Sustained copper-releasing adhesive hydrogel patch promotes optimized scarless tongue wound healing via antioxidative, angiogenic and antifibrotic synergy

Abstract

Tongue injuries are among the most common acute soft tissue injuries encountered in oral clinical practice. Suturing, as the conventional treatment, often leads to tissue distortion and tension-induced collagen over-deposition, resulting in scarring and restricted tongue mobility. Sutures are also prone to microbial accumulation and provoke inflammation as foreign bodies. In this study, a hydrogel patch specifically designed for the tongue's moist, frequently moving, and richly vascularized environment was developed for better healing of tongue injuries. Briefly, a chitosan–polyacrylic acid–tannic acid hydrogel matrix is loaded with Cu-MOF to form a CPTCu hydrogel, ensuring stable binding and controlled release of copper, while maintaining mechanical compliance with tongue tissue for continuous protection. The CPTCu hydrogel exhibited excellent biocompatibility, achieving a free radical scavenging rate of 73.6% and a bactericidal rate of over 99%. Compared with suturing, the CPTCu hydrogel significantly reduced the collagen volume by 35.9% and enhanced angiogenesis by 103.2%, and also effectively promoted regeneration of local muscle fibers in a rat tongue wound model. These results demonstrate that the CPTCu hydrogel is a promising candidate for optimized scarless tongue healing.

---

1. Introduction

Acute tongue injuries caused by accidental biting, sports trauma, or other accidents are frequently confronted in oral clinical practice, especially among children.^1,2 Due to the dense vascular network of the tongue, these injuries typically provoke rapid bleeding and swelling, posing risks of airway obstruction.³ Current standard clinical management relies primarily on suturing to achieve rapid hemostasis and primary wound closure.⁴ Suturing provides strong fixation of wound edges and precise anatomical tissue alignment, with clinical versatility and highly predictable treatment results. However, it has several inherent limitations that may compromise long-term healing outcomes. The mechanical tension it creates, amplified by the tongue's continuous movements, could lead to local tissue distortion and excessive collagen deposition, resulting in scar formation and impaired mobility and functionality.^5–7 The tongue's soft, delicate structure also makes it susceptible to secondary tearing under suture stress. The gaps between sutures and tissue may disrupt epithelial regeneration through contact inhibition.⁸ Moreover, the presence of sutures inevitably triggers inflammatory responses and fosters microbial accumulation, both of which can delay wound healing.^9,10 Additional drawbacks include the need for local anesthesia, patient cooperation, and the inability to fully isolate the wound from external irritants. Collectively, these limitations highlight the need for a noninvasive, easy-to-apply, and protective wound dressing tailored to the unique physiological characteristics of the tongue, which could serve as an effective alternative to traditional suturing and improve the overall healing quality of tongue injuries.

Designing a specialized wound dressing for the tongue is highly challenging, as it should not only address the difficulties of intraoral applications, but also consider the tongue's distinctive anatomical and functional features. The tongue resides in the saliva-rich oral environment, is richly innervated and vascularized, and functions as a highly movable muscular organ with variable mechanical properties.¹¹ Therefore, in addition to the robust wet tissue adhesion and healing-promoting properties required for conventional intraoral wound dressings, a tongue-specific dressing must also provide efficient hemostasis and mechanical flexibility to accommodate tongue movements.¹² Guided by these requirements, we developed a hydrogel patch using chitosan, polyacrylic acid, and tannic acid as the matrix, with Cu-MOF integrated as a functional component.

Chitosan is a biocompatible natural polysaccharide with intrinsic hemostatic, antibacterial, and mucoadhesive properties. These biofunctions are mainly attributed to its protonated amino groups that interact with blood components, bacterial membranes, and negatively charged mucosal surfaces.^13–15 However, native chitosan has limited mechanical strength, which can be effectively enhanced by blending or crosslinking with synthetic polymers.¹⁶ Polyacrylic acid (PAA) is an FDA-approved, low-cost, and nontoxic polymer that has been widely used in commercial wound dressings.¹⁷ Combining chitosan with PAA yields flexible hydrogels with favorable biocompatibility and strong wet-tissue adhesion.^18,19 Tannic acid (TA), a natural plant polyphenol and multifunctional crosslinker, interacts with the hydrogel network and tissue surfaces via abundant mechanisms. Incorporating TA increases the crosslinking density of the hydrogel network and further enhances its tissue adhesive, antibacterial, antioxidant, and hemostatic performances.^20,21

To further augment the bioactivity and pro-healing potential of the chitosan-PAA/TA (CPT) hydrogel, we focused on copper (Cu) as a functional element. Compared with metals like zinc and silver, Cu exhibits superior antibacterial activity alongside high biocompatibility.²² It also plays pivotal roles in angiogenesis, cell migration, and extracellular matrix (ECM) remodeling, making it highly promising for wound healing applications.²³ Despite these benefits, excessive Cu could induce cytotoxicity, necessitating precise dosage control.^24,25 For the safe incorporation of Cu, we employed a Cu-based metal–organic framework (Cu-MOF) as a controlled source of Cu. MOFs are crystalline coordination materials built from metal ions and organic ligands, forming ordered clusters with tunable degradability and excellent capability of metal-ion storage and sustained release.^26–28 In our previous work, we synthesized L-Asp-Cu MOF, which demonstrated excellent biocompatibility, high structural stability, and a controlled, slow-release profile for Cu.²⁹ Notably, L-Asp also contributes to cellular proliferation and redox homeostasis.³⁰ In the present study, we utilized L-Asp-Cu MOF as the Cu source, incorporating Cu into the hydrogel by immersing the preformed CPT hydrogel in an L-Asp-Cu MOF suspension to obtain a Cu-ligand–hydrogel composite (CPTCu). Compared with alternative loading strategies, this immersion method is gentler, simpler, and more efficient, capable of preserving the structural integrity and original mechanical properties of the hydrogel, thus facilitating product transformation and batch production.

The resulting CPTCu hydrogel retains softness and stretchability, while exhibiting enhanced wet-tissue adhesion, improved hemostatic efficacy, and robust antibacterial, antioxidant, pro-migratory, and pro-angiogenic activities—all tailored to tongue tissue applications. It retains the key benefits of suturing, including firm tissue fixation, fast hemostasis, and broad applicability, while effectively addressing the major limitations. In a rat tongue wound model, it achieved significantly reduced collagen deposition and minimized scar formation compared with suturing, resulting in excellent healing outcomes that closely resembled the native tissue architecture. To conclude, this study presents a noninvasive and scarless alternative to suturing for tongue wound treatment, providing new insights for future optimization of tongue wound management strategies.

2. Experimental section

2.1 Materials

Chitosan (degree of deacetylation ≥95%, viscosity 100–200 mPa s) was purchased from Macklin Biochemical Technology Co., Ltd (Shanghai, China). Acrylic acid, N-succinimidyl acrylate, tannic acid and lithium phenyl (2,4,6-trimethylbenzoyl) phosphinate (LAP) were purchased from Aladdin Scientific (Shanghai, China). L-Asp and copper acetate monohydrate (CuCH₃COOH·H₂O, 99%) were purchased from Acmec Biochemical Co., Ltd (Shanghai, China). Artificial saliva was purchased from Yuanye Bio-Technology Co., Ltd (Shanghai, China). Cell Counting Kit-8 (CCK-8) was purchased from ApexBio (USA). Calcein-AM/PI live/dead cell staining kit and DPPH free radical scavenging assay kit were purchased from Solarbio Science & Technology Co., Ltd (Beijing, China). Reactive oxygen species (ROS) assay kit was purchased from Beyotime Biotechnology (Shanghai, China). Matrigel was purchased from Corning (USA). Phalloidin-TRITC, total RNA extraction kit, cDNA synthesis kit and qPCR SYBR Green Master Mix was purchased from Yeasen Biotechnology Co., Ltd (Shanghai, China).

2.2 Preparation of the CPTCu hydrogel

To prepare the CS-PAA (CP) hydrogel, 200 mg of chitosan was added to 3 mL of acrylic acid and dispersed evenly. 7 mL of ultrapure water was added into the solution, followed by the addition of 100 mg of N-succinimidyl acrylate and 20 mg of lithium phenyl (2,4,6-trimethylbenzoyl) phosphinate (LAP) as the photoinitiator. The above mixture was stirred at room temperature for 1 h and cured under UV light (365 nm, 10 W) for 20 min. The obtained hydrogel was thoroughly washed with PBS 3 times for further use. For the preparation of the CS-PAA/TA (CPT) hydrogel, 7 mL of ultrapure water was replaced by the same volume of 1% (w/v) TA solution.

L-Asp-Cu MOF was synthesized according to the method reported in our previous study.²⁹ To fabricate CS-PAA/L-Asp-Cu MOF (CPCu) and CS-PAA/TA/L-Asp-Cu MOF (CPTCu), L-Asp-Cu MOF was ultrasonically dispersed in ultrapure water to prepare uniform suspensions with concentrations of 0.1%, 0.2%, and 0.3% (w/v). The CP and CPT hydrogels were immersed in L-Asp-Cu MOF suspensions of different concentrations at 10 times their own volume under constant shaking for 2 h. Subsequently, the hydrogels were washed 3 times for further use. Hydrogels prepared by immersing CP and CPT in 0.1%, 0.2%, and 0.3% (w/v) L-Asp-Cu MOF suspensions were named CPCu-1, CPCu-2, CPCu-3, CPTCu-1, CPTCu-2 and CPTCu-3, respectively.

