Sesamin Promotes Diabetic Wound Healing by Inhibiting Ferroptosis via the SIRT1/Keap1/Nrf2 Pathway

ABSTRACT

Diabetes mellitus frequently leads to diabetic wounds (DW), a serious complication for which current treatment options remain limited. This work examined the pro‐healing effects of sesamin (Sea), a major sesame lignan, on DW and the mechanisms underlying these effects. In vitro, human umbilical vein endothelial cells (HUVECs) were cultured under high glucose (HG) conditions to mimic diabetic dysfunction. Cell viability, lipid peroxidation, Fe²⁺ accumulation, mitochondrial function, and expression of ferroptosis‐related proteins were assessed. Sirt1 knockdown was performed to verify target specificity. In vivo, a streptozotocin‐induced diabetic mouse model with full‐thickness skin wounds was established. Sea was administered, and wound healing rates, reactive oxygen species (ROS) levels, GPX4 expression, and transcriptomic profiles were analyzed. In vitro, Sea dose‐dependently ameliorated HG‐induced dysfunction in HUVECs and suppressed ferroptosis. Mechanistically, Sea upregulates Sirt1 expression, which promotes the dissociation of Keap1/Nrf2, thereby facilitating Nrf2 nuclear translocation and upregulating anti‐ferroptosis proteins expression. These effects were abrogated by Sirt1 knockdown. In vivo, transcriptomic analysis revealed ferroptosis inhibition as a key mechanism underlying Sea‐mediated DW healing. Sea treatment accelerated STZ‐induced DW closuring, reduced ROS levels, and upregulated GPX4 expression, effects that were diminished by Sirt1 silencing. Collectively, our findings demonstrate that Sea promotes DW healing by activating the Sirt1/Keap1/Nrf2 pathway to inhibit ferroptosis, positioning Sea as a promising therapeutic candidate for DW treatment.

Sesamin activates the Sirt1/Keap1/Nrf2 pathway to inhibit ferroptosis and promote diabetic wound healing.

1. Introduction

The global burden of type 2 diabetes mellitus (T2DM) is escalating rapidly; by 2045, more than 783 million individuals are expected to be affected, with annual healthcare expenditures projected to exceed $966 billion. Among the most debilitating consequences of T2DM are diabetic wounds (DW), which occur in approximately 19%–34% of patients [1, 2, 3]. The clinical outcomes of DW are alarming: half to three‐fifths of these wounds become infected, and one‐fifth of such infections progress to the point of requiring lower limb amputation [2]. Patients with DW face a 5‐year mortality rate nearing 30%, significantly impacting clinical practice and socio‐economic systems [4]. Hence, gaining a deeper understanding of why diabetic wound (DW) healing is impaired and identifying new therapeutic interventions represent critical priorities in metabolic disease investigation.

Our knowledge of DW pathogenesis has been significantly advanced by recent studies, especially the association between diabetic wounds and ferroptosis [5, 6]. Within diabetic individuals, the wound setting is chronically pathological, with ongoing hyperglycemia, inflammation, ischemia, and hypoxia as key features [7, 8, 9]. Such a microenvironment renders tissues highly prone to acute reactive oxygen species (ROS), causing a sharp increase in oxygen species and subsequent lipid peroxidation damage [10]. Persistent high blood glucose levels disturb systemic iron balance, resulting in increased circulating non‐transferrin‐bound iron, which in turn promotes the onset of ferroptosis [11]. Particularly notable in this pathological environment is ferroptosis, a type of regulated cell death relying on iron and propelled by lipid peroxidation, drawing substantial scientific interest [12]. A complex interplay exists within this process, encompassing dysregulation of iron metabolism, mitochondrial dysfunction, and membrane lipid peroxidation. While ferrostatin‐1 (Fer‐1), a small molecule, efficiently suppresses ferroptosis via lipid peroxide scavenging, the precise role of ferroptosis in endothelial cells (e.g., human umbilical vein endothelial cells, HUVECs) within the diabetic microenvironment remains to be defined [13]. Thus, intervention in ferroptosis within HUVECs under diabetic conditions offers a highly promising strategy to enhance the management of DW as well as to combat other complications associated with diabetes.

Sesame seeds are widely consumed in Middle Eastern and Asian cuisines. Sesamin (Sea) is the most abundant lignan in sesame and serves as its primary active component [14]. Numerous in vitro and in vivo studies have demonstrated that Sea possesses diverse pharmacological properties, including antioxidant, anti‐inflammatory, anti‐apoptotic, anti‐ferroptotic, and immunomodulatory effects [15, 16, 17]. Furthermore, as reviewed by Majdalawieh et al. [18], Sea also exerts well‐documented immunomodulatory and anti‐inflammatory activities, and has been reported to lower lipids by regulating fatty acid synthesis, oxidation, and cholesterol metabolism [17]. While recent work indicates that sesamin inhibits ferroptosis and promotes tissue survival via the Nrf2/SLC7A11/GPX4 axis in skin flaps and oxygen–glucose‐deprivation‐stressed HUVECs, whether it directly promotes diabetic wound healing and through which upstream regulator (e.g., Sirt1/Keap1) remains to be elucidated [19]. To identify the potential molecular targets of sesamin in DW healing, we first performed network pharmacology analysis. The results identified Nrf2, Keap1, and Sirt1 as hub targets, with enrichment of the ferroptosis and nicotinate/nicotinamide metabolism pathways. Subsequent transcriptomic analysis of DW tissues treated with Sea further confirmed the activation of nicotinate/nicotinamide metabolism, which provides NAD⁺ for Sirt1 activation. Taken together, these findings led us to propose the following hypothesis: Sea may promote DW healing by upregulating the Sirt1/Keap1/Nrf2 pathway to suppress ferroptosis.

This study evaluated the effects of Sea on high glucose (HG)‐exposed HUVECs and confirmed its pro‐healing activity in a diabetic mouse model of full‐thickness skin defects. Sea exerts its anti‐ferroptotic effect by alleviating oxidative stress and preserving mitochondrial function, a process accompanied by upregulation of the Sirt1/Keap1/Nrf2 axis. Beyond extending the known biological functions of Sea to diabetic wound healing, these data uncover a previously unrecognized mechanism, namely that the Sirt1/Keap1/Nrf2 pathway mediates its anti‐ferroptotic activity. Therefore, Sea holds promise as an effective therapeutic agent for diabetic wound treatment.

2. Materials and Methods

2.1. Ethics Declaration

All animal procedures were approved by the Wenzhou Medical University Animal Ethics Committee (wydw2026‐0170) and conducted in compliance with NIH guidelines, with all efforts made to minimize suffering.

2.2. Reagents and Antibodies

MedChemExpress (Shanghai, China) supplied sesamin (HY‐N0121R, 98% purity), Fer‐1 (a selective ferroptosis inhibitor), and mannitol (used as an osmotic control). The primary antibodies used in study were listed as follows: VEGF‐A (Proteintech, 19003‐1‐AP, 1:1000), CD31 (Proteintech, 11265‐1‐AP, 1:1000), Nrf2 (Proteintech, 16396‐1‐AP, 1:1000), and Lamin B (Proteintech, 12987–1‐AP, 1:2000), ACSL4 (Abcam, ab155282, 1:1000), SLC7A11 (Abcam, ab307601, 1:1000), GPX4 (Abcam, ab252833, 1:1000), Keap1 (Abcam, ab119403, 1:1000), PTGS2 (Abcam, ab179800, 1:1000), Sirt1 (Abcam, ab110304, 1:1000), and β‐actin (Abcam, ab6276, 1:3000). DAPI, used as a nuclear counterstain, was purchased from Beyotime (Shanghai, China). Additional reagents, including C11‐BODIPY (Invitrogen, Carlsbad, CA, USA), FerroOrange (Tongren, Beijing, China), JC‐1 (Beyotime, Shanghai, China), cycloheximide (CHX) (Sigma‐Aldrich, St. Louis, MO, USA), and MG132 (Selleck Chemicals, Houston, TX, USA) were obtained as indicated. siRNA transfections were performed using Lipofectamine RNAiMAX (Thermo Fisher Scientific, Waltham, MA, USA). Horseradish peroxidase (HRP)‐conjugated secondary antibodies, anti‐rabbit IgG (#7074) and anti‐mouse IgG (#7076), were from Cell Signaling Technology (Danvers, MA, USA), phosphate‐buffered saline (PBS), fetal bovine serum (FBS), 0.25% trypsin, and Dulbecco's Modified Eagle's Medium (DMEM) (Gibco, Grand Island, NY, USA) were procured from their respective manufacturers as specified. The remaining chemicals—comprising streptozotocin (STZ), penicillin, streptomycin, sodium dodecyl sulfate (SDS), paraformaldehyde (PFA), osmium tetroxide, uranyl acetate, lead citrate, bovine serum albumin (BSA), diaminobenzidine (DAB), hematoxylin, hydrogen peroxide (H₂O₂), fibronectin, pentobarbital sodium, and citrate buffer—were all sourced from Sigma‐Aldrich (St. Louis, MO, USA).

