A structure-based drug design approach for the identification of antiviral compounds targeting the chikungunya virus RdRp protein

Abstract

There are an estimated 3 million Chikungunya virus (CHIKV) infections per year, but no vaccines or specific pharmaceutical treatments are available. RNA-dependent RNA polymerase (RdRp) of CHIKV, encoded by nonstructural protein 4 (nsP4), was targeted to identify potential antiviral compounds in this study. A 3D model of the nsP4 was generated by homology modeling technique and Pockdrug-Server was used to predict ligand binding sites. The pharmacophore hypothesis was developed followed by virtual screening. 300 most potential ligands were selected from a library of 3000 compounds, after filtering for the molecules satisfying Lipinski's Rule of Five. Using molecular docking followed by ADMET analysis, 27 leads were then selected. We used MD simulations of the three best compounds that bind to the catalytic, palm, and thumb domains to find out how the ligand-protein interactions change over time. RMSD, RMSF, radius of gyration, free energy landscapes, and principal component analyses revealed that complexes nsP4-ZINC-12820763 and nsP4-ZINC-33280972 were more stable. Simulated trajectories were further used to compute MMGBSA binding energies of PubChem-135638918 (-22.33 kcal/mol), ZINC-12820763 (-50.19 kcal/mol) and ZINC-33280972 (-32.69 kcal/mol) bound to catalytic, palm and thumb domains, respectively. HOMO and LUMO of potential molecules were also investigated using density functional theory computations. With the support of computational simulations, three compounds were shown to have the potential to inhibit RdRp of CHIKV and need further evaluation using in-vitro and in-vivo analysis.

Keywords

• Chikungunya
• Homology modeling
• E-pharmacophore
• Molecular docking
• Molecular dynamics