2.3 Characterization of L-Asp-Cu MOFs

The morphology and particle size of L-Asp-Cu MOFs were observed using scanning electron microscopy (SEM, Zeiss G300, Germany). The crystal structure was identified using an X-ray diffractometer (XRD, PANalytical, Netherlands). The functional groups of L-Asp-Cu MOF particles were detected using a Fourier transform infrared spectrophotometer (FTIR, SHIMADZU IRTrace-100, Japan) with the scan range of 400–4000 cm⁻¹.

2.4 Characterization of the CPTCu hydrogel

2.4.1 Structural characterization of the CPTCu hydrogel. FTIR was employed to identify the functional groups and molecular interactions within the hydrogels. The valence state and coordination environment of Cu in the hydrogels were characterized by X-ray photoelectron spectroscopy (XPS, Thermo Scientific Nexsa, USA). The cross-sectional microstructure of the hydrogels was examined by Cryo-SEM (Hitachi SU8100, Japan) and element mapping analysis was performed.

2.4.2 Water contact angle. The surfaces of the hydrogel samples in each group were cleaned and dried. Droplets of ultrapure water were dispensed onto the hydrogel surface using a microsyringe. Photographs were taken at the moment the water droplets contacted the surface. The test was conducted at 25 ± 1 °C and 50 ± 5% relative humidity.

2.4.3 Swelling ratio. 30 mg hydrogel of each group was immersed in 30 mL artificial saliva at 37 °C. At 2 h, 4 h, 6 h, 8 h, 10 h, 12 h, 24 h, 48 h and 72 h, the hydrogels were taken out, and their surfaces were dried and weighed. The swelling ratio the hydrogels was calculated as follows:
W_t represents the weight of hydrogels recorded at different time points, and W₀ represents the initial weight of the hydrogels.

2.4.4 Copper ion release profile of the CPTCu hydrogel. The release of copper ions from CPCu and CPTCu was analyzed by immersing 30 mg of each hydrogel in 30 mL of PBS (pH = 7.4) under constant shaking at 37 °C, 100 rpm. The solution was collected at 5 h, 10 h, 1 d, 3 d, 5 d, and 7 d, and the released copper ion was quantified using an inductively coupled plasma optical emission spectrometer (ICP-OES, PE Avio200, USA).

2.4.5 Mechanical experiments. Tensile tests and lap-shear tests of the hydrogels were performed with a universal testing machine (SHIMADZU AGX-V, Japan). For tensile tests, standard tensile samples were prepared and stretched at the speed of 30 mm min⁻¹ until breaking. For lap-shear tests, hydrogel samples of 10 mm × 10 mm × 1 mm in size were prepared. Strips of porcine tongue epithelium (5 cm × 1 cm) were trimmed from the dorsal surface of the tongue. The hydrogel samples were placed between two epithelial surfaces, gently pressed for 10 s, and then the tissue samples were stretched at a speed of 1 mm s⁻¹ until adhesive failure occurred. The experiment was repeated 3 times for each group.

2.4.6 Adhesive performance of CPTCu. Hydrogel samples of 10 mm × 10 mm × 1 mm in size were prepared. The heart, liver, spleen, lung, and kidney tissue were excised from SD rats and rinsed with normal saline. The hydrogel samples were gently pressed onto the tissue surfaces for 10 s to achieve adhesion. To demonstrate the universal adhesiveness of the hydrogels to different substrates, the same procedure was applied to glass, metal, and plastic surfaces.

2.5 Hemocompatibility of CPTCu

Hydrogel extracts were prepared by immersing 10 mg dried hydrogel in 1 mL normal saline at 37 °C for 24 h. Freshly collected whole blood of SD rat was washed with normal saline and centrifuged at 3000 rpm for 5 min and repeated 5 times to obtain purified erythrocytes. The concentrated erythrocytes were resuspended in normal saline to a final concentration of 4% (v/v). Equal volume of hydrogel extracts and erythrocyte suspension were gently mixed and incubated at 37 °C for 1 h. The mixture was then centrifuged at 1500 rpm for 15 min, and the absorbance of the supernatant at 540 nm was measured. Normal saline and Triton X-100 were used as the negative and positive control, respectively. The hemolysis rate was calculated as follows:

2.6 Cytocompatibility of CPTCu

2.6.1 Extraction of primary tongue fibroblasts. Primary fibroblasts from rat tongue tissue were extracted using the tissue block method. The rat tongue was excised, trimmed and washed multiple times with HBSS until no blood residues remained. It was then immersed in culture medium containing 10% fetal bovine serum (FBS) and 10% penicillin–streptomycin (P–S) for 5 min. The tongue tissue was finely minced into approximately 1 mm³ pieces using sharp scissors, washed several times with HBSS, and then seeded into culture flasks for even distribution. The flasks were incubated upside down overnight for tissue adherence, and fresh culture medium was added the next day to cover the tissue blocks. The medium was replaced every two days. After the cells had migrated out and confluency was reached, the tissue blocks were removed, and the cells were treated with 0.25% trypsin-EDTA for 3 min. Fibroblasts were observed to detach, while epithelial cells remained adhered. The collected cells were passaged twice cultured with high glucose DMEM containing 10% FBS and 1% P–S to purify the fibroblasts for further culturing.

2.6.2 Cell viability and morphology. Filtered hydrogel extracts (1 mg mL⁻¹) obtained by incubation in culture medium at 37 °C for 24 h were used for in vitro cell experiments. Rat primary tongue fibroblasts were cultured with hydrogel extracts for 72 h. To assess cell viability, CCK-8 assay was performed at 24 h, 48 h and 72 h. The culture medium was replaced with 10% CCK-8 solution and incubated 37 °C for 1 h, then the OD value at 450 nm was measured to calculate cell viability (n = 5). At 24 h and 72 h, live/dead cell staining was performed using calcein-AM and PI, and fluorescence images were acquired with a fluorescence microscope. To further observe cell morphology after 72 h of treatment with hydrogel extracts, cells were fixed with 4% paraformaldehyde and subsequently stained with phalloidin-TRITC and DAPI for fluorescence imaging (n = 4).

2.7 Animal care

Sprague-Dawley (SD) rats (male, 6 weeks old) were housed in a specific pathogen-free (SPF) animal facility under standard conditions and a 12 h light/dark cycle. Animals had free access to food and water, and were acclimated to the environment for over one week before experiments were conducted. All experiments were performed in compliance with the Chinese national standard of general code of animal welfare (GB/T 42011-2022) and adhered to the ‘Animal Research: Reporting In Vivo Experiments’ (ARRIVE) 2.0 guidelines. All experiments received approval from the Ethics Committee of West China Hospital of Stomatology, Sichuan University (No. WCHSIRB-AT-2025-563).

2.8 Hemostasis experiments

2.8.1 In vivo hemostasis. Different groups of hydrogel patches of 10 mm × 10 mm × 1 mm were prepared. A rat liver hemorrhage model was used to test the hemostasis ability. Animals were anesthetized and an abdominal incision was made to expose the liver. A piece of filter paper was placed beneath the liver, and a puncture was created using a 4 mm biopsy punch. The control group received no treatment, while the experimental groups were treated with CP, CPCu, CPT and CPTCu patches, respectively. The hemostasis time was recorded. After complete hemostasis, the filter papers were weighed to determine the amount of blood loss (n = 3).

2.8.2 In vitro hemostasis. Rat whole blood and four groups of hydrogels were prewarmed at 37 °C. 30 mg of all hydrogels was placed into centrifuge tubes, and 100 µL recalcified whole blood (0.2 M CaCl₂, 10 mM in the blood) was added to the hydrogel samples. For the blank control, only 100 µL recalcified whole blood was added to the centrifuge tube. After incubation at 37 °C for 10 min, 10 mL of DI water was gently added to remove unbound blood, and the absorbance of the supernatant at 540 nm was measured using a spectrophotometer (n = 4). The blood-clotting index (BCI) was measured as follows:

2.9 In vitro antioxidant ability

2.9.1 DPPH free radical scavenging assay. To prepare the hydrogel extract, 10 mg of hydrogel samples were mixed with 1 mL of extraction solution and incubated in a 40 °C water bath for 30 min. The resulting hydrogel extract was then mixed with DPPH working solution in an EP tube (n = 3). The extraction solution was used to replace the hydrogel extract as the negative control, 10 mg mL⁻¹ vitamin C solution was used to replace the hydrogel extract as the positive control, and ethanol was used to replace the working solution as the blank control. The tubes were vortexed and incubated at room temperature for 30 min protected from light. The absorbance at 515 nm was then measured. The DPPH free radical scavenging rate was calculated as follows:

2.9.2 Intracellular reactive oxidant species (ROS) assay. HUVEC and primary tongue fibroblasts were previously cultured in 24-well plates for 2 days with hydrogel extracts. Cells were incubated with the DCFH-DA probe at 37 °C for 30 min to allow intracellular loading. After incubation, the cells were washed three times with serum-free medium to remove excess DCFH-DA. Subsequently, 600 µM hydrogen peroxide solution was added, and the cells were incubated for 2 h to induce oxidative stress. Fluorescence images were then captured using a fluorescence microscope (n = 4).

