An antioxidant and injectable hydrogel dressing for repairing MRSA biofilm–infected diabetic wounds

Abstract

Chronic non-healing wounds represent a severe complication of diabetes mellitus, which frequently progress to infection, limb amputation, and even mortality. The dysregulated wound microenvironment, marked by persistent inflammation and oxidative stress, severely impedes tissue repair, and the presence of MRSA biofilm infection further worsens these impairments and poses major clinical challenges. To address these challenges, we constructed a multifunctional injectable hydrogel (SOT) that integrates antibiofilm, antioxidant, and immunomodulatory properties. This hydrogel is formed through dynamic covalent crosslinking between thiolated hyaluronic acid (HA-SH) and dopamine-modified oxidized dextran (ODex-DA), which enables favorable injectability, self-healing, and in situ gelation. Tannic acid–silver nanoparticles (TA-Ag NPs) incorporated into the system impart antibiofilm and reactive oxygen species (ROS)-scavenging properties. In a diabetic MRSA biofilm infection model, the SOT hydrogel eradicated biofilms, reduced excessive ROS, and promoted wound closure. These findings suggest that this immuno-instructive hydrogel platform may offer a promising therapeutic approach for the treatment of MRSA biofilm-infected chronic diabetic wounds.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04093-y.

Introduction

Diabetes mellitus is a chronic metabolic disorder with a rapidly rising global prevalence, posing a significant threat to human health [1–3]. It is estimated that by 2045, the number of individuals affected by diabetes will reach approximately 780 million worldwide [4]. Among its numerous complications, chronic non-healing wounds are particularly prevalent and severe, characterized by delayed tissue repair, high susceptibility to microbial infection, and frequent recurrence, often leading to limb amputation or even death [5, 6]. These pathological wounds arise from a highly complex microenvironment marked by persistent hyperglycemia, chronic inflammation, impaired angiogenesis, and immune dysregulation, which synergistically impair the typical wound healing cascade [7, 8]. Despite notable advances in chronic wound management over the past decade, diabetic wounds, particularly those complicated by bacterial biofilm formation, continue to present formidable clinical challenges.

Biofilms play a central role in driving persistent and refractory infections [9–11]. Approximately 80% of chronic and hospital-acquired infections are associated with biofilm formation [12–14], and 13 million people worldwide died from bacterial infections in 2019 [15–17]. In diabetic ulcers, the most prevalent biofilm-forming pathogens include methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa. Once colonized on the wound surface, these bacteria secrete extracellular polymeric substances (EPS) that assemble into a dense matrix, facilitating nutrient exchange while creating robust physical and chemical barriers [18, 19]. This matrix substantially restricts the penetration of host immune components—such as macrophages and antimicrobial peptides—as well as antibiotics [9, 20]. Consequently, most antimicrobial agents fail to reach bactericidal concentrations beyond the EPS barrier, leading to persistent, recurrent infections and further delaying wound healing [21, 22]. Among these pathogens, MRSA is particularly representative in chronic diabetic wounds and is a major cause of nosocomial infections. MRSA exhibits high-level resistance to all β-lactam antibiotics and multiple other antimicrobial classes, making the infection exceedingly difficult to control and greatly increasing treatment costs [23]. Beyond its intrinsic drug resistance, mature MRSA biofilms further intensify pathogenicity by sustaining inflammatory responses, triggering excessive release of proinflammatory cytokines, activating immune complexes, and inducing local tissue damage—collectively exacerbating the impairment of diabetic wound healing [24]. Therefore, early inhibition of biofilm formation and thorough eradication of both deep-seated and dormant bacteria are critical to preventing subsequent excessive inflammation and promoting efficient wound resolution.

In addition to infection, chronic inflammation constitutes another central pathological feature of diabetic wounds and plays a pivotal role in impaired healing [25]. Under physiological conditions, inflammation is a transient and well-regulated phase of wound repair that transitions into proliferative and remodeling stages. However, in diabetic individuals, persistent hyperglycemia, accumulation of advanced glycation end-products, and immune dysfunction prolong the inflammatory phase and prevent its natural resolution [26]. This results in macrophages being trapped in a proinflammatory M1 phenotype, producing elevated levels of cytokines such as TNF-α, IL-1β, and IL-6. These mediators drive neutrophil infiltration and promote the accumulation of reactive oxygen species (ROS), which in turn cause oxidative stress, cellular damage, inhibition of fibroblast proliferation and collagen deposition, and impaired angiogenesis and epithelialization [27, 28]. Therefore, there is an urgent need for wound dressings capable of simultaneously addressing biofilm-associated infection, oxidative stress, and immune dysregulation to effectively treat chronic diabetic wounds.

As a physical barrier between the wound and the external environment, dressings play a crucial role in preventing further tissue damage and secondary infections [29]. Among various types, hydrogels have emerged as promising candidates owing to their structural similarity to the extracellular matrix, high water content, excellent biocompatibility, and superior moisture-retention capacity [30–34]. Among various natural polymers, dextran (Dex) is an electrically neutral polysaccharide with good biocompatibility and biodegradability and has been demonstrated to promote hemostasis and wound healing [35]. The adjacent hydroxyl groups along the Dex backbone can be oxidized to generate aldehyde functionalities, which subsequently undergo Schiff base reactions with amine-containing compounds, providing a chemical basis for constructing dynamically crosslinked hydrogels [36, 37]. In addition, hyaluronic acid (HA), as a major component of the skin extracellular matrix, possesses excellent biocompatibility, biodegradability, viscoelasticity, and abundant functional groups, making it an ideal material for the fabrication of bioactive hydrogels for diabetic wound treatment [1]. HA plays a critical role in maintaining tissue hydration and extracellular space homeostasis and can be readily chemically modified to endow hydrogels with tunable physicochemical and biological properties [38, 39]. Furthermore, incorporating bioactive agents—such as antimicrobial compounds, antioxidants, and regenerative factors—has further advanced the development of hydrogel-based dressings for chronic wound care [31, 40]. However, most current hydrogel systems are limited to a single function, such as promoting tissue regeneration. They are insufficient to address the complex and multifactorial challenges presented by the diabetic wound microenvironment [41, 42]. Tannic acid (TA), a naturally abundant plant-derived polyphenol, exhibits excellent antioxidant capacity and potent antibiofilm activity, and has recently attracted increasing attention in wound healing and infection control [43–45]. Moreover, it serves as a green reducing agent for the eco-friendly synthesis of metal nanoparticles (e.g., Au, Ag), expanding its applications in tissue engineering and regenerative medicine [46–48]. Therefore, a multifunctional strategy that integrates antibiofilm, antioxidant, and immunomodulatory properties may provide a more effective therapeutic approach for diabetic wound treatment.

To verify our hypothesis, we developed a multifunctional injectable hydrogel (SOT) that integrates antibiofilm, antioxidant, and immunoregulatory functionalities. The hydrogel network is constructed from thiolated hyaluronic acid (HA-SH) and dopamine (DA)-modified oxidized dextran (ODex-DA) via a thiol–aldehyde dynamic covalent reaction, imparting injectability, self-healing ability, and in situ gelation. In addition, we incorporated TA–silver nanoparticles (TA-Ag NPs), in which the natural polyphenol TA contributes potent antioxidant and antibiofilm properties. In contrast, silver ions are potent broad-spectrum antimicrobial agents (Fig. 1). The synergistic activity of TA and Ag enables the hydrogel to disrupt mature biofilms and eliminate excessive ROS, thereby modulating the wound microenvironment. In a streptozotocin-induced diabetic MRSA biofilm infection model, the SOT hydrogel effectively cleared bacteria and ROS, promoted macrophage polarization toward the M2 phenotype, reduced inflammation, and accelerated wound closure. These findings suggest that this immuno-instructive hydrogel platform holds great potential for treating infected chronic diabetic wounds and provides a promising strategy for next-generation wound therapeutics.

Fig. 1.