2.3. HUVECs Culture and Passaging

HUVECs were commercially procured from the American Type Culture Collection (ATCC; Rockville, MD, USA). All culture and subculture procedures were carried out under sterile, standard conditions. The cells were grown in DMEM culture medium that had been supplemented with 10% heat‐inactivated FBS, along with 100 U/mL penicillin and 100 μg/mL streptomycin. Incubation took place in a humidified chamber set at 37°C with a 5% carbon dioxide (CO₂) atmosphere. Culture medium was renewed every 2 days, and an inverted light microscope was used on a daily basis to assess cell growth status and percent confluence.

2.4. Assessment of Cell Viability

At the end of each treatment regimen, the spent culture medium was aspirated from all wells and substituted with 100 μL of a pre‐mixed working solution composed of basal medium plus CCK‐8 reagent at a final proportion of 10% (v/v). Following an initial incubation of 1–2 h, the plates were further maintained at 37°C for an additional 1 h, and the optical absorbance was subsequently measured at 450 nm on a microplate reader (Thermo Fisher Scientific, Waltham, MA, USA). The entirety of the cell viability testing was conducted with a Cell Counting Kit‐8 (CCK‐8) detection kit purchased from Elabscience (Wuhan, China), and every procedural step adhered rigorously to the manufacturer's specifications.

2.5. Fluorescent Staining for Lipid ROS and Iron

Intracellular lipid peroxidation was assessed with the C11‐BODIPY probe obtained from Invitrogen (Carlsbad, CA, USA), and detected ferrous iron (Fe²⁺) levels using FerroOrange (Dojindo, Tokyo, Japan). HUVECs were then exposed to 1 μM of each probe for 30 min at 37°C following the designated treatments. The cells were washed with PBS. For FerroOrange‐stained samples, the incubation was carried out under light‐protected conditions. Subsequently, fluorescence images were captured under a confocal microscope (Nikon).

2.6. Quantitative Real‐Time PCR (qPCR)

Total RNA was extracted from cell samples using TRIzol reagent (Invitrogen, Carlsbad, USA). Reverse transcription to generate cDNA was accomplished with the Tissue RNA Purification Kit Plus (AG, Guangzhou, China). qPCR reactions were run on a LightCycler 480 II platform (Roche, Switzerland) using SYBR Green Pro Taq HS Premix (AG, Guangzhou, China). Relative transcript levels were calculated by the 2^−ΔΔCt method. The primer sequences employed for amplification of the target genes are listed in Table S1.

2.7. Western Blot (WB)

Following an initial rinse with ice‐cold PBS, cells were subjected to lysis on ice for 2 h using RIPA buffer (Beyotime, Shanghai, China) that had been supplemented with protease and phosphatase inhibitors (MilliporeSigma, MO, USA). The resulting lysates were centrifuged at 12 000 × g for 10 min at 4°C, after which the supernatants were carefully collected while avoiding disruption of the pellets. Protein samples (15–20 μg per lane) were separated by electrophoresis on 12.5% SDS‐polyacrylamide gels and then transferred electrophoretically onto PVDF membranes. Following transfer, the membranes were placed on an orbital shaker at room temperature and blocked for 2 h in a solution of 5% non‐fat milk. Subsequently, the membranes were rinsed three times with PBS (5 min per wash) and then left to incubate overnight at 4°C with the appropriate primary antibodies. The next day, after three additional PBS washes, the membranes were incubated with HRP‐conjugated secondary antibodies for 1 h at room temperature. Chemiluminescent signals were generated using an ECL substrate, and images were acquired using a Bio‐Rad ChemiDoc XRS+ system. Densitometric quantification of band intensities was performed with Image Lab 6.0 software (Bio‐Rad).

2.8. Transmission Electron Microscopy (TEM) of Cellular Ultrastructure

Transmission electron microscopy (TEM) was employed to visualize mitochondrial ultrastructure in HUVECs. Cells were fixed in 2.5% glutaraldehyde at 4°C overnight, post‐fixed in 1% osmium tetroxide for 30 min, dehydrated through a graded ethanol series, and embedded in epoxy resin. Ultrathin sections were cut, stained with uranyl acetate and lead citrate, and examined under a field‐emission transmission electron microscope (Hitachi, Tokyo, Japan). For quantitative analysis, mitochondria were randomly selected from at least 10 cells per group per experiment by an investigator blinded to group allocation, and images were captured at 12000× magnification. Mitochondrial size (μm²) and cristae number per mitochondrial area were measured using ImageJ (NIH, Bethesda, MD, USA). Only mitochondria with clearly defined outer membranes were included. At least 10 mitochondria were analyzed per group in each of three independent experiments.

2.9. Evaluation of Mitochondrial Function

Mitochondrial membrane potential (MMP) was assessed in vitro using the JC‐1 probe obtained from Thermo Fisher Scientific (Waltham, MA, USA). Briefly, following the designated treatments, cells were incubated with 5 μM JC‐1 for 30 min at 37°C under dark conditions, then rinsed with PBS for 3 min, and immediately examined under a Nikon microscope (Tokyo, Japan). High MMP, corresponding to polarized mitochondria, was indicated by red fluorescence, whereas green fluorescence reflected low MMP, a characteristic of depolarized mitochondria.

2.10. Assessment of Angiogenic Capacity In Vitro

To assess the angiogenic potential of HUVECs in vitro, a tube formation assay based on matrix gel solidification was performed. In brief, chilled extracellular matrix (ECM) gel was spread onto culture slides and polymerized at 37°C for 1 h. Treated HUVECs were trypsinized and then seeded onto the solidified gel at a density of 1 × 10⁴ cells per well, taking care to avoid bubble formation. Following a 6‐h incubation at 37°C, the formation of capillary‐like tube networks was observed. Representative phase‐contrast images were taken, and the number of branching points per field was quantified using ImageJ software as an indicator of angiogenic activity.

2.11. Evaluation of HUVECs Attachment Ability

To evaluate HUVECs attachment ability, 6‐well culture plates were first coated with fibronectin at a concentration of 5 μg/mL. Following exposure to Sea, HG, or Fer‐1, cells were seeded onto the coated plates and given 30 min to adhere. Non‐adherent cells were then eliminated by gentle washing with PBS, while those that had attached were fixed in 4% paraformaldehyde (PFA) and subsequently stained with DAPI. Adherent cells were counted in three randomly selected microscopic fields per well using a fluorescence microscope (Nikon, Tokyo, Japan).

2.12. Assessment of Cell Migratory Capacity Using the Scratch Method

A scratch assay was performed to assess the migration ability of HUVECs. Following treatment, cells were allowed to grow until they reached 80%–90% confluence. A linear wound was then made across the cell monolayer in each well of 6‐well plates using a 50 μL sterile pipette tip. Subsequently, the cells were maintained in serum‐free DMEM. Photographs of the wounded areas were taken at 0, 6, and 24 h after scratching for all experimental groups. Migration was assessed by measuring scratch width at three points per wound using ImageJ, with results shown as width reduction over time.