2.10 In vitro cell migration assay

A scratch assay of primary tongue fibroblasts was conducted to evaluate cell migration ability. Cells were seeded in 24-well plates and cultured until reaching over 90% confluence. A scratch was created at the center of each well using a 200 µL pipette tip, followed by washing three times with serum-free medium to remove the detached cells. The cells were then incubated for 48 h with hydrogel extracts containing 1% serum and 1% P–S. Images were captured at 0 h, 24 h, and 48 h, and the wound closure rate of 24 h and 48 h was calculated (n = 4).

2.11 In vitro tube formation assay

The angiogenic potential of HUVECs treated with hydrogel extracts was assessed by tube formation assay. Matrigel was evenly distributed in pre-cooled 24-well plates and solidified at 37 °C for 30 min. Cells resuspended with different hydrogel extracts were seeded onto the Matrigel and incubated for 6 h. Images were captured at 6 h and the number of junctions and tubule length were quantified using the Angiogenesis Analyzer plugin in ImageJ software (n = 4).

2.12 RT-qPCR analysis

The expression of angiogenesis-related genes of HUVECs and fibrosis-related genes of tongue fibroblasts were analyzed by real-time quantitative polymerase chain reaction (RT-qPCR). The HUVECs and tongue fibroblasts were cultured with complete medium or hydrogel extracts for 48 h, and total RNA was extracted and reverse-transcribed into cDNA. Quantitative PCR was performed using SYBR Green Master Mix following the manufacturer's instructions. Gene expression levels were normalized to the housekeeping gene beta actin. The relative expression of target genes was calculated using the ΔΔCt method. Primer sequences for detected genes are provided in Table S1.

2.13 Transcriptome sequencing of primary tongue fibroblasts

Tongue fibroblasts were cultured with complete medium and CPTCu extract for 48 h, respectively (n = 4 per group). Total RNA was extracted with TRIzol reagent (ABclonal, China). RNA purity and quantification were measured with a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). RNA integrity was evaluated using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Sequencing libraries were constructed with the VAHTS Universal V10 RNA-seq Library Prep Kit (Premixed Version) according to the manufacturer's guidelines. Transcriptome sequencing and analytical services were provided by OE Biotech Co., Ltd (Shanghai, China).

The libraries were subjected to sequencing on the Illumina Novaseq 6000 platform, yielding 150 bp paired-end reads. The lean reads were mapped to the reference genome via HISAT2. FPKM of each gene was calculated and HTSeq-count was employed to obtain read counts for each gene. PCA analysis was performed with R (v 3.2.0) to assess the biological duplication of samples. Differential expression analysis was performed using the DESeq2, and the q value <0.05 and fold change >2 or <0.5 was set as the threshold for significantly differential expression genes (DEGs). A volcano plot and a clustered heatmap were generated to visualize up-regulated and down-regulated DEGs. Based on the hypergeometric distribution, GO and KEGG pathway enrichment analysis of DEGs were conducted for significantly enriched terms using R (v 3.2.0). Gene Set Enrichment Analysis (GSEA) was performed with the GSEA software.

2.14 In vitro antibacterial ability

2.14.1 Counting assay of colony forming units (CFUs). 10 mg of each hydrogel was immersed in 10 mL of Luria-Bertani (LB) broth and incubated at 37 °C for 24 h to obtain the corresponding hydrogel extract medium. For bacterial culture, Escherichia coli and Staphylococcus aureus were pre-cultured in 10 mL of LB broth at 37 °C for 24 h, and then diluted to a concentration of 1 × 10⁴ CFU mL⁻¹. A 24-well plate was prepared, and 500 µL of the diluted bacterial suspension was added to each well, followed by the addition of 500 µL of each hydrogel extract medium. The plate was then incubated at 37 °C overnight. After incubation, the bacteria suspension was serially diluted to an appropriate concentration, and 100 µL of the diluted bacterial suspension was plated on LB agar plates, which were incubated at 37 °C overnight. CFUs were counted using ImageJ software and the plates were photographed (n = 3). The bacterial survival rate was calculated as follows:

2.14.2 Live/dead bacterial staining. SYTO 9/PI dual fluorescence dye was applied according to the manufacturer's instructions to stain the above mentioned co-cultured bacteria. After centrifuging the bacterial suspension and removing the culture medium, the bacteria were resuspended in sterile normal saline. Then, 1.5 µL of SYTO 9/PI staining solution was added to 1 mL of resuspended bacterial suspension and incubated at 37 °C for 15 min. The bacterial fluorescence was then observed and captured using an inverted fluorescence microscope.

2.14.3 SEM morphological analysis. The bacterial suspension co-cultured with hydrogel extracts was fixed overnight with 2.5% glutaraldehyde solution. The next day, the fixative was removed, and the samples were washed three times with PBS. The samples were then dehydrated with gradient ethanol (30%, 50%, 70%, 80%, 90%, 95%, 100%) for 10 min each. After dehydration, the samples were gold-coated and observed under SEM.

2.15 In vivo tongue wound healing experiments

6-week-old male SD rats were acclimated for one week under standard housing conditions before being included in the experiment. The animals were randomly divided into three groups (n = 5). Under general anesthesia, a circular defect (3 mm in diameter, 2 mm in depth) was created in the central dorsal region of the tongue using a biopsy punch. In the blank group, the wound was left untreated. In the suture group, the wound was closed with 4-0 silk sutures. In the CPTCu group, a 0.5 mm-thick CPTCu hydrogel patch was trimmed to fit the wound area and applied to cover the defect.

Macroscopic images of the wounds were captured on days 0, 2, 4, 6, and 8. The hydrogel patch was replenished as needed during the experiment. Wound healing rates at each time point were quantified using ImageJ software. On day 8 post-surgery, after final photographic documentation, all animals were euthanized. Tongue tissues were collected, fixed in 4% paraformaldehyde, embedded in paraffin, and sectioned at a thickness of 5 µm for histological analysis, including hematoxylin and eosin (H&E) staining, Masson staining, and multiplex immunohistochemical (mIHC) staining. In addition, the heart, liver, spleen, lung, kidney, stomach, and intestine from each rat were collected, paraffin-embedded, sectioned and stained with H&E to observe the tissue microstructure and evaluate the systemic biocompatibility of the CPTCu hydrogel.

2.16 Statistical analysis

All quantitative data are presented as the mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism 10. Comparisons between two groups were conducted using the Student's t-test, while comparisons among multiple groups were analyzed by one-way analysis of variance (ANOVA) (*P < 0.05, **P < 0.01, and ***P < 0.001).

3. Results and discussion

3.1 Synthesis and characterization of L-Asp-Cu MOFs and CPTCu hydrogels

A schematic diagram of L-Asp-Cu MOF and CPTCu hydrogel synthesis is shown in Fig. 1A. L-Asp-Cu MOF was synthesized via a hydrothermal method with copper acetate and L-Asp based on our previous work.²⁹ The freeze-dried product exhibited a light blue appearance and was ground into powder. SEM images of raw L-Asp-Cu MOF powders exhibited uniformly aggregated rod-like particles of varying lengths (Fig. 1B). XRD analysis of L-Asp-Cu MOF revealed the same pattern and characteristic diffraction peaks as in previous reports, confirming the successful formation of the crystalline Cu-MOF (Fig. S1).^29,31

Fig. 1 Synthesis and characterization of the L-Asp-Cu MOF and CPTCu hydrogel. (A) Synthetic processes of the L-Asp-Cu MOF and CPTCu hydrogel. (B) SEM image of raw L-Asp-Cu MOF powders. (C) FTIR spectra of chitosan (CS), tannic acid (TA), CP and CPT. (D) Macroscopic changes in the color of CPCu and CPTCu hydrogels after Cu-MOF immersion. (E) Cryo-SEM images of cross-sectional microstructures of CP, CPCu, CPT, and CPTCu. (F) Element mapping of the cross-section of CPTCu. (G) XPS full spectrum of CPTCu and Cu 2p spectra of L-Asp-Cu MOF and CPTCu.