Schematic diagram of the construction of antibiofilm and antioxidant hydrogel for chronic wound repair

Results and discussion

Synthesis and characterization

As shown in Fig. 2a, transmission electron microscopy (TEM) revealed that the TA-Ag NPs were uniformly dispersed with an average diameter of approximately 20 nm. The nanoparticles exhibited a zeta potential of -37 mV and a hydrated diameter of around 70 nm (Fig. 2b–c), indicating good colloidal stability. Ultraviolet–visible (UV–Vis) spectroscopy displayed a distinct absorption peak at 405 nm, which corresponds to the characteristic surface plasmon resonance of the silver nanoparticles (Fig. 2d). The chemical structures of Dex, oxidized Dex(ODex), and dopamine-modified ODex (ODex-DA) were confirmed by ¹H nuclear magnetic resonance (¹H NMR) spectroscopy (Fig. 2e). Dex exhibited characteristic signals of dextran, including the anomeric proton at ~ 5.0 ppm and sugar ring protons distributed between 3.2 and 4.2 ppm. After periodate oxidation, ODex retained the typical polysaccharide backbone signals, indicating that the oxidation process did not disrupt the main chain structure. Notably, new aromatic proton signals appeared at 6.6–6.8 ppm in the ODex-DA spectrum, which were absent in both Dex and ODex. These peaks are attributed to the catechol moiety of DA, confirming the successful conjugation of dopamine onto the ODex backbone. Moreover, UV–Vis spectra showed an absorbance peak at 280 nm, consistent with the presence of aromatic rings in catechol groups, and the DA grafting degree was calculated to be approximately 1.2% (Figure S1). ¹H NMR analysis of HA, L-cysteine, 1-(3dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC·HCl), and HA-SH further confirmed successful modification of HA-SH (Fig. 2f). Native HA exhibited a characteristic singlet at ~ 2.0 ppm corresponding to the N-acetyl group and broad signals between 3.2 and 4.2 ppm arising from the sugar ring protons. L-cysteine displayed distinct resonances at ~ 4.0 ppm (α-CH) and ~ 2.9–3.1 ppm (–CH₂–SH). After thiolation, HA-SH retained the characteristic signals of the HA backbone, while new resonances appeared at ~ 2.7–3.1 ppm, which can be attributed to the methylene protons adjacent to thiol groups derived from cysteine. These results confirm the successful thiolation of HA, with a grafting degree estimated at ~ 23% [49, 50].

Fig. 2.

Characterization of TA-Ag NPs, ODex-DA, and HA-SH. (a) TEM image of TA-Ag NPs. (b) Zeta potential of TA-Ag NPs. (c) Hydrodynamic diameter of TA-Ag NPs measured by DLS. (d) UV–Vis absorption spectrum of TA-Ag NPs. (e) ¹H NMR spectra of Dex, ODex, and ODex-DA. (f) ¹H NMR spectra of EDC·HCl, L-cysteine, HA, and HA-SH

As shown in Fig. 3a, ODex-DA and HA-SH existed as flowable solutions before mixing. A stable hydrogel (SO) was rapidly formed upon mixing, indicating successful gelation through thiol-aldehyde crosslinking. Meanwhile, the SO hydrogel exhibited excellent flexibility, allowing it to be easily handled with tweezers (Fig. 3b). To investigate the influence of TA-Ag NPs loading on the internal microstructure of the hydrogel, hydrogels containing different amounts of TA-Ag NPs were prepared (Table S1). As shown in Fig. 3c, all hydrogels exhibited a typical three-dimensional porous network structure. With the incorporation of TA–Ag NPs, the pore size slightly decreased and the distribution became narrower, indicating that the nanoparticles acted as physical crosslinking points that enhanced the structural compactness of the network. The average pore size of SOT3 was smaller than that of the pristine hydrogel, suggesting a denser structure beneficial for mechanical stability and sustained drug release. Elemental mapping further confirmed the homogeneous dispersion of C, N, O, S, and Ag elements throughout the SOT3 hydrogel matrix, indicating the successful incorporation of TA–Ag NPs without noticeable aggregation (Fig. 3d). The uniform distribution of Ag suggests that the TA-Ag NPs were well anchored within the polymer network via coordination and hydrogen bonding with the catechol and thiol groups. XPS further confirmed the presence of Ag in the SOT3 hydrogel (Fig. 3e) [51–53]. Furthermore, thermogravimetric analysis (TGA) results demonstrated that both SO and SOT3 hydrogels exhibited typical two-step weight loss behavior (Fig. 3f). The weight loss below 100 °C was primarily due to the removal of bound water, while the rapid weight loss in the 100–150 °C range was attributed to the cleavage of thioether bonds and the decomposition of dopamine side groups. SOT3 showed a faster weight loss in this temperature range, which may be related to the catalytic oxidation of Ag and the disruption of the hydrogen bond network. Despite the accelerated initial decomposition, a higher residual weight was observed at elevated temperatures, indicating that the carbonization of TA and the inorganic residue of Ag collectively enhanced the overall thermal stability of the system. The swelling behavior of the hydrogels was evaluated in PBS at 37 °C (Fig. 3g). All samples exhibited rapid water uptake within the first 6 h and reached equilibrium thereafter. The equilibrium swelling ratios decreased slightly with increasing TA–Ag NPs content, which could be attributed to the denser crosslinking and reduced hydrophilicity caused by TA–metal coordination. All hydrogels maintained high swelling ratios (> 1200%), ensuring adequate moisture retention and exudate absorption for wound dressing applications. The release of silver ions from the hydrogel followed a typical biphasic pattern (Fig. 3h). A rapid initial release of SOT3 occurred within 1 day, with a cumulative amount of 0.42 mg/L, primarily attributed to the dissociation of surface-adsorbed ions. This was followed by a slower, sustained release, reaching 0.63 mg/L at 7 days. These results suggest that the hydrogel’s three-dimensional network effectively regulates silver ion diffusion, supporting stable and long-lasting antibacterial activity for wound infection treatment. To investigate the degradation behavior of the hydrogel, in vitro degradation experiments were conducted (Figure S2). The results showed that after 7 days of incubation in hyaluronidase solution, all four hydrogel groups (SO, SOT1, SOT2, and SOT3) were nearly completely degraded, indicating their good degradation performance. Furthermore, the degradation behavior of the hydrogels in hyaluronidase solution was further analyzed using SEM, with observations made on Days 1 and 3 of degradation. As shown in Figure S3, on Day 1 of degradation, both the central and peripheral areas of the hydrogel samples (SO, SOT1, SOT2, and SOT3) exhibited early signs of degradation, with the peripheral regions showing more significant structural disruption compared to the central areas. By Day 3, more pronounced degradation was observed, especially in the peripheral regions, where the integrity of the hydrogel was compromised, and pore formation became more evident. The central areas also showed increased degradation, characterized by expanded pores and further loss of structural integrity. This suggests that the structure of the hydrogel was disrupted and became more loose, further confirming the degradation process. This structural change is consistent with the enzymatic hydrolysis of the polymer chains, which contributes to the increased porosity and overall degradation of the material.

Fig. 3.

