Centripetal nuclear shape fluctuations associate with chromatin condensation in early prophase

Abstract

The nucleus plays a central role in several key cellular processes, including chromosome organisation, DNA replication and gene transcription. Recent work suggests an association between nuclear mechanics and cell-cycle progression, but many aspects of this connection remain unexplored. Here, by monitoring nuclear shape fluctuations at different cell cycle stages, we uncover increasing inward fluctuations in late G2 and in early prophase, which are initially transient, but develop into instabilities when approaching the nuclear-envelope breakdown. We demonstrate that such deformations correlate with chromatin condensation by perturbing both the chromatin and the cytoskeletal structures. We propose that the contrasting forces between an extensile stress and centripetal pulling from chromatin condensation could mechanically link chromosome condensation with nuclear-envelope breakdown, two main nuclear processes occurring during mitosis.

Fig. 1. Shape fluctuations of HeLa cell nuclei are cell-cycle dependent and increase towards early mitosis.

a Snapshots of a representative nucleus at 7 time points from the start at early G1 phase, throughout S, G2 and prophase (scale bar 5 μm). Arrowheads indicate the reference time-points to determine cell cycle phase (see “Methods” section). b Average spectra of wave vector-dependent fluctuation amplitudes (modes 6–34) for cells at different stages in the cell cycle. The number of nuclei considered for each cell-cycle stage are reported in the legend in brackets. The fluctuation amplitude $⟨u_q^2⟩$ exhibits a decrease with increasing time from G1 until G2, where the fluctuations are reduced by about three times. Instead, active nuclear fluctuations during mitosis become four times higher in early mitosis (green line) and 10 times in late mitosis (purple line). Inset: contour detection of NE (red line) with fluorescent label Emerin. The initial manual selection of the center (red dot) and an initial point on the NE define the annular region containing the cell boundary used in image analysis. c Effective tension vs radius scatterplot shows clusters from different cell-cycle stages forming an open counterclockwise trajectory. d–f Box plots of shape-fluctuation parameters throughout the cell cycle. The data show no significant changes (p-value > 0.05) in effective bending modulus across the cell cycle, while effective tension increases significantly during S phase and decreases up to one order of magnitude during mitosis. The cell radius increases from the starting point in G1 until G2 and then does not change much. g the characteristic relaxation time for mode 3 becomes longer during mitosis. Gray bands and markers represent RBC fluctuation parameters. P values are reported in Table S2 and were calculated using the two-sample t-test; significant relations are highlighted with brackets.

Fig. 2. Calyculin A and latrunculin A treatments recapitulate the joint radius/effective tension changes found in prophase.

a Radius-tension change after treatments compared to the control phase G2, and early and late prophase for cycling cells. Calyculin A causes a reduction of both radius and effective tension, while latrunculin A decreases effective tension to a lower value, which remains constant with treatment time (early, 20 min vs late, 50 min). b, c Details of radius and tension of the same cells before and after both treatments, compared to the values throughout the cell cycle. Shape changes of representative nuclei are highlighted in panel b. The insets below panel (c) report the respective averaged fluctuation spectra. d Relaxation time of mode 3 slightly increases after calyculin treatment, as for prophase cells. Number of cells: calyculin early (13) and late (16), latrunculin 20 min (13) and 50 min (18). P values are reported in Table S2 and were calculated using the two-sample t-test; significant relations are highlighted with brackets.

Fig. 3. Late-G2 and prophase deformations are dominated by inwards invaginations, compatible with the action of centripetal pinning forces.

a Representative examples of the localized inward invaginations that emerge in early prophase and in early stages of calyculin A treatment, but are not found with latrunculin A treatment. b Examples of the non-localized invaginations observed in late prophase and later stages of calyculin A treatment. The insets in panel a and b illustrate the dynamics by snapshots at equal time lags (Videos S2–S4),c Invagination width at the maximal deformation increases by 2-3 fold in late prophase while depth can increase up to 10 fold. Invaginations from early and late phases of calyculin A treatment resemble the ones in prophase, while latrunculin A treatment has mild effects on the invaginations (they remain within <1 μm in depth and 25 degrees in width). d The histograms (top) as well as polar plots (bottom) of signed shape fluctuations show the bias towards inward motion of prophase and calyculin A late nuclei (orange are inward and blue outward fluctuations). Histograms count all contour angles for 500 frames; inward fluctuations were defined as negative deformations <−0.5 μm (orange band), and outward fluctuations as positive deformations >0.5 μm (blue band). e Boxplots of the skewness of the signed shape fluctuation histograms (panel d) over cell-cycle stages and upon drug treatment. The centripetal asymmetry increases during prophase and calyculin A treatment, while latrunculin A does not affect it. ATR inhibitor VE822 increases the events with negative skewness. Number of cells for control (29) and VE822 condition (38). P-values are reported in Table S2 and were calculated using the two-sample t-test; significant relations are highlighted with brackets.