Result and discussion

Chemistry

To synthesize 5,6-diphenyl-1,2,4-triazine-piperazine derivatives, a synthetic route has been optimized, as depicted in Scheme 1.

Scheme 1.: Synthesis of 5,6-diphenyl-1,2,4-triazine-piperazine derivatives (FP1 to FP9). Reagents and conditions: (a) EDC.HCl, HOBt, TEA, DMF, r.t. 8-10 h; (b) Dioxane. HCl, DCM, r.t., 5-6 h (c) chloroacetyl chloride, DIPEA, DCM, r.t., 4-5 h; (d) K2CO3, KI, DMF, 80 °C, 8 h (e) Glacial acetic acid, reflux, 2-3 h

In the first step, N-Boc-piperazine (1) was reacted with various substituted benzoic acids (2 to 10) in the presence of coupling agents EDC·HCl and HOBt, using triethylamine as a base in DMF, to afford the corresponding amides (11 to 19). Subsequent deprotection of the Boc group using HCl in dioxane yielded the acetamides (20 to 28). In the next step, these acetamides (20 to 28) were coupled with chloroacetyl chloride in the presence of N,N-diisopropylethylamine to give the acetylated intermediates (29 to 37). In parallel, thiosemicarbazide reacted with benzil in glacial acetic acid to form intermediate (38), which was subsequently reacted in the final step with the acetylated intermediates (29 to 37) in the presence of potassium iodide and potassium carbonate in DMF to afford the final products (FP1 to FP9). The synthesized target compounds (FP1 to FP9) were characterized using spectral and elemental analyses.

Molecular docking studies

Molecular modelling studies were carried out for the designed compounds, as well as the standard compounds, to justify the rationale behind the design of the hybrid compounds. As the exact site of action of the synthesized compounds was not identified, molecular docking studies were carried out for the synthesized compounds targeting two major enzymes of Mycobacterium tuberculosis, i.e. DprE1 and Mtb-DHFR. The docking analysis was performed to predict potential modes of action for the designed compounds. The docking analysis was performed to elucidate the interactions between the designed compounds and the active sites of the targeted enzymes.

Figure 2 presents the docking results for the synthesized compounds (8 to 14), as well as the standard inhibitors, i.e. PBTZ169 and methotrexate, to predict the binding affinities of the compounds with the targeted enzymes, i.e. DprE1 and Mtb-DHFR. The docking analysis was carried out using AutoDock Tools 1.5.7 [50], which estimated the binding energy between the ligand and the target proteins.

Figure 2.: Molecular docking scores of the designed compounds (FP1 to FP9)

Figure 3 presents the results obtained for the docking analysis, which indicate the alignment of the proposed binding mode of the inhibitor with the active site of the targeted enzyme, i.e. Mtb-DHFR. The docking results were compared with the docking results of the standard inhibitor, i.e. methotrexate, which produced a root mean square deviation (RMSD) of 0.107 nm. The results were well within the acceptable range of 0.2 nm, which indicated the accurate docking results.

Figure 3.: (a) validation results of the binding modes of methotrexate obtained using the AutoDock software, (b) docking conformation and Mtb-DHFR protein-ligand interactions of reference drug methotrexate

In the present study, the synthesized compounds were docked into the DprE1 (PDB ID: 4NCR) and Mtb-DHFR (PDB ID: 1DF7) proteins to assess their binding affinity for the receptors. All the designed molecules were docked into the same active site, wherein the co-crystallized ligand was lodged. It was observed that the inbuilt ligands PBTZ169 and Methotrexate showed binding affinity -38.45 and -45.19 kJ mol-1 (1 kJ = 0.239 kcal) towards the receptor, respectively. From that PBTZ169 that formed H, binding with amino acids like Lys134, Gly177, Gln336, CYS387, which are important amino acids. It was observed that CYS387, which is also involved in π-sulphur type of interaction and other amino acids, His132 with π-Cation, and Arg58, Trp16, Tyr60, Val365, Lys367 with Pi-Alkyl type of interaction was observed [51]. Designed compounds that showed binding affinity towards DprE1 ranged from -38.83 to -44.18 kJ mol-1 and -46.02 to -50.38 kJ mol-1 towards Mtb-DHFR. Based on the results, all compounds showed higher binding affinity for Mtb-DHFR than for DprE1, so a detailed molecular docking study is discussed here.