The chitosan-polyacrylic acid (CS-PAA, CP) hydrogel matrix was formed by LAP-initiated free radical polymerization of acrylic acid monomers and AA-NHS. The reactive NHS esters in the system could undergo amidation with free –NH₂ groups of chitosan and enhance chemical crosslinking density. On this basis, CPT hydrogel was constructed by incorporating tannic acid (TA) to introduce abundant phenolic hydroxyl groups, which can form multiple hydrogen bonds and electrostatic interactions with the amino groups of chitosan and the carboxyl groups of PAA. The final hydrogel network features both covalent and non-covalent multilevel crosslinking. FTIR spectra confirmed the successful synthesis of CP and CPT hydrogels (Fig. 1C). Both CP and CPT exhibited a broad absorption band from 3600 to 2500 cm⁻¹, attributable to the stretching vibrations of O–H and N–H. The strong peak at 1700 cm⁻¹ corresponded to carboxyl C=O stretching. The peak at 1069 cm⁻¹ was primarily assigned to the asymmetric stretching of C–O–C originating from chitosan.³² In CPT, the characteristic peak at 1320 cm⁻¹ was associated with the in-plane bending of phenolic O–H and the C–O vibration of the aromatic ring, indicating successful incorporation of TA. For CP, the absorption bands at 1630 and 1540 cm⁻¹ corresponded to C=O stretching (amide I) and N–H bending (amide II).³³ In CPT, the amide bands shifted to 1615 and 1520 cm⁻¹, suggesting enhanced hydrogen-bonding interactions following the introduction of TA.³⁴

Next, CP and CPT were loaded with L-Asp-Cu MOF via immersion in a L-Asp-Cu MOF suspension. Upon ultrasonic dispersion of L-Asp-Cu MOF in water, the original micron-scale rod-like crystals were sheared into nano-scale fragments and led to the partial liberation of Cu ions. During the immersion process, the nano-scale Cu-MOF fragments and the released Cu ions diffused into the hydrogel network driven by a concentration gradient and were subsequently immobilized through the formation of new coordination interactions with the abundant amino, carboxyl and phenolic hydroxyl groups in the hydrogel, resulting in a stable Cu-ligand-hydrogel composite structure.³⁵ As shown in Fig. 1D, after the immersion process, Cu-MOF suspension transitioned from an opaque light blue to a transparent state, while the CPT hydrogel turned from light-yellow to blue-green, indicating that Cu-MOF fragments and Cu ions shifted into the hydrogel. Cryo-SEM images of hydrated hydrogel cross-sections exhibited a porous network with irregular cavities, indicating a dense yet flexible crosslinked polymer framework (Fig. 1E). It could also be observed that the incorporation of Cu-MOF did not alter the microstructures of the hydrogel matrix.

The advantages of using a Cu-MOF suspension immersion strategy to load Cu ions, rather than directly adding Cu salts or Cu-MOF to hydrogel precursor solution, or immersing the hydrogel in Cu salt solutions, are stated as follows. First, direct incorporation of Cu salts into the hydrogel precursor solution disrupts polymerization and induces undesirable coordination or precipitation with tannic acid under acidic conditions, thereby compromising network formation. On the other hand, direct incorporation of Cu-MOF into the precursor solution may lead to structural damage of the MOF during photo-initiated gelation. In contrast, the immersion strategy is mild and preserves the integrity of the hydrogel structures. Second, when preformed hydrogels are immersed in Cu salt solutions, a large proportion of Cu ions is introduced in the form of unbound free ions, resulting in uncontrolled Cu release during application and increased the risk of cytotoxicity. In contrast, Cu-MOF serves as a highly stable Cu source. Only a small proportion of Cu ions is liberated under ultrasonication, which immediately form secondary coordination with functional groups in the hydrogel matrix. As a result, Cu ions predominantly present in a bound state rather than as free ions, enabling long-term controlled and sustained release with markedly reduced biological toxicity.

Energy-dispersive X-ray spectroscopy (EDS) mapping revealed a uniform distribution of C, O, and Cu in the CPTCu hydrogel, reflecting the uniform Cu-ligand-hydrogel composite architecture (Fig. 1F). Further analysis of the elemental composition and Cu valence states of the hydrogel was performed using XPS (Fig. 1G). The total Cu content of CPTCu was approximately 0.12 wt%. The Cu 2p spectra of L-Asp-Cu MOF revealed the presence of both Cu⁺ and Cu²⁺ states, while only Cu⁺ peaks were observed in that of CPTCu.³⁶ This confirmed that the coordination state of Cu ions had changed upon incorporation into the hydrogel. The hydrogel system inherently exhibits mild reducing capability due to the presence of TA and provides a high density of flexible coordination sites. A fraction of Cu was reduced and transitioned from the original coordination environment within the MOF to a secondary coordination environment of the MOF–hydrogel complex and stabilized in a Cu⁺-coordinated state, achieving a new Cu-organic ligand-hydrogel equilibrium.

To simulate an intraoral environment, the swelling behavior of the hydrogels was determined by immersing them in artificial saliva at 37 °C. As shown in Fig. 2A, the hydrogels reached swelling equilibrium after 24 h. The average swelling ratios of CP, CPCu, CPT, and CPTCu at 72 h were 325.9%, 235.9%, 305.6%, and 233.7%, respectively. The incorporation of TA exhibited no significant influence on swelling capacity, while the introduction of Cu ions markedly reduced the swelling ratio. This might be because of the dynamic and partially reversible nature of TA-mediated crosslinking in ion-rich artificial saliva. Competing ions weaken its hydrogen bonding and coordination interactions, while the intrinsic hydrophilicity of TA further offsets its swelling-restricting effect. In contrast, the introduction of Cu ions leads to the formation of more stable multidentate coordination, which substantially increased network intensity. It was also observed that the swelling profiles displayed a rise-fall transition within the first 24 h, which could be explained by the overshooting effect: initially, unprotonated carboxyl ions (–COO⁻) induced continuous expansion of the hydrogel network. As the proportion of protonated carboxyl groups (–COOH) increased in the weakly acidic artificial saliva, hydrogen bond crosslinking intensified, causing the swelling ratio to decline after reaching maximum and eventually stabilize at equilibrium.^37,38

Fig. 2 Physiochemical and mechanical properties of the CPTCu hydrogel. (A) The swelling behavior of CP, CPCu, CPT and CPTCu hydrogels in artificial saliva. n = 3 per group. (B) Cumulative release of Cu ions from the L-Asp-Cu MOF, CPCu and CPTCu. (C) Water contact angle of CP, CPCu, CPT and CPTCu. n = 3 per group. (D) Tensile tests and representative tensile stress–strain curves of each group. n = 3 per group. (E) Schematic diagram of the adhesive mechanisms of the CPTCu hydrogel to the tongue tissue. (F) Representative lap shear stress-displacement curves and lap shear strength of each group. n = 3 per group. (G) Adhesion and good compliance of the CPTCu patch on the rat dorsal tongue surface. (H) Universal adhesive ability of CPTCu to plastic, glass and metal.

To investigate the Cu release behavior of the hydrogels, the ion-release profiles of CPCu and CPTCu (100 µL) were measured and compared with those of L-Asp-Cu MOF powder contained in Cu-MOF suspension used for preparing an equivalent amount of hydrogel (2 mg) (Fig. 2B). Due to the dynamic equilibrium between Cu dissociation and re-coordination, L-Asp-Cu MOF exhibited an initial Cu release followed by a slight decline, ultimately reaching a plateau.³⁹ On the other hand, CPCu and CPTCu exhibited a gradual, sustained release profile. This is attributed to the tight anchoring of Cu ions by both the hydrogel network and L-Asp ligands, which collectively restricted Cu mobility.

The hydrophilicity of wound dressings is critical, as it can significantly influence cell adhesion and proliferation.⁴⁰ The water contact angles of CP, CPCu, CPT, and CPTCu were measured to be 51.9 ± 0.7°, 49.1 ± 4.3°, 49.8 ± 3.6°, and 47.8 ± 3.9°, respectively (Fig. 2C). The hydrogels contain abundant hydrophilic functional groups, which allow water molecules to rapidly spread and infiltrate upon contact, conferring excellent hydrophilicity and surface wettability.

3.2 Mechanical and adhesive performance of the CPTCu hydrogel

The tongue is characterized by its soft nature and high mobility. Therefore, wound dressings designed for the tongue must have mechanical properties that closely match those of the tongue, allowing them to deform synchronously with the tongue's movements while maintaining stable adhesion to provide continuous protection.⁴¹ Tensile tests were performed using a universal testing machine. All hydrogels exhibited excellent stretchability (>300% strain), with a maximum elongation of approximately 8 times their original length (Fig. 2D). The tensile strength of CP, CPCu, CPT and CPTCu showed a progressive increase, which is in agreement with FTIR and XPS results, further corroborating that the incorporation of TA and Cu ions enhances the hydrogel strength by increasing crosslinking density. The Young's modulus was calculated by linear fitting of the stress–strain curves within the 0–10% strain region. Human tongue tissue has been reported to exhibit an apparent Young's modulus of 12.2 ± 4.2 kPa in the relaxed state and 122.5 ± 58.5 kPa under tension.^42,43 CP, CPCu, CPT and CPTCu displayed a modulus of about 20.7 kPa, 35.7 kPa, 31.4 kPa and 37 kPa, respectively, which all fall within the functional mechanical range of tongue tissue. Such modulus provides sufficient mechanical support and fatigue resistance while retaining softness to avoid discomfort or stress concentration. In addition, hydrogels with stiffness slightly higher than that of resting soft tissue have been proved to facilitate fibroblast alignment and ECM remodeling.⁴⁴

Lap shear tests were performed with porcine tongue epithelium strips. The average lap shear strength of CP, CPCu, CPT and CPTCu was 106.5 kPa, 86.6 kPa, 113.8 kPa and 96.7 kPa, respectively (Fig. 2F). CPT hydrogel adheres firmly to moist tissues through a synergistic multimodal mechanism dedicated by covalent bonding from NHS esters, electrostatic interactions from chitosan, and π–π stacking and hydrogen bonding from TA and PAA chains, therefore exhibiting the highest adhesive capacity (Fig. 2E).¹⁹ The incorporation of Cu ions partially occupied available adhesion sites, leading to a slight reduction in adhesive strength. Nevertheless, the adhesive strength of CPTCu is sufficient to ensure its robust adhesion to the tongue surface. As shown in Fig. 2G, the CPTCu patch remained firmly attached to the tongue during curling, stretching and twisting, deforming synchronously with the tissue. It could also adhere tightly to major visceral organs (heart, liver, spleen, lung, and kidney), as well as plastic, glass and metal substrates, highlighting its universal adhesion capabilities (Fig. 2H and Fig. S2).