2.13. Pharmacological Network Analysis

To identify the molecular targets through which sesamin (Sea) exerts its pro‐healing effects on diabetic wounds, we queried three independent databases for drug target prediction: the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP; similarity threshold ≥ 0.85), the Similarity Ensemble Approach (SEA; maxTC > 0.5, E‐value < 0.01), and SuperPRED (probability score > 0.8). Disease‐related targets for diabetic wounds were sourced by querying the Online Mendelian Inheritance in Man (OMIM) resource, with retrieval restricted exclusively to entries documenting experimentally validated gene‐disease associations, as well as from GeneCards using a relevance score cutoff of ≥ 10. A Venn diagram generated via the Venny 2.1 online tool was used to identify overlapping targets shared by Sea and DW, which were then considered candidate therapeutic targets. A protein–protein interaction (PPI) network was built using the STRING database (minimum interaction score = 0.400, medium confidence for Homo sapiens ), and the resulting network was visualized with Cytoscape software. Gene Ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed using the DAVID database, applying significance thresholds of FDR‐corrected p < 0.05 and a minimum of two genes per term. Manually curated removal of redundant or excessively broad terms was carried out, and the final data were visualized using the ggplot2 package within the R environment.

2.14. Docking Analysis

Molecular docking was performed to predict the interaction between sesamin and Sirt1. The crystal structure of the human Sirt1 catalytic domain ( Homo sapiens ; PDB ID 4I5I, solved in complex with NAD and an EX527 analog) was retrieved from the RCSB Protein Data Bank and prepared with AutoDock Tools (ADT, version 1.5.6). Water molecules and the co‐crystallized ligand were deleted; polar hydrogen atoms were added and Gasteiger charges were assigned; the receptor was then saved in PDBQT format. The three‐dimensional structure of sesamin (C₂₀H₁₈O₆, PubChem CID 10618) was drawn in ChemBioDraw, converted to a 3D conformation, and energy‐minimized in ChemBio3D using the MM2 force field; the ligand was likewise converted to PDBQT with added polar hydrogens and Gasteiger charges. Docking was carried out with AutoDock Vina (version 1.1.2). The docking pocket was defined as the Sirt1 catalytic deacetylase pocket and was centered on the centroid of the co‐crystallized inhibitor in 4I5I, at coordinates (x, y, z) = (26.62, −15.36, 18.54) Å, enclosed within a grid box of dimensions (size_x, size_y, size_z) = [25, 25, 25] Å (spacing 0.375 Å). Docking was executed with an exhaustiveness of 8 and 9 output binding modes. Resulting poses were ranked by the Vina binding free‐energy score (kcal/mol), and the conformation with the most favorable (lowest) score was selected as the predicted binding pose; the final complex was visually inspected and rendered in PyMOL, and the mean binding energy was calculated.

2.15. Cellular Thermal Shift Assay (CETSA)

CETSA was performed to evaluate the binding of Sesamin to Sirt1 in intact cells. HUVECs were treated with Sea (20 μM) or PBS for 4 h, harvested, and resuspended in cold PBS with protease inhibitor cocktail. Cell suspensions were heated at indicated temperatures (37°C–67°C) for 3 min, cooled for 3 min, and lysed by three freeze–thaw cycles in liquid nitrogen. Soluble fractions were collected by centrifugation at 12000 × g for 15 min at 4°C and analyzed by Western blot using anti‐Sirt1 antibody. Band intensities were quantified, normalized to 37°C, and plotted against temperature to generate thermal denaturation curves.

2.16. Co‐Immunoprecipitation (Co‐IP) for Protein–Protein Interaction Analysis

Co‐IP assays were carried out to investigate PPI. Following the indicated treatments, HUVECs were lysed using a mild lysis buffer, and cellular debris was removed by centrifugation at 12000 × g for 15 min at 4°C. Protein concentrations in the cleared lysates were determined using a BCA protein assay kit. For each immunoprecipitation reaction, equal quantities of total protein were incubated overnight at 4°C with either 1–2 μg of specific primary antibody or normal rabbit IgG (serving as a negative control). Subsequently, 20–30 μL of protein A/G magnetic beads or agarose beads were introduced and the incubation continued for an additional 2–4 h at 4°C. After washing the beads four to five times with ice‐cold lysis buffer to eliminate non‐specific binding, bound proteins were eluted by heating the beads in 2× SDS loading buffer at 100°C for 10 min. The resulting eluates, together with whole‐cell lysates (input controls), were finally analyzed by Western blotting using antibodies directed against Nrf2 and Keap1.

2.17. Transfection of Small Interfering RNA (siRNA)

To specifically silence Sirt1 gene expression, HUVECs grown to 30%–50% confluence were transfected with siRNA targeting Sirt1 (Invitrogen, USA). The transfection process caused minimal toxicity to the cells; cell viability remained above 95% at 12 h after transfection, suggesting that the transfection conditions were well tolerated. Following medium exchange, the cells were further cultured for 72 h to achieve adequate depletion of Sirt1 protein. WB analysis was then performed to verify the knockdown efficiency, confirming that Sirt1 expression was markedly reduced in the cell model used for subsequent experiments.

2.18. Recombinant Adeno‐Associated Virus Serotype 9 (AAV9) Production and In Vivo Injection

For in vivo knockdown of Sirt1, an AAV9 vector encoding a U6 promoter‐driven short hairpin RNA (shRNA) targeting mouse Sirt1 (target sequence 5′‐GCCATGTTTGATATTGAGTAT‐3′) was procured from NovoPro (Shanghai, China). A scrambled shRNA of equal length was used as the negative control. The viral stock was provided at a titer of 1.0 × 10¹² vector genomes per milliliter (vg/mL). One week before wound creation, the AAV9‐shSirt1 construct (or the scrambled control) was delivered intradermally into a shaved dorsal area measuring 2.5 cm × 2.5 cm using a 33‐gauge needle arranged in a five‐point grid pattern. Each animal received a total dose of 5 × 10¹⁰ vg, divided equally across five injection sites (50 μL per site, with each site containing 1.0 × 10¹⁰ vg), at an injection depth of 1–2 mm. Successful intradermal administration was confirmed by the appearance of a transient wheal at each injection site, after which gentle pressure was applied with a sterile cotton swab to prevent leakage of the injectate.

2.19. Transcriptome Profiling by High‐Throughput Sequencing

RNA was extracted with TRIzol and assessed by NanoDrop/gel electrophoresis. mRNA was enriched, fragmented, and reverse‐transcribed. Libraries were sequenced on NovaSeq 6000 (150 bp PE). Reads were aligned with HISAT2, and FPKM was calculated. DEGs (|log₂FC| > 1, adj.p < 0.05) were identified by DESeq2, followed by GO/KEGG analysis.

2.20. Diabetes Induction Using Streptozotocin (STZ)

Experimental protocols involving animals were approved by the Animal Ethics Committee of Wenzhou Medical University (wydw2026‐0170). Eight‐week‐old male C57BL/6J mice (20–25 g body weight) were used throughout this study. All animals were housed in a pathogen‐free facility for 1 week of acclimatization before any intervention. To induce diabetes, mice received daily intraperitoneal injections of streptozotocin (STZ, 100 mg/kg) for five consecutive days, following a previously published protocol. Blood glucose was monitored both at baseline (prior to STZ administration) and after the injection course. A threshold of 16.7 mmol/L was used to define hyperglycemia. Only mice that consistently maintained blood glucose levels above this cutoff were enrolled in further experiments. This STZ‐based induction protocol produces a predominantly insulin‐deficient, type 1‐like diabetic phenotype, and the model is therefore referred to throughout as a streptozotocin‐induced diabetic mouse model rather than a type 2 diabetic model.

