SARS-CoV-2 Glycosylated Spike Activation Mechanism - Simulations of the Full Unbiased Pathway

Abstract

The SARS-CoV-2 virus is responsible for the COVID-19 pandemic and has taken over one million lives do date. Infection occurs through glycosylated spike trimers, which protrude the surface of viral particles and bind to human ACE2 receptors. For the spikes to bind ACE2, the receptor binding domains (RBD)s must transition from a closed to open state. To access the biological timescales of this transition, we used the weighted ensemble (WE) enhanced sampling method to generate unbiased, all-atom molecular dynamics simulations of the glycosylated spike undergoing RBD activation. Rather than adding an external biasing force, the WE method relies on running many short simulations in parallel along chosen reaction coordinate(s). The trajectories that rarely sample high energy regions are replicated, while the trajectories that frequently sample low energy regions are merged, focusing computational resources on sampling rare events. The trajectories also carry probabilities or weights, which are continuously updated, and there is no statistical bias added to the system. Hence, we are able to directly obtain both thermodynamic and kinetic properties from the WE simulations, which are not possible to accurately predict from most other enhanced sampling methods. From these simulations, we have uncovered the primary glycans and residues responsible for facilitating the activation mechanism, identified the internal barriers for initiating RBD opening, and obtained extensive sampling of spike conformations. These conformations can be used to train machine learning algorithms, test the effect of perturbations (mutations, small molecules, etc.) and for in silico design of SARS-CoV-2 Spike targeting therapeutics.