Fig. 4: Cytoplasmic forces accelerate molecular-scale kinetics in nuclear condensates and regulate mRNA processing.

a Scheme of droplet surface fluctuations as a function of cytoplasmic stirring (left); representative nuclear SRSF2-GFP droplet surface fluctuations in orange (above) and surface fluctuation intensity heatmaps of comparably-sized droplets (below) in SN oocytes with Control, amplified, or disrupted cytoplasmic forces; droplet radii, 2 to 2.7 μm; droplet number, Control = 25; Nocodazole = 12; FMN2^−/− = 21; FMN2^−/− + Nocodazole = 15. b FRAP sequences of SRSF2-GFP droplets in Control, FMN2^−/−, and FMN2^−/− + FMN2 SN oocytes. c Fluorescence recovery curves (mean ± s.e.m.) with simple exponential fits of SRSF2-GFP droplets in Control, FMN2^−/−, and FMN2^−/− + FMN2 SN oocytes; insets, apparent diffusion coefficients (D_app); droplet number, Control = 28, FMN2^−/− = 13, FMN2^−/− + FMN2 = 11. d Representative co-immunostainings of nuclear speckles and phosphorylated SF3b155 (pT313) in SN Control, FMN2^−/−, and FMN2^−/− + FMN2 oocytes with magnifications of single droplets. e Quantifications of droplet-specific pT313-SF3b155 intensities in growing Control and FMN2^−/− oocytes, and in FMN2^−/− + FMN2 SN oocytes; Droplet number, Control NSN = 34, Trans = 37, SN = 38, FMN2^−/− NSN = 31, Trans = 33, SN = 56, FMN2^−/− + FMN2 SN = 26. f, g Differential mRNA exon usage versus mean abundance or differential mRNA isoform usage versus isoform switch P_adj in SN FMN2^−/− oocytes relative to Control SN FMN2^+/− oocytes; Colored dots are over/underrepresented exons or isoforms in FMN2^−/− oocytes with a BH-adjusted (exons) or FDR-adjusted (isoforms) P value P_adj < 0.05. h, i About 10 × 10 dot plots showing percentages of mRNA alternative splicing patterns detected in FMN2^−/− (n = 3565 patterns) and percentages of predicted consequences of mRNA isoform switches in FMN2^−/− (n = 2178 consequences). j Jaccard measures of overlaps between sets of genomic sites, through computation of the ratio of their intersections to their union; differentially used exon sites and alternative splice sites of transcripts of FMN2^−/− oocytes were tested bidirectionally for spatial correlation with SRSF1, SRSF2, MBNL3, YY1 RNA-binding sites, and against in silico controls corresponding to the 50 first (Prom50) or last (Term50) nucleotides of all RefSeqNCBI transcripts. Violin plots with median ± quartiles (c, e); P values derived from two-tailed Mann–Whitney U-tests (c, e) and Kruskal–Wallis tests (e), ns not significant, P > 0.0678, **P < 0.0099, ***P < 0.0006, ****P < 0.0001; scale bars, 5 μm. Source data are provided as a Source Data file.