Computational Design of Multitargeted Ligands to Counter Drug Resistance of Mycobacterium Tuberculosis

Background Tuberculosis (TB) is a significant global health challenge due to drug resistance. Furthermore, tuberculous meningitis (TBM), which affects the central nervous system, has a particularly high mortality rate. TBM drugs have low efficacy because of their low blood-brain barrier (BBB) permeability. Many institutions that treat tuberculosis lack the infrastructure to identify specific drug-resistance mutations. The development of drugs with the capability of treating multiple strains would contribute considerably to the advancement of TB control in countries with limited resources. Therefore, there is an urgent requirement for novel therapeutics that can target native and drug-resistant strains. Objective This study aimed to design a novel drug to target native as well as drug-resistant Mycobacterium tuberculosis (MTB) strains associated with pulmonary TB and TBM. Methods RNA Polymerase beta-subunit (rpoB) was chosen because it is a validated target for MTB. Pharmacophore features, core moiety analysis, and docking scores were used for ligand screening. Deep neural networks (DeepFrag) were used for structural optimization, and binding affinity was evaluated using AutoDock Vina. Custom scoring schemes, STWMM for TB and STWMMM for TBM, met the requirements of high binding affinity for multiple targets, optimal pharmacokinetic profiles, and chemical synthesizability. Results M1, M2, and M3 were the molecules with the highest STWMM and STWMMM scores, indicating their potential for TB and TBM therapy. The average binding energy of M1 was -8.83 kcal/mol for native and mutant rpoB. The average binding energy for M2 and M3 was -9.63 and -9.83 kcal/mol, respectively). Conclusion In this study, novel ligands for native and drug-resistant TB and TBM therapy were obtained by multi-target drug design. A major challenge for current therapeutic regimens for TB and TBM is the rise of drug-resistant strains of Mycobacterium tuberculosis and the necessity of distinguishing them from the native strains. The multitargeted ligands developed in this study have the potential to overcome these limitations.