Abstract
Tuberculosis (TB) remains a global health threat, with increasing resistance to first-line drugs like isoniazid necessitating the discovery of novel therapeutics. This study focuses on the in silico evaluation of 11 novel compounds of oxindole-triazole tethered derivatives (NV1-NV11) against the catalase-peroxidase enzyme of Mycobacterium tuberculosis (PDB ID: 1SJ2). Molecular docking was performed to evaluate binding scores, revealing that compound NV10 exhibited the strongest interaction with a high binding affinity of −10.262 (kcal/mol), showing a more favorable docking score than isoniazid (−6.331 kcal/mol). Molecular dynamics simulations (MDS) over 100 ns confirmed the structural stability and compactness of the NV10-1SJ2 complex, supported by stable RMSD, RMSF, and radius of gyration profiles. MM-GBSA binding free energy calculations further validated NV10’s binding energy, with a favorable ΔG_bind of −68.21 ± 3.10 kcal/mol, primarily driven by van der Waals and Coulombic interactions. These results suggest that NV10 is a computationally promising candidate for developing new anti-TB agents.
Introduction
Tuberculosis (TB) is an infectious disease caused by Mycobacterium tuberculosis (MTB) that primarily affects the lungs but can impact any tissue. It spreads through the air when a person inhales droplet nuclei containing sputum contaminated with MTB. 1 Although TB became epidemic during the Industrial Revolution, it is preventable and controllable following the discovery of the Bacillus Calmette-Guérin (BCG) vaccine and antibiotics such as streptomycin, isoniazid, and rifampicin. According to the World Health Organization (WHO), a statistical report from 2024 estimates that TB disease is alarmingly severe, with a mortality rate approaching 50% without treatment. Furthermore, approximately 470,000 individuals develop multidrug-resistant (MDR) or rifampicin-resistant tuberculosis (RR-TB) annually, resulting in around 180,000 deaths. 2 Therefore, accurate identification of antimicrobial resistance is essential for effective patient management, optimizing antibiotic use, and mitigating the emergence of antimicrobial-resistant strains such as MDR, extensively drug-resistant (XDR), and drug-resistant (TDR) TB. 3 Patients diagnosed with MDR or XDR TB infections further complicate the situation, rendering treatment more challenging due to the practicality of drug sensitivity tests, prolonged treatment periods (8–24 months), and a lack of proper diagnosis. Additionally, the drugs rifampicin and isoniazid exhibit severe side effects, including hepatotoxicity. 4
Oxindole and its derivatives are among the most privileged heterocyclic compounds, attracting significant interest in organic synthesis due to the presence of both electrophilic (amide carbonyl at C2) and nucleophilic (active methylene at C3 and NH group) centers. This makes chemical modification feasible and highlights numerous applications in medicinal chemistry. 5 Some oxindole-based molecules are reported as key scaffolds in natural alkaloids with a broad spectrum of biological activities, such as antidiabetic, 6 anti-cancer, 7 anti-convulsant, 8 anti-fungal, 9 anti-HIV (human deficiency virus) activity, 10 anti-inflammatory, 11 antimalarial, 12 antimicrobial, 13 antitumor, 14 and antiviral. 15 On the other hand, triazoles are nitrogen-containing, five-membered heterocyclic skeletons that have gained special emphasis in medicinal chemistry and organic synthesis due to their unique structural and electronic properties, which demonstrate exceptional stability toward hydrolysis and redox conditions, making them resistant to metabolic degradation. 16 Additionally, triazoles exhibit a high dipole moment and the ability to form hydrogen bonds, enabling interaction with several biological targets. 17 In this study, molecular docking and MD simulations are employed to explore the interaction of triazole and oxindole derivatives with catalase peroxidase (KatG), a key enzyme involved in oxidative stress regulation and cellular detoxification. 18 KatG is a bifunctional catalase-peroxidase enzyme that plays a dual role in M. tuberculosis. It protects the bacterium against oxidative stress by decomposing hydrogen peroxide and organic peroxides, and it is also responsible for the bioactivation of the prodrug isoniazid into its active radical species. 19 Mutations in KatG are a primary cause of isoniazid resistance. While KatG represents a biologically relevant target due to its role in bacterial defense, complete inhibition of KatG could theoretically impair isoniazid activation and potentially reduce the efficacy of combination therapy. 20 Therefore, inhibition of KatG may not be a successful approach, but we can target this enzyme for oxidative stress. Given that, oxidative stress plays a significant role in M. tuberculosis survival and pathogenesis, targeting this enzyme could be a promising strategy for anti-TB drug development. 21