2.21. Mouse Wound Model and Compound Administration

To evaluate the therapeutic effects of Sea and Fer‐1 on diabetic wound healing, five experimental groups of mice (n = 5 per group) were established: a non‐diabetic control group, a diabetic control group (DM), and three diabetic treatment groups receiving either low‐dose Sea (10 mg/kg/day), high‐dose Sea (20 mg/kg/day), or Fer‐1 (5 mg/kg/day). To investigate whether Sirt1 is required for the pro‐healing action of Sea, an additional set of four groups (n = 5 each) was included: Control, DM, DM treated with high‐dose Sea alone (DM + Sea), and DM treated with high‐dose Sea combined with AAV‐shSirt1 to achieve Sirt1 knockdown (DM + Sea + AAV‐shSirt1). All animals were anesthetized by intraperitoneal injection of pentobarbital sodium (100 mg/kg). Upon verification that the corneal reflex had been completely abolished, a pair of full‐thickness circular skin defects measuring 8 mm across were surgically induced on both sides of the dorsal midline. Subsequently, Sea or Fer‐1 was delivered by means of percutaneous injection into the periwound region, with the first dose given 24 h after surgery followed by once‐daily treatments thereafter. Wound closure was monitored by photography on days 0, 4, 7, 10, and 14 post‐wounding, and wound areas were quantified using ImageJ software (NIH ImageJ, Bethesda, MD, USA) for subsequent statistical analysis. For each mouse, two symmetrical 8‐mm full‐thickness wounds were created on the dorsum; both wounds received identical treatment within a group. Wound area measurements from the two wounds of the same mouse were averaged to generate one value per animal, and the mouse (n = 5 per group) was treated as the independent experimental unit for all statistical analyses. The five averaged values per group were used for statistical comparison; the two wounds of an individual mouse were not treated as independent biological replicates.

2.22. Morphological and Collagen Evaluation Using H&E and Masson's Staining

Tissue samples harvested from wound sites underwent primary fixation, followed by paraffin embedding, after which sections of 4 μm were prepared. To assess the general histological architecture of the wounded skin, routine Hematoxylin and eosin (H&E) staining was carried out. Collagen distribution within the wound bed was evaluated using Masson's trichrome staining kit (Beyotime, Shanghai, China). All stained sections were examined and photographed under a light microscope (Olympus, Tokyo, Japan).

2.23. Protein Expression Detection by Immunohistochemistry (IHC)

For antigen retrieval, tissue sections were immersed in citrate buffer and heated at 95°C for 20 min. Endogenous peroxidase activity was then blocked by incubation with 3% hydrogen peroxide (H₂O₂) for 10 min. Following a 30‐min blocking step with 5% bovine serum albumin (BSA), the sections were exposed to primary antibodies overnight at 4°C. Thereafter, HRP‐conjugated secondary antibodies were applied for 1 h at room temperature. Diaminobenzidine (DAB) was used as the chromogenic substrate for signal development over a 5‐min period, and nuclei were subsequently counterstained with hematoxylin. After dehydration and mounting, the stained sections were examined and imaged using a Nikon bright‐field microscope.

2.24. Immunofluorescence (IF)

IF staining of wound tissue sections was performed using the same pretreatment steps described for IHC. After permeabilization and blocking, the sections were incubated overnight at 4°C with primary antibodies directed against the target proteins. Following thorough washing with PBS, Alexa Fluor‐conjugated secondary antibodies were applied for 1 h at room temperature under light‐protected conditions. Nuclear counterstaining was carried out with DAPI, and fluorescence images were captured using a confocal microscope. To confirm staining specificity, negative controls omitting the primary antibodies were included in each experiment.

2.25. Data and Statistical Analysis

Normality of the data was assessed using the Shapiro–Wilk test, and homogeneity of variances was evaluated by Levene's test. The results showed that all experimental groups met the assumptions of normal distribution and equal variance (p > 0.05). Accordingly, parametric statistical methods were applied for subsequent analyses. Comparisons among multiple groups were performed using one‐way analysis of variance (ANOVA) followed by Tukey's post hoc test for multiple comparisons. All data were obtained from at least three independent experiments (defined for in vitro assays as three biologically independent cell‐culture runs performed on separate days, each analyzed in triplicate) and are presented as the mean ± standard deviation (SD). A significance threshold of p < 0.05 was adopted to indicate statistically significant differences.

3. Results

3.1. Sesamin Accelerates Diabetic Wound Healing and Exhibits No In Vivo Toxicity

To evaluate whether Sea could promote diabetic wound repair, we generated a full‐thickness skin defect model using STZ‐induced diabetic mice. The molecular structure of Sea is shown in Figure 1A, while the experimental schedule and group design are illustrated in Figure 1B. Diabetes was elicited by means of intraperitoneal STZ injections (100 mg/kg) administered once daily for five consecutive days. Animals exhibiting blood glucose concentrations exceeding 16.7 mmol/L were classified as diabetic and included in subsequent experiments. One week after wound creation, Sea was administered daily via subcutaneous injection around the wound site at either a low dose (10 mg/kg) or a high dose (20 mg/kg). The ferroptosis inhibitor Fer‐1 (5 mg/kg) was injected subcutaneously as a positive control (Figure 1B). Wound healing was assessed at 0, 4, 7, 10, and 14 days after injury (Figure 1C), and representative wound images from each group are displayed in Figure 1D. Quantitative analysis of wound areas showed that Sea treatment significantly accelerated wound closure in a dose‐dependent manner relative to the diabetic control group (Figure 1E). Histological examination using H&E staining on Day 14 revealed enhanced re‐epithelialization and granulation tissue formation in Sea‐treated wounds. Masson's trichrome staining further indicated increased collagen deposition and more organized extracellular matrix remodeling (Figure 1F,G). Importantly, H&E staining of major organs (heart, liver, spleen, lung, and kidney) from Sea‐treated mice revealed no obvious morphological abnormalities or tissue damage, suggesting a favorable safety profile with no systemic toxicity (Figure 1H). Taken together, these findings demonstrate that Sea promotes diabetic wound healing in vivo without detectable adverse effects.

FIGURE 1.

Sesamin accelerates diabetic wounds healing in vivo. (A) Molecular structure of Sesamin (Sea). (B) Schematic diagram of the animal experimental design. (C‐D) Wound images during the healing process and a schematic diagram of the wound healing process. (E) Quantification of wound closure rate. (F) Hematoxylin and eosin (H&E) staining wound tissue sections. (G) Masson staining wound tissue sections. (H) H&E staining of histopathological sections of the heart, liver, spleen, lung, and kidney in different groups. Data are presented as the mean ± SD from 5 independent experiments (n = 5). Statistical significance is denoted as *p < 0.05, **p < 0.01, ***p < 0.001; ns indicates no significant difference.

3.2. Sea Rescues HG‐Induced Dysfunction of HUVECs In Vitro

To elucidate the molecular mechanisms underlying Sea's therapeutic effects on diabetic wounds, we performed network pharmacology analysis. Using the TCMSP, SEA, and SuperPRED databases, we identified 138 potential targets of Sea. A total of 6241 targets associated with diabetic wounds were compiled from the OMIM and GeneCards databases. Overlap analysis using a Venn diagram identified 103 common targets shared between Sea and diabetic wounds, which were considered potential mediators of the wound‐healing effects of Sea (Figure 2A). PPI analysis of these overlapping targets revealed that genes such as NFE2L2, Keap1, NOX1, Sirt1, and PI3k were closely interconnected (Figure 2B). Next, KEGG pathway enrichment analysis of these overlapping targets identified several significantly enriched pathways, including the PI3K‐Akt signaling pathway, HIF‐1 signaling pathway, VEGF signaling pathway, ferroptosis, focal adhesion and so on (Figure 2C). Of particular note, both the HIF‐1 and VEGF signaling pathways showed marked enrichment, implying that Sea may possess pro‐angiogenic properties. To establish the most suitable concentration of Sea for subsequent in vitro studies, HUVECs were exposed to a range of Sea concentrations for either 24 or 48 h, after which cell viability was promptly measured. The results showed that Sea promoted HUVECs proliferation at concentrations ranging from 0 to 40 μM. The CCK‐8 assay guided the selection of 20 μM and 40 μM Sea as the low and high concentrations for further in vitro studies. The higher concentration (40 μM) represented the maximal effective dose devoid of cytotoxicity, and the lower concentration (20 μM) produced an intermediate effect, together enabling dose‐dependent comparisons. The 24‐h time point was chosen for mechanistic experiments to capture early cellular responses. Subsequent analysis focused on two critical angiogenic markers, VEGF‐A and CD31. WB demonstrated that HG (25.5 mM) significantly reduced their expression, an effect that was dose‐dependently counteracted by Sea treatment (Figure 2E,F) [20]. Furthermore, immunofluorescence staining of the wound microvascular protein α‐SMA showed that Sea promotes microvascular formation in vivo (Figure 2G,H). Subsequently, the cellular functions of HUVECs were assessed. An osmotic control (mannitol) was used to verify that HG‐induced damage resulted from glucose metabolism, not osmotic pressure. HUVECs exposed to HG conditions exhibited severe functional impairment. By contrast, cells treated with mannitol (5.5 mM glucose plus 27.8 mM mannitol) showed no signs of dysfunction, and their status was indistinguishable from that of the normal control group (5.5 mM glucose) (Figure 2I–N) [21]. These findings effectively ruled out the influence of hyperosmolarity on HUVECs dysfunction. HG exposure severely compromised vascular network formation and cell–cell connections, as shown by tube formation assays. In contrast, Sea treatment markedly restored tubular integrity and promoted endothelial network assembly (Figure 2I,J). Scratch assays revealed that wound closure rates were substantially higher in the HG + Sea groups than in the HG‐alone group (Figure 2K,L). To examine the effect of Sea on HUVECs adhesion, we performed a fibronectin adhesion assay at the end of the experiment. Statistical analysis indicated that HG significantly suppressed cell adhesion, an effect effectively reversed by Sea treatment (Figure 2M,N). Collectively, these in vitro results indicate that Sea ameliorates the functional impairments caused by HG exposure in HUVECs in a concentration‐dependent manner, primarily through promoting angiogenic activity and facilitating cell migration.

