Identification of Concomitant Inhibitors against Glutamine Synthetase and Isocitrate Lyase in Mycobacterium tuberculosis from Natural Sources

Tuberculosis (T.B.) is a disease that occurs due to infection by the bacterium, Mycobacterium tuberculosis (Mtb), which is responsible for millions of deaths every year. Due to the emergence of multidrug and extensive drug-resistant Mtb strains, there is an urgent need to develop more powerful drugs for inclusion in the current tuberculosis treatment regime. In this study, 1778 molecules from four medicinal plants, Azadirachta indica, Camellia sinensis, Adhatoda vasica, and Ginkgo biloba, were selected and docked against two chosen drug targets, namely, Glutamine Synthetase (G.S.) and Isocitrate Lyase (I.C.L.). Molecular Docking was performed using the Glide module of the Schrӧdinger suite to identify the best-performing ligands; the complexes formed by the best-performing ligands were further investigated for their binding stability via Molecular Dynamics Simulation of 100 ns. The present study suggests that Azadiradione from Azadirachta indica possesses the potential to inhibit Glutamine Synthetase and Isocitrate Lyase of M. tuberculosis concomitantly. The excellent docking score of the ligand and the stability of receptor-ligand complexes, coupled with the complete pharmacokinetic profile of Azadiradione, support the proposal of the small molecule, Azadiradione as a novel antitubercular agent. Further, wet lab analysis of Azadiradione may lead to the possible discovery of a novel antitubercular drug.


Introduction
Tuberculosis is an air-borne disease affecting the lungs, presenting symptoms like severe cough, chest pain, and fever. Tuberculosis (T.B.) is a contagious bacterial infection resulting in 48.3 cases per 100,000 inhabitants per year [1]. As per periodic reports published by World Health Organization (WHO), every year 2-3 million people lose their lives due to active Tuberculosis, while billion others carry dormant T.B. throughout the world. In 1882, German Microbiologist Robert Koch revealed that M. tuberculosis is the disease-causing agent of T.B. [2]. Over the decades, different antibiotics have been intro- has resistance to all first-line and second-line T.B. drugs [3,4]. Tuberculosis caused by MDR/XDR strains requires long-term treatment spanning over 9-20 months or even more, with a success rate below 50% globally [3]. The principal reason behind the requirement of long-term treatment in MDR/XDR TB is the smaller subpopulations of nonreplicating cells, which are not affected by standard antibiotics used and hence, require particular drugs to combat the deadly disease [5,6].
In comparison to developed countries, T.B. is much more prevalent in developing countries. About 10.5 million people were infected with T.B. in 2015, and more than 1.4 million T.B. patients died, with 95% of deaths occurring in developing countries [7]. Human Immunodeficiency Virus (H.I.V.) coinfection with T.B. and the development of strains resistant to the most effective, 1 st line T.B. drugs are the two main factors responsible for the existing epidemic caused by T.B.
For the diagnosis and treatment of T.B., different strategies have been developed and are still being developed. In 1993, Directly Observed Treatment Short-Course (DOTS) was one of the most significant strategies among them. In 1998, DOTS plus was introduced to combat the multidrug-resistant (M.D.R.) strain [2,7]. In the 1800s, "just sleep and eat nutritious food" was the advice given to T.B. patients [2]. In the beginning of the twentieth century, the Bacillus Calmette-Guerin (B.C.G.) vaccine was developed, saving many lives from this deadly disease. Until the discovery of Streptomycin in 1944, there was not even a single antitubercular drug [8]. But within a short period of time, Streptomycin also became ineffective in treating T.B., due to the emergence of Streptomycin-resistant strains. After that, different antitubercular drugs were introduced, including isoniazid (1952), pyrazinamide (1952), cysteine (1952), ethambutol (1962), rifampin (1957), and pyrazinamide (1957) [9]. However, due to the emergence of MDR-TB and XDR-TB strains, most commonly used antitubercular drugs became ineffective. Additionally, with the advancement of medical research in the last few years, scientists have obtained valuable insights into T.B. pathogenesis and treatment, yet further extensive research is required to decrease the incidence and subsequently eradicate T.B. successfully [2,10]. Hence, there is an urgent need to identify new drug candidates in T.B. therapeutics [11].
According to the World Health Organization (WHO), "a medicinal plant" is any plant in which one or more of its organs contain substances that could be used for therapeutic purposes or as precursors for synthesizing suitable drugs. Plants are a repository of secondary metabolites which possess different medicinal properties. Plant secondary metabolites have caught much attention as drug candidates because of their less toxic nature [11]. Around 80% of the population in developing countries relies on traditional medicines derived from plants for their primary health care. In Ayurveda, different diseases, including Tuberculosis are treated using various plants based on the bioactive compounds available [12].
Microorganisms are usually controlled by blocking specific essential enzymes, thereby inhibiting the various biosynthetic pathways in them. Glutamine Synthetase (G.S.) and Isocitrate Lyase (I.C.L.) are two essential enzymes for >the growth of M. tuberculosis, and are effective drug targets [13,14]. G.S. enzyme (E.C. 6.3.1.2) of M. tuberculosis, catalyses the ATP-dependent reaction between ammonium and glutamate, resulting in glutamine, A.D.P., phosphate, and a proton. Four G.S. homologues are found in M. tuberculosis, of which only one, encoded by the gene glnA1, is expressed highly for the survival of M. tuberculosis in both in vitro and in vivo conditions [15][16][17]. This enzyme can alter the ammonia concentration in the host cells where the infection is prevalent; it affects the pH of the phagosome resulting in the prevention of the phagosome fusion with the lysosome [18,19].
In the Glyoxylate shunt, Isocitrate Lyase (I.C.L.) catalyses the conversion of Isocitrate to Succinate and Glyoxylate. Malate Synthetase condenses Glyoxylate with acetyl-CoA to form malate [20]. Two T.C.A. cycle steps involving decarboxylation are reportedly by-passed by I.C.L. and Malate Synthase, which is considered an essential phenomenon for fungal and bacterial pathogenesis [21]. ICL1 and ICL2, two forms of the enzyme Isocitrate Lyase encoded by genes icl11 and icl12, respectively, are also crucial for the growth, survival, and virulence of M. tuberculosis [14].  Based on their usage in traditional medicine and their documented antitubercular activity, four plant species, namely, Azadirachta indica, Camellia sinensis, Adhatoda vasica, and Ginkgo biloba were selected for the current study [21][22][23][24][25]. From the ligand database, that was prepared based on the selected medicinal plants, Computer-Aided Drug Discovery (CADD) protocols were used to identify compounds with inhibitory properties against the two targeted enzymes, Glutamine Synthetase and Isocitrate Lyase of M. tuberculosis.