In this context, the NV1-NV11 series was rationally designed through bioisosteric substitution and pharmacophore modeling to integrate oxindole and 1,2,3-triazole motifs into a single hybrid framework. Unlike previously reported oxindole-triazole derivatives, the NV series incorporates a distinct C-3 methylene-linked phenyl-1,2,3-triazole substitution pattern, which is absent in earlier KatG-targeting scaffolds. This modification introduces enhanced conformational flexibility while maintaining optimal electronic distribution for catalytic-site engagement. NV10, in particular, features a C-3 methylene-bridged triazole bearing a 4-chlorobutyl side chain, designed to facilitate simultaneous interaction with polar catalytic residues and adjacent hydrophobic pockets within KatG. The linker architecture was optimized to improve spatial orientation and interaction persistence, addressing limitations observed in earlier analogues that demonstrated weak hydrogen-bond stability or insufficient hydrophobic anchoring during molecular dynamic simulations (MDS). Molecular docking was employed to predict binding modes and interaction energies within the KatG active site (PDB ID: 1SJ2), while MDS was conducted to assess the conformational stability and persistence of ligand-enzyme interactions under dynamic conditions. In the present study, the designed oxindole-triazole hybrids are not structural analogues of isoniazid and were not intended to block the isoniazid activation pathway. Rather, they were computationally evaluated for their ability to interact with catalytic or adjacent regulatory residues that may disrupt bacterial oxidative balance. The findings aim to provide structural insights that may guide the rational development of next-generation anti-TB agents targeting oxidative stress pathways in M. tuberculosis.
Materials and Methods
Ligand Selection
The 11 ligands (Table 1) were selected based on prior structure-activity relationship (SAR) data for oxindole and triazole analogue with antimicrobial potential.22–25 Substituents were chosen to enhance hydrogen bonding, π–π stacking, and hydrophobic interactions with the KatG active site.
2D Structures of Oxindole-triazole Derivatives.
Ligand Preparation
The minimized structures were optimized using the MMFF94 force field, with the lowest energy conformers saved in .mol2 format for docking. 26 Chemical structures were retrieved from ChemSpider, maintained by the Royal Society of Chemistry (Cambridge, UK), accessed between August 10 and 12, 2025.
NV10 → ChemSpider ID (CSID): 4445106
Isoniazid → CSID: 3638
Only NV10 possessed an existing ChemSpider record at the time of access. The remaining compounds (NV1-NV9 and NV11) were novel, designed derivatives and were not individually registered in ChemSpider (Table 1). These structures were manually constructed and energy-minimized using MarvinSketch v22.19.0 (ChemAxon Ltd., Budapest, Hungary) under Windows 10 Pro (64-bit), applying the MMFF94 force field with an RMS gradient convergence threshold of 0.01 kcal/mol•Å.
Target Protein Retrieval and Preparation
The x-ray crystallographic structure of the target protein (KatG) was downloaded from the RCSB Protein Data Bank (PDB ID: 1SJ2, accessed on August 15, 2025). 18 The KatG of M. tuberculosis is essential for bacterial survival, as it neutralizes reactive oxygen species and activates the frontline prodrug isoniazid. Mutations in katG, especially S315T, are a major cause of isoniazid resistance. Targeting KatG’s active site can disrupt its oxidative stress defense offering a route to combat resistant strains. 27 In this study, KatG was selected to evaluate oxindole-triazole hybrids, aiming to identify stable binders with potential anti-tubercular activity. Key validation parameters such as resolution, mutation and Ramachandran plot were verified to ensure protein suitability. The protein suitability details are shown in Table 2. Structural validation was performed via PROCHECK (PDBsum server), accessed August 16, 2025.
Comparison of Standard Values and Retrieved Protein to Validate the Protein Chosen for Docking Investigation.
