Designing novel inhibitors for Mycobacterium tuberculosis by targeting inhA and kasA genes using ligand-based drug discovery

Abstract

Tuberculosis (TB) affects almost one-third of the world’s population and is caused by the Mycobacterium tuberculosis (M. tuberculosis) bacteria. One of the problems in treating tuberculosis is the emergence of multidrug-resistant tuberculosis (MDR-TB) and extensively drug-resistant tuberculosis (XDR-TB). The disease and the emergence of multidrug-resistant variants of M. tuberculosis have opened the doors for researchers to find novel drug targets. Since both fatty-acid enoyl-acyl carrier protein reductase (InhA) and β-ketoacyl-ACP synthase (KasA) enzymes are the primary targets of action for isoniazid (INH) and thiolactomycin (TLM) in M. tuberculosis, the aim of the study was to design novel drug candidates that target and bind to InhA and KasA enzymes responsible for the synthesis of mycolic acid of the M. tuberculosis cell wall from TLM derivatives and cause inhibition of the cell wall synthesis. The research study was approved by The Environmental and Biosafety Research Ethics Committee of the University of the Free State, with Ethics approval number UFS-ESD2019/0064. The study was applied experimental research conducted in three stages to develop drug candidates for novel drugs by using the ligand-based bioinformatics tool PubChem to find the native ligand binding sites for TLM derivatives that have a strong binding capacity to the InhA and KasA proteins. We created a construct sequence for the inhA and kasA genes of M. tuberculosis H37Rv strain using nucleotide Basic Local Alignment Search Tool (BLAST) search on the National Center for Biotechnology Information (NCBI) and performed drug susceptibility testing on Mycobacterium smegmatis (M. smegmatis) MC2155 strain to surrogate M. tuberculosis H37Rv strain due to its non-pathogenic characteristic and similarity to M. tuberculosis in genome and structure. M. smegmatis strain was isolated utilizing the Becton Dickinson (BD) BBL Bactec Mycobacteria Growth Indicator Tube (MGIT) bottles and then inoculated on Mueller Hinton agar (MHA) plates infused with different concentrations of drug compounds of interest. The inhA and kasA gene sequences were subjected to BLAST analysis at the NCBI against M. tuberculosis whole genome ribosomal ribonucleic acid (rRNA) sequence to identify the closest homologs. Among resulting hits, the sequences with 100% identity homologs were selected. These strains were selected, and the H37Rv rRNA sequences were obtained. Phylogenetic analysis was carried out using the maximum likelihood method based on the Tamura–Nei model. All positions in the sequence containing gaps and missing data were eliminated. The phylogenetic analysis showed the branch lengths and the bacterial isolates identified in this study. Both receptors InhA and KasA were submitted to Protein Database Summaries (PDBsum) to determine their interaction, and coordinates of the interaction between the protein and the ligand noted. 5-fluoro-1-(glucosyl) pyrimidine-2,4(1H,3H)-dione (GPK) was derived from TLM and used as a ligand. After multiple runs, a cluster analysis was performed. The hydrophobic residues-binding positions for InhA and KasA proteins were acquired, and the ligands docked in the active position to determine the interactions with the highest binding affinity. The results of M. tuberculosis strain H37Rv sequences with 100% identity homologs were selected from the resulting hits. The obtained sequence reads were aligned to the reference genome (CP003248) using the NCBI BLAST. There were no open gaps found within the nucleotide sequences, meaning that there were 100% identities and 0% gaps between the reference genome subject (Sbjct) and the query genes (Query) inhA “GeneID: 886523” (CP072765.1) with 831/831 bits, and kasA "GeneID: 887269" (NC_000962.3; CP072765.1; CP007027.1; CP053903.1; CP003248.2) with 1251/1251 bits. There were seven synthesized TLM derivative compounds, GPK91, GPK92, GPK93, GPK94, GPK95, GPK96, and GPK97, but there were only two potential candidates for the novel antimycobacterial drugs, GPK91 and GPK96. The designed ligand compounds were docked with InhA PDB ID 3FNE and KasA PDB ID 4C6X imported from PDBsum. TLM was used as a reference ligand in both proteins. Our study showed that all the designed drug compounds had a good affinity toward the binding pockets of InhA and KasA. When docked with