3.3 Biocompatibility of the CPTCu hydrogel

Given the requirement for safe intraoral application, the biocompatibility of the CPTCu hydrogel was first evaluated. A hemolysis assay was conducted to assess its blood compatibility (Fig. 3A).⁴⁵ The results showed that all hydrogels exhibited hemolysis ratios below 5%, suggesting that CPTCu has good hemocompatibility and can be applied as a blood-contacting material.⁴⁶

Fig. 3 Biocompatibility of the CPTCu hydrogel. (A) Hemolysis assay of CP, CPCu, CPT and CPTCu. n = 4 per group. (B) Viability of HUVECs treated with CP, CPCu, CPT, CPTCu and the control for 24 h, 48 h, and 72 h. n = 5 per group. (C) Representative images of live/dead staining of tongue fibroblasts after incubation with hydrogel extracts for 24 h and 72 h. (D) Representative images of immunofluorescence staining of tongue fibroblasts after incubation with hydrogel extracts for 72 h. The nucleus was stained with DAPI (blue) and F-actin was stained with Phalloidin-TRITC (red).

Next, the cytocompatibility of CPTCu was explored. To better simulate the native tongue microenvironment, rat tongue-derived primary fibroblasts and human umbilical vein endothelial cells (HUVECs) were used as the main cell models. The fibroblasts were isolated using a tissue block method and subsequently purified through two rounds of passaging prior to experiments. The acquired cells exhibit a typical fibroblast spindle-shaped morphology. Immunofluorescence showed positive Vimentin staining, confirming their mesenchymal origin (Fig. S3).^47,48

It is well established that Cu is an essential trace element involved in cellular function and redox homeostasis, but excessive Cu accumulation can induce cytotoxicity.⁴⁹ Likewise, TA has been reported to suppress fibroblast viability at elevated concentrations.⁵⁰ Accordingly, a CCK-8 assay was conducted to determine the appropriate and non-toxic composition of the CPTCu hydrogel for subsequent biological applications (Fig. 3B and Fig. S4). The results indicated that the increase of Cu concentration in CPCu hydrogels leaded to a trend of decreased HUVECs and tongue fibroblasts viability. For the CPTCu hydrogel, CPTCu-2 (loaded with 0.2% (w/v) L-Asp-Cu MOF) enhanced the viability of HUVECs without notably compromising tongue fibroblasts viability, and therefore it was selected for further applications. We also observed that the TA concentration used in our system (1% (w/v)) did not exert any appreciable cytotoxic effect on either cell type. Via live/dead staining and cytoskeletal imaging of tongue fibroblasts, we further confirmed the good biocompatibility of the selected Cu and TA concentrations. After 24 h and 72 h culturing with hydrogel extracts, all groups exhibited high cell survival rates, and cytoskeleton staining showed well-spread cell morphology with no observable differences compared with the control (Fig. 3C and D).

3.4 Hemostasis ability of the CPTCu hydrogel

The tongue is highly vascularized, and injuries often result in substantial bleeding. Therefore, dressings designed for tongue wounds must provide instant hemostatic capability. We first studied the in vivo hemostasis ability of CPTCu with a rat liver hemorrhage model. Bleeding was generated using a biopsy punch, after which CP, CPCu, CPT and CPTCu patches were immediately applied to the puncture, while no treatment was applied to the control group. Macroscopic images of final hemostasis, mass of total blood loss, and hemostatic time were recorded for each group (Fig. 4A and B). CPT exhibited the least blood loss and the shortest hemostatic time, and all hydrogel groups helped to reduce bleeding. Compared with the control, the application of CPTCu substantially decreased blood loss by 76.4% and shortened hemostatic time by 48.7%. The hemostatic ability is mainly contributed by immediate adhesion and sealing of hydrogel patches to the bleeding wound. The incorporation of TA further enhanced hemostasis, which is attributed both to strengthened adhesion and to interactions between polyphenol groups and blood to activate coagulation factor XII upon contact.^51,52 The blood clotting index (BCI) was calculated to evaluate in vitro hemostatic properties (Fig. 4C). A lower BCI indicates better hemostatic performance. The results showed a consistent trend with the in vivo experiments: the BCI of CPTCu was 28.6 ± 2.3%, approximately 4 times lower than that of the control group. These results indicate that the CPTCu hydrogel achieves instant effective hemostasis and is an excellent candidate for acute tongue wound management.

Fig. 4 Hemostatic and antioxidant properties of the CPTCu hydrogel. (A) Representative photographs of the in vivo liver hemostasis model. (B) Total blood loss and hemostatic time of each group in the in vivo hemostatic experiment. n = 3 per group. (C) Blood clotting index (BCI) of CP, CPCu, CPT and CPTCu. n = 4 per group. (D) Macroscopic images of the solution color of each group in the DPPH free radical scavenging assay. (E) Representative fluorescence images of intracellular ROS (marked with DCFH-DA probe) of HUVECs and tongue fibroblasts after incubation with CP, CPCu, CPT and CPTCu. (F) DPPH radical scavenging rate of CP, CPCu, CPT and CPTCu. n = 3 per group. (G) and (H) Relative fluorescence intensity of intracellular ROS of the HUVECs (G) and tongue fibroblasts (H); n = 4 per group.

3.5 Antioxidant ability of the CPTCu hydrogel

As part of the host defense against invading pathogens, a large amount of reactive oxygen species (ROS) is generated in the wound microenvironment.⁵³ While this process benefits wound healing, excessive ROS production can damage cell structures, delay wound healing, and may also contribute to fibrotic scar formation.^54,55 The antioxidant potential of CPTCu was first examined via DPPH free radical scavenging assay. The DPPH solution exhibits a purple color, which gradually fades upon reduction by antioxidants, accompanied by a decrease in absorbance. As shown in Fig. 4D, the solution with CP and CPCu remained dark purple, CPTCu caused a noticeable decrease in color intensity, whereas CPT turned the solution yellow, comparable to the vitamin C positive control. CPT and CPTCu showed a radical scavenging rate of 95.5 ± 0.3% and 73.6 ± 7.6%, respectively, indicating that CPT and CPTCu hydrogels exhibit remarkable radical scavenging capabilities (Fig. 4F).

We further validated the antioxidant effect of the hydrogels at cellular level using an oxidative stress model. HUVECs and tongue fibroblasts were cultured with hydrogel extracts for 2 days, followed by loading with the DCFH-DA fluorescent probe and incubation with hydrogen peroxide (H₂O₂) for 2 h to induce oxidative stress. DCFH-DA is widely used to monitor redox processes in cells. After cellular uptake, it is deacetylated into a non-fluorescent compound, which could be subsequently oxidized by ROS to form 2′,7′-dichlorofluorescein (DCF), emitting green fluorescence under microscopy. As shown in Fig. 4E, the fluorescence intensity in the blank group was significantly higher than the control group untreated with H₂O₂. By culturing with CP and CPCu extracts, fluorescence was partially reduced, while the CPT and CPTCu group exhibited a significant decrease in fluorescence, approaching control levels, further confirming their ability to restore physiological ROS levels (Fig. 4G and H).

