Targeting SARS-CoV-2 main protease (3CLpro) with Paeonia-derived phytochemicals

Abstract

Coronavirus Disease 19 (COVID-19), caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), stimulated intensive drug-development efforts throughout the world. As COVID-19 has become endemic in many regions of the world, phytochemicals used in ethnomedicine may have utility in the prevention or treatment of COVID-19. Here, we employed machine learning/deep learning drug discovery tools to evaluate compounds previously identified in roots of plants of the Paeonia genus for their drug-like properties and potential to inhibit 3CLpro, the main protease of SARS-CoV2. Our results identify Paeonia-derived compounds that have favorable drug-like properties and are predicted to have a high binding affinity for 3CLpro. Molecular dynamics simulations supported the binding affinity of Paeonia-derived compounds for 3CLpro. We validated these in silico computational results by experimentally determining the 50% inhibitory concentration (IC50) of two Paeonia-derived compounds, paeoniflorigenone and 3-O-methylquercetin, using purified 3CLpro with baicalein as a control. The IC50 values for paeoniflorigenone, 3-O-methylquercetin and baicalein were 9.33 µM, 43.94 and 43.85 µM, respectively. Root extracts from five Paeonia species were found to have minimal cytotoxicity against three different human cell types. Our results represent a proof-of-concept study demonstrating that in silico techniques, including machine and deep learning methods, can be used to identify phytochemicals as starting points for the discovery of antiviral compounds against SARS-CoV-2.