Preparation, microstructure, and physicochemical properties of hydrogel. (a) Representative photographs showing the formation of the SO hydrogel by mixing ODex-DA and HA-SH solutions. (b) Photograph of the stable formed SO hydrogel lifted with forceps. (c) SEM images of different hydrogels and corresponding pore size distribution (scale bar: 200 μm). (d) SEM elemental distribution map of SOT3 hydrogel (scale bar: 200 μm). (e) XPS full spectra of SO and SOT3 hydrogels. (f) TGA curves of SO and SOT3 hydrogels. (g) Swelling behavior of different hydrogels (n = 3). (h) Ag⁺ release profile from hydrogel (n = 3)

Mechanical properties of hydrogels

The self-healing ability and injectability of the SOT hydrogel were further evaluated. As shown in Fig. 4a-b, the SOT3 hydrogel was cut into two halves and then brought into contact without any external stimuli. After 10 min, the separated pieces fused together and could be stretched without fracture (Fig. 4c), confirming the hydrogel’s favorable self-healing performance. This remarkable property can be attributed to the dynamic reversible interactions within the hydrogel network, such as thiol-aldehyde linkages and hydrogen bonding, which enable rapid reformation of the crosslinked structure after damage. Moreover, the SOT3 hydrogel displayed good injectability, allowing smooth extrusion through a dual-barrel syringe to form designed letters (Fig. 4d). The combination of injectability and self-healing capability endows the SOT3 hydrogel with great potential for minimally invasive administration and conformal filling of irregular wound sites. The rheological and mechanical behaviors of the SO and SOT3 hydrogels were systematically evaluated to access the effects of TA-Ag NPs incorporation. As shown in the time sweep curves (Fig. 4e), the storage modulus (G′) of all samples was consistently higher than the loss modulus (G″), indicating the formation of stable and elastic crosslinked networks. with increasing TA-Ag NPs content, G′ values gradually increased, suggesting enhanced crosslinking density and improved structural integrity of the hydrogels. The strain sweep tests (Fig. 4f) revealed that G′ remained constant under low strain but decreased sharply when the strain exceeded a critical value, indicating the breakdown of the internal network. In the step–strain measurements (Fig. 4g), G′ dropped below G″ under high strain (800%), corresponding to the destruction of the hydrogel network and transition to a sol-like state. When the strain was reverted back to 1%, G′ rapidly recovered to its initial level, indicating the rapid reconstruction of the elastic network and the good self-healing capability of the hydrogels [54]. This dynamic covalent network was established through a rapid and reversible thiol–aldehyde addition reaction between the thiol groups of HA-SH and the aldehyde groups of ODex-DA, endowing the hydrogels with excellent self-healing and injectability properties [55]. As shown in Fig. 4h, SO and SOT3 hydrogels exhibited pronounced shear-thinning behavior, characteristic of injectable systems. This feature allows facile extrusion through a syringe and ensures effective filling of irregular wound geometries. The similar rheological behaviors of SO and SOT3 suggest that TA-Ag NPs incorporation had negligible effects on the overall viscoelastic of the hydrogels. Furthermore, the compressive stress–strain curves (Fig. 4i) and the corresponding elastic modulus values (Fig. 4j) demonstrated that all hydrogels possessed good compressive strength and elasticity. The incorporation of TA-Ag NPs significantly increased the compressive stress and elastic modulus, indicating that TA-Ag NPs acted as effective physical crosslinkers that reinforced the hydrogel network and enhanced its mechanical stability. These results suggest that the SOT hydrogels exhibit favorable viscoelasticity, self-healing ability, and good injectability, making them suitable for wound dressing applications.

Fig. 4.

Self-healing, injectability, and mechanical properties of the hydrogel. (a) The SOT3 hydrogel was cut into two halves, with the red sample stained using rhodamine. (b) The separated halves were brought into contact without external stimuli. (c) Demonstration of the stretching of hydrogel fragments after self-healing. (d) Injectability of the SOT3 hydrogel. (e) Time-dependent storage modulus of different hydrogels. (f) Storage modulus and loss modulus of SO and SOT3 hydrogels under strains ranging from 1% to 1000%. (g) Alternating strain sweep (1% and 800%) of SO and SOT3 hydrogels showing dynamic recovery behavior. (h) Shear-thinning behavior of SO and SOT3 hydrogels under varying shear rates. (i) Compressive stress–strain curves of different hydrogels. (j) Comparison of compressive elastic modulus among different hydrogels (n = 3)

In vitro antibacterial effects of hydrogel

Chronic diabetic wounds are typically characterized by polymicrobial infections. MRSA is well known for its robust biofilm-forming ability and persistent induction of inflammatory, whereas E. coli and other Gram-negative bacteria contribute to polymicrobial interactions that exacerbate tissue damage and delay epithelial regeneration [56]. Therefore, MRSA and E. coli were selected as representative strains to evaluate the broad-spectrum antibacterial performance of the hydrogels. Agar plate assays demonstrated that the survival rate of MRSA was markedly reduced after treatment with hydrogels containing TA-Ag NPs (Fig. 5a), with the SOT3 group showing only 1.4% viability (Fig. 5b). Similarly, the number of E. coli colonies was significantly decreased in the SO2 and SOT3 groups compared with the control and SO groups (Fig. 5c), with quantitative analysis showing survival rates of 16.09% and 11.96%, respectively (Fig. 5d). In contrast, the SO hydrogel alone provided only partial inhibition, with visible bacterial colonies remaining (Fig. 5a, b). SYTO-9/PI fluorescence staining further confirmed the bactericidal activity of the hydrogels. Samples treated with SOT3 showed strong red fluorescence, indicating extensive bacterial death, whereas the control and SO groups displayed predominantly green fluorescence, corresponding to viable bacteria (Fig. 5e, f). Statistical analysis showed that MRSA and E. coli survival rates after SOT3 treatment were 26.9/100 and 36/100, respectively, further verifying the dose-dependent antibacterial efficacy of the TA-Ag NPs–loaded hydrogels (Fig. 5g–h). SEM was used to investigate bacterial membranes integrity and structural alterations following various treatments (Fig. 5i-j). MRSA and E. coli in the untreated control group exhibited normal morphology, with smooth surfaces and intact membranes. In contrast, treatment with SO, SOT1, SO2, and SOT3 hydrogels induced evident morphological damage. The SO group showed relatively mild disruption, whereas severe wrinkling and membrane compromise were observed in the SOT1, SO2, and SOT3 groups. These SEM observations indicate that the hydrogels exert antibacterial effects by disrupting the structural integrity of bacterial membranes.

Fig. 5.

In vitro antibacterial and antibiofilm effects of hydrogel. (a) Representative photographs of MRSA colonies grown on agar plates after incubation with different treatments. (b) Quantitative analysis of MRSA bacterial survival rate (n = 3). (c) Representative photographs of E. coli colonies grown on agar plates after incubation with different hydrogels. Data are presented as mean ± SD (n = 3). ****P < 0.0001. (d) Quantitative analysis of E. coli bacterial survival rate. Data are presented as mean ± SD (n = 3). ****P < 0.0001. (e, f) Live/dead fluorescence staining images of MRSA and E. coli, respectively. (g, h) Live/dead ratio of MRSA and E. coli, respectively. (i, j) SEM images of MRSA and E. coli after different treatments

In vitro antibiofilm effects of hydrogel

Bacterial biofilms are highly organized communities embedded in self-produced EPS and firmly attached to wound surfaces. Unlike planktonic bacteria, biofilm-encased bacteria exhibit markedly increased antibiotic resistance and enhanced immune evasion, contributing significantly to delayed healing in chronic wounds [2]. Recent studies have highlighted the potent antibacterial and antibiofilm activities of TA-Ag NPs, a natural polyphenol with strong metal-chelating and protein-binding capacities [57]. Given the robust antibacterial properties of TA-Ag NPs, the antibiofilm efficacy of the hydrogels against MRSA and E. coli was systematically evaluated. Crystal violet staining was used to quantify biofilm removal (Fig. 6a, b). After incubation, control wells showed dense and continuous violet staining, indicative of thick and well-established biofilms. In contrast, hydrogels containing TA-Ag NPs exhibited a dose-dependent inhibitory effect. For MRSA, the biofilm removal increased from 10.7% in the SO to 24.7%, 38.9%, and 46.6% in the SOT1, SOT2, and SOT3, respectively (Fig. 6c). Similarly, E. coli biofilms removal rates were 5.0%, 23.1%, 28.8%, and 36.2% across the same groups, with all differences statistically significant (Fig. 6d). These results demonstrate that increasing TA-Ag NPs content enhances antibiofilm activity, likely via EPS disruption through chelation and generation of reactive oxygen species that compromise bacterial membranes.

Fig. 6.

