Insights into protein aggregation revealed by hybrid-resolution simulations and kinetic network analysis


Uncontrollable aggregation of proteins is a central hallmark of numerous neurodegenerative diseases. During the aggregation, soluble proteins are converted into aggregates rich in beta structures through a series of molecular events of protein-protein interactions and structural transitions. Understanding these events in atomic details is invaluable for devising disease treatment. Molecular simulations are in general useful in characterizing atomic details of biomolecular processes but remains limited in study of protein aggregation due to the need to perform simulations at very long timescales. To overcome this challenge, we have developed a multiscale model that mixes representations at different resolutions and can simulate efficiently structural variations of proteins with atomic accuracy. Analyses of these simulations with kinetic network models further enable characterization of complex free energy landscapes of protein aggregation. In this talk, I will present our efforts in applying this approach to understand two key events of aggregation, namely nucleation and elongation, for amyloid-beta proteins associated with the Alzheimer’s Disease. Our studies reveal full atomic pictures of structural transitions during these processes, explaining puzzling experimental observations regarding regional selectivity in Abeta for nucleation and unidirectional growth of Abeta aggregates. In addition, I will also discuss our recent computational results based on which a potential inhibitor of aggregation has been identified.