Protein condensates formed via liquid-liquid phase separation (LLPS) show potential as biomaterials due to their dynamic nature and their ability to highly concentrate specific biomolecules. However, uncontrolled coalescence and interfacial instability hinder their broader application. Here, we demonstrate that engineering the condensate surface provides a direct route to designing well-defined biomaterials by controlling their size, stability, and functionality. First, we stabilize condensate surfaces using protein cages that act as surface-stabilizing agents, a concept inspired by Pickering emulsions. By adjusting the cage-condensate interaction strength and the cage-to-condensate ratio, we developed a one-pot method to produce coalescence-resistant droplets with defined diameters from the nano- to micro-scale. Stabilized droplets retain dynamic liquid properties, including selective molecular exchange. These features support their use as size-controlled, surface-engineered protein condensate materials. Second, we use condensates as two-dimensional interfacial platforms to assemble higher-order protein shells. The condensate surface first recruits specific proteins and then templates their cross-linking into a continuous network through strong lateral interactions. Subsequent dissolution of the internal liquid core yields hollow protein shells with clear compartmentalization, offering opportunities as nano-reactors or delivery vehicles. Our work shows that the condensate interface can be rationally engineered for materials design. We present two distinct strategies—stabilizing dynamic liquid droplets and templating the assembly of hollow protein shells. This approach presents a new direction for protein condensates, expanding their role from biological phenomena to a platform for materials engineering.
Oh et al. (2026) studied this question.