The antioxidant activity of CPTCu is primarily dedicated by TA, a classical primary antioxidant capable of directly scavenging free radicals through hydrogen or electron donation mechanisms.^56,57 Cu ions at appropriate concentrations were reported to exhibit weak antioxidant activity by mimicking superoxide dismutase (SOD), while excessive Cu ions could generate cytotoxic ROS.^58,59 In our study, CPTCu displayed a slightly reduced antioxidant capacity compared with CPT, which can be explained by the interaction between phenolic hydroxyl groups and Cu ions that partially compromised TA's ROS-scavenging activity.³⁵ Nevertheless, the Cu concentration used in our system did not negate the overall antioxidant performance of CPTCu and still demonstrated a satisfactory antioxidant performance. Excessive ROS can significantly alter the biology of ECM through oxidative reactions and is a critical driver for scar formation.^60,61 The ROS-scavenging and antioxidant activity of CPTCu may also contribute to reduced scarring during wound repair.⁶²

3.6 In vitro wound healing and angiogenesis effects of the CPTCu hydrogel

An in vitro scratch assay was used to mimic the in vivo wound healing process, and the effect of CPTCu hydrogels on the migration of tongue fibroblasts was studied.⁶³ As shown in Fig. 5A and C, the CPTCu group achieved the fastest wound closure at both 24 h and 48 h, with the scratch closure rate being 29.6% and 25.8% higher than that of the control, respectively. This effect is primarily linked to the role of Cu ions, which have been proved to modulate ECM remodeling as well as directly promote fibroblast migration and proliferation.^64,65 While the CPCu group demonstrated some inapparent wound closure enhancement, the effect was much more pronounced in the CPTCu group. This is likely due to the antioxidant capacity of TA, which creates an optimal redox environment for cell migration, synergistically enhancing Cu-mediated activation of cell migration signaling.

Fig. 5 Cell-migration and angiogenesis effects of the CPTCu hydrogel and relevant RT-qPCR and transcriptome sequencing analysis. (A) Representative images of the scratch assay of tongue fibroblasts cultured with different hydrogel extracts at 0 h, 24 h, and 48 h. (B) Representative images of the tube formation assay of HUVECs cultured with different hydrogel extracts at 6 h. (C) Statistical analysis of the wound closure rates in the scratch assay. n = 4 per group. (D) Software analyzed results of the number of junctions and total tubule length of tube formation assay. n = 4 per group. (E) Gene expressions of angiogenesis-related genes in HUVECs and fibrosis-related genes in tongue fibroblasts with beta actin as the reference gene. n = 3 per group. (F) Volcano plot and clustered heatmap illustrating DEGs of interest in the CPTCu group compared with the control. (G) Top 20 up- and down-regulated KEGG enrichment pathways in the CPTCu group compared with the control. n = 4 per group.

The angiogenic potential of Cu is well recognized, primarily through mechanisms including stimulating endothelial cell proliferation and migration, modulating the expression and activity of various angiogenic factors, and activating nitric oxide synthase.^66,67 Promoting angiogenesis helps reconstruct local blood supply and restore oxygen and nutrient delivery to the wound site, thereby supporting ordered ECM remodeling and contributing to reduced fibrosis and scar formation.⁶⁸ The effect of CPTCu on angiogenesis was evaluated via an HUVECs tube formation assay on Matrigel. After 6 h of culturing with hydrogel extracts, bright-field images were captured and analyzed using ImageJ software (Fig. 5B and D). Both the CPCu and CPTCu groups formed distinct network-like structures, with image analysis revealing the highest number of junctions and total tubule length, reflecting the contribution of Cu ions.

Subsequent RT-qPCR analysis demonstrated that CPTCu up-regulated the expressions of Vegfa and Fgf2 in HUVECs, deciphering its role in promoting endothelial cell proliferation and angiogenesis at the transcriptional level.⁶⁹ In tongue fibroblasts, CPTCu down-regulated Tgfβ1, Col1, and Col3 expression. Critically, the Col1/Col3 ratio was maintained, and the suppression of the key profibrotic driver Tgfβ1 indicates that CPTCu mitigates fibrotic scarring while preserving normal collagen proportion (Fig. 5E).^70,71

To further elucidate the mechanisms underlying its pro-healing and antifibrotic effects, we performed RNA-Seq on tongue fibroblasts treated with CPTCu. The treatment modulated 320 differentially expressed genes (78 up-regulated, 242 down-regulated). The volcano plot and clustered heatmap revealed a marked up-regulation of genes linked to wound healing and oxidative stress response (e.g., Hmox1, Nqo1, Gstp1), concurrent with the down-regulation of key pro-fibrotic genes (e.g., Egr1/2/3, Fos) (Fig. 5F).^72–77 Subsequent Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of the top 20 pathways indicated that CPTCu up-regulated pathways related to antioxidation, epithelial regeneration, and protective stress responses (e.g., glutathione and retinol metabolism), while down-regulated core pathways driving fibrosis and chronic inflammation (e.g., MAPK, Wnt signaling pathways) (Fig. 5G).^78–80 Gene Ontology (GO) term analysis corroborated these findings, showing up-regulation of terms related to healing promotion and antioxidant stress, and down-regulation of terms linked to inflammation and ECM deposition (Fig. S5). GSEA analysis revealed that the CPTCu group exhibited significant positive enrichment of pathways related to protein synthesis, energy metabolism, and antioxidant and detoxification processes, while showing significant negative enrichment of signaling pathways associated with fibroblast overactivation and fibrosis (Fig. S6).^81–83 These results strongly suggest that CPTCu successfully directs wound healing toward a more regenerative and less fibrotic phenotype, significantly increasing the potential for rapid healing with minimal scarring.

3.7 Antibacterial performance of the CPTCu hydrogel

The severe hindrance of tissue regeneration due to wound infection highlights the importance of an ideal wound dressing exhibiting antimicrobial properties to prevent infection.⁸⁴ We selected the most common bacterial species associated with oral mucosal infections—aerobic Gram-negative rods (Escherichia coli, E.coli) and Gram-positive Staphylococcus aureus (S. aureus)—to evaluate the antibacterial performance of the CPTCu hydrogel.⁸⁵ After culturing with hydrogel extracts, the bacterial suspensions from each group were diluted, plated onto agar plates, and evenly spread, followed by inverted incubation overnight. Colony forming units (CFUs) of each group were counted and photographed the next day (Fig. 6A). The results showed that the survival rate for E. coli cultured with CPTCu was only 0.27 ± 0.12%, and for S. aureus the survival rate was 0.64 ± 0.49% (Fig. 6B). Logarithmic transformation of the CFUs in original bacterial suspensions revealed that culturing with CPTCu reduced the lg CFU value of E. coli by 2.587 and that of S. aureus by 2.272, suggesting strong antibacterial properties (Fig. S7).

Fig. 6 Antibacterial ability of the CPTCu hydrogel. (A) Representative photographs of the agar plates in the CFU counting assay. (B) Relative E. coli and S. aureus viability calculated based on CFU counting results. n = 3 per group. (C) Live (green) stained percentage of E. coli and S. aureus of each group in the live/dead bacterial staining assay. n = 3 per group. (D) Representative fluorescence images of live/dead E. coli and S. aureus staining. (E) SEM images of E. coli and S. aureus treated with different hydrogels. White arrows indicate morphologically changed bacteria.

Both TA and Cu exhibit inherent antibacterial activity. TA, as a multidentate ligand, interacts with bacterial proteins via strong hydrophobic interactions and hydrogen bonding, penetrates the bacterial cell wall, and interferes with bacterial metabolism.^86,87 Cu exhibits broad-spectrum antibacterial activity, primarily through interactions with bacterial cell membranes, redox reactions, and disruption of bacterial metabolic processes.⁸⁸ With the incorporation of TA and Cu, the antibacterial efficacy of all hydrogel groups was enhanced. However, it could be observed that the antibacterial effect against E. coli was primarily attributed to Cu, whereas the effect against S. aureus was mainly driven by TA. This difference arises because TA has limited ability to penetrate the double-membrane structure of Gram-negative bacteria, while Cu can induce membrane damage, lipid peroxidation, and DNA degradation in Gram-negative species.^89,90

To visualize the effect of CPTCu on bacterial viability, SYTO-9 and PI were used for live/dead bacterial staining. As shown in Fig. 6C and D, the CPTCu group exhibited the fewest green fluorescent signals representing live bacteria, consistent with the result of the CFU assay. However, live/dead staining did not demonstrate an antibacterial effect of CPTCu as pronounced as that reflected by CFU counts (75.1% versus 99.7% for E. coli; 66.2% versus 99.4% for S. aureus). This discrepancy arises partly from the distinct principles of the two experiments: CFU quantifies only culturable bacteria capable of forming colonies under culture conditions, whereas live/dead staining is mainly based on membrane integrity, which may not fully reflect metabolic activity or proliferative ability. Another contributing factor is the possible presence of CPTCu-induced viable but non-culturable (VBNC) bacteria, which maintain membrane integrity and thus appear “alive” in staining assays but fail to form colonies on agar plates and therefore are not counted in CFU measurements.^91,92

To further understand the morphological effects of hydrogels on E. coli and S. aureus, SEM images were captured. Fig. 6E shows that after culturing with CPCu, CPT and CPTCu, significant damage and noticeable folding of the bacterial outer membrane occurred, some bacteria even completely lost their normal shape. After membrane damage, the cytoplasm leaked out, leading to bacterial death. These findings collectively demonstrate the superior antibacterial performance of CPTCu.