Compound FP1, which had a binding affinity of -49.16 kJ mol-1 towards Mtb-DHFR, formed hydrogen bonding with the Ala7 amino acid and π-sigma type of interaction with Ile14 and Thr46 amino acids. The phenyl ring of the structure also formed a π-alkyl interaction with Ile20. The docked conformation FP3 showed that the ligand formed key stabilizing interactions within the active site (Figure 4).

Figure 4.: Docking conformations and Mtb-DHFR protein-ligand interactions of designed molecules. (a) FP1, (b) FP3, (c) FP8. Dark green colour indicates conventional hydrogen bonding, light green colour indicates Van der Waals interactions, purple colour indicates π-sigma interactions, pink colour indicates hydrophobic interactions (π-π stacking, π-π T-shaped)

The triazine ring engaged in π-π stacking with Phe31 and π-alkyl interactions with Ile14, whereas the aromatic ring formed interactions with Ile20 and Ile94 residues. The carbonyl oxygen atom of the compound is engaged in hydrogen bonding with Arg32 and Arg60. The piperazine ring formed an alkyl interaction with Leu57 and Val54, with a binding affinity of -48.37 kJ mol-1 towards Mtb-DHFR. Compound FP3 also has a binding affinity of -39.12 kJ mol-1 towards DprE1 and forms hydrogen bonding with Tyr60 and Lys134 amino acids. Compound FP4, FP5, FP6, FP7 and FP9 that formed a common type of interaction with Ala7, Ile20, Arg32, and Thr46 amino acid, and having binding affinity -49.87, -46.02, -47.91, -46.48, -49.62 and -46.65 kJ mol-1, respectively. Meanwhile, compound FP8, which has a binding affinity of -50.38 kJ mol-1 and a triazine ring that forms hydrogen bonding with Ala7 and pi-pi T-shaped interaction with Tyr100, a further biphenyl ring that forms π-sigma type interaction with Ile14, and Thr46 amino acid. This compound FP8 also exhibited a binding affinity of -9.62 towards DprE1 and interacted with key amino acids such as Pro116 (π-π), Lys134 (π-Alkyl), Leu363 (π-Alkyl), Val365 (π-Alkyl), Lys367 (π-Alkyl), Cys387 (π-s), and Tyr415 (π-Cation).

Molecular dynamics simulation

Following molecular docking, the stability and dynamic behaviour of the protein-ligand complexes were examined through molecular dynamics simulations. The backbone RMSD and RMSF parameters were used to assess structural deviations and flexibility relative to the unbound protein. The RMSD profiles of the protein in its apo form and in complex with the reference molecule, as well as FP1, FP3 and FP8 ligands, are shown in Figure 5. The apo structure exhibited greater fluctuations throughout the simulation, indicating a more flexible conformation in the absence of a bound ligand. In contrast, the ligand-bound systems exhibited quicker stabilization during the initial equilibration phase, followed by relatively stable RMSD trajectories. Among the complexes, FP1 and FP3 maintained consistent RMSD values around ~0.18 to 0.21 nm with only minor fluctuations, suggesting stable accommodation within the binding cavity (Table 1). The FP8 complex showed slightly higher deviation (0.24 to 0.27 nm), though still within an acceptable range, reflecting moderate structural stability of the complex.

Figure 5.: RMSD of protein backbone on binding of the ligands (reference, FP1, FP3, FP8 and apo) to the target enzyme Mtb-DHFR

No. | Parameters | DprE1
Reference | FP1 | FP3 | FP8 | apo form
1 | RMSD backbone, nm | 0.1979 | 0.1982 | 0.1795 | 0.2295 | 0.2871
2 | RMSD Cα, nm | 0.1956 | 0.1973 | 0.1799 | 0.2248 | 0.1843
3 | RMSF, nm | 0.1001 | 0.0872 | 0.1070 | 0.1308 | 0.1094

The RMSF profiles further supported these observations, where most residues fluctuated below 0.20 nm, indicating limited local flexibility. Only a few loop-region residues displayed fluctuations approaching 0.30 to 0.35 nm, which is typical for solvent-exposed or non-structured regions. Importantly, key residues within the binding site, including hydrophobic and aromatic residues such as Trp22 and Phe31, remained conformationally stable throughout the simulation, preserving crucial protein-ligand contacts. Hydrogen-bonding interactions, including water-mediated contacts involving Trp22, persisted for a substantial portion of the simulation timeframe. These interactions, together with stable hydrophobic packing, contributed to the effective retention of FP1, FP3 and FP8 within the active site, thereby reinforcing their potential as promising inhibitory candidates.

