Fusion protein condensate formation via coiled-coil domains

While recent research shows that biomolecular condensates play important roles in normal cellular processes and diseases, the driving forces in condensate formation are not well understood, especially regarding the role of structured self-associative protein domains. In this work, we study the contribution of a model structured domain, coiled-coil domain, in promoting condensate formation of fusion proteins (FPs). Starting from a large set of ~50,000 FPs, we systematically narrowed down to investigate 14 FPs and their corresponding 18 coiled-coil domains. We showed that all 14 FPs are capable of assembling condensates with high potency. When isolated from the rest of the protein contexts, 11 of the 18 coiled-coil domains can induce condensation on their own, despite their short length compared to their full-length counterparts. To understand the differences between coiled-coils that can drive condensate formation and those that cannot, we developed a "triad-extension" model and found the condensate-sufficient coiled-coil domains have a higher propensity to extend beyond perfectly end-to-end matched dimer/oligomer to promote condensate formation.