Network-Based Identification and Experimental Validation of Drug Candidates Toward SARS-CoV-2 via Targeting Virus–Host Interactome

Fang, Jiansong; Wu, Qihui; Ye, Fei; Cai, Chuipu; Xu, Lvjie; Gu, Yong; Wang, Qi; Liu, Ai-lin; Tan, Wenjie; Du, Guan-hua

Despite that several therapeutic agents have exhibited promising prevention or treatment on Coronavirus disease-2019 (COVID-19), there is no specific drug discovered for this pandemic. Targeting virus–host interactome provides a more effective strategy for antivirus drug discovery compared with targeting virus proteins. In this study, we developed a network-based infrastructure to prioritize promising drug candidates from natural products and approved drugs via targeting host proteins of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2). We firstly measured the network distances between drug targets and COVID-19 disease module utilizing the network proximity approach, and identified 229 approved drugs as well as 432 natural products had significant associations with SARS-CoV-2. After searching for previous literature evidence, we found that 60.7% (139/229) of approved drugs and 39.6% (171/432) of natural products were confirmed with antivirus or anti-inflammation. We further integrated our network-based predictions and validated anti-SARS-CoV-2 activities of some compounds. Four drug candidates, including hesperidin, isorhapontigenin, salmeterol, and gallocatechin-7-gallate, have exhibited activity on SARS-COV-2 virus-infected Vero cells. Finally, we showcased the mechanism of actions of isorhapontigenin and salmeterol via network analysis. Overall, this study offers forceful approaches for in silico identification of drug candidates on COVID-19, which may facilitate the discovery of antiviral drug therapies.