Size scaling of phase-separated domains and mesoscale clusters that precede liquid-liquid phase separation (LLPS): theory and experiment

We discuss observations* of large (10’s-100 nm’s) protein assemblies in the one-phase regime that precedes LLPS, using an analytical theory of these protein assemblies based on analogies with other mesoscale structures in amphiphilic systems with core-shell assemblies. What is unique about intrinsically disordered proteins is that the same protein can – via its different conformations -- act as both the “inner phase” and “amphiphilic surface layer.” This is consistent with the experiments that identified two types of dynamics associated with the “clusters”. Thus, relatively large assemblies can be stable for even a single protein with no amphiphile or “internal phase” . We formulate a statistical mechanics model of such core-shell assemblies to predict the size distribution of the observed “clusters” in the one-phase region and compare it with the results of light scattering experiment. The data for relatively large clusters is well-fit by a model with interfacial tension, while the fits for smaller clusters must also account for the bending energy and geometric corrections. In addition, electron spin-resonance experiments estimate the core-shell volume ratios, indicating that at the LLPS transition, there is no sharp change in the rotational time scales of the proteins in the core and shell. This may suggest that, in these cases, the LLPS may arise from attraction-induced phase separation of the “clusters,” similar to phase separations in microemulsions.

*Manas Seal, Maria Oranges, Alexey Bogdanov, Daniella Goldfarb: Dept. Chemical and Biological Physics, Weizmann Institute of Science