Antibiofilm properties of hydrogel. (a, b) Representative crystal violet-stained images of MRSA and E. coli biofilms after different treatments (n = 3). (c, d) Quantitative analysis of the inhibition rate of MRSA and E. coli biofilms based on crystal violet staining (n = 3, ****P < 0.0001). (e, f) Live/dead fluorescence images of MRSA and E. coli biofilms following different treatments (length, width, height: 350 × 350 × 5.7 μm) (n = 3). (g, h) Quantification of live/dead ratios for MRSA and E. coli biofilms (n = 3). (i, j) SEM images of MRSA and E. coli biofilms after different treatments (scale bar: 10 μm; enlarged view: 2 μm)

Live/dead staining with SYTO-9 and PI further visualized biofilm viability (Fig. 6e and f). In the control and SO groups, MRSA and E. coli biofilms exhibited strong green fluorescence, reflecting viable bacterial populations within an intact matrix. Treatment with SOT2 and particularly SOT3 resulted in pronounced red fluorescence, suggesting substantial bacterial membrane damage and death. Quantitative analysis showed the MRSA live/dead ratio decreased from 73.2/26.8 (SO) to 46.9/53.1 (SOT1), 43.1/56.9 (SOT2), and 34.7/65.3 (SOT3) (Fig. 6g). Similarly, for E. coli, the ratio decreased from 63.3/36.7 (SO) to 52.3/47.7 (SOT1), 44.5/55.5 (SOT2), and 33.8/66.2 (SOT3) (Fig. 6h). SOT3 exhibited the strongest antibiofilm effect, consistent with its highest TA-Ag NPs content. SEM imaging provided direct morphological evidence of biofilm disruption (Fig. 6i and j). Control and SO groups showed dense, multilayered bacterial clusters forming continuous biofilm layers. In contrast, SOT2 and SOT3 treatments markedly disrupted biofilm integrity, leaving only scattered bacterial debris. High-magnification images revealed wrinkled and collapsed bacterial surfaces, confirming severe biofilm damage and compromised cell integrity. These morphological observations, consistent with fluorescence and quantitative analyses, indicate that the hydrogels effectively eradicate mature biofilms. The enhanced antibiofilm efficacy of TA-Ag NPs–loaded hydrogels is attributed to the abundant phenolic hydroxyl groups on TA-Ag NPs, which interact with negatively charged bacterial cell walls via hydrogen bonding and electrostatic attraction, increasing membrane permeability [58]. Once internalized, Ag⁺ ions bind to thiol groups of cell wall and membrane proteins, disrupting protein structure and membrane function, ultimately causing intracellular content leakage and cell collapse. Overall, the SOT2 and SOT3 hydrogels exhibited excellent biofilm clearance and bactericidal activity against both MRSA and E. coli, highlighting the potential of TA-Ag NPs composite hydrogels as multifunctional dressings for eradicating biofilm-associated infections and promoting chronic wound healing.

Mechanism of anti-MRSA biofilm activity

To further investigate the mechanism of action of SOT3 hydrogel against MRSA biofilm, transcriptomic analysis was performed on MRSA treated with either the control or SOT3 group. The Venn diagram clearly shows the gene distribution between the control and SOT3-treated groups (Fig. 7a). PCA analysis demonstrated good reproducibility among the samples (Figure S4). As shown in the volcano plot (Fig. 7b), a cutoff criterion of a 2-fold up/down regulation with a p-adjusted value < 0.05 was applied, resulting in 398 differentially expressed proteins (DEPs), including 233 upregulated genes and 165 downregulated genes. A heatmap was used to further demonstrate the consistency of gene variation, highlighting the differences between the two groups (Figure S5). Gene Ontology (GO) functional and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed to further understand the mechanism by which SOT3 exerts its anti-biofilm activity against MRSA. The GO analysis focused on categorizing the functional spectrum of differentially expressed genes with upregulated and downregulated GO terms. The study encompassed three major categories: biological processes, cellular components, and molecular functions (Fig. 7c). The results emphasize that upregulated genes are involved in potassium ion transmembrane transporter activity, biogenic amine metabolism and synthesis processes, and amino acid biosynthesis (Figure S6a). Previous reports have shown that bacterial cells activate certain amino acid biosynthesis pathways to sustain survival after undergoing various stress treatments [59]. Microbial biological systems are highly dependent on the availability of metal ions in the environment [60]. After SOT3 treatment, genes related to Fe ion transport (isdG, yclO) were upregulated, leading to a restricted intracellular iron content (Figure S6b). This may be due to TA reducing the bioavailability of extracellular iron through complexation, prompting MRSA to enhance iron uptake-related genes to maintain its survival [61, 62]. In contrast, downregulated genes are primarily associated with the negative regulation of ribosomal subunits, L-amino acid, and arginine catabolic processes (Figure S6c). Ribosomes play a central role in regulating protein translation, which is crucial for bacterial growth. After SOT3 treatment, the expression of numerous ribosomal subunit-related genes was downregulated. For example, genes associated with the 30 S small subunit (such as rpsC, rpsD, rpsS, and rplJ) and the 50 S large subunit (such as rplB, rplC, rplA, and rplL) were inhibited, indicating that a decrease in MRSA’s after treatment (Figure S6d). KEGG analysis revealed that the upregulated differentially expressed proteins were significantly enriched in multiple metabolism-related pathways, including quorum sensing (QS), phenylalanine, tyrosine, and tryptophan biosynthesis, as well as biofilm formation (Fig. 7d). QS is a bacterial communication mechanism that enables cells to sense population density and coordinate collective behaviors, thereby regulating biofilm growth and development and mediating intercellular interactions [57, 63]. As shown in Fig. 7e, SOT3 treatment markedly perturbed the QS response of MRSA, as evidenced by the upregulation of several stress-related genes, including sspA, comK, and secY [64, 65]. In addition, genes involved in phenylalanine and tryptophan metabolism (such as trpA, trpC, and trpD) were significantly upregulated, suggesting a metabolic reprogramming response that may partially compensate for the suppression of arginine biosynthesis (Fig. 7f). In contrast, key genes associated with arginine biosynthesis, including argF, arcA, and arcC, were markedly downregulated, resulting in impaired arginine production and consequently reduced biofilm formation and virulence of MRSA (Fig. 7g and h) [66, 67]. These metabolic disturbances further exacerbated intracellular oxidative stress and induced DNA damage, thereby synergistically enhancing the antibacterial efficacy of SOT3 [68]. Previous studies have demonstrated that tannins primarily inhibit MRSA growth by suppressing ribosome-related pathways, disrupting the translation process, and inhibiting protein synthesis [69]. Consistently, SOT3 treatment significantly reduced the abundance or gene expression levels of proteins constituting the 50 S and 30 S ribosomal subunits of Staphylococcus aureus, such as the 50 S ribosomal proteins L2 (rplB) and L16 (rplP) (Fig. 7i), further highlighting a multi-target mechanism underlying the antibacterial activity of SOT3. These results suggest that SOT3 hydrogel may exert its anti-MRSA biofilm activity by regulating multiple metabolic pathways, particularly by influencing key processes such as amino acid synthesis, quorum sensing, and iron ion transport.

Fig. 7.

Transcriptomic analysis of MRSA under different treatments. (a) Venn diagram depicting the intersecting target genes between SOT3 and control-treated MRSA. (b) Volcano plot illustrating the differential gene expression profiles, with non-significant genes shown in gray, upregulated genes in red, and downregulated genes in blue. (c) GO enrichment analysis. (d) KEGG pathway enrichment analysis for upregulated genes. (e) Heatmap and cluster analysis of quorum sensing genes. (f) Heatmap and cluster analysis of phenylalanine and tryptophan metabolism genes. (g) KEGG pathway enrichment analysis for downregulated genes. (h) Heatmap and cluster analysis of ribosome-related genes. (i) Heatmap and cluster analysis of arginine biosynthesis-related genes

Biocompatibility evaluation of the hydrogel

Good biocompatibility is essential for hydrogels intended for clinical wound dressing applications. The biocompatibility of the SOT hydrogel was systematically evaluated through a series of in vitro assays. L929 fibroblasts cultured with various hydrogel formulations for 1 day (Fig. 8a) and 2 days (Fig. 8b) maintained high viability across all groups, indicating minimal cytotoxicity and good cytocompatibility. Hemolysis assays revealed negligible red blood cell lysis, with hemolysis rates well below the established safety threshold of 5%, confirming the hydrogel’s good blood compatibility (Fig. 8c). Furthermore, live/dead staining after 24 and 48 h of incubation showed predominantly viable cells and minimal cell death, demonstrating the hydrogel’s capacity to support cell survival and maintain a favorable cellular microenvironment (Fig. 8d-e). These results demonstrate that the SOT hydrogel possesses strong cytocompatibility and hemocompatibility, underscoring its potential suitability for biomedical applications.

