RNA is a critical element for the sizing and the composition of phase-separated RNA-protein condensates

Abstract

Liquid-liquid phase separation is thought to be a key organizing principle in eukaryotic cells to generate highly concentrated dynamic assemblies, such as the RNP granules. Numerous in vitro approaches have validated this model, yet a missing aspect is to take into consideration the complex molecular mixture and promiscuous interactions found in vivo. Here we report the versatile scaffold ArtiG to generate concentration-dependent RNA-protein condensates within living cells, as a bottom-up approach to study the impact of co-segregated endogenous components on phase separation. We demonstrate that intracellular RNA seeds the nucleation of the condensates, as it provides molecular cues to locally coordinate the formation of endogenous high-order RNP assemblies. Interestingly, the co-segregation of intracellular components ultimately impacts the size of the phase-separated condensates. Thus, RNA arises as an architectural element that can influence the composition and the morphological outcome of the condensate phases in an intracellular context.

Fig. 1

FFm scaffold forms concentration-dependent condensates in living cells. a Schematic of POI-FFm multivalent self-interacting scaffold. Genetically engineered ferritins assemble into ArtiG^POI through a phase separation process (Fm = F36M-FKBP, POI = protein of interest, F = Ferritin, ArtiG = ArtiGranule). b Representative confocal image of ArtiG^mCh (red) in HeLa cells, 24 h after transfection of mCherry-FFm construct. Nuclei were stained with Hoechst (blue). Scale bar, 10 µm. c Representative time-lapse confocal images of ArtiG^mCh (white) nucleation and growth in HeLa cells expressing mCherry-FFm construct. Scale bar, 10 µm. d Representative time-lapse confocal images of the growth of an isolated condensate (white). Scale bar, 2 µm. e Representative time-lapse confocal images of several ArtiG^mCh (white) coalescing into a larger condensate. Scale bar, 2 µm. f Comparison of the temporal evolution of the dilute cytosolic mCherry-FFm fluorescence (dilute phase), with the total cytoplasmic fluorescence (dilute + dense phases) for the lower cell shown in c. The violet dots represent the number of granules measured as a function of time. For the purpose of representation, the images in c, d, and e are slightly saturated

Fig. 2

ArtiG^mCh recapitulate the main properties of phase-separated liquid droplets. a Ultrastructure of an individual ArtiG^mCh (i, black arrow) and of two ArtiG^mCh (ii, black arrows) undergoing fusion in HeLa cells, 24 h after transfection of mCherry-FFm construct. ER endoplasmic reticulum, M mitochondrion, LD lipid droplet, PM plasma membrane. Scale bar, 500 nm. b Temporal evolution of the distance between two ArtiG^mCh fusing with each other and relaxing into a single droplet within a characteristic time of 40 min. Insert: time-lapse of the corresponding fusion event. Scale bar, 2 µm. c Plots of the characteristic relaxation time of different fusion events as a function of the diameter, with the slope giving the capillary velocity (ratio between the surface tension γ and the viscosity η). d Example of recovery of fluorescence intensity after photobleaching of a 2 µm ArtiG^mCh. e Kymograph representation of the fluorescence recovery of the ArtiG^mCh analyzed in d. f Dissolution of ArtiG^mCh upon addition of 2.5 µM of FK506 to the culture medium. Comparison of the temporal evolution of the total cytoplasmic fluorescence (dilute and dense phases) with the dilute cytosolic mCherry-FFm fluorescence (dilute phase) as a function of time

Fig. 3

ArtiG^PUM recruit specific endogenous RNAs. a Representative time-lapse confocal images of HeLa cells expressing mCherry-FFm and PUM.HD-FFm constructs. Scale bar, 10 µm. For the purpose of representation, the images are slightly saturated. b Representative confocal image of ArtiG^mCh/PUM (red) in HeLa cells, 24 h after transfection of mCherry-FFm and PUM.HD-FFm constructs. Nuclei were stained with Hoechst (blue). Scale bar, 10 µm. c To visualize polyadenylated RNA, a PABP-EGFP fusion (white) was co-transfected with mCherry-FFm and PUM.HD-FFm (red) into HeLa cells 24 h before fixation. Confocal imaging. Scale bar, 10 µm. Zoom, 2 µm. d Epifluorescence imaging of NORAD lncRNA following smFISH-Cy3 (red in merge) in HeLa cells expressing ArtiG^EGFP (green in merge, upper row) and ArtiG^EGFP/PUM (green in merge, lower row). Nuclei were stained with DAPI (blue). Scale bar, 10 µm. Zoom, 2 µm

Fig. 4

RNA modifies ArtiG^PUM composition, nucleation incidence, and morphology. a HeLa cells expressing ArtiG^mCh and ArtiG^mCh/PUM (red) were fixed and analyzed by immunofluorescence using antibodies recognizing endogenous DDX6 (white). Confocal images. Scale bar, 10 µm. Zoom, 2 µm. b HeLa cells expressing ArtiG^mCh and ArtiG^mCh/PUM (red) were treated or not with arsenite for 30 min, fixed, and analyzed by immunofluorescence using antibodies recognizing endogenous ATXN2L (white). Enlarged regions of the confocal images presented in Supplementary Fig. 4c. Scale bar, 2 µm. c As indicated, ferritin and endogenous DDX6 were identified in thin sections of ArtiG^mCh/PUM-expressing cells, using a secondary antibody coupled to 10 nm gold particles. Ferritin is concentrated all-over ArtiG^mCh/PUM (upper panel), whereas DDX6 accumulates at the edge of the condensates (lower panels, black arrows). Scale bar, 500 nm. ER endoplasmic reticulum, M  mitochondrion, Nu nucleus, PM plasma membrane. d Percentage of transfected cells displaying ArtiG condensates 24 h after transfection of mCherry-FFm and PUM.HD-FFm constructs. Data represent the mean ± SD of three independent experiments. e Quantification of the size distribution of ArtiG^mCh and ArtiG^mCh/PUM. Dots represent single measurements of ArtiG size and are pooled from three independent experiments. Means and SDs are superimposed

Fig. 5

RNA is a critical element for ArtiG^PUM size scaling. a Representative confocal images of ArtiG^mCh/PUM (white) in HeLa cells, 24 h after transfection of mCherry-FFm and PUM.HD-FFm constructs in a plasmid ratio of 1:1 (i), 5:1 (ii), 10:1 (iii), and 1:0 (iv). Nuclei were stained with Hoechst (blue). Scale bar, 10 µm. b Quantification of the size distribution of ArtiG^mCh/PUM and ArtiG^mCh described in a. Dots represent single measurements for ArtiG size and are pooled from three independent experiments. Means and SDs are superimposed. ****p < 0.0001. For representation, the two largest ArtiG^mCh ( > 15 µm, as shown in Fig. 4e) were not included. c Epifluorescence images of NORAD lncRNA following smFISH-Cy3 (red) in HeLa cells expressing ArtiG^EGFP/PUM, ArtiG^EGFP, and ArtiG^EGFP/PUMdCT (green). The panels show enlarged regions of the cells represented in Supplementary Fig. 5b. Nuclei were stained with DAPI (blue). Scale bar, 2 µm. d Representative confocal image of ArtiG^mCh/PUMdCT (white) in HeLa cells, 24 h after transfection of mCherry-FFm and PUMdCT-FFm constructs in a plasmid ratio of 1:1. Nuclei were stained with Hoechst (blue). Scale bar, 10 µm. e Schematic of the model for ArtiG^PUM nucleation, growth, and co-segregation of endogenous RNPs