FIGURE 2.

Sesamin restores HUVECs dysfunction impaired by high glucose in vitro. (A) Venn diagram of Sesamin (Sea) active ingredients and targets related to diabetic wounds. (B) Primary clusters extracted from the protein–protein interaction (PPI) network illustrate the main functional gene clusters of potential targets. (C) Kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment analysis of core targets of Sea against diabetic wounds. (D) Cell viability percentages after exposure to different concentrations of Sea for 24/48 h. (E, F) Western blot analysis of VEGF‐A and CD31 expression in HUVECs. (G, H) Immunofluorescence staining and quantification analysis of α‐SMA in wound tissues. (I, J) Tube formation assay in HUVECs with quantification of total tube per field. (K, L) Scratch assay in HUVECs with quantification of width of area (%). (M, N) Evaluation of the effect of Sea on HUVECs adhesion ability using the cell‐matrix adhesion assay. Data are presented as the mean ± SD from 3 independent experiments. Statistical significance is denoted as *p < 0.05, **p < 0.01, ***p < 0.001; ns indicates no significant difference. β‐actin, beta‐actin.

3.3. Transcriptomic Profiling Reveals That Sea Suppresses Ferroptosis in DW

To gain deeper insight into the transcriptional changes underlying the pro‐healing effect of Sea on DW, RNA‐sequencing (RNA‐seq) was performed on wound tissue samples collected on Day 10 post‐wounding from diabetic mice treated with or without Sea (Figure 3A). Applying thresholds of |log₂FC| > 1 and adjusted p < 0.05, volcano plot analysis uncovered 486 differentially expressed genes (DEGs) when comparing the diabetic group with the Sea‐treated group. Among these, 217 genes showed increased expression, whereas 267 exhibited decreased expression (Figure 3B). KEGG pathway analysis further demonstrated that Sea intervention affected several signaling cascades associated with tissue repair. Within the diabetic wounds receiving Sea treatment (DW + Sea group), a diverse array of reparative signaling cascades—exemplified by the Wnt/β‐catenin and FOXO regulatory networks—displayed pronounced upregulation. Conversely, biological processes intimately linked to inflammatory responses and programmed cell death were substantially enriched (Figure 3C). In contrast, pathways closely associated with inflammation and cell death, such as the IL‐17 signaling cascade, ferroptosis, and TNF‐mediated signaling, were markedly suppressed (Figure 3D). GO enrichment analysis of the DEGs indicated substantial enrichment of biological processes closely linked to wound healing and ferroptosis, including cell adhesion, glutathione metabolism, lipid metabolism, inflammatory responses, and cellular responses to oxidative stress (Figure 3E). Gene set enrichment analysis (GSEA) further confirmed significant enrichment of the ferroptosis pathway (Figure 3F).

FIGURE 3.

Transcriptomic analysis reveals Sesamin's inhibition of ferroptosis in diabetic wounds. (A) Schematic of RNA sequencing design and sample preparation. (B) Volcano plot depicting up/down‐regulated genes. (C) Pathway enrichment analysis of up‐regulated genes. (D) Pathway enrichment analysis of down‐regulated genes. (E) Pathway enrichment of highly expressed genes. (F) Gene Set Enrichment Analysis (GSEA) of ferroptosis‐related pathways. (G–J) Immunofluorescence staining and quantification analysis of GPX4 and ACSL4 in wound tissues. (K, L) Western blot analysis of ACSL4, PTGS2, SLC7A11, and GPX4 in wound tissues. Data are presented as the mean ± SD from 3 independent experiments. Statistical significance is denoted as *p < 0.05, **p < 0.01, ***p < 0.001; ns indicates no significant difference. β‐actin, beta‐actin.

To validate the transcriptomic findings, the expression of ferroptosis‐related proteins was systematically examined. IF staining showed that, compared with controls, DW tissues expressed higher levels of the pro‐ferroptotic marker ACSL4 and lower levels of the anti‐ferroptotic marker GPX4. These pathological changes were effectively reversed by Sea treatment (Figure 3H–J). WB analysis further corroborated these results, demonstrating that Sea administration dose‐dependently restored the reduced expression of GPX4 and SLC7A11 and attenuated the elevated levels of ACSL4 and PTGS2 in diabetic wound tissues (Figure 3K,L).

To elucidate the biological relevance of the anti‐ferroptotic effect of Sea, the commercially available ferroptosis‐specific inhibitor Fer‐1 was used as a positive control. Continuous postoperative monitoring (on Days 0, 4, 7, 10, and 14) revealed that Fer‐1 treatment significantly accelerated wound closure, as evidenced by gross images (Figure 1C,D). Histological staining further demonstrated that both Sea and Fer‐1 interventions promoted wound contraction and enhanced collagen deposition (Figure 1F,G). Moreover, neither treatment induced obvious histopathological alterations in major organs (heart, liver, spleen, lung, and kidney) of mice, indicating favorable in vivo safety profiles (Figure 1H). Given that the functional status of HUVECs is critical for wound repair, the protective effects of the two agents on endothelial cell function were compared. HUVECs were treated with Fer‐1 (12 μM) and subjected to scratch, tube formation, and adhesion assays. The results showed that Fer‐1 successfully rescued HG‐elicited dysfunction in HUVECs, achieving a therapeutic potency similar to that observed with Sea (Figure 2I–M). Taken together, both transcriptional profiling and phenotypic findings indicate that the wound‐promoting action of Sea in diabetic contexts stems predominantly from its capacity to modulate ferroptotic processes.

3.4. Sea Dose‐Dependently Suppresses HG‐Induced Ferroptosis in HUVECs

To test whether Sea directly counteracts ferroptosis in HUVECs under diabetic conditions, cells were cultured in high glucose (HG, 25.5 mM) either in the absence or presence of Sea. The specific ferroptosis blocker Fer‐1 served as a positive control. Because ferroptosis is defined by lipid peroxidation together with iron accumulation, we first evaluated intracellular lipid ROS levels using the C11‐BODIPY probe. HG stimulation strongly elevated lipid peroxidation as visualized by C11‐BODIPY staining; this increase was markedly reduced by Sea in a concentration‐dependent fashion, an effect comparable to that seen with Fer‐1 (Figure 4A,B). In parallel, FerroOrange staining indicated that HG‐triggered ferrous iron (Fe²⁺) buildup was considerably diminished when cells were co‐treated with Sea (Figure 4C,D). Mitochondrial dysfunction is another recognized feature of ferroptosis. We therefore measured MMP using the JC‐1 probe. HG caused a substantial drop in MMP, reflected by a reduced red/green fluorescence ratio, whereas Sea treatment dose‐dependently preserved MMP (Figure 4E,F). Immunofluorescence staining further confirmed that Sea restored GPX4 protein levels in HG‐exposed HUVECs (Figure 4G,H). Western blot analysis revealed that HG markedly suppressed the expression of the anti‐ferroptotic proteins GPX4 and SLC7A11, while elevating the pro‐ferroptotic markers ACSL4 and PTGS2. These changes were reversed by Sea, with the higher concentration of Sea yielding effects similar to those of Fer‐1 (Figure 4I,J). Transmission electron microscopy (TEM) showed that HG induced typical ferroptotic mitochondrial alterations, characterized by shrunken mitochondria, increased membrane density, and reduced or absent cristae. Sea treatment, in contrast, effectively maintained normal mitochondrial ultrastructure, an effect comparable to that observed with Fer‐1 (Figure 4K). Taken together, these results demonstrate that Sea dose‐dependently inhibits HG‐induced ferroptosis in HUVECs by lowering lipid peroxidation and iron accumulation, preserving mitochondrial function, and modulating key ferroptosis‐associated proteins.