Fig. 8.

Biocompatibility evaluation of the hydrogel. (a-b) Viability of L929 cells after 24 and 48 h of co-culture with hydrogels (n = 6). (c) Hemolysis assay results for hydrogels (n = 3). (d-e) Live/dead staining images of L929 cells cultured with hydrogels for 24 h and 48 h (scale bar: 200 μm)

Migration and angiogenesis ability assessment of hydrogel

The human umbilical vein endothelial cells (HUVEC) migration and angiogenesis effects of the hydrogels were evaluated. HUVEC were first seeded onto matrigel-coated plates, followed by incubation with extracts from different hydrogels. After 4 h of treatment, HUVEC exposed to the hydrogel extracts migrated toward each other and formed capillary-like tubular structures (Fig. 9a). As shown in Fig. 9c, the number of junctions formed after treatment with SO, SOT1, SOT2, and SOT3 hydrogels reached 164.25%, 192.25%, 203.5%, and 216.0% of the control group, respectively, which was higher than that of the control group (90.75%). Similarly, the quantitative analysis of meshes demonstrated a comparable trend following hydrogel treatment (Fig. 9d). In addition, to exclude the influence of cell proliferation on migration, a monolayer of HUVEC was pretreated with mitomycin B for 2 h, followed by incubation with hydrogel extracts (Fig. 9b). Images of the scratch area were captured at 0 h and 12 h. As shown in Fig. 9b, the migration of HUVEC treated with SO hydrogel was slightly inhibited compared with the control group. In contrast, treatment with SOT1, SOT2, and SOT3 hydrogel extracts significantly enhanced cell migration, with migration rates of 62.44%, 67.74%, and 82.75%, respectively (Fig. 9e).

Fig. 9.

Migration and angiogenesis ability assessment of hydrogel. (a) Representative images of HUVEC migration after treatment with different hydrogel extracts (scale bar = 400 μm). (b) Representative images of capillary-like tube formation by HUVECs following treatment with different hydrogels (scale bar = 400 μm). (c, d) Quantitative analysis of junction and mesh numbers in HUVECs after 4 h of treatment with various hydrogels (n = 4). ****p < 0.0001. (e) Quantification of HUVEC migration rates following treatment with different hydrogels (n = 3). ****p < 0.0001, ***p < 0.001, and *p < 0.05

In vitro antioxidant and modulation of macrophage M2 polarization of hydrogel

Excessive levels of reactive nitrogen species (RNS) and ROS in the chronic wound microenvironment can disrupt intracellular signaling pathways, leading to cellular dysfunction and tissue damage, thereby hindering the wound healing process [70, 71]. Therefore, efficient scavenging of these free radicals is crucial for modulating oxidative stress and promoting tissue regeneration. This study developed a hydrogel system incorporating TA and catechol moieties, both of which possess phenolic hydroxyl groups known for their antioxidant capacity. To assess the antioxidant properties of the hydrogels, three representative radical models—1,1-diphenyl-2-picrylhydrazine (DPPH), 2,20-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS), and hydroxyl radicals (·OH)—were used in vitro. DPPH and ABTS are stable nitrogen-centered radicals commonly used to assess RNS scavenging efficiency, while ·OH, generated via the Fenton reaction, is representative of ROS. The underlying mechanisms of these scavenging reactions are illustrated in Figure S7, respectively. As shown in Fig. 10a, the characteristic absorption peak of DPPH solution at 517 nm gradually decreased upon incubation with the hydrogels, accompanied by a visible color change from deep purple to pale yellow, indicating reduced radicals. Quantitative analysis (Fig. 10b) revealed that the DPPH scavenging efficiency increased significantly with the concentration of TA-Ag NPs, suggesting that their incorporation effectively enhanced the RNS scavenging capability of the hydrogel. Figure 10c and d demonstrate the hydrogel’s scavenging ability against ABTS⁺ radicals. All four hydrogel formulations exhibited notable ABTS⁺ scavenging activity, with those containing TA-Ag NPs displaying superior performance, further confirming the antioxidant contribution of TA. The ·OH scavenging capacity was also evaluated using the hydroxyl radical model (Fig. 10e and f). The SOT3 hydrogel exhibited significantly higher ·OH scavenging efficiency compared to the SO hydrogel, indicating that the presence of TA-Ag synergistically improved the ROS scavenging capacity. The synergistic antioxidant effect of TA and catechol endows the hydrogel with strong free radical scavenging ability, which shows great potential for regulating oxidative stress, alleviating chronic inflammation, and promoting skin regeneration in wound healing applications.

Fig. 10.

In vitro free radical scavenging properties of hydrogels. (a) Absorbance curves of the ABTS assay, and (b) scavenging efficiencies of different hydrogels. ****p < 0.0001. (c) Absorbance curves of the DPPH assay, and (d) scavenging efficiencies of different hydrogels. ****p < 0.0001. (e) Absorbance curves of the ·OH assay, and (f) scavenging efficiencies of different hydrogels (n = 3). ****p < 0.0001. (g) Bright-field and fluorescence images showing ROS levels in RAW264.7 cells (scale bar: 100 μm). (h) Flow cytometry quantification of ROS levels in RAW264.7 cells. (i) Representative immunofluorescence images of RAW264.7 cells after 24 h exposure to different hydrogels, stained for CD206 (M2 marker) and CD86 (M1 marker) (scale bar: 20 μm). (j) M2 to M1 macrophages ratio in RAW264.7 cells after treatment. Data are presented as mean ± SD (n = 3). ***P < 0.001 and ****P < 0.0001

To assess the intracellular ROS-scavenging ability of the hydrogels, H₂O₂-stimulated RAW264.7 cells were co-incubated with hydrogel extracts and subsequently stained with DCFH-DA. As shown in Fig. 10g, bright green fluorescence was observed in the control group, indicating excessive intracellular ROS generation. However, this oxidative stress was alleviated upon treatment with the hydrogel extracts, as further confirmed by flow cytometry analysis (Fig. 10h). Among the tested formulations, the SOT3 hydrogel exhibited the most vigorous ROS-scavenging activity, followed by SOT2 and SOT1, while the SO hydrogel showed the weakest effect, though still superior to the control. These results demonstrate that the incorporation of TA-Ag NPs effectively enhances the antioxidant performance of the hydrogels, which can be attributed to the high TA content in the TA-Ag system.

In diabetic wounds, the imbalance in macrophage polarization between M1 and M2 phenotypes plays a critical role in impaired healing. Persistent activation of pro-inflammatory M1 macrophages leads to chronic inflammation, while the impaired transition to the pro-regenerative M2 phenotype restricts tissue repair, resulting in prolonged inflammation. During the inflammatory phase, macrophages polarize into M1 or M2 phenotypes in response to distinct cytokine signals. To evaluate the anti-inflammatory potential of the hydrogel under oxidative stress, immunofluorescence staining for CD86 (M1 marker) and CD206 (M2 marker) was performed. As shown in Fig. 10i, LPS-stimulated control cells exhibited strong CD86-positive (green) M1 polarization. In contrast, cells treated with the various hydrogels showed a marked reduction in CD86 expression and increased CD206 expression. Quantitative analysis of fluorescence intensity confirmed that the hydrogels promoted the polarization of RAW264.7 macrophages toward the M2 phenotype, thereby modulating the inflammatory response (Fig. 10j).