Results and Discussions
2.1. Molecular Docking. The binding affinity of the ligands towards the two selected receptors, 2WGS and 1F61, was determined via the Molecular Docking study. The binding affinity was determined based on the docking score; the lower the docking score and the greater the affinity. It was observed that the bioactive compound Azadiradione showed the highest binding relationship towards 2WGS (binding affinity score -10.2) followed by 7-desacetyl-7- benzoylazadiradione, 7-desacetyl-7-benzoylgedunin, alphaamyrin, and Gedunin with binding affinities, -10.1, -9.1, -8.8, and -8.7, respectively (Table 1). For the receptor 1F61, 7-desacetyl-7-benzoylazadiradione displayed the highest binding affinity (binding affinity score -9.8), followed by 7desacetyl-7-benzoylazadiradione, Azadiradione, Gedunin, and alpha-amyrin with binding affinities, -9, -8.5, -8.2, and -8, respectively (Table 1). Based on the binding affinity and the ability to effectively inhibit more than one receptor, the ligand, Azadiradione, was selected for further analysis.  Additionally, the hydrogen bonds between Azadiradione, a bioactive compound, and residues of the active site of the proteins were examined throughout the simulation. In the complex formed by 2WGS and Azadiradione, H-bonds were observed between the ligand, Azadiradione, and the residues SER-146, SER-151, TRY-153, and PHE-262 of 2WGS ( Figure 5(a)). In the complex formed by 1F61 and Azadiradione, H-bonds were observed between Azadiradione and the residues, GLN-80, ARG-379, and ARG-386 of 1F61 ( Figure 5(b)).
The ADME-Tox analysis performed for the selected compound Azadiradione indicates that the compounds have an acceptable range of important physicochemical parameters such as Mw and Consensus Lipophilicity Score. In addition, it passed all the three major drug discovery rules: Lipinski, Ghose, and Veber rules. This ADME-Tox study revealed that the compound Azadiradione should not have any issue with becoming a successful drug candidate from the biochemical point of view. The computational docking study of the selected plant-based ligand against the target enzymes G.S. and I.C.L., followed by Molecular Dynamics Simulation studies and ADME-Tox study, is sufficient to predict the antitubercular nature of Azadiradione. Azadiradione is a triterpenoid compound used traditionally in several diseases and has some reports against malaria, respiratory disorders, cancer, intestinal helminthiasis, leprosy, constipation, inflammation, blood morbidity, dermatological complications, rheumatism, and so on [32,33], which also supports its existing acceptability as a drug molecule.

