AI-assisted identification of a novel Orthopoxvirus inhibitor targeting F13

Abstract

Treatment of mpox virus (MPXV) is crucial for public health. However, research into drugs for MPXV has fallen behind, particularly in anticipation of sudden outbreaks. This study aimed to identify new inhibitors of Orthopoxvirus using artificial intelligence (AI)-assisted methods. We explored AlphaFold v2.0 to simulate the F13 protein structures of MPXV, vaccinia virus (VACV), and variola virus (VARV). Utilizing MOE2019 software, we identified amino acid binding sites suitable for small molecule docking, focusing on a phosphodiesterase active site pocket in F13. Our efforts led to the identification of JCS-2022, a promising new inhibitor that exhibited docking similarities with the known anti-poxvirus drug tecovirimat. In vitro experiments demonstrated that JCS-2022 had a half maximal effective concentration (EC₅₀) of 0.05430 μmol/L (μM), comparable to tecovirimat’s EC₅₀ of 0.04794 μM. At a dosage of 1.6 μM, JCS-2022 significantly reduced VACV plaque size, indicating effective inhibition of extracellular enveloped virus (EEV) formation. Immunofluorescence analysis confirmed a reduction in VACV-induced actin tail formation. Our findings suggest that JCS-2022 is a strong candidate for development as a small molecule inhibitor against Orthopoxvirus, highlighting the potential of AI-assisted methods in accelerating drug discovery for infectious diseases.