3.8 Therapeutic effects of the CPTCu hydrogel patch on the rat tongue defect model

Based on the above experimental evidence, we established a rat tongue defect model to further evaluate the integrated therapeutic effect of the CPTCu hydrogel under physiological conditions (Fig. 7A). A 2 mm deep circular defect was created in the central dorsum of the rat tongue using a 3 mm biopsy punch. Due to the longitudinal tension of the tongue, the created defect exhibited an oval shape. The animals were randomly divided into 3 groups: the blank group received no treatment. The suture group, which followed the principles of current clinical standard management and acts as a positive control, applied transverse interrupted suturing with a 4-0 silk suture. For the CPTCu group, sterile CPTCu patch was trimmed to the wound size and applied to the tongue surface with gentle pressure for 10 s to seal the defect. Macroscopic photographs of each group were taken every two days and wound areas were quantified using ImageJ (Fig. 7B). The wound healing rates revealed markedly accelerated wound closure in the CPTCu group, which reached 98.3 ± 2.9% closure by day 8, 24% faster than the blank group (79.3 ± 5.9%). The suture group was not included in wound closure quantification, because the sutures impeded accurate delineation of the wound area from the images, and complete primary closure was already deemed to be achieved on day 0 by suturing. However, it is noteworthy that despite initial complete closure, new wound surfaces occurred and sutures were observed to detach from day 2, and by day 8 only about 50% of the sutures remained. This was likely caused by repeated tongue movements and friction from food during feeding, which generated tension from the sutures and resulted in secondary tissue injury. By day 8, the suture group exhibited swelling and partial tissue deformation, whereas the CPTCu group displayed smooth, well-healed surfaces without visible defects.

Fig. 7 Therapeutic effects of the CPTCu hydrogel in a rat tongue defect model. (A) Schematic diagram of the tongue defect model construction and treatment schedule. (B) Representative photographs of the tongue wound in the blank, suture and CPTCu group and quantitative analysis of the wound healing rate of the blank and CPTCu group on day 0, 2, 4, 6 and 8. n = 5 per group. (C) H&E staining, Masson staining, and multiplex immunohistochemical staining of CD68, CD31, and α-SMA of the wound area on day 8. CD68: purple, CD31: red, α-SMA: green, nuclei: blue. (D) Quantitative analysis of collagen volume fraction (CVF) based on Masson staining. n = 5 per group. (E) Relative fluorescence intensity of CD68, CD31, and α-SMA in each group. n = 5 per group.

Subsequently, the tongues were harvested and fixed on day 8, coronally cut at the center of the wound and sectioned for histological analysis (Fig. 7C). H&E staining showed unevenly thick epithelium in blank and suture groups. The blank group exhibited disorganized rete pegs, while the suture group lacked distinct rete peg structures. In contrast, the CPTCu group demonstrated well-organized epithelium and uniformly distributed rete pegs. Higher-magnification images revealed densely infiltrated basophilic cells with scattered eosinophils and loosely organized light-stained matrix in both the blank and suture groups, indicating that the wound is transitioning from the inflammatory phase to the proliferative phase with unsolved inflammation and immature matrix remodeling. Conversely, the CPTCu group displayed markedly reduced inflammatory cell infiltration, densely aligned mature collagen fibers, and organized regenerated muscle structures, suggesting that the wound had progressed to the remodeling phase.

Masson staining was further conducted to differentiate collagen and muscle fibers. Irregular red and blue-stained regions were observed in the blank and suture groups, whereas the CPTCu group had typical red-stained muscle fiber structures. Quantification of collagen volume fraction (CVF) showed significantly less collagen deposition in the CPTCu group, which decreased by 35.9% compared with the suture group (Fig. 7D). To evaluate treatment effects at cellular and functional levels, multiplex immunohistochemical staining of CD68, CD31, and α-SMA was performed (Fig. 7C). CD68 labels macrophages, CD31 marks endothelial cells and vessel structures, and α-SMA marks vascular smooth muscle cells and myofibroblasts. The CPTCu group exhibited the lowest CD68 expression and the highest CD31 and α-SMA expression, pointing to attenuated inflammation, enhanced angiogenesis, and improved ECM remodeling and wound healing (Fig. 7E). Meanwhile, it is noticed that the suture group showed the strongest CD68 signal, which is possibly due to food debris retained around the sutures that exacerbated local inflammation. Compared with the suture group, the macrophage infiltration in the CPTCu group decreased by 62.5%, and angiogenesis increased by 103.2%. These findings are consistent with the in vitro results and jointly confirm the superior wound healing efficacy of the CPTCu hydrogel.

In addition, major organs including the heart, liver, spleen, lung, kidney, stomach, and intestine were collected for H&E staining to evaluate potential systemic toxicity related to CPTCu local application and possible ingestion. As shown in Fig. S8, all tissues in the CPTCu group displayed normal morphology, demonstrating good tissue biocompatibility and safety for in vivo CPTCu application.

4. Conclusion

To address clinical complications associated with suturing tongue injuries, such as tissue deformation, scar formation, and chronic inflammation, we designed a tongue-customized hydrogel patch composed of a chitosan–polyacrylic acid–tannic acid matrix incorporating L-Asp-Cu MOF (CPTCu) as a noninvasive substitute for suturing. The hydrogel exhibits a Young's modulus of 37 kPa and an adhesion strength of 96.7 kPa, closely matching the mechanical characteristics of native tongue tissue while maintaining firm adhesion to its highly movable surface to ensure continuous wound protection. Within the hydrogel, Cu ions engage in dynamic coordination with polymer chains and L-Asp ligands, enabling long-term sustained Cu release in the oral environment. CPTCu demonstrated robust bioactivities in vitro, including a 76.4% reduction in bleeding, 73.6% ROS scavenging efficiency, and 99.7% antibacterial activity. In a rat tongue defect model, CPTCu accelerated wound closure by 24% at day 8 compared with the control, and reduced the collagen volume fraction by 35.9%, while enhancing angiogenesis by 103.2% compared with suturing. These results highlight the potent hemostatic, antioxidant, pro-migratory, angiogenic, and antibacterial capabilities of CPTCu, which function synergistically to establish a regenerative wound microenvironment that accelerates wound closure, limits collagen deposition, and minimizes scarring.

The limitations of this study mainly lie in the following aspects. First, although the rat tongue wound model is representative in terms of anatomy and healing mechanisms, it still differs from the clinical human tongue in size, mechanics, movements, and microenvironment. Second, the long-term stability of the hydrogel under complex intraoral conditions has not been fully tested. It is necessary to validate its performance in real clinical scenarios for further applications. In future studies, fine-tuning of the MOF structure, ligand chemistry, and the composition and density of the hydrogel network may enable precise control over ion release kinetics and biological responses, allowing for adaptation to different types and severities of tongue injuries.

Overall, the CPTCu hydrogel represents a compelling option for suture-free, minimum-scarring, and physiological repair of tongue wounds, offering valuable insight for future tongue wound management strategies.

Conflicts of interest

The authors listed declare no conflicts of interest.

Data availability

The data supporting this article have been included within the article and as part of the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d5tb02753e.

Acknowledgements

This study was supported by the National Natural Science Foundation of China (82201128, 82271034), the China Postdoctoral Science Foundation (2022M722250), the Natural Science Foundation of Sichuan Province (2023NSFSC2000), the Key Research Project Foundation of State Key Laboratory of Oral Diseases (SKLOD-2025KP007), and the Research and Development Program of West China Hospital of Stomatology (RD-03-202107). The schematic diagrams (Table of Contents, Fig. 1A, 2E, and 7A) were created in Biorender.com.

