Computational and artificial intelligence-guided discovery of potential therapeutic candidates targeting Ml0099 protein of Mycobacterium leprae.

Abstract

Despite a global decline in incidence, leprosy remains a major public health concern, with India accounting for over 55% of cases worldwide, highlighting the unmet need to characterize hypothetical proteins of M. leprae as novel therapeutic targets against emerging drug-resistant strains. The three-dimensional homology structure and characteristics of ML0099 was predicted using an artificial intelligence-driven approach implemented through the AlphaFold 3 server and deepTHMM. Molecular docking with phospholipase substrates with different head moieties, coupled with large-scale virtual screening of ∼3600 FDA-approved drugs, was performed to identify promising therapeutic candidates followed by molecular dynamics analysis to validate the affinity for bound compounds. This study demonstrated that ML0099 is likely a member of the CULP protein family, containing an N-terminal signal peptide and a conserved pentapeptide motif (GxSxG), a hallmark of lipolytic enzymes. Structural analysis further identified conserved catalytic residues-Ser175, Asp268, and His299-forming a putative catalytic triad. The phosphatidyl-choline exhibited the most favourable substrate binding profile upon docking, predicting it's optimal enzymatic substrate. Virtual screening of FDA-approved compounds identified Zileuton and Glipizide as promising candidates, exhibiting binding energies of -8.5 to -9.1 kcal/mol along with stable interaction networks. Subsequent molecular dynamics simulations, MM-PBSA/GBSA and PCA analysis confirmed the stability of the protein-ligand complexes. Computational and AI-driven approaches demonstrated the potential biological role of ML0099 and identified the repurposed drugs Zileuton and Glipizide as promising inhibitors, providing a robust foundation for future experimental validation and therapeutic development.