In addition, flow cytometry analysis further validated the macrophage polarization profiles. As shown in Figure S8, after 24 h of LPS stimulation, the control group exhibited a high proportion of CD86-positive M1 macrophages (28.74%) and a low proportion of CD206-positive M2 macrophages (6.69%). Treatment with SO, SOT1, SOT2, and SOT3 hydrogels significantly reduced CD86 expression and increased CD206 expression compared to the control group, indicating that these hydrogels effectively promoted M2 polarization. The SOT3 group showed the most pronounced effect, with the highest CD206 expression (29.68%) and the lowest CD86 expression (6.92%), suggesting a more favorable inflammatory microenvironment. Additionally, ELISA results showed that the level of anti-inflammatory cytokine IL-10 was elevated in the hydrogel-treated group (Figure S9). These results demonstrate that the prepared hydrogels effectively promote M2 polarization and have the potential to modulate the inflammatory microenvironment in diabetic wounds, highlighting their promise for wound healing applications.

In vivo wound healing properties

An ideal wound dressing should possess the ability to effectively control infection and facilitate tissue repair. To simulate the high risk of infection associated with chronic diabetic wounds, an infected full-thickness skin defect model was established to evaluate the therapeutic efficacy of the hydrogels (Fig. 11a). Wounds were treated with different hydrogel formulations or 3 M Tegaderm™ (a commercial film dressing), and wound progression (Fig. 11b) and healing kinetics (Fig. 11c) were assessed at days − 1, 3, 6, 9, and 15. Throughout the 15-day treatment period, the SO and SOT2 groups exhibited significantly smaller wound areas than the control group (Fig. 11d). Among all groups, the SOT2 hydrogel demonstrated the most effective healing performance and the fastest wound closure rate at each time point (Fig. 11b and d). Notably, on day 6, the residual wound area in the SOT2 group was approximately 27.3%, while the 3 M group still showed 57.8%. By day 15, wounds in the SOT2 group had nearly completely closed (residual area: 3.6%), whereas the SO group had a wound area of 10.5%, and other groups exhibited substantial unhealed regions (Fig. 11d). These findings suggest that the SOT2 hydrogel offers significant therapeutic benefits, likely attributed to its ability to maintain a moist wound environment, suppress inflammation, and deliver antibacterial TA-Ag NPs. Furthermore, the SO hydrogel still outperformed the control group, possibly due to the intrinsic pro-healing properties of natural polymers such as dextran and sodium hyaluronate. Hematoxylin and eosin (H&E) staining and Masson staining results further confirmed the good wound repair performance of the prepared hydrogel (Fig. 11e and f). Compared to the control group, the SOT2 hydrogel-treated wounds exhibited markedly improved healing outcomes, characterized by well-formed epidermal and dermal layers, as well as newly formed blood vessels and hair follicles. In addition, the most abundant deposition of deep blue collagen fibers was observed in the SOT2-treated tissues, indicating enhanced tissue regeneration and remodeling at the wound site.

Fig. 11.

In vivo evaluation of wound healing performance. (a) Schematic illustration of the establishment of an infected full-thickness skin wound model and the corresponding healing process. (b) Representative wound images from each treatment group on days − 1, 3, 6, 9, and 15. (c) Schematic depiction of wound healing progression across treatment groups from day − 1 to day 15. (scale bar: 2 mm). (d) Quantitative analysis of wound contraction over time in each group from day − 1 to day 15. ****P < 0.0001. (e–f) Histological assessment of tissue regeneration on day 15 using H&E (e) and Masson’s trichrome staining (f). (scale bar: 500 μm; enlarged views: 200 μm). (g) Quantification of granulation tissue length on day 15 based on H&E staining (n = 3). **P < 0.01 and ****P < 0.0001. (h) Quantification of relative collagen deposition on day 15 based on Masson’s trichrome staining (n = 3). ***P < 0.001 and ****P < 0.0001

To further elucidate the mechanisms by which the hydrogel accelerates wound healing, its antibacterial, anti-inflammatory, and pro-angiogenic properties were systematically evaluated. As shown in Figure S10, residual bacterial colonies at the wound site were assessed on days 0 and 6 following treatment with PBS or different hydrogel formulations. By day 6, the SOT2 group exhibited a significant reduction in bacterial colonies, demonstrating excellent antibacterial activity and suggesting that the hydrogel effectively inhibits bacterial growth at the wound site. Inflammation is a significant impediment to the healing of chronic diabetic wounds, and the phenotypic polarization of macrophages at the injury site is considered a crucial indicator of the inflammatory status. As shown in Fig. 12a and b, minimal ROS fluorescence was observed in the wound tissues treated with SOT2 hydrogel on day 6 post-treatment, whereas strong red fluorescence signals—indicative of abundant ROS-positive cells—were evident in the control group. These results demonstrate the superior in vivo ROS-scavenging capability of the SOT2 hydrogel. Additionally, M1 macrophages primarily mediate pro-inflammatory responses, whereas M2 macrophages contribute to inflammation resolution and promote tissue repair and regeneration. To investigate the effect of the hydrogel on macrophage polarization, immunofluorescence staining was performed using CD86 and CD206 as specific markers for M1 and M2 macrophages, respectively. As shown in Fig. 12c, treatment with SOT2 markedly increased the proportion of CD206-positive (Fig. 12c and e, green fluorescence) macrophages at the wound site, while significantly reducing the proportion of CD86-positive (Fig. 12d and e, red fluorescence) cells, compared to the control group. These results suggest an effective reduction of local inflammation. Quantitative CD206/CD86 ratio analysis further confirmed that the SOT2 hydrogel promotes macrophage polarization toward the pro-regenerative M2 phenotype (Fig. 12f).

Fig. 12.

ROS levels, macrophage polarization, and angiogenesis in hydrogel-treated diabetic wounds. (a) Representative fluorescence images showing ROS levels in wound tissues on day 6 after treatment. (b) Representative fluorescence images showing quantitative analysis of ROS levels in wound tissues on day 6 after treatment. ****P < 0.0001. (c) Immunofluorescence co-staining of CD206 in newly formed tissues across various treatment groups (n = 3). (d) Immunofluorescence co-staining of CD86 in newly formed tissues across different treatment groups (n = 3). (e) Merged image of (c) and (d). (f) Quantitative analysis of the CD206/CD86 ratio in newly formed tissues (n = 3). ****P < 0.0001. (g) Immunofluorescence staining of CD31 and α-SMA in newly formed tissues from different treatment groups (n = 3). (h) Quantification of CD31 expression levels in newly formed tissues (n = 3). ****P < 0.0001. (i) Quantification of α-SMA expression levels in newly formed tissues (n = 3). ****P < 0.0001

Neovascularization plays a vital role in the repair of chronic diabetic wounds by improving the hypoxic and nutrient-deficient microenvironment, thereby promoting tissue regeneration and accelerating healing. To evaluate the pro-angiogenic effects of the hydrogels, immunostaining for CD31 and α-SMA was performed on newly formed tissue at day 15 post-treatment. As shown in Fig. 12g, compared to the control group, hydrogel-treated wounds exhibited a markedly increased density of newly formed vessels, with more extensive CD31- and α-SMA-positive regions. Quantitative analysis revealed that the SO and SOT2 groups showed approximately 200% CD31 expression compared to the control group (Fig. 12h), significantly higher than the control. Notably, the SOT2 group exhibited the largest α-SMA-positive area (~ 730%) (Fig. 12i), indicating its superior angiogenic capability and potential to accelerate wound closure.

In addition, the secretion of representative anti-inflammatory cytokines by macrophages serves as a critical indicator for evaluating the inflammatory status. Immunohistochemical staining was performed to assess the inflammatory status of the tissue following hydrogel treatment. The expression levels of proinflammatory cytokines TNF-α (Fig. 13a and c) and IL-1β (Fig. 13b and d) in the tissues after treatment with the SOT2 hydrogel were significantly reduced compared with the control group, indicating that the hydrogel effectively reduce local inflammation at the wound site.

Fig. 13.

