Lassa virus (LASV) is a hemorrhagic fever arenavirus of significant public health concern, infecting millions of people per year in Africa. Here, we developed a computational strategy to identify specific inhibitors of LASV glycoprotein-mediated virus cell entry, leveraging a previous screen of 297,156 small molecules from the MLPCN library, with the results deposited in the PubChem database. Data mining methods were developed to efficiently select small molecules prioritized for both potency and specificity in inhibiting Lassa virus infection. Cheminformatics classification then identified diverse chemical scaffolds that had not been previously reported. Representatives were evaluated against authentic LASV infection, yielding potencies as low as 10 nM. Investigation of the target mechanism compared vesicular stomatitis virus bearing the GPs of LASV, the distantly related Junín virus, and unrelated Ebola virus. The results identified 3 distinct chemical scaffolds that demonstrated strong LASV selectivity, each acting at the membrane fusion stage of virus cell entry. Sensitivity was mapped to regions of GP2 known to coordinate pH-triggered conformational rearrangements needed for membrane fusion. Time-of-addition experiments demonstrated loss of activity coincident with endosomal escape, and cell-cell fusion assays confirmed direct inhibition of GP-mediated syncytia formation. Together, these findings characterize each scaffold as an effective LASV fusion inhibitor and highlight the effectiveness of our combined computational and experimental approaches in identifying mechanistically informative antiviral scaffolds.
Close et al. (Mon,) studied this question.