Fig. 4. Arc selectively disperses Stg and TARP γ-8, but not TARP γ-3, from the PSD condensates.

a ITC-based measurements showing Arc’s binding to γ-8, γ-3 WT, and γ-3 P226S. 200 µM TARP_CT in syringe was titrated to 20 µM Arc GAG. b Fluorescence images showing the phase separation of PSD-95 (5 µM) and γ-8 (15 µM) with or without addition of Arc (20 µM). c Quantification data showing the droplet numbers formed by γ-8 and PSD-95 in b. Results were from 3 independent batches of imaging assays and presented as means ± SD. d Fluorescence images showing the phase separation of PSD-95 (5 µM) and γ-3 WT or γ-3 P226S (15 µM) with or without addition of Arc (20 µM). e Quantification data showing the droplet numbers formed by PSD-95 and γ-3 WT or γ-3 P226S in d. Results were from 3 independent batches of imaging assays and presented as means ± SD. f Schematic diagram illustrating the imaging assay to simultaneously visualize Stg WT, γ-8, and γ-3 with increasing concentration of Arc. 2.5 µM Stg, 2.5 µM γ-8 and 2.5 µM γ-3 were mixed with 4× PSD condensates formed by PSD-95, GKAP, Shank3 and Homer3 (each at 5 µM). As high concentration of total TARPs led to heterogeneous distributions of different PSD proteins inside each single droplet (data not shown here), we used a special form of GKAP (with Ser-phosphorylation on the third GK-binding-repeat to enhance GKAP’s association to PSD-95) to stabilize our reconstitution system thus allowing all the PSD proteins to perfectly colocalize.⁵³ g Confocal microscope images showing the droplet enrichment changes of Stg WT, γ-8, γ-3 and Shank3 in PSD droplets during the Arc titration. h Statistics analysis showing the enrichment changes of Stg WT, γ-8, γ-3 and Shank3 during the Arc titration in g. Results were from 3 independent batches of imaging assays and presented as means ± SD.