Identification and optimization of novel inhibitors against proteinaceous drug targets from pathogenic species using in silico approaches

Abstract

A drug discovery project typically starts with a pharmacological hypothesis: that the modulation of a specific biological mechanism would be beneficial in the treatment of the targeted disease. Being a part of a Structural Biology lab, the author’s research work was integrated into the larger goals of the host laboratory where proteins of therapeutic interest are purified and characterized, assays are developed, 3D structures are solved and these are exploited for the identification of novel inhibitors with therapeutic potential. The proteinaceous targets for inhibitor discovery were chosen from the ongoing projects in the lab from human pathogens like M. tuberculosis. The objectives of the work were to use mainly rational structural techniques like virtual screening to identify and optimize inhibitors of the proteins. The work was carried out collaboratively with the Medicinal Chemistry groups in the Institute who synthesized many of the predicted compounds. In addition, the objectives also included the creation and maintenance of a comprehensive resource that can be used for in silico drug discovery and exploration against M. tuberculosis and other pathogens. The first chapter gives an introduction to the concepts of rational drug design and virtual screening and places the current work in the backdrop of the present state of the art: Computer-aided drug design with particular emphasis on virtual screening principles and strategies. It also gives a brief introduction to the algorithms available in evaluating protein-ligand interactions. The second chapter deals with the identification of diverse compound scaffolds as novel inhibitors of M. tuberculosis NAD+-dependent DNA ligase through automated docking approaches. The protein has recently drawn a lot of attention as a novel therapeutic target. Virtual screening using automated docking tools combined with an in house small molecule library has been reported. The efforts have yielded over 5 novel classes of inhibitors of this protein whose efficacy has been verified experimentally. In vitro and antibacterial assays support the efficacy of the inhibitors. These compounds inhibit M. tuberculosis NAD+ ligase with higher specificity compared to bacteriophage T4 ATP ligase as well as human DNA ligase I. The virtual screening model with Autodock has been validated with a good correlation coefficient (r2 = 0.544). The third chapter elaborates the ongoing strategies that are being used for the identification of new inhibitors; the challenges faced, and the proposed strategies to improve both the potency and the specificity of the designed inhibitor for M. tuberculosis NAD+ ligase. The optimization of the established leads, reported in Chapter 2, has been detailed. In order to find inhibitors with different modes of LigA inhibition, the BRCT domain was modeled. Additionally strategies like fragment library based lead identification, exploitation of conserved water clusters etc. which are being employed for improving specificity of inhibitors towards LigA, has been detailed. These studies have resulted in the prediction of new inhibitors. The fourth chapter deals with the identification of Mycobacterium tuberculosis Lysine ε-aminotransferase inhibitors, an important enzyme of latency, which has been ranked amongst the top-3 targets against tuberculosis persistence by the Structural genomics consortium (http://www.webTB.org). The identification was carried out through a combination of pharmacophore based-receptor and ligand modeling with 3-dimensional flexible database searches; and prioritization with molecular docking approaches. These efforts have resulted in identification of two novel scaffolds as Lysine ε-aminotransferase inhibitors and are the very first inhibitors of the enzyme from any source. The co-crystal structure of one of the inhibitors with the enzyme supports the in silico inhibitor identification strategy. The second half of the chapter reports the ligand-based and structure-based virtual screening strategies that have been employed for inhibitor optimization. The fifth chapter reports the identification of inhibitors of Mycobacterium tuberculosis Isocitrate lyase (ICL) through virtual library design and molecular docking approaches. The molecular docking results with Autodock and Gold for both the homolog’s icl1 and icl2 have been reported. The nitrobutanoates, galactosylpropanolamines and cyclopropyl derivatives have been identified reported as novel Isocitrate lyase inhibitors. We have identified that pairs of inhibitors, each apparently targeting a specific 3D-conformation of the enzyme results in better inhibition of the same as compared to individual compounds. The results serve as a new paradigm for the design of potent ICL inhibitors. The sixth chapter deals with the development of IS-IT?: a web-based docking server for evaluation of ligand-protein interaction using Autodock4 as the backend program. The server was developed to complement automated docking with AutoDock against a set of well-known drug targets from pathogenic sources. The description of the web interface, the number of potential drug targets etc. has been detailed. The web server has been evaluated by the author with several small molecule compounds for potential binders (receptors) and two test cases has been reported in the chapter; one with aryl amino derivative and other with hydroxymate derivatives. Several groups within CDRI and outside have used the web-server already. The site can be accessed at the URL: http://myc.cdri.res.in/new2/login_homepage.html. The seventh chapter reports the comprehensive genome analysis of M. tuberculosis for identification of transcription regulatory genes. The Bioinformatics analysis reveals a total of 207 transcription regulators in the genome which were analyzed in term of their abundance in other pathogenic mycobacterial species, family distribution, domain organization and gene regulatory network. The reported analysis is the first detailed study towards the understanding of the global and local structure of M. tuberculosis transcription regulatory network against the backdrop of other mycobacteria. The insights gained into the organization and evolution of the M. tuberculosis gene regulatory network will expectedly provide an alternative framework for further studies on the pathogen. The data obtained as the outcome of this transcription regulatory analysis is incorporated into a web server and can be accessed at the URL: http://myc.cdri.res.in/mycdb/homepage.htm.