Discovery and activity evaluation of fungal chitin synthase inhibitors via virtual screening

Chitin synthase is a fungal-specific and highly selective antifungal target, and its inhibitors represent a promising direction for antifungal drug development. In this study, a structure-based high-throughput virtual screening was performed against a compound library comprising 1.6 million molecules, using the high-resolution cryo-electron microscopy structure of Candida albicans chitin synthase 2 (CaChs2) as the molecular target. Based on the virtual screening results, a machine learning-based classification model was constructed to identify key structural features associated with high CaChs2 binding potential. The top three compounds with the most favorable binding energies were subsequently selected for in vitro enzyme inhibition assays. All three compounds exhibited appreciable inhibitory activity against CaChs2 and showed improved inhibitory potency compared with the nucleoside inhibitor Polyoxin B (PolB). Among these, Mol_1 exhibited the strongest inhibitory activity, with an IC₅₀ value of 20.89 μM. Additionally, it effectively restored the susceptibility of resistant strains to fluconazole (FZ), reducing the minimum inhibitory concentration (MIC) of FZ in clinically derived resistant strains from >512 μM to 32 μM. The molecular docking and molecular dynamics (MD) simulations shows that Mol_1 could stably bind to the catalytic site of CaChs2, suggesting its potential as a competitive enzyme inhibitor. Overall, this study highlights the feasibility and utility of a chitin synthase structure-based virtual screening strategy for the identification of novel antifungal potentiators and adjunctive scaffolds.