InhA, GPK91 showed the hydrogen bond interaction with GLY96, ALA198, and ILE294 with the best docking score of -8.88 binding energy. In the study done by Sharma et al. (2015), InhA 2X23 showed a hydrogen bond interaction with VAL65:HN1 with the docking scores of -12.6 binding energy for ZINC01777652, and -12.7 for ZINC03831448. In this study, GPK96 was docked with KasA 4C6X and showed a high affinity towards the binding pocket compared to the reference TLM, with a docking score of -7.746 binding energy. GPK96 also showed the hydrogen bond with MET213, THR315, THR313, and additional Pi-Pi stacking with the PHE404. The study by Sharma et al. (2015), KasA 4C72, showed hydrogen bond interaction with HIS311:HE21 with -10.7 binding energy for ZINC01530603 and -10.8 for ZINC01532344. In the same study by Sharma et al. (2015), all the compounds made several of interactions at the binding site with HIS345 and GLN171 that were similar to TLM. Similarly, in our study, there were other ligands with good affinity towards the binding positions of InhA and KasA, but none of the compounds displayed a simultaneous targeting of InhA and KasA. Previous studies have demonstrated that TLM is an inhibitor of the FAS-II condensing enzyme β-ketoacyl-ACP synthase (Slayden et al., 1996). This activity of TLM makes it a proficient reagent for designing novel drugs that could inhibit the activity of the InhA and KasA proteins. A series of articles that have been published state that, InhA is not the primary target of INH but rather that of KasA, which is the primary target of INH in M. tuberculosis (Mdluli et al., 1996; 1998; Barry, 1997; 2001; Slayden et al., 2000; Slayden and Barry, 2000). A study by Slayden et al., 2000, postulated that both InhA and KasA enzymes were targets for INH but that KasA was the primary target. Mycobacterium smegmatis showed drug sensitivity on MHA plates for GPK91 and GPK96 concentrations down to 0.781 μg/ml and 3.125 μg/ml, respectively. For GPK91 there was detectable growth with partial inhibition observed at [0.391 μg/ml] and no inhibition observed at [0.195 μg/ml] and less. Similarly, for GPK96, there was detectable growth with partial inhibition observed at [1.563 μg/ml], and no inhibition observed at [0.781 μg/ml] and less. The minimal inhibitory concentration (MIC) of GPK91 and GPK96 for M. smegmatis was determined at 0.781 μg/ml and 3.125 μg/ml, respectively (Table 4.3). The MIC is designated when a 99.9% reduction in colony-forming units (CFU) occurs at a specific compound’s concentration compared to the control. The MIC results of GPK91 were lower compared to the previous results observed by Slayden et al. (2000) for TLM, and the range was between 2.5 μg/ml – 5.0 μg/ml on the agar plate. GPK96 MIC result was comparable with the previous results observed by Slayden et al. (2000) for TLM. Slayden et al. (1996) also showed that TLM possesses in vivo antimycobacterial activity against the saprophytic strain M. smegmatis MC2155 and the virulent strain M. tuberculosis Erdmun, resulting in complete inhibition of growth on solid media at 75 and 25 μg/ml respectively (Slayden et al., 1996). The results showed that GPK91 might act as an InhA protein inhibitor by targeting the GLY96, ALA198, and ILE294 amino acid residuals that could result in the downregulation of the inhA gene. The results also predicted that GPK96 might act as a KasA protein inhibitor by targeting the MET213, THR315, THR313, and Pi-Pi:PHE404 amino acid residues that could result in the downregulation of the kasA gene. With similarity to TLM in the structure, together GPK91 and GPK96 showed selective activity on the specific binding sites of InhA and KasA proteins therefore proving to be the only two that could be developed into potential novel anti-tuberculosis drug candidates due to their high binding energy towards these amino acids. Both compounds had the highest binding energy at the binding sites of InhA and KasA proteins. The minimum inhibitory concentration values showed that GPK91 can induce inhibition of M. smegmatis at a lower concentration than GPK96. Further studies could determine the mechanism of action and conduct clinical trials to assess the antimicrobial efficacy and toxicity of the drug candidates to render personalized patient treatment with a high potential of improving treatment outcome and prognosis, effectively reducing drug burden that would otherwise cause adverse side effects of the current treatment.