References

1. C. W. Lamell, G. Fraone, P. S. Casamassimo and S. Wilson, Pediatr. Dent., 1999, 21, 34–38.
2. U. Mohan Das and P. Gadicherla, Int. J. Clin. Pediatr. Dent., 2008, 1, 39–41.
3. C. L. Gonsalves, J. W. Zhu, G. Y. Kim, C. F. Leveille and A. J. Kam, Paediatr. Child Health, 2022, 27, 32–42.
4. H. Q. Guo, X. Yang, X. T. Wang, A. P. Ji and J. Bai, Surg. Infect., 2022, 23, 298–303.
5. A. Ravindra, D. R. Nayak, K. Devaraja, N. M. Matthew and S. Tiwari, Indian J. Otolaryngol. Head Neck Surg., 2022, 74, 6296–6306.
6. M. R. Jamalpour, N. Pouresmaeil and S. Khazai, BMC Oral Health, 2024, 24, 1215.
7. H. E. desJardins-Park, S. Mascharak, M. S. Chinta, D. C. Wan and M. T. Longaker, Front. Physiol., 2019, 10, 322.
8. A. Faris, L. Khalid, M. Hashim, S. Yaghi, T. Magde, W. Bouresly, Z. Hamdoon, A. T. Uthman, H. Marei and N. Al-Rawi, Int. Dent. J., 2022, 72, 278–287.
9. K. N. Leknes, I. T. Røynstrand and K. A. Selvig, J. Periodontol., 2005, 76, 34–42.
10. L. Mahesh, V. Kumar, A. Jain, S. Shukla, J. Aragoneses, J. Martínez González, M. Fernández-Domínguez and J. Calvo-Guirado, Materials, 2019, 12, 2848.
11. K. D. R. Kappert, N. Connesson, S. A. Elahi, S. Boonstra, A. J. M. Balm, F. van der Heijden and Y. Payan, J. Biomech., 2021, 114, 110147.
12. Y. Ding, Z. Zhu, X. Zhang and J. Wang, Adv. Healthcare Mater., 2024, 13, 2400912.
13. C. H. Wang, J. H. Cherng, C. C. Liu, T. J. Fang, Z. J. Hong, S. J. Chang, G. Y. Fan and S. D. Hsu, Int. J. Mol. Sci., 2021, 22, 7067.
14. M. A. Matica, F. L. Aachmann, A. Tøndervik, H. Sletta and V. Ostafe, Int. J. Mol. Sci., 2019, 20, 5889.
15. N. Desai, D. Rana, S. Salave, R. Gupta, P. Patel, B. Karunakaran, A. Sharma, J. Giri, D. Benival and N. Kommineni, Pharmaceutics, 2023, 15, 1313.
16. G. D. Mogoşanu and A. M. Grumezescu, Int. J. Pharm., 2014, 463, 127–136.
17. L. Wang, L. Duan, G. Liu, J. Sun, M. Shahbazi, S. C. Kundu, R. L. Reis, B. Xiao and X. Yang, Adv. Sci., 2023, 10, 2207352.
18. R. Lv, X. Cao, T. Zhang, W. Ji, U. Muhammad, J. Chen and Y. Wei, Carbohydr. Polym., 2025, 351, 123111.
19. H. Yuk, C. E. Varela, C. S. Nabzdyk, X. Mao, R. F. Padera, E. T. Roche and X. Zhao, Nature, 2019, 575, 169–174.
20. H. Jafari, P. Ghaffari-Bohlouli, S. V. Niknezhad, A. Abedi, Z. Izadifar, R. Mohammadinejad, R. S. Varma and A. Shavandi, J. Mater. Chem. B, 2022, 10, 5873–5912.
21. H. Cao, D. Xiang, X. Zhou, P. Yue, Y. Zou, Z. Zhong, Y. Ma, L. Wang, S. Wu and Q. Ye, Carbohydr. Polym., 2023, 307, 120609.
22. Y. Liu, Y. Zhao, S. Guo, D. Qin, J. Yan, H. Cheng, J. Zhou, J. Ren, L. Sun, H. Peng, X. Wu and B. Li, Carbohydr. Polym., 2024, 346, 122656.
23. S. Alizadeh, B. Seyedalipour, S. Shafieyan, A. Kheime, P. Mohammadi and N. Aghdami, Biochem. Biophys. Res. Commun., 2019, 517, 684–690.
24. M. Harrison, C. Jones, M. Solioz and C. Dameron, Trends Biochem. Sci., 2000, 25, 29–32.
25. S. Tury, L. Chauveau, A. Lecante, V. Courgnaud and J. L. Battini, Cell Rep., 2023, 42, 113065.
26. X. Wen, J. Wang, X. Pei and X. Zhang, J. Mater. Chem. B, 2023, 11, 11405–11425.
27. X. Zhang, J. Chen, X. Pei, J. Wang, Q. Wan, S. Jiang, C. Huang and X. Pei, ACS Appl. Mater. Interfaces, 2017, 9, 25171–25183.
28. J. Chen, X. Zhang, C. Huang, H. Cai, S. Hu, Q. Wan, X. Pei and J. Wang, J. Biomed. Mater. Res., Part A, 2017, 105, 834–846.
29. Y. Zhang, J. Wu, B. Zhang, A. Shi, X. Pei, X. Zhang, Z. Zhu, J. Chen and J. Wang, Chem. Eng. J., 2025, 505, 159617.
30. M. Holeček, Nutrients, 2023, 15, 4023.
31. G. Gizer, M. Sahiner, Y. Yildirim, S. Demirci, M. Can and N. Sahiner, Curr. Res. Green Sustainable Chem., 2021, 4, 100110.
32. M. Božič, S. Gorgieva and V. Kokol, Carbohydr. Polym., 2012, 89, 854–864.
33. S. Mallakpour, Z. Radfar and M. Feiz, Carbohydr. Polym., 2023, 305, 120528.
34. M. H. Farag, M. F. Ruiz-López, A. Bastida, G. Monard and F. Ingrosso, J. Phys. Chem. B, 2015, 119, 9056–9067.
35. N. Sahiner, O. Guven, S. Demirci, S. S. Suner, M. Sahiner, B. Ari and M. Can, Colloids Surf., B, 2025, 252, 114669.
36. Z. Jiang, W. Zhang, C. Zhu, Z. Hu, X. Xiang, D. Cao and X. Wang, ACS Nano, 2025, 19, 26170–26192.
37. E. Díez-Peña, I. Quijada-Garrido and J. M. Barrales-Rienda, Macromolecules, 2003, 36, 2475–2483.
38. Y. Yin, X. Ji, H. Dong, Y. Ying and H. Zheng, Carbohydr. Polym., 2008, 71, 682–689.
39. I. Christodoulou, T. Bourguignon, X. Li, G. Patriarche, C. Serre, C. Marlière and R. Gref, Nanomaterials, 2021, 11, 722.
40. S. S. Nasiri, Z. Ahmadi and F. Afshar-Taromi, Int. J. Biol. Macromol., 2023, 245, 125521.
41. W. Xue, R. Yang, S. Liu, Y. Pu, P. Wang, W. Zhang, X. Tan and B. Chi, Biomater. Sci., 2022, 10, 2417–2427.
42. T. Funami, in Rheology of Biological Soft Matter: Fundamentals and Applications, ed. I. Kaneda, Springer Japan, Tokyo, 2017, pp. 171–198.
43. L. Huang, C. Du, R. B. N. Scharff, L. Beccai, J. Wang and D. Du, Food Res. Int., 2025, 203, 115786.
44. F. Zhao, M. Zhang, M. Nizamoglu, H. J. Kaper, L. A. Brouwer, T. Borghuis, J. K. Burgess, M. C. Harmsen and P. K. Sharma, Acta Biomater., 2024, 182, 67–80.
45. I. Sæbø, M. Bjørås, H. Franzyk, E. Helgesen and J. Booth, Int. J. Mol. Sci., 2023, 24, 2914.
46. J. Park, T. Y. Kim, Y. Kim, S. An, K. S. Kim, M. Kang, S. A. Kim, J. Kim, J. Lee, S. Cho and J. Seo, Adv. Sci., 2023, 10, e2303651.
47. P. Laurila, I. Virtanen, V. P. Lehto, T. Vartio and S. Stenman, J. Cell Biol., 1982, 94, 308–315.
48. Z. Ostrowska-Podhorodecka, I. Ding, M. Norouzi and C. A. McCulloch, Front. Cell Dev. Biol., 2022, 10, 869069.
49. F. Tisato, C. Marzano, M. Porchia, M. Pellei and C. Santini, Med. Res. Rev., 2010, 30, 708–749.
50. R. G. Daré, C. V. Nakamura, V. F. Ximenes and S. O. S. Lautenschlager, Free Radical Biol. Med., 2020, 160, 342–355.
51. H. Geng, Q. Dai, H. Sun, L. Zhuang, A. Song, F. Caruso, J. Hao and J. Cui, ACS Appl. Bio Mater., 2020, 3, 1258–1266.
52. J. Park, T. Y. Kim, Y. Kim, S. An, K. S. Kim, M. Kang, S. A. Kim, J. Kim, J. Lee, S. Cho and J. Seo, Adv. Sci., 2023, 10, 2303651.
53. Y. Peng, D. He, X. Ge, Y. Lu, Y. Chai, Y. Zhang, Z. Mao, G. Luo, J. Deng and Y. Zhang, Bioact. Mater., 2021, 6, 3109–3124.
54. Z. Tang, I. F. Chowdhury, J. Yang, S. Li, A. K. Mondal and H. Wu, Adv. Colloid Interface Sci., 2025, 339, 103425.
55. B. C. Carney, J. H. Chen, R. A. Kent, M. Rummani, A. Alkhalil, L. T. Moffatt, D. S. Rosenthal and J. W. Shupp, J. Surg. Res., 2019, 244, 312–323.
56. N. Sahiner, S. Sagbas, M. Sahiner, C. Silan, N. Aktas and M. Turk, Int. J. Biol. Macromol., 2016, 82, 150–159.
57. A. Zeb, J. Food Biochem., 2020, 44, e13394.
58. H. Ghanbari, H. Derakhshankhah, K. Bahrami, S. Keshavarzi, K. Mohammadi, P. Hayati, R. Centore and E. Parisi, Sci. Rep., 2024, 14, 21325.
59. F. Bellia, V. Lanza, I. Naletova, B. Tomasello, V. Ciaffaglione, V. Greco, S. Sciuto, P. Amico, R. Inturri, S. Vaccaro, T. Campagna, F. Attanasio, G. Tabbì and E. Rizzarelli, Antioxidants, 2023, 12, 1632.
60. M. Yang, P. Liu, P. Cheng, C. Li, J. Liu, H. Sun, F. Guan and M. Yao, Theranostics, 2025, 15, 9838–9861.
61. Y. Dong and Z. Wang, