Data mining guided affinity ultrafiltration for rapid screening of SARS-CoV-2 3CL inhibitors in medicinal herbs.

Clinical studies have demonstrated that many traditional Chinese medicines (TCM) exhibit not only antiviral effects but also efficacy in alleviating clinical symptoms of Coronavirus Disease 2019 (COVID-19). However, the pharmacologically active constituents responsible for their anti-COVID-19 efficacy remain unclear. This study aimed to establish a novel strategy for identifying active ingredients from herbs clinically used for COVID-19. An integrated approach combining data mining with Affinity Ultrafiltration (AUF) was developed. Initially, using a data mining strategy, high-frequency herbs were selected from the herbs clinically used in COVID-19. Furthermore, AUF technology was used to screen for potential bioactive components from the high-frequency herbs. The anti-COVID-19 potential of active compounds was assessed through enzyme activity assays and molecular docking, followed by validation in cellular and animal models. Data mining revealed that Glycyrrhiza uralensis Fisch., Lonicera japonica Thunb. and Forsythia suspensa (Thunb.) Vahl were the high-frequency herbs used in COVID-19. Five potential 3-Chymotrypsin-like protease (3CL) inhibitors were screened from three herbs via AUF and identified using high-resolution mass spectrometry. Further enzymatic and cellular assays demonstrated that Licochalcone C (LCC) and Forsythiaside A (FTA) inhibited Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) replication at micromolar concentrations. Notably, FTA treatment significantly suppressed the elevation of pulmonary parameters and inflammatory mediators induced by SARS-CoV-2 nucleocapsid protein in mice. In summary, this study proposes a novel strategy integrating data mining with AUF to discover active compounds from TCM. Two components, LCC and FTA, were identified as dose-dependent inhibitors of SARS-CoV-2 Omicron strain replication in vitro.