FIGURE 4.

Sesamin inhibits high glucose‐induced ferroptosis in HUVECs. (A, B) C11‐BODIPY staining and quantification analysis in HUVECs (red: Reductive C11‐BODIPY; green: Oxidative C11‐BODIPY). (C, D) FerroOrange staining and quantification analysis in HUVECs. (E, F) JC‐1 staining and quantification analysis in HUVECs (red: JC‐1 aggregate; green: JC‐1 monomer). (G, H) Immunofluorescence staining and quantification analysis of GPX4 in HUVECs. (I, J) Western blot analysis of ACSL4, PTGS2, SLC7A11, and GPX4 in HUVECs. (K) Transmission electron microscopy (TEM) images of HUVECs. Data are presented as the mean ± SD from three independent experiments. Statistical significance is denoted as *p < 0.05, **p < 0.01, ***p < 0.001. β‐actin, beta‐actin.

3.5. Sea Upregulates the Sirt1/Nrf2 Pathway by Disrupting Keap1‐Nrf2 Interaction in HG‐Induced HUVECs

To explore the molecular mechanism through which Sea inhibits ferroptosis, we re‐examined the KEGG pathway enrichment results from the transcriptomic analysis. Notably, the nicotinate and nicotinamide metabolism signaling pathway was found to be highly expressed and significantly altered in the DW + Sea group (Figure 3C). This metabolic pathway supplies NAD⁺, an essential substrate for the sirtuin family, thereby directly influencing sirtuin deacetylase activity and linking cellular energy status to the regulation of aging and related functions. Accordingly, we next focused on the sirtuin family. qPCR screening of Sirt1–Sirt6 revealed that, among all family members, Sea most markedly upregulated Sirt1 mRNA expression under HG conditions in a dose‐dependent manner, whereas no substantial changes were observed in the expression of other sirtuins (Figure 5A). WB analysis further confirmed that Sea treatment significantly elevated Sirt1 protein levels, with the maximal effect detected at 48 h (Figure 5B–E). Molecular docking predicted that Sea binds to Sirt1 at residues including ARG‐266, VAL‐258, and SER‐434 via hydrogen bonds, hydrophobic interactions, and π‐stacking, with a calculated binding energy of −7.9 kcal/mol (Figure 5F). Given that Sirt1 is known to regulate Nrf2, a master transcription factor controlling antioxidant and anti‐ferroptotic gene expression, we next investigated whether Sea activates Nrf2 signaling. Nuclear and cytoplasmic fractionation assays showed that Sea promoted Nrf2 nuclear translocation in a dose‐dependent manner, as evidenced by increased nuclear Nrf2 protein levels (Figure 5G,H). Concurrently, Sea treatment reduced the protein level of Keap1, the endogenous inhibitor of Nrf2 (Figure 5G,I). To gain deeper insight into the regulatory interplay between Sea and the Nrf2/Keap1 signaling axis, a cycloheximide (CHX) chase experiment was conducted. CHX is a well‐known inhibitor of de novo protein synthesis, and its application allowed us to assess the stability of Nrf2 protein in HUVECs following Sea treatment. The results demonstrated that, relative to the control group, Sea treatment markedly prolonged the half‐life of Nrf2, as reflected by a significantly slower degradation rate of the Nrf2 protein (Figure 5J). In parallel Co‐IP assays were performed to examine the physical interaction between Keap1 and Nrf2. The data revealed that Sea treatment diminished the binding affinity between Keap1 and Nrf2, leading to a reduced amount of Keap1‐associated Nrf2. Notably, this effect was highly reminiscent of that observed upon exposure to the proteasome inhibitor MG132, which is known to block ubiquitin‐proteasome‐mediated degradation (Figure 5K). To further validate the direct binding of Sea to Sirt1, a CETSA was performed in intact HUVECs. Sea treatment markedly enhanced the thermal stability of endogenous Sirt1, as shown by the rightward shift of the melting curve, with significant stabilization observed at 49°C–61°C (Figure 5L,M). IF staining further revealed that Sea promoted Nrf2 nuclear translocation under HG conditions, with quantitative analysis confirming a significant increase in nuclear Nrf2 intensity (Figure 5N,O). Consistently, qPCR analysis demonstrated that Sea dose‐dependently restored the mRNA expression of Nrf2 target genes GPX4, SLC7A11, NQO1, and HO‐1, which were markedly suppressed by HG exposure (Figure 5P). To validate the functional requirement of Nrf2 in Sea‐mediated ferroptosis inhibition, we employed ML385, a selective Nrf2 inhibitor. Notably, ML385 treatment largely abrogated the protective effects of Sea against HG‐induced ferroptosis (Supplementary Figure S1A–J). Collectively, these results establish that Sea directly binds to and stabilizes Sirt1, disrupts the Keap1‐Nrf2 interaction, promotes Nrf2 nuclear translocation, and transcriptionally activates anti‐ferroptotic genes, forming a coherent Sirt1/Nrf2 signaling cascade that underlies the ferroptosis‐inhibitory effect of Sea in HG‐stressed HUVECs.

FIGURE 5.

Sesamin attenuates high glucose‐induced ferroptosis in HUVECs via Sirt1 signaling. (A) Quantitative real‐time PCR for mRNA levels of Sirts family members (Sirt1/2/3/4/5/6) in HUVECs. (B–E) Western blot analysis of Sirt1 in HUVECs. (F) Molecular docking analysis of Sea‐Sirt1 binding interactions. (G–I) Western blot analysis of n‐Nrf2, c‐Nrf2 and Keap1 in HUVECs. (J) Cycloheximide chase assay to assess Nrf2 protein stability: HUVECs were incubated with 30 μM cycloheximide with or without Sea for the indicated times. (K) Co‐immunoprecipitation and WB to detect the interaction between Nrf2 and Keap1: HUVECs were treated with MG‐132 (5 μM) with or without Sea, followed by cell lysis. (L) Representative western blot images of Sirt1 in HUVECs treated with Sea or PBS at the indicated temperatures (37°C, 43°C, 49°C, 55°C, 61°C, and 67°C). (M) Quantification of relative Sirt1 band intensity normalized to the 37°C PBS control. (A) Representative immunofluorescence images of Nrf2 (red) and DAPI (blue) in HUVECs. (B) Quantification of nuclear Nrf2 fluorescence intensity. (C) Quantitative real‐time PCR analysis of relative mRNA expression of the Nrf2‐responsive genes GPX4, SLC7A11, NQO1, and HO‐1. Data are presented as the mean ± SD from three independent experiments. Statistical significance is denoted as *p < 0.05, **p < 0.01, ***p < 0.001; ns indicates no significant difference. β‐actin, beta‐actin.

