Impact of Early Pandemic Stage Mutations on Molecular Dynamics of SARS-CoV-2 M ^pro

Abstract

A new coronavirus (SARS-CoV-2) is a global threat to world health and economy. Its dimeric main protease (M^pro), which is required for the proteolytic cleavage of viral precursor proteins, is a good candidate for drug development owing to its conservation and the absence of a human homolog. Improving our understanding of M^pro behavior can accelerate the discovery of effective therapies to reduce mortality. All-atom molecular dynamics (MD) simulations (100 ns) of 50 mutant M^pro dimers obtained from filtered sequences from the GISAID database were analyzed using root-mean-square deviation, root-mean-square fluctuation, <i>R</i>_g, averaged betweenness centrality, and geometry calculations. The results showed that SARS-CoV-2 M^pro essentially behaves in a similar manner to its SAR-CoV homolog. However, we report the following new findings from the variants: (1) Residues GLY15, VAL157, and PRO184 have mutated more than once in SARS CoV-2; (2) the D48E variant has lead to a novel "TSEEMLN"" loop at the binding pocket; (3) inactive apo M^pro does not show signs of dissociation in 100 ns MD; (4) a non-canonical pose for PHE140 widens the substrate binding surface; (5) dual allosteric pockets coinciding with various stabilizing and functional components of the substrate binding pocket were found to display correlated compaction dynamics; (6) high betweenness centrality values for residues 17 and 128 in all M^pro samples suggest their high importance in dimer stability-one such consequence has been observed for the M17I mutation whereby one of the N-fingers was highly unstable. (7) Independent coarse-grained Monte Carlo simulations suggest a relationship between the rigidity/mutability and enzymatic function. Our entire approach combining database preparation, variant retrieval, homology modeling, dynamic residue network (DRN), relevant conformation retrieval from 1-D kernel density estimates from reaction coordinates to other existing approaches of structural analysis, and data visualization within the coronaviral M^pro is also novel and is applicable to other coronaviral proteins.