The research, published in Current Research in Food Science, was led by Chengcheng Han and Peng Xue of Shandong Second Medical University, together with colleagues in China. The team set out to answer a question that has long puzzled food microbiologists: plant saponins are famous for punching holes in the membranes of eukaryotic cells by binding sterols such as cholesterol, but bacterial membranes contain no sterols at all. What, then, makes certain saponins so lethal to bacteria like B. cereus?

The answer, it turns out, may lie in hopanoids — ancient pentacyclic triterpenoid lipids that bacteria synthesize as structural and functional analogues of sterols. These lipids modulate membrane fluidity, maintain the integrity of the lipid bilayer, and help cells resist environmental stresses. The researchers hypothesized that the right saponin could target these sterol stand-ins, and they designed an unusually comprehensive experiment to find out, combining untargeted metabolomics, lipidomics, molecular docking, and gene knockout for the first time in a single investigation of saponin antibacterial action.

The raw material was quinoa husk, the outer seed coat of Chenopodium quinoa Willd., which is exceptionally rich in triterpenoid saponins. Using reversed-phase silica chromatography and preparative liquid chromatography, the team isolated eight saponin monomers, Q1 through Q8. Structural characterization by HPLC-MS and nuclear magnetic resonance revealed that Q7 was 3-O-β-D-arabinose-phytolaccagenic acid, while Q8 — phytolaccagenic acid itself — had never before been isolated from quinoa husks.

When the eight monomers were tested against B. cereus, only Q7 and Q8 showed antibacterial activity. Q8 was clearly the stronger of the two, with a minimum inhibitory concentration of 125 micrograms per milliliter and a minimum bactericidal concentration of 1000 micrograms per milliliter. Crucially, the activity was not strain-dependent: a food-derived outbreak isolate, an emetic-type strain, and a certified food-inspection reference strain all showed identical MIC and MBC values to the laboratory type strain, suggesting that Q8 could work broadly against the diverse B. cereus populations that contaminate rice and cereals.

The structure-activity relationship that emerged was strikingly consistent. The more heavily glycosylated, more polar saponins Q1 through Q6 were inactive, as was the polar ginsenoside Re. By contrast, less polar compounds — the aglycones hedera and phytolaccagenic acid, and the deglycosylated ginsenoside Rk1, which showed an MIC of just 15.625 micrograms per milliliter — were far more potent. The pattern suggests that removing sugar groups reduces polarity and exposes the hydrophobic triterpenoid scaffold, allowing the molecule to interact directly with bacterial membranes. Protoplast experiments reinforced this idea: when the peptidoglycan cell wall was stripped away, Q8 remained just as effective, indicating that the membrane, not the wall, is the primary target.

A cascade of biophysical assays then traced the lethal sequence of events. SYTO 9/PI live-dead staining showed that Q8 destroyed selective membrane permeability. Rhodamine 123 measurements revealed rapid, dose-dependent membrane depolarization, an early sign of functional collapse. SYTOX Green uptake confirmed that the barrier function was breached, and Bradford protein assays documented massive cytoplasmic protein leakage — nearly 6 milligrams per milliliter — comparable to the levels caused by the potent reference saponin Rk1. Transmission electron microscopy delivered the visual proof: cells treated with Q8 were deformed, with ruptured envelopes and spilled cytoplasm, while cells treated with the inactive analogues Q3 and Re remained perfectly intact.

Q8 also triggered a concentration-dependent accumulation of intracellular reactive oxygen species, and untargeted metabolomics identified 98 significantly altered metabolites, with major perturbations in glutathione metabolism, cysteine and methionine metabolism, and fatty acid degradation. Because B. cereus depends on these thiol-based redox buffers to survive processing stresses, the findings suggest that Q8 compromises the bacterium’s oxidative defenses. Lipidomics added a second layer of damage: 22 significantly altered lipid species, with marked depletion of unsaturated lipids such as linolenyl palmitatoleate, MGDG (16:0/18:2), and PE (P-16:0/18:4) — precisely the acyl chains most vulnerable to ROS-mediated oxidation.

The most novel discovery concerned hopanoid biosynthesis. Molecular docking showed that Q8 binds squalene-hopene cyclase, the enzyme that catalyzes a key step in hopanoid production, with a binding energy of -8.51 kilocalories per mole — substantially stronger than Q7’s -6.19. Quantitative PCR confirmed that Q8 suppressed shc transcription by up to 2.81-fold, while the inactive analogues had no effect. To prove causation, the team used a CRISPR-Cas9 system to knock out shc in B. cereus. The mutant, which produced no detectable hopanoids, became dramatically more susceptible to Q8: its inhibition zone grew from 1.07 to 1.57 centimeters, and its MIC and MBC halved to 62.5 and 500 micrograms per milliliter. Reintroducing shc restored wild-type susceptibility, confirming that hopanoids are the membrane-stabilizing lipids that shield B. cereus from saponin attack.

Before celebrating, the researchers checked biosafety. At 250 micrograms per milliliter, Q8 maintained 100 percent viability of GES-1 human gastric mucosal cells, indicating that its membrane-disrupting activity is selectively restricted to prokaryotes — likely because saponin irritancy depends on binding affinity with membrane cholesterol, which Q8 apparently lacks. In a practical rice preservation model, Q8 sprayed onto rice at one MIC per gram significantly delayed B. cereus-induced yellowing and spoilage over four days at 30 degrees Celsius, keeping bacterial counts at 9.2 log CFU per gram versus 11.4 in untreated controls, and preserving the fresh white appearance of the grain. The authors caution that sensory properties, thermal stability, in vivo safety, and comparisons with approved preservatives such as Nisin still require systematic evaluation. But the core message is already viral-worthy: a molecule purified from agricultural waste, at low cost and with good biocompatibility, kills a major foodborne pathogen by simultaneously rupturing its membrane, flooding it with oxidative stress, and sabotaging the ancient hopanoid lipids that bacteria have used to stabilize their membranes for billions of years.

Subject of Research: Structure-activity relationship and antibacterial mechanism of saponins from quinoa husks against Bacillus cereus and their application in rice

Article Title: Structure-activity relationship and antibacterial mechanism of saponins from quinoa husks against Bacillus cereus and their application in rice

Article References: Han, C., Wang, Y., Yang, L., He, Y., Song, L., Zhang, R., Lu, H., & Xue, P. (2026). Structure-activity relationship and antibacterial mechanism of saponins from quinoa husks against Bacillus cereus and their application in rice. Current Research in Food Science, Article 101600. https://doi.org/10.1016/j.crfs.2026.101600

Image Credits: AI Generated

DOI: 10.1016/j.crfs.2026.101600

Keywords: Structure-activity, relationship, antibacterial, mechanism, saponins, quinoa, husks, against, Bacillus, cereus, application, rice