Computational Studies of Protein Aggregation Kinetics: From Stochastic Kinetic Modeling to Understanding Molecular Mechanisms

Yuan-Wei Ma¹, Jia-Liang Shen¹, Tong-You Lin¹, Min-Yeh Tsai^1*

¹Chemistry, Tamkang University, New Taipei City, Taiwan

* Presenter:Min-Yeh Tsai, email:victorleaf@gmail.com

Protein aggregation includes several mechanistic steps, including nucleation, elongation, and secondary processes. These processes are important to understanding the pathogenesis of many diseases. However, the nucleation and growth of protein aggregates are difficult to quantify efficiently due to their intrinsically stochastic nature. This is often encountered in vivo where the number of protein copies is small (N≈O(10²-10³)). Therefore, the computation of stochastic aggregation is a challenging task. In this talk, I will present our recent progress in modeling stochastic aggregation kinetics. Using the chemical Langevin equation approach, we introduce noise into already statistically averaged equations obtained using mathematical moment closure schemes. The resulting stochastic moment equations allow the efficient calculation of the moments of the polymeric species distributions for all times. The simulation reveals a scaling law that correlates the size of fluctuation with the total number of monomers. In addition to the kinetic modeling, we also carry out coarse-grained molecular dynamics simulations to investigate the molecular mechanisms of fibril surface-mediated amyloidogenesis. Our molecular simulations reveal several monomeric amyloid precursors that stick along the fibril surface. They are potential precursors responsible for fibril elongation and secondary nucleation. Notably, the configurations of these precursors display different diffusive dynamics along the surface, indicating multiple timescales of fibril growth. We also find that fibrillar twisting is an emergent property. Our simulation results support the notion that twisted fibrils utilize the orientation of monomers to tune their diffusive dynamics to biologically relevant timescales.

Keywords: Protein aggregation, MD simulation, Nucleation, Stochastic models, Amyloid-beta