(89d) A Systematic Multiscale Coarse-Grained Model of Actin Filament Networks

Actin cytoskeletal network, the fundamental framework of a cell, is critical for a cell’s ability to move, divide, and transport cargo. Actin filaments constantly assemble, disassemble, sever, bundle and form branches in the presence of a number of actin-binding proteins (ABPs) giving rise to a dynamic network. The ABP mediated remodeling of these networks is governed by conformation-dependent interactions between actin and the ABPs. These conformational changes typically involve multiple timescales and therefore, simulating the entire network at a cellular level in atomistic resolution is computationally infeasible. In this regard, a multiscale modeling framework was previously developed in our group [Katkar et al., Biophysical Journal, 115, 1589-1602, 2018] to gain insights into nucleotide hydrolysis in a single actin filament.

In this work, we extend the framework to systematically simulate the dynamic remodeling of actin filament networks, both in the absence and presence of ABPs such as the arp23 complex. We adopt the Ultra-Coarse-Grained (UCG) methodology where the coarse-grained (CG) models undergo stochastic transitions between different states based on CG particle conformations during the simulations. Here, the transition between states is used to mimic processes that actin subunits undergo, such as polymerization, depolymerization, nucleotide hydrolysis, binding to ABPs etc. The model is parameterized using short atomistic simulations corresponding to each state and the average transition rates are tuned to agree with known macroscopic experimental rates. Using this model, we investigate the filament assembly under varying conditions of actin and ABP concentrations. Our systematic parameterization strategy allows us to make predictions pertaining to the cooperative effects between nucleotide hydrolysis and filament network remodeling.