Immunohistochemical analysis of inflammatory factors in wound tissue on day 15 of treatment. (a–b) Immunohistochemical staining of pro-inflammatory cytokines IL-1β (a) and TNF-α (b) in wound tissues on day 15. (c–d) Quantitative analysis of IL-1β (c) and TNF-α (d) expression in wound tissues on day 15 based on immunohistochemistry (n = 3). **P < 0.01, ***P < 0.001, and ****P < 0.0001

Biosafety is a critical criterion for evaluating the clinical translation potential of biomaterials. To assess the in vivo biocompatibility of the hydrogel, histological analyses of major organs—including the liver, heart, kidneys, spleen, and lungs—were conducted using H&E staining after treatment. No apparent pathological alterations or inflammatory responses were observed in any of the examined organs across treatment groups, indicating that the hydrogel did not induce long-term tissue damage or systemic toxicity (Figure S11). These results underscore the excellent biosafety and biocompatibility of the hydrogel, highlighting its promise for clinical application as a wound dressing.

Conclusions

We developed a multifunctional injectable hydrogel for the treatment of infected diabetic wounds, capable of addressing major pathological challenges such as biofilm formation, excessive oxidative stress, and immune dysregulation. The hydrogel, constructed from HA-SH, ODex-DA, and TA–Ag NPs, exhibits excellent injectability, self-healing ability, and in situ gelation. The incorporation of TA–Ag NPs endows the hydrogel with antibiofilm and antioxidant activities while promoting macrophage polarization toward a pro-regenerative M2 phenotype. In an MRSA biofilm–infected diabetic wound model, the hydrogel effectively reduced bacterial burden and excessive ROS, modulated the inflammatory microenvironment, and significantly accelerated tissue regeneration and wound closure. Collectively, these findings demonstrate that the hydrogel represents a promising therapeutic strategy for the treatment of MRSA biofilm–infected chronic diabetic wounds. Despite the promising therapeutic outcomes observed in this study, several limitations remain. Wound-healing efficacy and biosafety were evaluated only in small animals, and validation in large-animal (e.g., porcine) models is required to further evaluate therapeutic efficacy. In addition, the long-term in vivo fate, accumulation, and clearance of silver species remain unclear despite systematic investigation of short-term Ag⁺ release and antibacterial performance.

Materials and methods

Materials

Hyaluronic acid (HA, Mw ≈ 300 kDa) was purchased from the Bloomage Freda Biotechnology Co., Ltd. (China). Dopamine hydrochloride (DA), sodium periodate (NaIO₄), cysteamine hydrochloride, and TA were purchased from Sigma-Aldrich unless otherwise specified. DPPH and ABTS were obtained from Shanghai Macklin Biochemical. Silver nitrate was purchased from Chengdu Kelong Chemical Reagent. CCK-8 was obtained from Targetmol. MRSA (ATCC43300) and E. coli (ATCC25922) was obtained from Luwei Technology. Matrigel was provided by Xiamen Mogengel Biotechnology Co., Ltd (Xiamen, China). Enzyme-linked immunosorbent assay (ELISA) kits were purchased from Dakewe. Cell culture reagents were obtained from Gibco (Thermo Fisher Scientific).

Synthesis of HA-SH

HA-SH was synthesized via carbodiimide-mediated coupling between the carboxyl groups of HA and the amine groups of cysteamine. Briefly, HA (2 g) was dissolved in deionized water (100 mL) under stirring at room temperature. Subsequently, EDC (0.95 g) and NHS (0.57 g) were added to activate the carboxyl groups of HA, and the reaction mixture was stirred for 30 min at 25 °C with the pH adjusted to 5.5. Cysteamine hydrochloride (0.78 g) was then added, and the reaction was allowed to proceed for 12 h at 25 °C protected from light. The resulting solution was purified by dialysis against deionized water using a dialysis membrane for 3 days, followed by lyophilization to obtain HA-SH for long-term storage.

Preparation of ODex–DA

Dextran was oxidized using sodium periodate (NaIO₄) to obtain oxidized dextran (ODex) bearing aldehyde groups. Subsequently, dopamine (DA) was conjugated to ODex via Schiff base formation between the aldehyde groups of ODex and the amine groups of DA. Briefly, dextran (5 g) was dissolved in deionized water, followed by the addition of NaIO₄ (4 g). The reaction was allowed to proceed for 1 h at 25 °C in the dark. The resulting solution was dialyzed and freeze-dried to obtain ODex. ODex (2 g) was then dissolved in deionized water and reacted with dopamine hydrochloride (1 g) for 1 h under a nitrogen atmosphere. The final reaction mixture was purified by dialysis and subsequently freeze-dried to yield ODex-DA.

Synthesis of TA–Ag NPs

10 mg of TA was dissolved in 40 mL of deionized water. A 5 wt% ammonium hydroxide solution was added dropwise to adjust the solution to a clear state, followed by the addition of 1 mL of silver nitrate solution (10 mg/mL). The mixture was stirred at 50 °C for 30 min, followed by centrifugation and washing to obtain the TA-Ag NPs.

Fabrication of hydrogel

To obtain the without TA-Ag NPs hydrogel (SO), equal volumes of HA-SH (50 mg/mL) and ODex-DA (100 mg/mL) solutions were mixed thoroughly. For the preparation of TA-Ag NPs–loaded hydrogels, varying amounts of TA-Ag NPs were first dispersed in the ODex-DA solution, which was then mixed with an equal volume of HA-SH solution to form a series of hydrogels, designated as SOTx (where x indicates the sample number), as detailed in Table S1. Specifically, SOT1, SOT2, and SOT3 represent hydrogels containing TA-Ag NPs at final concentrations of 0.625, 1.25, and 2.5 mg/mL, respectively, while maintaining the same HA-SH/ODex-DA formulation and mixing ratio.

Rheological properties

Rheological measurements were performed using a rheometer (MCR302, Anton Paar) under a constant shear strain of 0.1% (γ), at an angular frequency of 10 rad/s and a temperature of 25 °C. All experiments were conducted in triplicate to ensure reproducibility. To evaluate the self-healing capability of the hydrogel, rheological measurements were conducted using an alternating step-strain test with low (γ = 1%) and high (γ = 800%) strain at a constant angular frequency (ω = 1 rad/s). The shear-thinning behavior of the hydrogel was evaluated by performing rheological measurements over an angular frequency range of 0.1 to 100 rad/s.

Mechanical properties

The compressive properties of the hydrogels were evaluated using a dynamic mechanical analyzer (Q800, TA Instruments, USA) at a loading rate of 2 N/min.

Degradation performance

Degradation experiments were conducted using different hydrogels (100 µL per group), which were freeze-dried, and their initial weights (W₀) were recorded. The hydrogels were then immersed in a hyaluronidase solution (1 U/mL, Aladdin, derived from bovine testes) and incubated at 37 °C. The enzyme solution was refreshed every two days to maintain enzyme activity. At predetermined time points, the hydrogels were removed from the enzyme solution, rinsed with ultrapure water, freeze-dried, and weighed (W_t). Additionally, the degraded hydrogels were subjected to SEM imaging after freeze-drying. The degradation rate was calculated using the following equation:

Degradation rate = (W₀ - W_t) / W₀ × 100%.

Antioxidant activity assay

ROS scavenging ability was evaluated using 2,2-diphenyl-1-picrylhydrazyl (DPPH), 2,20-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS), and hydroxyl radical (•OH) scavenging assays. In the DPPH scavenging assay, hydrogel samples were incubated with 0.1 mM DPPH ethanol solution under dark conditions at room temperature, and the absorbance was measured at 517 nm. For the ABTS assay, the ABTS⁺ working solution was prepared by mixing 7 mM ABTS with 2.45 mM potassium persulfate and incubating in the dark for 24 h. After dilution to an appropriate concentration, it was co-incubated with the hydrogel samples, and the absorbance was recorded at 736 nm. The hydroxyl radical scavenging assay was conducted based on the Fenton reaction system, which contained FeSO₄, H₂O₂, salicylic acid, and the hydrogel samples. The absorbance was measured at 536 nm after the reaction.

