Evaluation of antibacterial activity of hydroethanolic seed extract of Descurainia sophia against multidrug-resistant Salmonella typhi isolated from typhoid patients: an integrated in silico and in vitro study

Typhoid fever caused by Salmonella typhi remains a major public health problem, and multidrug-resistant (MDR) and extensively drug-resistant strains have left few oral treatment options in endemic regions such as Pakistan. Seeds of Descurainia sophia (L.) are used in Unani and folk medicine for febrile illness and contain cardenolides, glucosinolate-derived isothiocyanates and long-chain fatty acids, but their activity against MDR S. typhi and a plausible molecular target have not been examined. We therefore combined in silico and in vitro approaches to evaluate a hydroethanolic seed extract against clinical MDR S. typhi isolates. A library of 14 literature-reported seed compounds was screened by ADMET profiling (admetSAR 3.0), which selected three drug-like cardenolides (helveticoside, evomonoside and strophanthidin); these were docked against DNA gyrase A with moxifloxacin and gemifloxacin as references, and the top-ranked lead, helveticoside, was then examined in a 50 ns all-atom molecular dynamics (MD) simulation. The extract was characterised by GC-MS and FTIR and tested against five MDR isolates (MDR-ST1 to MDR-ST5) by disc diffusion and broth microdilution (CLSI M07). GC-MS confirmed 13 of the 14 library compounds (91.4% of total ion chromatogram area), including the three leads; erysimoside was the only compound not detected. Helveticoside (- 9.7 kcal/mol) and evomonoside (- 9.5) showed higher predicted affinities than moxifloxacin (- 7.9) and gemifloxacin (- 8.2); strophanthidin scored - 8.5. In the MD simulation, the helveticoside complex remained stable (ligand RMSD ~ 1.8 Å without dissociation; protein Cα RMSD ~ 2.2-2.5 Å after ~ 15 ns), and 12 residues, including ARG31, SER185 and GLN288, contacted the ligand in > 95% of frames, six of them also present in the docked pose. The extract inhibited all five isolates in a dose-dependent manner (mean zone of inhibition 11.5-17.15 mm at 25-100 mg/mL; 19.3 ± 0.9 mm for moxifloxacin, reported descriptively because the extract is crude), with a geometric mean MIC of 8.4 mg/mL (range 6.25-12.5) and MBC/MIC = 2 for all isolates, consistent with bactericidal activity. FTIR showed functional groups consistent with the major constituent classes.Docking and MD together generate the hypothesis that DNA gyrase A inhibition by cardenolide constituents may contribute to the antibacterial effect of the extract. The MD simulation supports the structural stability of the predicted helveticoside binding mode but does not demonstrate enzyme inhibition, and fatty acids and isothiocyanates in the extract may also contribute. Overall, the crude extract showed in vitro antibacterial activity against MDR S. typhi at high concentrations, and the study provides a testable mechanistic hypothesis that requires bioactivity-guided fractionation, pure-compound testing and target-engagement assays.