A Computational Framework for Preferential Switching of Competing Aβ Aggregation Pathways Based on Game Theory Approach

Abstract

Polymorphism among amyloid-β (Aβ) aggregates are emerging as the main contributors to the observed phenotypic diversity in sporadic Alzheimer's disease (sAD). Aggregation of Aβ is susceptible to several heterotypic factors, and lipids (LDs) are the most important physiological agents to interact with, and modulate Aβ aggregation. Therefore, understanding the mechanism of how LDs modulate Aβ aggregation to generate polymorphic strains along competing pathways is crucial for understanding AD pathogenesis. Here, we demonstrate that steady-state switching dynamics between the competing aggregation pathways thar arise due to heterotypic Aβ-LD interactions can be modeled by a novel game theory approach. The game theory model abstracts the mass-action based non-linear, coupled dynamical system of competing pathways of aggregation and the corresponding Nash equilibrium points can predict the preferred pathways as a function of LD parameters and the level of dilution of the medium. While higher LD concentration leads to the generation of more micelle-bound oligomers (promoting the off-pathway of aggregation), higher levels of dilution of the medium result in dissociating micelle bound oligomers enabling higher conversion to fibrils (promoting the traditional on-pathway of aggregation). These models were subsequently validated by our previously developed ensemble kinetics simulations (EKS) that use ordinary differential equations based models of the competing pathways to estimate the rate constants of converting oligomers between on- and off-pathway aggregates; the combined computational framework accurately predicts the temporal rates and stability of oligomers formed under conditions predicted by controlled in vitro biophysical experiments. Hence, our results demonstrate the possibility of switching oligomeric strains between different competing pathways of Aβ aggregation besides providing a novel computational framework to model its dynamics.