Amyloid Aggregation under the Lens of Liquid-Liquid Phase Separation

Increasing experiments suggest that amyloid peptides can undergo liquid-liquid phase separation (LLPS) before the formation of amyloid fibrils. However, the exact role of LLPS in amyloid aggregation at the molecular level remains elusive. Here, we investigated the LLPS and amyloid fibrillization of a coarse-grained peptide, capable of capturing fundamental properties of amyloid aggregation over a wide range of concentrations in molecular dynamics simulations. On the basis of the Flory-Huggins theory of polymer solutions, we determined the binodal and spinodal concentrations of LLPS in the low-concentration regime, phi(BL) and phi(SL), respectively. Only at concentrations above phi(BL), peptides formed metastable or stable oligomers corresponding to the high-density liquid phase (HDLP) in LLPS, out of which the nucleated conformational conversion to fibril seeds occurred. Below phi(SL), the HDLP was metastable and transient, and the subsequent fibrillization process followed the traditional nucleation and elongation mechanisms. Only above phi(SL), the HDLP became stable, and the initial fibril nucleation and growth were governed by the high local peptide concentrations. The predicted saturation of amyloid aggregation half-times with increasing peptide concentration to a constant, instead of the traditional power-law scaling to zero, was confirmed by simulations and by a thioflavin-T kinetic assay and the transmission electron microscopy of islet amyloid polypeptide (IAPP) aggregation. Our study provides a unified picture of amyloid aggregation for a wide range of concentrations within the framework of LLPS, which may help us better understand the etiology of amyloid diseases, where the amyloid protein concentration can vary by similar to 9 orders of magnitude depending on the organ location and facilitate the engineering of novel amyloid-based functional materials.

Automated summary

This study investigates the role of liquid-liquid phase separation (LLPS) in amyloid aggregation at the molecular level through molecular dynamics simulations, providing insights into the critical concentrations and saturation behavior of the process. The results offer a unified picture of amyloid aggregation across a wide range of concentrations within the LLPS framework, enhancing the understanding of amyloid diseases and facilitating the development of novel amyloid-based functional materials.