The structural refinement of the protein was carried out using Chimera v1.16, 28 applying the AMBER force field (ff14SB) for standard residues and AM1-BCC for nonstandard residues. No unresolved or missing residues were present in the final selected chain. When present, alternate conformations were resolved by retaining the conformation with the highest occupancy. Hydrogen atoms were added consistent with a physiological pH of 7.4. Non-essential components like co-crystal ligands, water molecules, and unnecessary chains were removed. 29
Grid Generation
The binding site was predicted using the CASTp server (accessed August 18, 2025), which identifies surface pockets and internal cavities based on protein topology. 30 The grid box for docking was defined to encompass the predicted catalytic pocket using AutoDock Tools 1.5.6. 31 For the above protein receptor 1SJ2, active residues are shown in Table 3. The size of the Grid box was made short so that it would be consistent with the protein’s active site and with the ligand expected to be docked (Table 4).
List of Active Sites of Amino Acids.
Grid Parameter to Cover Active Site Amino Acids.
Molecular Docking of Target Protein with Ligands
Docking simulations were performed using AutoDock Vina v1.2.0 on Ubuntu 22.04 LTS (64-bit), with ligand and receptor preparation carried out using AutoDock Tools v1.5.6. Protein and ligands were converted into pdbqt format using in-house scripts. The scripts are given in the supplementary file. Protein was treated as rigid, while ligands retained freely rotatable bonds. For docking studies, we used AutoDock Vina v1.2.0 using the defined grid dimension. 32 The grid box was centered on the target’s active site, allowing the program to identify additional potential interaction sites between ligands and the receptor. All other settings were kept at the default. The XYZ coordinates are provided in Table 4. Additional parameters were set as follows: CPU to 23, exhaustiveness to 32, number of modes to 9, and energy range to 3.33,34 The docking reliability was assessed by examining interaction consistency with catalytically relevant residues and by performing MDS of the top-ranked compound–protein complex to evaluate binding stability over time.
To ensure computational reproducibility, standardized parameters were applied throughout the study. A fixed random seed (123456) was used to enable deterministic behavior across runs. Docking reproducibility was validated by confirming ranking consistency across all five independent runs, ensuring that the top-ranked binding poses exhibited an RMSD of less than 1.5 Å between runs, and observing stable clustering of binding conformations. These measures ensured deterministic reproducibility and robustness of the docking predictions.
Visualization
The docking results were analyzed using BIOVIA Discovery Studio 2024 and Maestro 12.3 to map 2D and 3D interactions. 35
MD Simulation
MD simulation was performed on the docked complex of 1SJ2 using Desmond 2020.1 software (Schrödinger, LLC), where the NV10 compound was labeled as NV10_1SJ2. The simulation system included an explicit solvent model with TIP3P water molecules and was parameterized using the OPLS-2005 force field. A cubic periodic boundary box with dimensions of 10 Å × 10 Å × 10 Å was applied. A 0.15 M NaCl solution was added to simulate physiological conditions, and Na⁺ ions were included to neutralize the overall system charge. 36
A preliminary equilibration phase lasting 10 ns was conducted under the NVT ensemble to achieve stabilization of the protein-ligand complex. An additional equilibration period of 12 ns was conducted under the NPT ensemble, utilizing the Nose–Hoover thermostat to sustain a pressure of 1 bar, incorporating variable temperature control and a relaxation time of 1.0 ps. Simulations employed a 2 fs integration time step. 37 Pressure regulation was achieved through the use of a barostat featuring a relaxation time of 2 ps, and the Martyna–Tuckerman–Klein chain coupling method was implemented. Short-range Coulomb interactions were computed with a cutoff of 9 Å, while long-range electrostatics were addressed using the particle mesh Ewald (PME) method. The production MDS was conducted for a duration of 100 ns. 38 “The evaluation of system stability and behavior was conducted using multiple parameters, which included root mean square deviation (RMSD), root mean square fluctuation (RMSF), radius of gyration (Rg), hydrogen bonding, salt bridge analysis, and solvent-accessible surface area (SASA).” 39
Binding Free Energy Analysis
Protein-ligand interaction profiling was performed using BIOVIA Discovery Studio 2024 and Maestro v12.3. To ensure methodological consistency and reproducibility, explicit geometric criteria were applied for all noncovalent interaction analyses. Hydrogen bonds were defined by a donor-acceptor distance ≤3.5 Å and a donor-hydrogen-acceptor angle ≥120°. Hydrophobic contacts were assigned when nonpolar heavy atoms were within 4.0 Å. π–π stacking interactions were identified based on an aromatic ring centroid–centroid distance ≤5.0 Å. π–cation interactions were defined by a maximum distance of 6.0 Å between the aromatic ring centroid and the cationic center. Salt bridges were assigned when oppositely charged atoms were within 4.0 Å. Water-mediated hydrogen bonds (water bridges) were considered when the hydrogen-bond distance criterion was ≤3.5 Å. These criteria were uniformly applied across docking pose evaluation and molecular dynamics trajectory analysis.
