In silico validation of allosteric inhibitors targeting Zika virus NS2B–NS3 protease

Abstract

The Zika virus (ZIKV), a member of the Flaviviridae family, poses a major threat to human health because of the lack of effective antiviral drugs. Although the NS2B–NS3 protease of ZIKV (NS2B–NS3pro) is regarded as a major target for antiviral inhibitors, viral mutations can lead to ineffective competitive inhibitors. Allosteric inhibitors bind to highly conserved nonprotease active sites, induce conformational changes in the protease active site, and prevent substrate binding. Currently, no molecular simulation techniques are available for accurately predicting and analysing conformational changes in the protease catalytic domain. In this study, we developed a combined approach that involves blind docking, Gaussian accelerated molecular dynamics, two-dimensional potential of mean force profiling, density functional theory (DFT) calculations, and interaction region indicator (IRI) analysis and employed it to examine the allosteric inhibitor-01 molecule and its interaction with ZIKV NS2B–NS3pro. Our results indicated that the binding of inhibitor-01 to NS2B–NS3pro resulted in two major conformational states, state I and state II, which in turn changed the volume of the protease active site from 1014 ų to 710 and 820 ų, respectively. These two states had an inactive catalytic domain (residues His116, Asp140, and Ser200). DFT and IRI analyses revealed that, in state I, Lys138 and Gln139 formed hydrogen bonds with inhibitor-01, whereas Lys138, Leu214, Asn217, Val220, and Ile221 engaged in van der Waals interactions with inhibitor-01. Advancements in computational techniques and power are expected to facilitate further progress in overcoming challenges associated with designing allosteric inhibitors for viral proteases.