Fig. 5: Effect of mechanical deformation on local protein concentration.

a Schematics showing how concentrating molecules can induce phase separation by pushing across the phase boundary. b Schematics of quantifying the effect of concentration increase. Relative average intensity measures how much material is disproportionately distributed in the trailing compartment. c Increase in trailing-end 53BP1 intensity as cells squeeze through constrictions. Error bars are standard deviations of 14 cells for 2-µm and 12 cells for 15-µm. ***p < 0.001 by one-way ANOVA test (two-tailed). d Increase in trailing-end BRD4 intensity as cells squeeze through constrictions. Error bars are standard deviations of 33 cells and 10 cells for 15-µm. ***p < 0.001 by one-way ANOVA test (two-tailed). e eYFP tag alone does not show material built-up associated with mechanical deformation. All progression bins have non-significant differences compared to Progression = 0. The error bar is the standard deviation of n = 4 cells. f–h Testing various constructs known to phase-separate in the cell nucleus: RNPS1-eYFP, SART1-eYFP, and SRRM1-eYFP. Error bars are SD of 17, 6, and 27 cells for RNPS1-eYFP, SART1-eYFP, and SRRM1-eYFP, respectively, for the 2-µm group, and 10, 8, 14 cells for RNPS1-eYFP, SART1-eYFP, and SRRM1-eYFP, respectively, for 15-µm group. Insets show the change in the averaged partition coefficient of pre-formed condensates during migration progression. RNPS1-eYFP *p = 0.01, all subsequent ***p < 0.001 by one-way ANOVA test (two-tailed). Statistical analysis was done at each progression time point comparing 2-µm vs 15-µm channel groups. i Schematics showing how a shift in phase boundary manifests in an increase in partition coefficient, the ratio of C_den to C_dil. j–l Measured partition coefficient of RNPS1-eYFP, SART1-eYFP, and SRRM1-eYFP. Error bars are SEM of 17, 7, and 27 cells for RNPS1-eYFP, SART1-eYFP, and SRRM1-eYFP, respectively. *p = 0.04 (RNPS1 P = 0.3), ***p < 0.001 (RNPS1 P = 0.8), *p = 0.02 (SART1 P = 0.2), *p = 0.02 (SART1 P = 0.3), **p = 0.003 (SART1 P = 0.4), ***p < 0.001 (SART1 P = 0.5, 0.6, 0.7, 0.8), **p = 0.008 (SRRM1 P = 0.5) by one-way ANOVA test compared to P = 0 (two-tailed).