Conclusion
In view of the wide-spectrum resistance of the currently existing M.D.R. and XDR T.B. strains display against the presently used antitubercular drugs, the present study was designed to dock 1778 compounds from four   (Table 3) with the two targeted enzymes. Furthermore, the ADME-Tox study of Azadiradione exhibited noble drug-like properties of the ligand. The collective findings of the present in silico study, have been interpreted to propose that Azadiradione possesses significant potential as a candidate antitubercular molecule which is subject to future validation.

Databases
Used. Multiple databases were used at different stages of the study; one such significant database is the K.E.G.G. pathway, which includes the metabolic and regulatory pathways sections. Utilizing data from K.E.G.G. pathway database, the present study compared host metabolism with critical metabolic pathways of M. tuberculosis, and the necessary enzymes were identified [41]. Essential enzymes of the pathogen, M. tuberculosis, were also identified with the help of the Database of Essential Genes (D.E.G.). D.E.G. has reported 614 such essential genes [42]. 3D structures of targeted proteins, that is, Glutamine Synthetase (Chain A, PDB ID: 2WGS) and Isocitrate Lyase (Chain A, PDB ID: 1F61) were retrieved from RCSB PDB database [14,[43][44][45][46][47]. Other databases like Drug Bank, PDB Bind, ZINC, and so on were also used in executing different steps of the research study, mainly to get the 2D and 3D information of the selected ligands.  format. The protein structures were prepared by adding polar hydrogen bonds and removing hetero atoms, water molecules, and bound ligand(s) using AutoDock Tools utility. Only Chain A from 2WGS and 1F61 was selected for studying the docking interactions.

Molecular Docking Analysis.
Protein-Ligand Docking study was carried out to understand the interaction of the two molecules with the ligands and determine the ligand(s) that would form complexes possessing minimal energy with the receptor. In docking analysis, a flexible receptor molecule has to be achieved. So by using the GRID module of Auto Dock v4.0, the flexible receptor molecule was achieved by setting a little residue. In contrast, AutoDock Vina with PyRx platform was used for calculating the affinity of small molecules against the targeted proteins Glutamine Synthetase and Isocitrate Lyase [48] (https://pyrx.sourceforge.io). The best protein-ligand complexes were then selected based on the energy conformation of the complexes. PyMOL software in PYTHON platform was used to analyse complex interactions, which illustrated hydrophobic interactions, ionic interactions, and hydrogen bonding between the receptors and the respective ligands.  D.) Simulation. The Desmond software module of Schrödinger suite 2020-3 was used for carrying out the M.D. Simulation. Simulations of 100 ns were performed to analyze the in silico stability of the complexes formed by 2WGS and 1F61 with G.S. and I.C.L. OPLS3 force field was used to study the 2WGS and 1F61 complexes in the explicit solvent system during M.D. Simulation. The atomic framework was solvated with crystallographic water (TIP3P) particles under orthorhombic intermittent limit conditions for 10 Å buffer region. In addition to the deletion of the overlapping water molecules, the entire structure of 34,255 atoms were neutralized by adding Na + as counter ions. An ensemble of a Nose-Hoover thermostat and barostat was utilized to maintain a constant temperature of 300 K and a constant pressure of 1 bar (N.P.T.) [26]. Hybrid energy minimization algorithms were used, consisting of 1000 steps of steepest descent followed by conjugate gradient algorithms. M.M.G.B.S.A. analysis was done for both the complexes to calculate ligand binding energies using the Prime module of Schrodinger Suite 2020-3.

4.7.
In Silico ADME-Tox Study. ADMET-Tox plays a vital role in the field of drug testing and design. Servers and software are available that use data from previously reported drugs to predict the ADMET properties of ligand molecules. In silico ADME-Tox studies were performed using a Swiss ADME server to identify the best-performing ligands [27][28][29][30]. Various Lipinski parameters were used to evaluate ligand(s)' drug-likeness and ADMET properties.

Data Availability
The supporting data to the results reported in the article can be found with the corresponding authors, Dr. Saurov Mahanta and Dr. Saravanan Muthupandian.