Biomolecular condensates bridge experiment and theory of mass-conserving reaction-diffusion systems in phase separation

Abstract

Phase separation is crucial in biological processes and ecological resilience, yet experimental evidence supporting validity of mass-conserving reaction-diffusion (MCRD) models in describing phase separation remains rare. Here, we identified one type of biomolecular condensates - double-stranded DNA (dsDNA) and the human transcription factor p53 as forming dsDNA-protein interactive co-condensates (DPICs) as an experimental model, where dynamic DPICs evolve into droplet-like patterns through reversible autocatalytic biochemical reactions. Thus, we can use the MCRD model to describe this experimental system. The pattern formation depends on concentration of protein and dsDNA and their local nonlinear interactions, which were experimentally tested at meso-scale and integrated into an MCRD model. Our results provide direct evidence that the experimental data-driven MCRD models can reproduce the observed phase diagram and scaling-invariance coarsening patterns, thus, providing a compelling mechanism for phase-separation pattern in other systems.