Born to Condense: Polysomes Drive Co-Translational Condensation of Biomolecular Condensate Proteins

Biomolecular condensates formed by protein liquid-liquid phase separation (LLPS) are ubiquitous in cells and play crucial roles in cellular regulation. While the physics and functions of LLPS are well studied, its interplay with protein synthesis - translation - remains largely unexplored. Here we introduce a theoretical framework for Co-Translational Condensation (CTC), in which nascent protein chains of polysomes - multiple ribosomes on one mRNA - interact with condensates, localizing translation to condensate surfaces. Using coarse-grained simulations, we show that protein domain architecture dictates the thermodynamics of CTC, consistent with a Langmuir adsorption model. Bioinformatic analysis of more than 7,500 proteins reveals that most condensate-associated proteins have architectures favoring CTC, with strong interaction regions of nascent chains exposed on polysomes. At the dynamical level, simulation and reaction-diffusion modeling reveal that CTC is kinetically feasible within typical polysome lifetimes, either through large polysomes nucleating new condensates or via diffusion to pre-existing condensates. As a case study, we demonstrate that CTC enhances post-translational modifications by minimizing unmodified intermediates. More broadly, we anticipate CTC may also influence protein folding, misfolding, and signal-integration latency. Together, our results establish CTC as a general mechanism coupling translation with phase separation, with broad implications for protein evolution, cellular organization, and synthetic biology.