DFT calculation

Theoretical studies of the synthesized compounds were conducted using the 6-31G** basis set within the Schrödinger suite. This approach facilitated the estimation of essential thermochemical descriptors, namely ionization potential (IP = -_E_HOMO), electron affinity (EA = -E_LUMO), electronegativity [χ = (IP + EA)/2], chemical potential (μ = -χ), global hardness [η = (IP − EA)/2], softness (σ = 1/2_η), and electrophilicity index [ω = μ_2/2_η]. Collectively, these parameters offer important insights into the molecular reactivity profile. Prior to property evaluation, the designed compounds (FP1 to FP9) were geometrically optimized to obtain _E_HOMO and _E_LUMO values, which are presented in Table 2.

Compound | HOMO, eV | LUMO, eV | ΔE (ELUMO - EHOMO), eV
FP1 | -0.2249 | -0.074 | 0.1510
FP2 | -0.2286 | -0.079 | 0.1500
FP3 | -0.2255 | -0.076 | 0.1493
FP4 | -0.2262 | -0.097 | 0.1297
FP5 | -0.2292 | -0.101 | 0.1285
FP6 | -0.2240 | -0.075 | 0.1489
FP7 | -0.2281 | -0.087 | 0.1410
FP8 | -0.2244 | -0.075 | 0.1492
FP9 | -0.2231 | -0.071 | 0.1521

The molecular reactivity descriptors calculated for the selected compounds FP1, FP3 and FP8 (Table 3) indicate closely related but distinguishable electronic characteristics (Figure 6). All three compounds exhibited comparable ionization potential values (0.2249 eV for FP1, 0.2255 eV for FP3 and 0.2244 eV for FP8), suggesting a similar tendency to donate electrons. The electron affinity values also fall within a narrow range, reflecting the balanced electron-accepting capabilities of these molecules.

Compound | IP, eV | EA, eV | Χ / eV | μ / eV | η / eV | σ / eV-1 | ω
FP1 | 0.2249 | 0.074 | 0.2618 | -0.2618 | 0.1879 | 0.0940 | 0.00644
FP2 | 0.2286 | 0.079 | 0.2678 | -0.2678 | 0.1893 | 0.0946 | 0.00679
FP3 | 0.2255 | 0.076 | 0.2636 | -0.2636 | 0.1874 | 0.0937 | 0.00651
FP4 | 0.2262 | 0.097 | 0.2745 | -0.2745 | 0.1779 | 0.0890 | 0.00670
FP5 | 0.2292 | 0.101 | 0.2795 | -0.2795 | 0.1788 | 0.0894 | 0.00699
FP6 | 0.2240 | 0.075 | 0.2615 | -0.2615 | 0.1865 | 0.0932 | 0.00637
FP7 | 0.2281 | 0.087 | 0.2716 | -0.2716 | 0.1846 | 0.0923 | 0.00681
FP8 | 0.2244 | 0.075 | 0.2620 | -0.2620 | 0.1868 | 0.0934 | 0.00641
FP9 | 0.2231 | 0.071 | 0.2586 | -0.2586 | 0.1876 | 0.0938 | 0.00627

Figure 6.: The computed HOMO, LUMO energies and energy gap for compounds FP1, FP3 and FP8

In terms of electronegativity, FP3 (0.2636 eV) and FP8 (0.2620 eV) show marginally higher values than FP1 (0.2618 eV), indicating a slightly stronger pull toward electron density. The negative electrostatic potential values further support the stability of the electronic environment of these compounds. Notably, FP3 exhibits the lowest hardness (0.1874 eV), followed closely by FP8 (0.1868 eV), indicating greater softness and greater adaptability to electronic redistribution. This behaviour is consistent with their relatively higher softness values, which are often associated with improved chemical reactivity.

Overall, the combined analysis suggests that FP3 and FP8 may possess slightly higher reactivity than FP1, owing to their lower hardness and favourable softness parameters. However, the proximity of all the calculated values highlights that these compounds share a similar electronic framework, making them promising candidates with balanced stability and reactivity profiles.

In silico drug likeness and computed pharmacokinetics parameters

All the compounds were evaluated in silico for their physicochemical properties (Table 4). Furthermore, the surface area and number of rotatable bonds were found to be within the limits. Each compound displayed varying degrees of water solubility, ranging from moderate to high. Additionally, a range of pharmacokinetic properties was computed using the SwissADME server.

