Computational design and optimization of Traditional Chinese Medicine compounds as potential inhibitors of Escherichia coli gyraseB

Abstract

The increasing prevalence of antibiotic resistance among major pathogens poses a significant threat to public health, necessitating innovative control measures and novel antibacterial drugs. This study employs a computational design and lead optimization approach to discover new drug candidates by virtually screening the Traditional Chinese Medicine (TCM) compound library against Escherichia coli DNA GyraseB (PDB ID: 6F86), a validated bacterial enzyme absent in humans. In silico screening was conducted using machine learning models, pharmacophore filtering, Lipinski rule-based assessment, molecular docking, and Molecular Mechanics/Generalized Born Surface Area (MM/GBSA) calculations. Bioisosteres of the top-ranking TCM compound, lithospermic acid, were generated and evaluated for ADMET properties, pharmacokinetics, physicochemical properties, and molecular dynamics simulations to assess predicted binding stability. Lithospermic acid and its bioisosteric Analog-1 exhibited more favorable Glide XP docking scores (-10.382 kcal/mol) and MM/GBSA binding free energies (-76.41 kcal/mol) than the ATP-site reference novobiocin (-5.49/-36.71 kcal/mol) and ciprofloxacin docked under the same protocol (-4.51/-41.65 kcal/mol). Molecular dynamics simulations (150 ns) indicated that Analog-1 maintained more persistent interactions with key ATP-site residues than ciprofloxacin, though both ligands showed conformational flexibility. All findings represent computational predictions requiring experimental validation, particularly given the analogs’ Lipinski violations predicting poor oral bioavailability. This study identifies lithospermic acid and Analog-1 as promising computational leads warranting synthesis and in vitro testing against E. coli GyraseB.