BPS2026 – Inferring the surface properties of protein clusters that precede LLPS from their size distributions

Proteins with intrinsically disordered domains often display complex self-assembly behavior, including the formation of large, reversible clusters in regimes where bulk liquid-liquid phase separation (LLPS) is absent. These clusters, typically 30–300 nm in size, are markedly different from the transient oligomers found in small-molecule systems (e.g., oil-water LLPS). Instead, proteins such as CPEB4 and FUS assemble into mesoscale clusters comprising hundreds to thousands of monomers (Kar et al. PNAS 2022), raising fundamental questions about their stability and role in phase separation. We have proposed that protein conformational heterogeneity underlies this phenomenon, with distinct conformations stabilizing either the dense cluster core or its surrounding shell (Golani et al. Comm. Phys. 2025). Guided by insights from electron paramagnetic resonance (EPR) spectroscopy (Oranges et al. Biophys. J. 2024), we developed a mesoscopic theoretical framework that treats protein clusters analogously to amphiphilic systems, where core-shell assemblies are well-established. By fitting size distributions measured with dynamic light scattering (DLS) and nanoparticle racking analysis (NTA), we extracted key physical parameters governing the mesoscale cluster organization and size distribution: interfacial tension, solubility energy, bending rigidity, and spontaneous curvature. We now consider polymeric models of these proteins that include both hydrophilic and less hydrophilic regions to provide more molecular insight into the mesoscale cluster properties. We suggest that the surface properties and conformational protein states within these clusters are important for understanding the aging behavior of biomolecular condensates in cells and may also play an important role in regulating their functions within the cellular environment.