The binding free energies of the ligand–protein complexes were estimated using the Molecular Mechanics Generalized Born Surface Area (MM-GBSA) approach. Each molecular dynamics trajectory was 100 ns in duration, and RMSD analysis indicated structural stabilization after approximately 40–50 ns. Therefore, the analysis time window was defined as 50–100 ns. Frames were extracted at a frequency of one frame per 1 ns, resulting in a total of 50 frames per replicate. 40 This analysis was performed using the Python-based script thermal_mmgbsa.py, with a single-step sampling protocol. The binding free energy values (in kcal/mol) were obtained based on the principle of energy component additivity. The entropic contributions were not included in the binding free energy estimation. Various energy terms—including coulombic, covalent, hydrogen bonding, van der Waals interactions, self-contact corrections, lipophilic contributions, and solvation energies for both protein and ligand—were calculated and summed to yield the final MM-GBSA energy estimates. The equation used to calculate ΔGbind is Equation 1. 41
Repeatability and Reproducibility
To ensure full transparency and reproducibility of the computational workflow, all custom scripts and primary output data generated in this study are archived and available from the corresponding author upon reasonable request. In-house docking preparation scripts developed using AutoDock Tools for batch pdbqt conversion, as well as the Python-based MM-GBSA analysis script (thermal_mmgbsa.py, Python 3.8 compatible) employed with the Prime MM-GBSA module of Prime, are fully documented and version-controlled.
All raw docking outputs generated using AutoDock Vina—including log files, ranked binding poses, clustering results, grid parameter files, and random seed documentation—are preserved in their original formats. Additionally, complete molecular dynamics trajectory files (.dtr), system configuration files (.cms and .cfg), checkpoint files, and per-frame MM-GBSA energy decomposition tables derived from simulations performed in Desmond are securely stored in institutional research data repositories.
Results and Discussion
Protein Structure Validation
Protein structure validation using the Ramachandran plot revealed that 88.2% of residues were located in the most favored regions, 11.3% in additionally allowed regions (total = 99.5%), and 0.5% in generously allowed regions, with no residues in disallowed regions (Figure 1). These results confirmed that the protein model possessed good stereochemical quality, with the majority of residues adopting energetically favorable conformations and no evidence of steric clashes.
Ramachandran Plot of Protein 1SJ2 from the PDBsum Server (https://www.ebi.ac.uk/thornton-srv/databases/pdbsum /).
