Panax Notoginseng Saponins Repair Diabetic Wounds in Mice via Mitochondrial Protection

Panax notoginseng saponins accelerated repair of diabetic wounds in mice by preserving the lipid mediator 11,12-epoxyeicosatrienoic acid, protecting mitochondria in wound-associated macrophages and suppressing an inflammatory form of cell death called pyroptosis. The experiments linked this response to the Ephx2/11,12-EET pathway, which regulates the availability of 11,12-EET at the wound site. Because the work used diabetic mice, cultured macrophages and a public human single-cell dataset, it identifies a preclinical mechanism rather than demonstrating an established treatment for people with diabetic foot ulcers.

Diabetic wounds are difficult to repair because persistent hyperglycemia disrupts metabolism, prolongs inflammation and limits the activity of cells involved in collagen deposition and tissue reconstruction. Macrophages are central regulators of this environment. They can adopt a pro-inflammatory phenotype associated with markers such as iNOS and CD86, release cytokines including interleukin-1 beta, interleukin-6 and interleukin-18, and activate the NLRP3 inflammasome. That pathway activates caspase-1 and cleaves gasdermin D, enabling pyroptosis and the release of inflammatory signals. In the mouse study, diabetic wounds contained more macrophages with pro-inflammatory markers and elevated levels of proteins associated with inflammasome activation and gasdermin D cleavage.

Researchers induced diabetes in male C57BL/6 mice with five daily injections of streptozotocin and then created a single 6-millimeter full-thickness excisional wound on each animal’s back. Untreated diabetic mice healed more slowly than mice with non-diabetic wounds. Oral treatment with purified Panax notoginseng saponins, abbreviated PNS, accelerated closure in a dose-dependent manner. Histological analysis showed that treated diabetic wounds had smaller epithelial gaps, less inflammatory-cell infiltration and greater collagen deposition. PNS did not significantly change body weight or random blood-glucose concentrations, although it improved performance in glucose-tolerance and insulin-tolerance tests. The wound findings therefore cannot be explained simply by a measured reduction in random blood glucose.

PNS also changed the inflammatory profile within the wounds. The saponin treatment reduced the abundance of macrophages carrying iNOS, a marker associated with the pro-inflammatory M1 phenotype, and increased cells carrying Arg1, a marker associated with the reparative M2 phenotype. Serum concentrations of interleukin-1 beta, interleukin-6 and interleukin-18 fell, while interleukin-10 increased. At the molecular level, PNS lowered NLRP3, inactive and active caspase-1, and both full-length and cleaved gasdermin D in wound tissue. Transmission electron microscopy showed swollen mitochondria and other ultrastructural abnormalities in diabetic wounds; PNS-treated tissues had fewer of these abnormalities. The treatment also shifted mitochondrial dynamics away from fission and toward fusion by lowering the activation of dynamin-related protein 1 and increasing mitofusin 1, mitofusin 2 and optic atrophy 1.

The team used untargeted metabolomics to search for metabolic changes connecting PNS treatment to the cellular response. Sixty-seven metabolites differed significantly between normal and diabetic wound tissue, and PNS restored 17 of those metabolites toward normal levels. Among metabolites altered by both diabetes and treatment, 11,12-EET emerged as a key candidate. This lipid is produced from arachidonic acid by cytochrome P450 epoxygenases and is degraded by soluble epoxide hydrolase, a protein encoded by Ephx2, into dihydroxyeicosatrienoic acid metabolites with substantially reduced activity. The ratio of 11,12-EET to its degradation product changed after treatment, supporting the interpretation that reduced degradation contributed to the metabolic effect.

Human foot-skin data in the public single-cell dataset GSE165816 supplied an additional, although nonexperimental, comparison. Macrophages from people with non-healing diabetic foot ulcers expressed more Ephx2 than macrophages from healthy skin, whereas Ephx1 showed the opposite pattern. No genes from the CYP2J family were detected in that dataset, limiting what could be inferred about the enzymes that produce EETs. In RAW 264.7 macrophages exposed to 30 millimolar glucose and 100 nanograms per milliliter of lipopolysaccharide, PNS selectively reversed Ephx2 messenger RNA and protein expression rather than changing Ephx1 or the tested CYP2J-family genes. It also increased the thermal stability of Ephx2 protein, which suggests a possible intracellular interaction but does not establish a direct binding mechanism. PNS restored 11,12-EET in the culture medium in a concentration-dependent fashion without reducing macrophage viability at the tested concentration.

