ABSTRACT
COVID-19 is a persistent public health concern due to the emergence of more virulent and contagious variants resulting from mutations in the spike protein. The spike protein in newer variants, including Delta and Omicron, may be less sensitive to neutralizing antibodies and have a more favorable binding environment to the human ACE2 receptor. In the interest of identifying anti-COVID-19 allosteric drugs, a network-based approach based on coarse-grained molecular dynamics (CGMD) simulations, in complement to pocket-based analysis, is used to identify the possible allosteric pathways of the wild-type, Delta, and Omicron BA.1 spike proteins. Three pockets around 30 Å away from the spike-ACE2 interface are identified underneath the three receptor-binding domain (RBD) chains, which are potentially druggable due to favorable hydrophobicity and surface accessibility. Meanwhile, the network-based approach reveals intrinsic changes within the coupling between the three RBD chains, which could affect the overall communication between the spike-ACE2 interface active site and the three pockets, in particular between the stronger coupling between RBDA and RBDB for the wild type, versus the stronger coupling between RBDA and RBDC in Omicron BA.1. These results are to be used in subsequent drug discovery studies in targeting the spike protein allosterically as part of the search for COVID-19 drugs and as part of the toolbox against future pandemics.