Molecular Docking Analysis
A total of 12 ligands, including the reference compound isoniazid, were docked against the target protein (PDB ID: 1SJ2). KatG is a validated target relevant to INH activation and resistance. The mechanistic role of KatG in activating INH and the contribution of katG mutations to INH resistance are well documented. Designing ligands that bind KatG—either to inhibit protective catalase-peroxidase activity or to bind alternate sites—is a plausible approach to overcome resistance mechanisms, but requires careful experimental validation because perturbing KatG function may have complex effects on activation of prodrugs and cellular oxidative balance. 42
Binding scores ranged from –6.331 to –10.262 kcal/mol, with compound NV10 showing the highest binding score (–10.262 kcal/mol), followed by NV2 (–9.146 kcal/mol) and NV4 (–9.384 kcal/mol), as shown in Table 5. The reference drug INH exhibited the least favorable binding score of –6.331 kcal/mol. NV10 formed hydrogen bonds (six in total), including key interactions with ARG42A and GLN36A, as well as a π-cation interaction with LYS46A. The extended 4-chlorobutyl arm contributes hydrophobic interactions within the enzyme’s binding pocket, while the phenyl-triazole unit enables π–π stacking and hydrogen-bond formation with key catalytic residues. The previous reports on oxindole and spiro-oxindole scaffolds have demonstrated measurable anti-TB activity, including efficacy against katG and inhA mutant strains. 43 These prior findings reinforce the value of oxindole derivatives in addressing resistance-related challenges. 44 Moreover, the integration of 1,2,3-triazole moieties—well known for their metabolic stability, high dipole moments, and ability to form multiple polar interactions—follows established medicinal chemistry strategies that have yielded potent anti-TB leads in earlier studies. 45 By combining oxindole and triazole pharmacophores, NV10 capitalizes on the complementary interaction profiles of both scaffolds. However, we conducted in silico prediction; further experimental validation is required to measure the activity.
The Binding Score of Ligands and the Reference Drug Isoniazid.
Similarly, NV2 demonstrated strong binding affinity and formed hydrogen bonds with LEU43B, GLU289A, and ASN701B in addition to hydrophobic contacts. NV4, with a binding score of –9.384 kcal/mol, engaged in multiple interactions, including π-cation interaction with LYS488B and hydrogen bonds with GLN190B. The reference isoniazid formed five hydrogen bonds, mainly involving ASN133A, ARG145A, and ARG146A, along with a single π-cation interaction with ARG146A, which contributed to its relatively weaker binding affinity. The binding interaction were shown in Table 6.
Interactions and Docking Score of Ligands with 1SJ2 Protein.
Compounds like NV2, NV4, and NV8 also displayed promising profiles, balancing strong binding affinities with diverse interaction types. Interestingly, NV4 and NV8 both interacted with THR618B and GLN36A, reinforcing the idea of conserved interaction hotspots. This prediction indicates that the novel compounds, especially NV10 and NV2, have the potential to be developed as potent inhibitors. NV10 binds to GLY34A, ASN35A, GLN36A, ARG42A, and LYS46A through multiple hydrogen bonds and a π-cation interaction (Figure 2). In contrast, isoniazid mainly interacts with ASN133A, ARG145A, ARG146A, and GLU289A. These findings are summarized in Table 7. The 2D and 3D interactions are visualized in the supplementary file (Figures S1 to S22).
2D Interactions of NV10 and Reference Drug Isoniazid Against Protein 1SJ2.
Comparative Interaction Between NV10 and Isoniazid.
Molecular Dynamics Simulation
In the docking set, NV10 scored –10.262 kcal/mol. We therefore interpret the NV10 rank as provisional and place greater weight on convergent evidence. To further support this decision, we conducted an MD simulation. MDS was performed exclusively for compound NV10 because it demonstrated the most promising overall profile among the screened candidates. Specifically, NV10 exhibited:
The lowest binding energy in molecular docking studies, The most favorable interaction pattern with key active-site residues.
Since MD simulation is computationally intensive and primarily used to validate the stability of the most promising protein-ligand complex under dynamic physiological conditions, NV10 was selected as the computationally significant candidate for further dynamic validation.
RMSD and RMSF
Figure 3A represents an RMSD analysis of the NV10-1SJ2 over a simulation for 100 ns. The trajectory begins with an initial sharp increase within the first 10 ns, stabilizing around 1 to 1.2 Å. This rapid rise with minor fluctuations indicates minor adjustments in the conformational state. Between 20 and 40 ns, a more significant fluctuation in the RMSD values is observed, reaching peaks close to 2 Å. This may indicate a dynamic rearrangement or slight flexibility. However, after 40 ns, the RMSD stabilizes, maintaining values within the range of 1.8 to 2.0 Å until the end of the simulation. The average RMSD value (1.67 Å) shows structural integrity of the system. Figure 3B represents the RMSF analysis of the NV10-1SJ2. The plot exhibits low RMSF values (0.85 Å), indicating that the backbone atoms remain stable throughout simulation. However, distinct peaks are observed at residue regions around 0, 340–347, and 716, with fluctuations exceeding 3 Å. These stability indicators parallel observations from other computational anti-TB studies, where lead candidates with steady RMSD, limited backbone fluctuations, and persistent hydrogen bonds often translated into favorable in-vitro potency. 46 However, we presented complex stability as a predictive approach rather than a translative strategy.
