A combination of MD-derived pharmacophores and structure-based strategy identifies dual allosteric inhibitors of NS2B-NS3 proteases in Zika and West Nile viruses

Starting from the crystal structure of a Zika virus (ZIKV) NS2B-NS3 protease with bound allosteric inhibitor, we generated dynamic pharmacophore models derived from representative molecular dynamics (MD) conformations to capture key interaction patterns. These models guided a pharmacophore-based virtual screening of commercial libraries, including commercial compounds and approved drugs, enabling the rapid selection of virtual hit compounds. Biochemical evaluation identified several inhibitors, with the antiretroviral drug Nelfinavir emerging as the most promising compound. Nelfinavir displayed the strongest inhibition of ZIKV and West Nile virus (WNV) proteases and showed antiviral activity in cell-based assays, with EC50 values of 1.79 ± 0.01 μM on ZIKV and 3.01 ± 1.25 μM in WNV, supporting its relevance as a repositionable scaffold for modulation of ZIKV and WNV infections. The activity observed across flavivirus proteases highlights the value of targeting conserved allosteric regions. Overall, this integrated computer-aided drug design (CADD) strategy demonstrates the power of dynamic, structure-based pharmacophores to uncover novel allosteric inhibitors and accelerate antiviral drug repurposing.