Virtually Designed Triclosan-Based Inhibitors of Enoyl-Acyl Carrier Protein Reductase of Mycobacterium tuberculosis and of Plasmodium falciparum

摘要

We report here new chemical structures of predicted nanomolar triclosan-based inhibitors (TCLs) of Mycobacterium tuberculosis enoyl-acyl carrier protein reductase (InhA) virtually proposed by computer-assisted molecular design. 3D models of InhA-TCL complexes were prepared by in situ modifications of the reference crystal structure (PDB entry 1P45) for a training set of 15 TCLs with known InhA inhibitory activities. A QSAR model was built leading to linear correlation between the calculated free energies of complexation (Delta Delta G(com)) and experimental values IC50exp : pIC(50) = -0.0657 x Delta Delta G(com) + 3.0502, R-2 = 0.96. In addition, ligand-based quantitative pharmacophore model (PH4) was built from bound conformations of the training set compounds and confirmed the correlation between molecular models and observed activities: pIC(50)(exp) =0.8929 x pIC(50)(pre) - 0.441, R-2 = 0.95. Structural information from both models helped us to propose new TCL analogues. A virtual library of TCLs with known predicted activities against enoyl-acyl carrier protein reductase of Plasmodium falciparum (PfENR) was evaluated, revealing dual target TCLs. Moreover, analysis of binding site interactions suggested enriching substitutions, which led to more potent TCLs with predicted pIC(50)(pre) as low as 7 nM. The computational approach, which used both free energy estimated from molecular modeling and 3D-QSAR pharmacophore model, was helpful in virtually proposing the dual-targeted drugs and provided valuable information for the design of novel potential antituberculotic agents.