3.6. Sirt1/Nrf2 Pathway Activation Is Required for the Anti‐Ferroptotic Effect of Sea

To determine whether Sirt1 is required for the anti‐ferroptotic action of Sea in HUVECs subjected to high glucose stress, we employed specific siRNA to suppress Sirt1 gene expression. Three independent siRNA sequences targeting Sirt1 were transfected into HUVECs. Prior to loss‐of‐function experiments, we first confirmed that scrambled negative control siRNA (si‐NC) did not affect any of the assessed ferroptosis‐related parameters, including lipid peroxidation, intracellular Fe²⁺, mitochondrial membrane potential, GPX4 expression, or mitochondrial ultrastructure in HUVECs (Figure S2A–I). WB analysis then identified si‐Sirt1 #3 as the most efficient siRNA sequence, reducing Sirt1 protein levels by approximately 84% relative to si‐NC (Figure 6A,B). This sequence was therefore selected for all subsequent experiments. To assess whether Sirt1 is required for the anti‐ferroptotic activity of Sea, cellular lipid peroxidation levels were measured using the C11‐BODIPY probe. The results showed that Sirt1 knockdown dramatically compromised the ability of Sea to counteract HG‐induced lipid peroxidation. In particular, the fluorescence signal corresponding to oxidized species detected within the HG + Sea + si‐Sirt cohort exhibited a significant increase relative to the level recorded for the HG + Sea cohort, indicating that the protective action of Sea was largely abolished when Sirt1 was absent (Figure 6C,D). Consistently, FerroOrange staining showed that Sirt1 silencing reversed the Sea‐mediated reduction of intracellular Fe²⁺ accumulation under HG conditions (Figure 6E,F). MMP assessed by JC‐1 staining demonstrated that Sirt1 knockdown reversed the protective effect of Sea on mitochondrial function, with a pronounced decrease in the red/green fluorescence ratio compared with the HG + Sea group (Figure 6G,H). IF staining of GPX4 further showed that the Sea‐induced upregulation of this key anti‐ferroptotic protein was substantially diminished following Sirt1 silencing (Figure 6I,J). Finally, TEM provided supportive morphological evidence for ferroptosis, showing that Sirt1 knockdown abrogated the protective effect of Sea on mitochondrial ultrastructure (Figure 6K). Quantitative analysis confirmed that Sirt1 knockdown reversed Sea‐induced increases in both mitochondrial size and cristae number per mitochondrial area (Figure 6L,M). Collectively, these findings indicate that Sirt1 is indispensable for the anti‐ferroptotic activity of Sea in HUVECs exposed to high glucose conditions.

FIGURE 6.

Sesamin suppresses ferroptosis through the Sirt1/Keap1/Nrf2 pathway. (A, B) Western blot analysis of Sirt1 protein expression in differentially treated HUVECs. (C, D) C11‐BODIPY staining and quantification analysis in HUVECs (red: Reductive C11‐BODIPY; green: Oxidative C11‐BODIPY). (E, F) FerroOrange staining and quantification analysis in HUVECs. (G, H) JC‐1 staining and quantification analysis in HUVECs (red: JC‐1 aggregate; green: JC‐1 monomer). (I, J) Immunofluorescence staining and quantification analysis of GPX4 in HUVECs. (K) Transmission electron microscopy (TEM) images of HUVECs. (L) Quantification of relative mitochondrial size. (M) Quantification of cristae number per mitochondrial area. Data are presented as the mean ± SD from three independent experiments. Statistical significance is denoted as *p < 0.05, **p < 0.01, ***p < 0.001.

3.7. Sirt1 Knockdown Abrogates the Therapeutic Effect of Sea on Diabetic Wounds

To further validate the essential role of Sirt1 in Sea‐mediated diabetic wound healing, we performed loss‐of‐function experiments using AAV9‐sh‐Sirt1 administered via intradermal injection 1 week prior to wound creation (Figure 7A). As shown in Figure 7B–D, diabetic wounds exhibited delayed closure compared to controls, while Sea treatment significantly accelerated wound healing. Notably, this pro‐healing effect was largely abolished upon Sirt1 knockdown, with wound closure rates in the DW + Sea + sh‐Sirt1 group resembling those of untreated diabetic wounds (Figure 7B–D). Histological analysis by H&E staining on Day 14 post‐wounding revealed that Sea‐enhanced re‐epithelialization and granulation tissue formation were markedly diminished by Sirt1 silencing (Figure 7E). Masson's trichrome staining demonstrated that Sea‐induced collagen deposition and extracellular matrix remodeling were also reversed in Sirt1‐knockdown wounds (Figure 7F). IHC staining confirmed efficient Sirt1 knockdown in the DW + Sea + sh‐SIRT1 group, with significantly reduced Sirt1‐positive signals compared to the DW + Sea group (Figure 7G,H). IF staining for α‐SMA, a marker of myofibroblast differentiation and angiogenesis, showed that Sea treatment increased α‐SMA expression in wound tissues, an effect that was attenuated upon Sirt1 knockdown (Figure 7I,J). IF staining further confirmed that Sea‐induced GPX4 upregulation and ACSL4 downregulation were diminished in Sirt1‐knockdown wounds (Figure 7K–N). WB analysis of wound tissue lysates revealed that Sea treatment upregulated the anti‐ferroptotic proteins GPX4, SLC7A11, and Sirt1 signaling while downregulating pro‐ferroptotic markers ACSL4 and PTGS2. These regulatory effects were largely reversed by Sirt1 silencing, indicating that Sirt1 is required for Sea‐mediated modulation of ferroptosis‐related proteins in vivo (Figure 7O,P). Collectively, these in vivo results demonstrate that Sirt1 is indispensable for the therapeutic effects of Sea on diabetic wound healing, and that Sirt1 knockdown abrogates Sea‐mediated ferroptosis inhibition and wound repair (Figure 8).

FIGURE 7.

Sirt1 knockdown attenuates the therapeutic effects of Sesamin in diabetic wounds. (A) Schematic diagram of the animal experimental design. (B, C) Wound images during the healing process and a schematic diagram of the wound healing process. (D) Quantification of wound closure rate. (E) Hematoxylin and eosin (H&E) staining wound tissue sections. (F) Masson staining wound tissue sections. (G, H) Immunohistochemistry staining and quantification analysis of Sirt1 in wound tissues. (I–N) Immunofluorescence staining and quantification analysis of α‐SMA, GPX4, and ACSL4 in wound tissues. (O, P) Western blot analysis of Sirt1, ACSL4, PTGS2, SLC7A11, and GPX4 in wound tissues. Data are presented as the mean ± SD from five independent experiments (n = 5). Statistical significance is denoted as *p < 0.05, **p < 0.01, ***p < 0.001. β‐actin, beta‐actin.

FIGURE 8.

Sesamin activates the Sirt1/Keap1/Nrf2 pathway to inhibit ferroptosis and promote diabetic wound healing.

4. Discussion

The present work offers experimental evidence that Sea significantly speeds up diabetic wound healing in mice. Through the integration of transcriptomic analysis with in vivo experimental verification, we show that Sea dose‐dependently inhibits ferroptosis within diabetic wound tissues. Sea treatment notably decreased lipid peroxidation and ferrous iron levels, while upregulating GPX4 expression. These observations align well with emerging studies suggesting that inhibition of ferroptosis is an effective approach to facilitate wound healing in diabetes. Our findings not only reinforce this emerging concept but also identify Sea as a previously unrecognized regulator of ferroptosis, thereby offering a potential therapeutic avenue for impaired diabetic wound healing [19].

Given the central role of HUVECs in maintaining vascular barrier integrity and regulating hemodynamic balance, they are particularly vulnerable to damage induced by the hyperglycemic environment in diabetes. Our study demonstrated that Sea treatment significantly alleviated high glucose‐induced dysfunction in HUVECs. Of note, under normal physiological conditions, HUVECs rely primarily on aerobic glycolysis as their main metabolic pathway, a metabolic feature that supports rapid adenosine triphosphate (ATP) production while limiting oxidative injury [22]. In HUVECs exposed to HG, metabolic reprogramming toward oxidative phosphorylation (OXPHOS) drives cellular dysfunction by promoting excessive ROS production, which disrupts redox homeostasis and impairs cellular function [23]. This finding underscores the need for deeper exploration into the signaling interplay between ferroptosis and metabolic reprogramming. A key question for future studies is whether pharmacological modulation of ferroptosis can effectively reverse the metabolic dysfunction in HUVECs under diabetic conditions.

