The cellular phase-separated condensates compartmentalize and concentrate biomolecules with distinct physicochemical properties, which has great potential for therapy purposes. However, the discovered phase separation phenomena in living organisms were restricted intracellularly, which limits the biomedical application of phase separation. Here, we designed a phase separation enhanced delivery system to specifically condensate molecules on the cell surface for efficient drug delivery. As a proof of concept, we demonstrated that an anti-cancer drug conjugate can selectively co-phase separate with the targeting component on cancer cell surface and efficiently kill the cells after internalization. Both cellular and in vivo assays showed more potency with this system than traditional antibody conjugates. The method provides insights in the application of phase separation as a powerful tool for therapeutic purpose.

Transthyretin (TTR) aggregation and amyloid formation are associated with several ATTR diseases, such as senile systemic amyloidosis (SSA) and familial amyloid polyneuropathy (FAP). However, the mechanism that triggers the initial pathologic aggregation process of TTR remains largely elusive. Lately, increasing evidence has suggested that many proteins associated with neurodegenerative diseases undergo liquid–liquid phase separation (LLPS) and subsequent liquid-to-solid phase transition before the formation of amyloid fibrils. Here, we demonstrate that electrostatic interactions mediate LLPS of TTR, followed by a liquid-solid phase transition, and eventually the formation of amyloid fibrils under a mildly acidic pH in vitro. Furthermore, pathogenic mutations (V30M, R34T, and K35T) of TTR and heparin promote the process of phase transition and facilitate the formation of fibrillar aggregates. In addition, S-cysteinylation, which is a kind of post-translational modification of TTR, reduces the kinetic stability of TTR and increases the propensity for aggregation, while another modification, S-sulfonation, stabilizes the TTR tetramer and reduces the aggregation rate. Once TTR was S-cysteinylated or S-sulfonated, they dramatically underwent the process of phase transition, providing a foundation for post-translational modifications that could modulate TTR LLPS in the context of pathological interactions. These novel findings reveal molecular insights into the mechanism of TTR from initial LLPS and subsequent liquid-to-solid phase transition to amyloid fibrils, providing a new dimension for ATTR therapy.