Computational discovery of dual potential inhibitors of SARS-CoV-2 spike/ACE2 and M: 3D-pharmacophore, docking-based virtual screening, quantum mechanics and molecular dynamics.

To find drugs against COVID-19, caused by the SARS-CoV-2, promising targets include the fusion of the viral spike with the human angiotensin-converting enzyme 2 (ACE2) as well as the main protease (M). These proteins are responsible for viral entry and replication, respectively. We combined several state-of-the-art computational methods, including, protein-ligand interaction fingerprint, 3D-pharmacophores, molecular-docking, MM-GBSA, DFT, and MD simulations to explore two databases: ChEMBL and NANPDB to identify molecules that could both block spike/ACE2 fusion and inhibit M. A total of 1,690,649 compounds from the two databases were screened using the pharmacophore model obtained from PLIF analysis. Five recent complexes of M co-crystallized with different ligands were used to generate the pharmacophore model, allowing 4,829 compounds that passed this prefilter. These were then submitted to molecular docking against M. The 5% top-ranked docking hits from docking result having scores -8.32 kcal mol were selected and then docked against spike/ACE2. Only four compounds: ChEMBL244958, ChEMBL266531, ChEMBL3680003, and 1-methoxy-3-indolymethyl glucosinolate (4) displayed binding energies 8.21 kcal mol (for the native ligand) were considered as putative dual-target inhibitors. Furthermore, predictive ADMET, MM-GBSA and DFT/6-311G(d,p) were performed on these compounds and compared with those of well-known antivirals. DFT calculations showed that ChEMBL244958 and compound 4 had significant predicted reactivity values. Molecular dynamics simulations of the docked complexes were run for 100 ns and used to validate the stability docked poses and to confirm that these hits are putative dual binders of the spike/ACE2 and the M.