Virtual screening and molecular dynamics simulation to identify potential SARS-CoV-2 3CL^pro inhibitors from a natural product compounds library

Based on the crystal structure of the 3C-like protease/Nsp5 (PDB ID 6W63), virtual hits were screened from a natural product compounds database-containing 407270 natural products-by using the high-throughput virtual screening (HTVS) module of Discovery Studio software, and then filtering by "Lipinski's rule of five" from the top 20 virtual hits. Two star-hits were selected by CDOCKER results and the protein-ligand interactions with the 3CL(pro )were analyzed. Finally, a 100 ns molecular dynamics simulation was carried out to verify the stability of the receptor-ligand complexes. We screened potent broad-spectrum non-covalent inhibitors that could bind to the SARS-CoV-2 3CL(pro )active binding site from the natural product compounds library through HTVS and molecular dynamics simulations methods. The LibDock scores and -CDOCKER energy value of the star-hits were higher than the original ligands (X77) bound to 3CL(pro). CNP0348829 and CNP0474002, as star-hits, can bind stably to the active site of 3CL(pro), which are promising candidate compounds for the treatment of SARS-CoV-2 and provide a theoretical basis for the development of antiviral drugs. The results of the present study may be useful in the prevention and therapeutic perspectives of COVID-19. However, further in vitro and in vivo validation tests are required in the future.