Secondary antibodies were added at 2 μg/mL, covered from light and incubated for 60 hr at 4°C before washing for 30 hr with PBT. Embryos were then gradually rehydrated into 100% methanol, stored overnight at –20°C, then cleared with Murrays solution (two parts benzyl benzoate, one part benzyl alcohol). Embryos were imaged with a ×10 air objective (NA 0.45) on a Zeiss LSM 800 confocal microscope using 488 and 568 laser lines. Once cells containing a mitotic spindle were identified, z-stacks were taken at 1 μm intervals. Using Imaris, we performed 3D visualization and segmentation of mitotic spindles and metaphase plates to directly measure volumes. Cell size was measured in FIJI by manually tracing the cell in the z-stack with the largest cell area and used this measurement to calculate cell diameter. Interphase nuclear volumes were calculated from a published dataset (Jevtić and Levy, 2015), which used very similar methods to obtain measurements of cross-sectional areas of nuclear and cell sizes of dissociated blastomeres.

Only extracts of high quality (determined by the quality of spindles, nuclei, or chromosomes formed) were used for quantifications shown in this study.

All experiments shown were performed with at least three biological replicates, unless otherwise stated in figure legends. Calculations of raw data were performed in R (version 3.3.0), and final data were plotted using the ggplot2 package (Wickham H, 2016). Negative values in fluorescence intensity measurements due to errors in imaging were removed from data before plotting. Box and violin plots depict the range of the data (first quartile, third quartile, and median), with individual datapoints in gray. All fold-differences stated in this work were calculated based on median values. As such, we used the Mann–Whitney U test to calculate significance values, which is best suited for comparisons based on medians rather than means.

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

We thank Yoshiaki Azuma (University of Kansas) and Susannah Rankin (OMRF) for their generous donation of antibodies against topo II and condensins, respectively. We would like to thank J Smolka, H Cantwell, G Cavin-Meza, X Liu, and S Coyle for critical feedback on the manuscript. Some illustrations were made by Rebecca Konte and Biorender.com. Imaris software was generously donated by A Dernburg. This work was supported by NIH MIRA grant R35GM118183 and the Flora Lamson Hewlett Chair in Biochemistry to RH and a Jane Coffin Childs Memorial Fund fellowship to CYZ, an R01 from NHGRI to JD (HG003143). JD is an investigator of the Howard Hughes Medical Institute.

This study was conducted within the guidelines for Care and Use of Laboratory Animals by the National Institutes of Health. All procedures were performed in accordance with our Animal Utilization Protocol (AUP-2014-08-6596-2) and under strict regulation by the UC-Berkeley Institutional Animal Care and Use Committee (IACUC, NIH Insurance #A4107-01).

© 2023, Zhou et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.