Hybrid dynamic pharmacophore models as effective tools to identify novel chemotypes for anti-TB inhibitor design: A case study with Mtb-DapB

Abstract

The current study examines the efficacy of dynamics-based hybrid pharmacophore models (DHPM) based on newly explored interaction features, as tools to screen potential inhibitors of Mycobacterium tuberculosis (Mtb)-DapB, a validated target essential for L-Lysin biosynthesis. Molecular dynamics (MD) simulations were performed on Mtb-DapB models, generated from a reported crystal structure by linking the cofactor (NADH) and substrate mimetic inhibitor (2,6-PDC), thereby creating a hybrid molecule (HM). Comparative investigation of the dynamics of interactions made by NADH, 2,6-PDC and HM from the MD trajectories revealed a number of stable interactions between HM and the hinge region residues leading to significant alterations of the Mtb-DapB structure. These newly identified interactions, translated into DHPM may be utilized as Mtb-DapB specific screening filters and provide new insights for structure activity relationships. To validate the applicability of the DHPM, about 10 million compounds compiled from the publicly available chemical space were screened using the DHPM as well as conventional pharmacophore models (based on the NADH/2,6-PDC). Systematic analyses of structures, physicochemical/ADMET properties and molecular docking of the 387 and 982 compounds screened by the DHPM and conventional models respectively demonstrate the capabilities of DHPM to screen structurally diverse chemical entities with better druglike properties and higher target binding potentials. The study demonstrates application of DHPM to predict 8 highly active anti-Mtb (MIC<=2µM)compounds whose mechanisms of action were previously unknown.3

Hybrid Dynamic Pharmacophore Models as Effective Tools to Identify Novel Chemotypes for Anti-TB Inhibitor Design: A Case Study With Mtb-DapB

“…AB and CC thank Schrodinger Inc. for providing short term license for some modules and Prof. Prasad V. Bharatham of National Institute of Pharmaceutical Education and Research, Mohali, for allowing to use software resources at his lab. This manuscript has been released as a pre-print at bioRxiv.org, (Choudhury and Bhardwaj, 2020).…”

“…Molecular dynamics (MD) based pharmacophore models have emerged as quite powerful tools which not only account for the flexibility of the target(s), but also help to identify novel key interaction features in the binding sites (Spyrakis et al, 2015) which is otherwise unexplored in the crystal structures and may be harnessed to design new inhibitors (Choudhury et al, 2014(Choudhury et al, , 2015Choudhury and Bhardwaj, 2020;Guterres and Im, 2020). Literature reports several studies that have used molecular dynamics to sample huge number of receptor conformations and generated pharmacophore models from these conformations which showed improved enrichment for specific interaction features (Bottegoni et al, 2011;Wieder et al, 2016;Culletta et al, 2020).…”