MD Simulation Analysis of 100 ns Trajectories of (A) RMSD of Cα Backbone of NV10-1SJ2 Complex, (B) RMSF of Cα Backbone of NV10-1SJ2 Complex, (C) Radius of Gyration (Rg) of Cα Backbone of NV10-1SJ2, (D) Formation of Hydrogen Bonds in NV10-1SJ2 Complex.
Structural Compactness and Stability
Figure 3C depicts the Rg analysis of NV10-1SJ2. The Rg was predicted to investigate the compactness and structural stability of the protein. The Rg plot shows a relatively stable value over the simulation period of 100 ns, with values fluctuating between approximately 35 Å and 35.71 Å. This indicates compactness in protein structure with no significant unfolding or conformational changes. Figure 3D illustrates the number of H-bonds formed by the NV10-1SJ2 complex. Hydrogen bonds play a vital role in maintaining the stability and structural integrity of complexes. The average number of hydrogen bonds (3.01) indicates stable intermolecular interactions in the complex, with fluctuations ranging between 0 and 5 contributing to its structural stability over time.
Figure 4A depicts the SASA analysis, highlighting the conformational changes and stability of NV10 when bound to protein; the lower SASA of the bound receptor compared to the unbound receptor suggests that NV10 binding effectively reduces the receptor’s solvent exposure. This reduction in SASA upon binding further implies a stable interaction between receptor and NV10.
The Solvent-accessible Surface Area of NV10-1SJ2 (A). Bar Graph of Protein-ligand Contacts of NV10-1SJ2 (B).
Protein–Ligand Interaction
Figure 4B depicts the protein–ligand interaction profile of the NV10–1SJ2 complex throughout the simulation. Hydrogen bonds were the predominant interaction type, observed consistently with residues GLN36A, ARG42A and GLU 703A, each exhibiting an interaction fraction close to or above 0.8. Hydrophobic contacts were notable with LYS27B, LYS45A and PRO195B, while water bridge interactions were less frequent, primarily involving ASN35A, GLU 195B and GLU 703A. Additionally, limited ionic interactions were detected with ARG63, suggesting that hydrogen bonding played the key role in ligand stabilization within the binding pocket. The protein-ligand percent interaction is given in the supplementary file (Figure S23). Strong and persistent hydrogen-bond interactions were observed with GLU703A (91%) and ARG42A (89%), while GLN36A (78%) also showed substantial interaction stability. Moderate contacts were maintained with LYS46A (19%), GLY34A (28%) and ASN35A (14%), indicating auxiliary stabilization roles (Figure S23). Ligand-centric stability metrics (ligand RMSD, pocket Cα RMSD/RMSF, and contact fractions) collectively support a single, persistent NV10 pose stabilized by GLN36A/ARG42A H-bonding and a LYS46A π-cation interaction, with hydrophobic contributions from the phenyl-triazole/alkyl-chloride arm; these orthogonal MD readouts strengthen the case for NV10 beyond docking scores. These interactions, involving charged and polar residues, computationally predicted that electrostatic and hydrogen bonding forces predominantly govern the stability of NV10 within the binding pocket.
MM-GBSA Energy Calculations
The binding free energy and other contributing energies in the form of MM-GBSA were calculated for the NV10-1SJ2 complex using the MD simulation trajectory. The results in Table 8 indicate that the binding of NV10-1SJ2 is driven by a combination of strong van der Waals and Coulombic interactions. These energies dominate over the solvation penalty, resulting in a highly favorable overall binding energy (−68.21 ± 3.10 kcal/mol). The modest contributions from H-bonding and lipophilic interactions further enhance the flexibility of the complex. The low covalent interaction energy suggests that NV10 does not form covalent bonds with the target but relies heavily on noncovalent interactions. The significant desolvation penalty indicates that water molecules in the binding site must be displaced to accommodate NV10, which may affect binding in a water-rich environment. The MM-GBSA binding free energies are presented as relative estimates, since entropic contributions were not explicitly calculated.