The Sirt1/Keap1/Nrf2 signaling axis is broadly recognized as a central regulator of cellular stress adaptation, and aberrations in this pathway have been linked to numerous pathological conditions [24, 25]. Sirt1, functioning as an NAD⁺‐dependent deacetylase, plays an essential role in the regulation of various metabolic processes, such as the cellular defense against oxidative stress and the control of apoptotic pathways [26]. Within the sirtuin family, Sirt1 contributes critically to mitochondrial homeostasis, inflammatory responses, programmed cell death, and redox balance [26, 27]. In mammalian systems, elevated Sirt1 expression has been shown to markedly lower intracellular oxidative stress levels, stimulate cell cycle progression and proliferation, while suppressing cellular aging and apoptosis [27]. Nrf2, the most important antioxidant transcription factor in cells and a key downstream target of Sirt1, protects cells from oxidative damage by upregulating antioxidant and detoxifying enzyme expression [28]. When Sirt1 expression decreases, Nrf2 expression levels correspondingly decline, thereby negatively regulating oxidative stress [29]. Notably, Nrf2 activity is tightly regulated by the Keap1 protein [30]. Furthermore, our transcriptomic analysis also identified significant enrichment of Wnt/β‐catenin and FOXO signaling pathways following Sea treatment. Importantly, these pathways do not act in isolation but exhibit crosstalk with the SIRT1/Keap1/Nrf2 axis. Sirt1 deacetylates FOXO, enhancing its transcriptional activity and promoting antioxidant gene expression. Nrf2 and Wnt/β‐catenin signaling also mutually regulate each other, contributing to cell proliferation and tissue repair. Thus, Sea may exert its pro‐healing effects through an orchestrated network involving Sirt1/Nrf2, FOXO, and Wnt pathways. Elucidating the precise molecular interactions among these pathways represents an important direction for future research.

Sea, a major lignin primarily found in sesame oil ( Sesamum indicum ) and sesame seeds, has demonstrated pro‐adipogenic, anti‐obesity, and anti‐diabetic effects [31, 32]. Drawing from the observations described above, we put forward a mechanistic framework in which Sea directly engages Sirt1, upregulates Sirt1 expression, and promotes the dissociation of Nrf2 from Keap1. This Sirt1‐mediated mechanism alleviates the Keap1‐dependent suppression of Nrf2, leading to enhanced mitochondrial function and ultimately counteracting high glucose‐induced ferroptosis. The core steps of this model are as follows: Sea upregulates Sirt1 expression, and the upregulated Sirt1 promotes the dissociation of Nrf2 from Keap1. The liberated Nrf2 then translocates to the nucleus and drives the transcription of anti‐ferroptotic genes such as GPX4 and SLC7A11. Loss‐of‐function experiments further validated the essential role of Sirt1. Knockdown of Sirt1 in HUVECs led to increased Keap1 levels (suggesting impaired Nrf2‐Keap1 dissociation), reduced expression of GPX4 and SLC7A11, and a marked abrogation of the protective effect of Sea. In parallel in vivo experiments, Sea treatment enhanced GPX4 expression, lowered ROS levels, and accelerated diabetic wound healing. Notably, these beneficial effects were largely abolished by AAV‐shSirt1‐mediated Sirt1 knockdown. Collectively, these results support that Sirt1 is a key mediator of the anti‐ferroptotic and pro‐healing actions of Sea.

Several limitations should be noted. Our iron assessment was limited to intracellular labile Fe²⁺ by FerroOrange staining rather than total iron quantification (e.g., ICP‐MS), and the in vivo Sirt1 knockdown experiment lacked an empty AAV vector control and a Fer‐1 monotherapy group in non‐diabetic mice, although Sirt1 knockdown specificity was verified and no overt toxicity was observed. In addition, we did not directly measure Sirt1 deacetylase activity or Nrf2 acetylation status; while our Co‐IP, cycloheximide‐chase, and nuclear/cytoplasmic fractionation results support a Sirt1/Keap1/Nrf2 mechanism, they do not in themselves provide direct enzymatic evidence that Sea enhances Sirt1 deacetylase function. Finally, although a negative control siRNA (si‐NC) confirmed the specificity of the transfection procedure under basal conditions, an HG + Sea + si‐NC group was not directly included in the functional assays. Additionally, because non‐splinted dorsal wounds in mice heal predominantly through wound contraction while human diabetic wounds rely more strongly on re‐epithelialization and granulation tissue formation, the murine wound model may not fully recapitulate the human healing process. These limitations are planned to be addressed in subsequent investigations.

In summary, the Sirt1/Keap1/Nrf2 signaling axis occupies a central position within the cellular antioxidant defense network, implying that pharmacological modulation of this pathway may serve as an effective means to control the magnitude of oxidative stress. Taken together, our findings demonstrate that Sea accelerates diabetic wound healing through a mechanism involving upregulation of Sirt1 expression, disruption of the Keap1‐Nrf2 interaction, enhanced Nrf2 nuclear translocation, and subsequent activation of anti‐ferroptotic gene expression (Figure 8).

5. Conclusion

In conclusion, this study provides the first evidence that Sea effectively inhibits ferroptosis in both HG‐exposed HUVECs in vitro and diabetic wound tissues in vivo through activation of the Sirt1/Keap1/Nrf2 signaling axis. Mechanistically, Sea upregulates Sirt1 expression, promotes the dissociation of Nrf2 from its negative regulator Keap1, and facilitates Nrf2 nuclear translocation, which in turn drives the transcription of anti‐ferroptotic target genes including GPX4 and SLC7A11. These mechanistic insights not only expand our understanding of the pharmacological actions of Sea but also provide a solid theoretical basis for its application in diabetic wound therapy. Collectively, our findings suggest that Sea may serve as a safe and effective natural product‐derived candidate for accelerating diabetic wound healing and ameliorating associated complications.

Author Contributions

Xin Wen: contributed to the original draft and methodology. Sheng‐tuo Zhou: methodology, investigation. Gao‐sheng Zhu: validation, supervision. Tao Xu and Zi‐huai Huang: resources, project administration. Yue‐ting Wang and Jian‐zuo Lu: project administration, funding acquisition.

Funding

This work was funded by the Wenzhou Science and Technology Bureau Project (Y20220925).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: ML385 reverses the anti‐ferroptotic effect of Sesamin in high glucose‐exposed HUVECs. (A, B) C11‐BODIPY staining and quantification analysis in HUVECs (red: Reductive C11‐BODIPY; green: Oxidative C11‐BODIPY). (C, D) FerroOrange staining and quantification analysis of intracellular labile Fe_2⁺ in HUVECs. (E, F) JC‐1 staining and quantification analysis in HUVECs (red: JC‐1 aggregate; green: JC‐1 monomer). (G, H) Western blot analysis of GPX4 and SLC7A11 in HUVECs. (I, J) Western blot analysis of ACSL4 and PTGS2 in HUVECs. HUVECs were exposed to high glucose (HG) and treated with Sesamin (Sea) in the presence or absence of the Nrf2 inhibitor ML385 (10 μM, 2 h pre‐treatment). Data are presented as the mean ± SD from three independent experiments. Statistical significance is denoted as *p < 0.05, **p < 0.01, ***p < 0.001; ns indicates no significant difference. β‐actin, beta‐actin.

Figure S2: Negative control siRNA does not affect ferroptosis‐related parameters in HUVECs. (A, B) C11‐BODIPY staining and quantification analysis in HUVECs (red: Reductive C11‐BODIPY; green: Oxidative C11‐BODIPY). (C, D) FerroOrange staining and quantification analysis of intracellular labile Fe²⁺ in HUVECs. (E, F) JC‐1 staining and quantification analysis in HUVECs (red: JC‐1 aggregate; green: JC‐1 monomer). (G, H) Immunofluorescence staining and quantification analysis of GPX4 in HUVECs. (I) Transmission electron microscopy (TEM) images of HUVECs. HUVECs were transfected with negative control siRNA (si‐NC) or left untransfected (Control). Data are presented as the mean ± SD from three independent experiments. Statistical significance is denoted as *p < 0.05, **p < 0.01, ***p < 0.001; ns indicates no significant difference.

Table S1: Primer pairs for qRT‐PCR of mouse target genes and internal control GAPDH.

Data Availability Statement

All data supporting the findings of this study are available within the article. The underlying raw data are available from the corresponding author upon reasonable request.