Compound | Molecular weight | Number of hydrogen bond acceptors | logP*
FP1 | 495.6 | 5 | 3.74
FP2 | 530.04 | 5 | 4.1
FP3 | 513.59 | 6 | 3.88
FP4 | 574.49 | 5 | 4.18
FP5 | 540.59 | 7 | 3.01
FP6 | 525.62 | 6 | 3.68
FP7 | 530.04 | 5 | 4.23
FP8 | 509.62 | 5 | 4.02
FP9 | 564.49 | 5 | 4.73

Anti-tuberculosis activity

All synthesized compounds were evaluated for their in vitro anti-tubercular activity against Mycobacterium tuberculosis H37Rv strain, and their efficacy was compared with standard first-line anti-TB drugs, namely isoniazid, ethionamide, and pyrazinamide (Table 5).

Compd. | R | MIC, μg mL-1H37RV | Cytotoxicity, IC50, μg mL-1RAW 264.7a
FP1 |  | 3.12 | 46.19 ± 0.18
FP2 |  | 25 | ND
FP3 |  | 1.6 | 96.59 ± 0.16
FP4 |  | 25 | ND
FP5 |  | 6.2 | ND
FP6 |  | 50 | ND
FP7 |  | 25 | ND
FP8 |  | 1.6 | 46.28 ± 0.15
FP9 |  | 3.12 | ND

Among the series, compounds FP3 (4-fluoro) and FP8 (4-methyl) exhibited the most promising activity, with MIC values of 1.6 μg mL-1. This level of activity was found to be comparable to that of isoniazid and ethionamide (MIC = 1.6 μg mL-1 for both), suggesting that these compounds possess strong potential as lead molecules for further development. Compounds FP1 (unsubstituted, -H) and FP9 (2,4-dichloro) also demonstrated noteworthy inhibition of M. tuberculosis, with MIC values of 3.12 μg mL-1, nearly equivalent to that of pyrazinamide (MIC = 3.13 μg mL-1), a standard second-line anti-TB agent. These results indicate that certain substitutions, such as small electron-donating groups or appropriately placed halogens, can lead to compounds with activity profiles closely resembling those of established drugs. On the other hand, several other derivatives, including FP2 (4-Cl), FP4 (4-Br), FP5 (4-NO₂) and FP7 (2-Cl), showed moderate to poor activity, with MIC values ranging from 25 to 50 μg mL-1, and were significantly less potent than the reference standards. These findings further reinforce the importance of substitution pattern and the electronic nature of functional groups in modulating anti-tubercular activity. Overall, the comparative data suggest that FP3 and FP8, in particular, exhibit a level of efficacy that is on par with first-line anti-TB drugs, making them promising candidates for further pharmacological evaluation.

Structure-activity relationship for anti-tubercular activity

The structure-activity relationship (SAR) study of the synthesized compounds (FP1 to FP9) against Mycobacterium tuberculosis H37Rv showed that the activity against tuberculosis was significantly affected by the type and size of the substituents on the phenyl ring. For instance, the unsubstituted phenyl derivative FP1 showed significant activity against the pathogen (MIC = 3.12 μg mL-1). This suggests that the phenyl ring itself serves as a privileged structure for the target pathogen. When the phenyl ring was slightly modified with smaller substituents, the activity was further increased, as demonstrated by FP3, where the para position on the phenyl ring was occupied by a fluoro group, and FP8, where the para position was occupied by a methyl group, showing the highest activity against the pathogen (MIC = 1.6 μg mL-1). In contrast, compounds bearing bulkier halogen substituents, such as chloro (FP2) and bromo (FP4), showed decreased activity, with an MIC of 25 μg mL-1. These observations suggest that bulkier substituents negatively affect the compounds’ activity. The presence of a powerful electron-withdrawing group, such as nitro in FP5, resulted in moderate activity, with an MIC of 6.2 μg mL-1. These observations suggest partial tolerance to electron-withdrawing groups. On the contrary, the presence of a powerful electron-donating group, such as methoxy, in FP6 resulted in decreased activity, with an MIC of 50 μg mL-1. These observations suggest negative effects of excessive electron donation. The positional and multiplicity effects were also observed in this series. For example, the monochloro-substituted FP7 showed moderate activity, with an MIC of 25 μg mL-1. The dichloro-substituted FP9 showed increased activity with an MIC of 3.12 μg mL-1. These observations suggest the positive effects of increased lipophilicity. The most active compounds in this series showed good cytotoxicity profiles in RAW 264.7 cells.