Binding Free Energy Components for the NV10-1SJ2 Calculated by MM-GBSA.
Chemical Synthesis Plan
NV10 can be synthesized via a base-catalyzed condensation between a substituted 1,2,3-triazole-4-carbaldehyde and oxindole in ethanol under reflux using piperidine as a catalyst (Figure 5). The reaction proceeds through deprotonation of the active C-3 methylene of oxindole, followed by nucleophilic attack on the aldehyde group of the triazole derivative and subsequent dehydration to form a C-3 methylene-linked oxindole-triazole hybrid. This strategy was selected because it preserves both pharmacophores, forms a stable carbon–carbon linkage, and allows structural diversification through variation of the triazole substituent. The mild reaction conditions and straightforward procedure support synthetic feasibility and make NV10 suitable for future experimental synthesis and biological validation.
Basic Synthetic Scheme of Compound NV10.
Limitations
Several previously published studies that combine docking/MD with in-vitro tests show that promising computational scores do not always equate to low MICs, due to permeability, efflux, metabolism, and cell-wall uptake differences in M. tuberculosis. Therefore, the next essential steps are synthesis of NV10 (and close analogues NV2/NV4/NV8), followed by enzymatic KatG assays, whole-cell MIC determinations (H37Rv and representative INH-resistant clinical isolates), and early ADME profiling. 47 Because KatG both defends against ROS and activates isoniazid, experimentally perturbing KatG activity may have complex consequences on drug activation and resistance. As other groups have emphasized, testing across katG-mutant clinical isolates is important to determine whether a new KatG-binding inhibitor would be broadly useful or selectively active only in WT backgrounds. 42 Because NV10’s binding relies on multiple polar contacts, we anticipate non-trivial desolvation costs that could attenuate whole-cell potency; we therefore propose polarity-tuning and pre-organization strategies (linker rigidification, triazole/oxindole edits) and outline a staged experimental plan spanning KatG target engagement, MIC in WT and katG-mutant strains, cytotoxicity/ADME, and PK to translate the computational signal into tractable, in vivo-relevant leads.
Conclusion
This in silico study predicted oxindole-1,2,3-triazole derivatives (NV1-NV11) against M. tuberculosis KatG using molecular docking, MD simulations, and MM-GBSA analysis. NV10 predicted the highest binding affinity (–10.262 kcal/mol), compared to isoniazid (−6.331 kcal/mol), and engaged key residues through multiple hydrogen bonds, hydrophobic, and π-cation interactions. MD simulations confirmed its structural stability, while MM-GBSA yielded a favorable ΔG_bind dominated by van der Waals and Coulombic forces. This is a computational finding, and experimental validation is required in future, such as synthesis, enzyme tests, MIC studies, and toxicity evaluation.
Supplemental Material
Supplemental material for this article is available online.
Footnotes
Acknowledgements
The authors are wholeheartedly thankful to the SRM Institute of Science and Technology, Kattankulathur, for providing facilities to carry out this research. We also thank the Dean of SRM College of Pharmacy, Kattankulathur, and the Head of Department, Department of Pharmaceutical Chemistry, SRM College of Pharmacy, for providing the facilities necessary for carrying out the research work.
Authors’ Contribution
Naveen Kumar R. was responsible for designing ligands, conducting molecular docking studies, performing molecular dynamics simulations, analyzing MM-GBSA data, interpreting results and preparing the original draft of the manuscript. Sundarrajan Thirunanasambandam contributed by supervising the project, validating methodologies, critically revising the manuscript for significant intellectual content and managing the overall project. Both authors participated in revising the manuscript, and they read and approved the final version for publication.
Consent to Participate
Not applicable.
Consent for Publication
Not applicable.
Data Availability Statement
Data will be provided by the corresponding author on request. All datasets are maintained in unprocessed form to preserve computational integrity and enable independent validation.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethical Approval
This study does not involve experiments on animals or human subjects.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Informed Consent
Not applicable.
Use of Artificial Intelligence-assisted Technology:
We declare that there is no use of any AI tools for writing or editing the submitted manuscript, and that none of the images were altered using AI.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