Pharmacological tests then asked whether soluble epoxide hydrolase inhibition and 11,12-EET activity were required for the cellular effects. TPPU, an inhibitor of soluble epoxide hydrolase, reproduced many effects of PNS in macrophages under high-glucose and lipopolysaccharide stress: it reduced CD86 expression, lowered inflammatory cytokines, decreased the Annexin V/propidium iodide double-positive population and suppressed markers of pyroptosis. It also restored the balance among mitochondrial fusion and fission proteins. Adding TPPU to PNS did not significantly strengthen these effects, a result consistent with the compounds acting through a shared pathway, although it does not by itself prove an identical molecular target. In contrast, 14,15-EEZE, an antagonist of 11,12-EET, reversed PNS’s protection of mitochondrial morphology, membrane potential and energy metabolism. It also eliminated the PNS-induced increases in mitofusin 1, mitofusin 2 and optic atrophy 1 and the associated reduction in dynamin-related protein 1.

The same loss-of-function pattern was observed in diabetic mice. Co-administration of TPPU did not significantly improve wound closure beyond PNS alone. By contrast, 14,15-EEZE prevented PNS from accelerating wound closure. The antagonist also disrupted the improvement in collagen organization caused by PNS and reversed the reduction in gasdermin D expression in wound tissue. PNS increased 11,12-EET and suppressed Ephx2 messenger RNA in diabetic wound tissue, but those changes were abolished by 14,15-EEZE and were not significantly potentiated by TPPU. Together, these pharmacological results support 11,12-EET as a key mediator of the wound-healing response, linking preservation of the lipid with reduced pyroptosis, improved collagen deposition and faster closure in this mouse model.

Analytical chemistry identified five relatively abundant saponins in the PNS extract: notoginsenoside R1, ginsenosides Rg1, Re and Rb1, and ginsenoside Rd. Molecular docking predicted strong binding of all five compounds to Ephx2, with calculated binding energies from approximately minus 9.0 to minus 8.5 kilocalories per mole. These predictions were followed by mouse experiments in which each saponin was administered individually at 10 milligrams per kilogram. Each compound accelerated diabetic wound closure, increased collagen deposition, reduced Ephx2 messenger RNA and raised 11,12-EET in wound tissue. The individual saponins also lowered interleukin-1 beta, interleukin-6 and interleukin-18 and reduced gasdermin D expression. Their observed wound-healing effects were comparable to those of metformin, which was used as a positive control at 200 milligrams per kilogram, but the experiment did not establish clinical equivalence or suitability for human use.

Several limitations constrain how the findings can be interpreted. The investigation centered on macrophages, whereas wound repair also involves keratinocytes, fibroblasts, endothelial cells and multiple signaling networks. Untargeted metabolomics offers broad coverage but can miss low-abundance metabolites or compounds absent from searchable databases. The pharmacological tools also have limitations: soluble epoxide hydrolase and 11,12-EET antagonists may have effects beyond the intended pathway, and gene-knockout or conditional-knockout experiments were not performed. Oral exposure is another issue because ginsenosides and notoginsenosides have relatively low systemic bioavailability owing to limited membrane permeability and extensive metabolism by gut microbes and hepatic enzymes. Intact saponins and their microbial derivatives may therefore contribute differently to the observed effects. The results show that PNS can act on the Ephx2/11,12-EET pathway in diabetic mice and stressed macrophages, but they do not establish that the same mechanism will produce safe or effective wound repair in humans.

Subject of Research: Preclinical mechanisms of Panax notoginseng saponins in diabetic wound repair

Article Title: Panax notoginseng Saponins Accelerate Diabetic Wound Repair by Minimizing Mitochondrial Dysfunction and Inhibiting Pyroptosis of Macrophage via Regulating Ephx2/11,12‐EET Axis

Article References: Chen, N., Zhu, B., Sun, C.-H., & Hou, T. (2026). Panax notoginseng Saponins Accelerate Diabetic Wound Repair by Minimizing Mitochondrial Dysfunction and Inhibiting Pyroptosis of Macrophage via Regulating Ephx2/11,12‐ EET Axis. Food Science & Nutrition, 14(10), Article e72494. https://doi.org/10.1002/fsn3.72494

Image Credits: AI Generated

DOI: 10.1002/fsn3.72494

Keywords: diabetic wounds, Panax notoginseng, saponins, macrophages, pyroptosis, mitochondria, 11,12-EET, Ephx2, inflammation, metabolomics, mouse model, Panax