Non-equilibrium modeling of flux through biomolecular condensates

Abstract

Cellular biomolecular condensates function far from equilibrium, sustained by continuous influx and efflux of molecular components. Notable examples include the nucleolus, P-bodies, and stress granules, in which regulated molecular transport is essential for their function. Experimental characterization of molecular flux through condensates remains challenging due to limitations in spatiotemporal resolution, making computational approaches a powerful alternative for systematic investigation. Here, we present a computational approach (TRACE) to model molecular transport through biomolecular condensates under non-equilibrium steady-state conditions at residue-level resolution. Using TRACE, we systematically probe physicochemical factors that govern molecular flux through model condensates. We find that protein sequence composition and patterning determine the internal `mesh size' of the condensates, which strongly dictates molecular transport efficiency, and the ballistic-to-diffusive dynamics crossover of the fluxing molecules. We also find that molecule stiffness can enhance their flux by exposing larger surface for intermolecular interactions. Furthermore, we find that modulating host-guest interactions can modulate solubility and friction which jointly governs molecular flux. Finally, we use our framework to show that crosslinking reduces condensate pore size, hindering molecular flux -- consistent with experimental observations of dextran diffusion through nucleolar phases. Collectively, these insights provide a basis for rationally tuning condensate-mediated molecular flux, with potential applications in therapeutics and soft material design.