LLPS of FXR1 drives DMV clustering. (A and B) FRAP results in cells show that after photobleaching the whole punctum (arrows), most of the fluorescent signal of GFP-FXR1 recovers, while the mCherry-Nsp3 signal fails to recover (A). Quantification data are presented as mean ± SEM (n = 12) in B. Bars: 2 μm. (C) APEX2-FXR1 signal is detected surrounding clustered DMVs in cells. Bars: 500 nm. (D) FXR1 undergoes LLPS in vitro. Bar: 5 μm. (E) The schematic shows the domain organization of Nsp3. Nsp3N is indicated by the dashed box. (F) Nsp3N partitions into FXR1 liquid droplets in vitro. Bar: 5 μm. (G) Top row: In the absence of Nsp3N, FXR1 droplets (cyan) are associated with a few liposomes (green) in vitro. Bottom row: FXR1 droplets capture high levels of Nsp3N-decorated liposomes. Bars: 5 μm. (H) In a GFP-Trap assay, FKBP-GFP-FXR1(L351P) immunoprecipitates comparable levels of FRB-mCherry-Nsp3 as WT proteins treated with 1.5 μM rapamycin. Quantifications of levels of Nsp3 (normalized by FXR1 levels) are shown. (I and J) Re-expressing RNAi-resistant FKBP-GFP-FXR1, but not FKBP-GFP-FXR1(L351P), induces big Nsp3/4⁺ puncta in siFXRs cells (I). Numbers of large and all Nsp3/4⁺ puncta are shown as mean ± SD (NC, n = 18; FKBP-GFP-FXR1, n = 18; FKBP-GFP-FXR1(L351P), n = 12) (J). ****, P <0.0001; N.S., not significant. Bars: 5 μm; insets, 2 μm. Source data are available for this figure: SourceData F3.