In vitro antimicrobial testing

MRSA and E. coli were selected to assess the antibacterial properties of the hydrogels. Prepolymer solution (100 µL) was transferred into each well of a 48-well plate and crosslinked. A bacterial suspension (100 µL, 10⁷ CFU/mL) was applied onto the hydrogel surface and incubated at 37 °C for 2 h. Afterwards, bacteria were collected by resuspension in 500 µL sterile PBS, serially diluted, and plated on Luria–Bertani agar for 12 h. A suspension of bacteria in PBS without hydrogel served as a negative control. Colonies were photographed and counted, and the survival rate was calculated using the formula:

Survival rate (%) = CFU after hydrogel treatment / CFU of control × 100%.

In vitro antibiofilm assay

MRSA and E. coli biofilms were established in 24-well plates by incubation for 48 h. Subsequently, hydrogels were applied and co-cultured with the biofilms for an additional 6 h. The biofilms were then fixed with paraformaldehyde and stained with crystal violet for quantification via spectrophotometry at 570 nm. To assess bacterial viability within the biofilms, the SYTO-9/PI bacterial viability kit was employed, and live/dead bacterial distribution was visualized using confocal laser scanning microscopy. The treated biofilms were subjected to the same procedure, including fixation, graded dehydration, and drying, followed by SEM observation to examine their microstructural morphology.

Transcriptome sequencing and analysis

MRSA was co-cultured with different hydrogels for 6 h, followed by resuspension, centrifugation, washing, and storage at -80 °C. The transcriptome sequencing, data preprocessing, and annotation of MRSA were performed by Majorbio Bio-Pharm Technology Co., Ltd. (www.majorbio.com).

Ag⁺ release assay

To evaluate the release behavior of Ag⁺, 100 µL of hydrogel was immersed in 3 mL of deionized water and incubated at 37 °C under static conditions. At predetermined time intervals, 1 mL of the supernatant was withdrawn and immediately replaced with an equal volume of ultrapure water to maintain a constant total volume. The collected samples were analyzed to determine the concentration of released Ag⁺ using inductively coupled plasma optical emission spectrometry (ICP-OES, ARCOS, Spectro, Germany). The cumulative release of Ag⁺ was calculated based on the measured concentrations at each time point. Each sample was tested in triplicate.

Cell migration assay

HUVECs were seeded in 24-well plates and cultured until a nearly confluent monolayer formed. To reduce the influence of cell proliferation, cells were incubated with hydrogel extracts and mitomycin B. A straight scratch was made on the monolayer using a 200 µL pipette tip, and the cells were then treated with different hydrogel extracts. Images of the scratch were captured at predetermined time points. The gap area was measured using ImageJ, and the migration rate was calculated as:

Migration rate (%) = (S₀−S_t) / S₀ × 100%,

where S₀ is the initial scratch area and S_t is the remaining gap at time t. Experiments were performed in triplicate for each condition.

Tube formation assay

Matrigel (100 µL, Mogengel, 082704) was added to each well of a 48-well plate and allowed to solidify for 30 min at 37 °C. HUVEC (3 × 10⁴ cells), pretreated as described, were then seeded onto the matrigel and exposed to different hydrogel extracts for 4 h. The formation of capillary-like structures was monitored using an optical microscope, and representative images were captured for analysis.

Macrophage polarization assay

RAW264.7 macrophages were seeded into 24-well plates and stimulated with lipopolysaccharide (LPS, 100 ng/mL) for 24 h to induce an inflammatory phenotype. Subsequently, hydrogels were added and co-incubated with the cells for 24 h. The expression of M1 (CD86) and M2 (CD206) markers was evaluated via immunofluorescence staining and flow cytometry. Cytokine secretion levels were quantitatively analyzed using ELISA kits.

In vivo MRSA-Infected diabetic wound model

All animals were housed under standard specific pathogen-free conditions with a controlled environment (temperature 24 ± 2 °C, relative humidity 60–70%, and a 12-hour light/dark cycle), with free access to food and water. Prior to treatment, animals were randomly assigned to different experimental groups using a random number table. All procedures were conducted in accordance with institutional guidelines and approved by the Animal Ethics Committee of the State Key Laboratory of Biotherapy, Sichuan University (Approval No. 20250313030). Diabetes was induced in male BALB/c mice (18–20 g) via a single intraperitoneal injection of streptozotocin (120 mg/kg). Blood glucose levels were monitored using a glucometer, and mice with blood glucose levels higher than 16.7 mM were confirmed as diabetic model mice. Full-thickness excisional wounds (8 mm in diameter) were created on the dorsal skin under general anesthesia. SOT2 struck an appropriate balance among antibacterial efficacy, macrophage polarization, sustained Ag⁺ release, and biosafety. Compared to SOT1, SOT2 demonstrated superior antibacterial performance. While SOT3 contained a higher concentration of TA-Ag NPs, it did not result in a proportional increase in therapeutic efficacy, which raised concerns about excessive Ag⁺ exposure. Therefore, SOT2 was considered the most suitable formulation for subsequent in vivo evaluations.

To establish a localized infection, 100 µL of MRSA (1 × 10⁸ CFU) was applied topically to the wound bed. After 24 h, treatments were administered, including PBS (control), sterile Tegaderm dressing, SO hydrogel (100 µL), and SOT2 hydrogel (100 µL). Wound closure was monitored at predetermined time points through digital photography, and wound areas were quantitatively analyzed using ImageJ software.

Histological and immunohistochemical analysis

At the designated time points, wound tissues were harvested and fixed with 4% paraformaldehyde. Tissue sections were stained with H&E and Masson’s trichrome to assess tissue morphology and collagen deposition. Immunofluorescence staining was also performed to evaluate oxidative stress (DHE), angiogenesis (CD31 and α-SMA), and macrophage polarization (CD86 and CD206). Fluorescence images were analyzed to quantitatively assess inflammatory responses and neovascularization.

Statistical analysis

Data are expressed as the mean ± standard deviation. Unless otherwise stated, n denotes the number of independent biological replicates. Statistical analysis was performed using GraphPad Prism 10.0 and Origin 9.0. For comparisons between two groups, one-way analysis of variance (ANOVA) was applied. In cases involving more than two groups, one-way ANOVA followed by Tukey’s post hoc test was used. A p-value below 0.05 was considered to indicate statistical significance.

Author contributions

Zhongwu Bei designed the study, developed the methodology, and wrote the original draft. Zhongwu Bei, Qi Tong, Yun Yang, and Lin Ye conducted experiments, curated data, validated results, performed visualization, and revised the manuscript. Shiyu Liang, Jianan Li, and Jie Liu assisted with software implementation and data processing. Zhongwu Bei, Shiyu Liang, and Jianan Li contributed to resources and investigation. Xuyue Liang and Wen Chen participated in methodology and visualization. Zhongwu Bei, Shiyu Liang supervised the revision. Wen Chen, Wenwen Liu, and Xiaorui Yu coordinated the research and supervised the project. Zhongwu Bei, Bi Shi, and Shiyu Liang contributed to resources/investigation, revised the manuscript, and perform the supplementary experiments. Zhongwu Bei, Lin Ye, and Zhiyong Qian performed formal analysis, acquired funding, and revised the manuscript. Zhongwu Bei, Lin Ye, and Zhiyong Qian coordinated the research, supervised the project, revised the manuscript, and acquired the funding.

Funding

The authors gratefully acknowledge financial support from the National Natural Science Foundation of China (32530059, U21A20417), the Nature Science Foundation of Sichuan Province (2024NSFSC0046), the Postdoctoral Fellowship Program of CPSF under Grant Number GZC20241165, the Sichuan Natural Science Foundation for Young Scholars (2024NSFSC1724), and the “1·3·5” Project for Disciplines of Excellence, West China Hospital, Sichuan University (ZYGD24003).

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The Animal Ethics Committee of the State Key Laboratory of Biotherapy, Sichuan University, approved all animal experiments.

Competing interests

The authors declare no competing interests.