The motivation is urgent. Typhoid fever, caused by Salmonella enterica serovar Typhi, still generates an estimated 9 to 11 million cases and more than 110,000 deaths each year according to the World Health Organization, with the heaviest burden in South Asia and sub-Saharan Africa. Pakistan sits at the epicenter: extensively drug-resistant strains, resistant to all standard first- and second-line oral agents, have accounted for more than 10,000 confirmed cases since the 2016–2019 outbreak in Hyderabad, Sindh, and a multi-laboratory network study covering 2017–2019 found XDR strains in 57 percent of culture-confirmed isolates. With azithromycin and carbapenems increasingly compromised, the search for new bioactive compounds has taken on real clinical weight.

The plant in question, Descurainia sophia (L.) Webb ex Prantl, known vernacularly as khubkalan or flixweed, is an annual Brassicaceae herb used in Unani, Ayurvedic and traditional Chinese medicine for febrile illness, gastroenteritis and respiratory inflammation. Its seeds contain cardenolide cardiac glycosides such as helveticoside, evomonoside, strophanthidin and erysimoside, glucosinolate-derived isothiocyanates, long-chain fatty acids and organosulfur compounds. Members of all these classes have independently reported antibacterial activity, but no prior study had examined the seeds against multidrug-resistant S. typhi using a combined computational and wet-laboratory approach.

The team first built a phytochemical library through systematic literature mining, retaining only compounds confirmed by chromatographic or spectroscopic methods in at least two independent peer-reviewed sources. Fourteen compounds survived. ADMET profiling with the admetSAR 3.0 server, using pre-defined thresholds for molecular weight, lipophilicity, predicted mutagenicity, cytochrome P450 inhibition, polar surface area and other drug-likeness parameters, narrowed the field to three cardenolides: helveticoside, evomonoside and strophanthidin. The six fatty acids failed on logP and absorption criteria, while the isothiocyanates and diallyl disulfide were excluded mainly on predicted Ames mutagenicity, properties that limit their development as systemic drugs but not their possible contribution to an extract’s in vitro activity.

Docking against Salmonella typhi DNA gyrase A, the validated target of fluoroquinolone antibiotics, produced striking numbers on paper. Helveticoside scored −9.7 kcal/mol and evomonoside −9.5, both higher than the reference drugs moxifloxacin (−7.9) and gemifloxacin (−8.2); strophanthidin scored −8.5. The reference drugs recovered the key pharmacophoric residues SER185, LEU32, VAL34 and ARG31, validating the grid placement. A 50 ns all-atom molecular dynamics simulation of the helveticoside–gyrase A complex then showed the ligand remaining in the pocket with a root-mean-square deviation of roughly 1.8 Å after an adjustment phase, while the protein backbone stabilized at 2.2–2.5 Å. Twelve residues, including ARG31, SER185 and GLN288, contacted the ligand in more than 95 percent of simulation frames, six of them also present in the docked pose.

The authors are explicit that these computational results generate a hypothesis rather than demonstrate enzyme inhibition. Docking scores from a single program tend to favor larger ligands, and the two glycosides are considerably heavier than the fluoroquinolones. The simulation used an apo, chain A-only structure lacking DNA, GyrB and metal ions, so it represents only part of the quinolone binding environment, and a ligand that stays in a pocket is not thereby shown to inhibit the enzyme. Only six of twelve persistent contacts coincided with the docked pose, and two docked hydrogen-bonding residues were not retained, suggesting partial rearrangement. No gyrase supercoiling or cleavage-complex assay was performed, so the DNA gyrase hypothesis remains entirely speculative.

On the laboratory side, seeds collected in Multan district, Punjab were macerated in 70:30 ethanol–water, yielding a dark greenish-brown semi-solid extract at 3.72 ± 0.04 percent. GC-MS confirmed thirteen of the fourteen library compounds, accounting for 91.4 percent of the total ion chromatogram area, including all three prioritized cardenolides; only erysimoside went undetected, attributed to its size and polarity. Allyl isothiocyanate was the largest peak at 18.6 percent relative area, followed by erucic acid at 14.7 percent and benzyl isothiocyanate at 12.4 percent. FTIR spectroscopy showed functional groups consistent with these major constituent classes, and a skin patch test under OECD Guideline 404 found no dermal irritation after 24 hours.

Five clinical isolates, each from a separate patient at a Multan referral hospital, were all resistant to chloramphenicol, ampicillin and co-trimoxazole, defining multidrug resistance, though all remained susceptible to ciprofloxacin, moxifloxacin and gemifloxacin. In disc diffusion assays the extract inhibited every isolate at every concentration, with pooled mean zones of inhibition rising from 11.5 ± 0.7 mm at 25 mg/mL to 17.15 ± 1.3 mm at 100 mg/mL; a two-way mixed-effects ANOVA confirmed a significant concentration effect, and all four concentrations differed significantly from one another. Broth microdilution per CLSI M07 gave MICs of 6.25 to 12.5 mg/mL, and MBC values were exactly twofold the MIC for every isolate, meeting the conventional criterion for bactericidal activity.

The authors frame the potency honestly. A geometric mean MIC of 8.4 mg/mL is high in absolute terms, roughly an order of magnitude or more above values reported for more active plant fractions against S. typhi, and the moxifloxacin and gemifloxacin control zones of 19.3 and 22.75 mm exceeded the extract’s best result. The activity is best described as preliminary and moderate-to-weak, and the zone-of-inhibition data are treated as qualitative confirmation, since agar diffusion is unreliable for lipophilic fatty acids and volatile isothiocyanates. Because the extract is an unfractionated mixture, the zone diameters were reported descriptively rather than tested against the antibiotic controls.

What actually drives the inhibition is unresolved. The fatty acids, particularly the abundant unsaturated species, are well-documented membrane-disrupting agents, and Gram-negative bacteria are typically less susceptible because the outer membrane limits access, which may explain why activity appears only at milligram-per-milliliter concentrations. Membrane perturbation could also sensitize the bacteria to other constituents, potentially helping the large cardenolide glycosides enter the cell. The isothiocyanates are intrinsically antibacterial in other organisms but volatile, so their effective concentration during the assays is uncertain. The authors call for bioactivity-guided fractionation, pure-compound MIC determination, checkerboard interaction assays, LC-MS/MS confirmation of the cardenolides, a gyrase enzyme assay, testing against fluoroquinolone-resistant and XDR isolates, and cytotoxicity evaluation in human cells, a critical gap given the narrow therapeutic index of cardiac glycosides, before any therapeutic relevance can be considered.

Subject of Research: Antibacterial activity of Descurainia sophia seed extract against multidrug-resistant Salmonella typhi

Article Title: 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

Article References: Qureshi, M., Fatima, R., Anjum, A., Khan, B., Khan, M. I. A., Rehman, A., Khan, M. K., Khan, I. A., & Hamed, H. (2026). 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. International Microbiology. https://doi.org/10.1007/s10123-026-00909-x

Image Credits: AI Generated

DOI: 10.1007/s10123-026-00909-x

Keywords: Descurainia sophia, Salmonella typhi, typhoid fever, multidrug resistance, cardenolides, molecular docking, molecular dynamics, DNA gyrase A, GC-MS, isothiocyanates, fatty acids, Pakistan