Infection by the novel coronavirus, SARS-CoV-2, involves the attachment of the receptor binding domain (RBD) of its spike proteins to the ACE2 receptors of the host cells. The binding process is initiated by a "down" to "up" conformational change in the spike protein, presenting the RBD to the receptor. Majority of computational and experimental studies for the design of therapeutics targeting the spike protein have focused on the RBD. However, the RBD region is highly prone to mutations, and therefore is likely to give rise to drug resistance. In contrast, we here focus on the correlations between the RBD and other domains of the spike protein. We performed molecular dynamics (MD) simulations to gain a deeper understanding of the role of the "distant residues", outside the RBD, in the molecular mechanism of the infection process. We, thereby, make predictions of key residues in distant allosteric binding sites which can be targeted in future drug design efforts. The model we propose, based on time-independent component analysis (tICA) and protein graph connectivity network, is able to identify multiple residues that exhibit long-distance coupling with the RBD opening. Mutation on these residues can lead to new strains of coronavirus with different degrees of transmissibility and virulence. The most ubiquitous D614G mutation, and the A570D mutation of the highly contagious B.1.1.7 lineage (the UK variant) of SARS-CoV-2, are predicted "ab-initio" from our model, without using any sequence or mutation profile data. Identifying these emerging mutants may prove to be an important step towards pre-designing vaccines for future outbreaks. Conversely, broad spectrum therapeutics like drugs and monoclonal antibodies can be generated targeting these key distant but conserved regions of the spike protein.<br/><br/>We also explore the dynamics of the RBD-antibody complexes, using MD simulation, for the wild type SARS-CoV-2 spike protein as well as two mutants: the B.1.1.7 (the UK variant), and the B.1.351 (the South African variant) strain. Using principal component analysis (PCA) and protein graph connectivity network, we quantify the differences of the interaction between the spike and the antibodies caused by the mutations in RBD, including N501Y, K417N, and E484K. This study is the initial step towards computer-aided rational design of monoclonal antibodies to combat the infections from future strains of coronavirus containing new mutations.

Probing the dynamical couplings in the SARS-CoV-2 spike protein using unsupervised learning and protein graph connectivity network

ACS Fall 2021

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.