Molecular Dynamics Simulations of G- and F-Actin Explain Aspects of Actin Polymerization

The proper regulation of actin filament (F-actin) dynamics is key for proper cellular and physiological function. Filament dynamics are regulated through a wide array of actin binding proteins, but the nucleotide state of actin itself is also an important factor. Actin filaments polymerize in a head-to-tail fashion, meaning that the filament is polar and the two filament ends are intrinsically different. The barbed end of the filament polymerizes more quickly, and the pointed end dissociates faster. As of yet, neither a F-actin crystal structure nor high resolution cryo-EM structures of either filament end are available, limiting our knowledge of important changes responsible for end-dependent dynamics. To explore filament dynamics, we performed extended molecular dynamics simulations of F-actin. We find that the filament's pointed end is significantly flatter than the remainder of the filament and is in a conformation not sampled by our G-actin simulations - this means that incoming monomers would need to undergo significant conformational changes (isomerization) that would both weaken the binding affinity and slow the rate of polymerization. Conversely, we find that the barbed end of the filament takes on a conformation nearly identical to that of the ATP monomer, enhancing ATP G-actin's ability to polymerize as compared to ADP G-actin. Applying inverse Boltzmann weighting, we can account for the thermodynamic cost imposed by isomerization, and the difference between ATP- and ADP-actin that we calculate exactly matches the values found in experiments.