De Novo Drug Design Using Computational Tools: Inhibition of CoaBC for Tuberculosis Treatment as a Case Study

Tuberculosis is a worldwide health problem that threatens to worsen as resistance to existing drugs emerges. Despite huge research efforts, there is a critical demand for the discovery of new drugs against tuberculosis (TB) with unique mechanisms of action. The Mycobacterium coenzyme A (CoA) pathway is a unique target for the discovery of new drugs for TB due to the lack of similarity of biosynthesis CoA enzyme sequences between prokaryotes and eukaryotes. Here, we aimed to develop new anti-TB drugs to inhibit MsmCoaBC, an enzyme involved in the synthesis of CoA using various computational methods. An e-pharmacophore hypothesis was developed using the MsmCoaBC ligand complex, which was then screened against a library of 629,774 fragments from the Enamine, Asinex, and FCH groups. The matched fragments were then docked into the active site MsmCoaBC. High affinity fragments were combined using molecular hybridization and further subjected to molecular docking, MM-GBSA calculations, and ADME prediction. This led to the discovery of seven compounds (breeds 1-7), with excellent docking scores (from −10.882 to −9.934 kcal/mol) and MM-GBSA free-binding energies (from −86.06 to −71.26 kcal/mol) compared to the reference ligand (docking score = −9.934 kcal/mol, binding free energy = −59.6 kcal/mol). The three best protein—ligand complexes were subjected to molecular dynamics to study their interaction stability. They showed stable complexes with the MsmCoaBC enzyme. These results open the way for additional in vitro and clinical testing of these compounds as novel anti-tuberculosis agents.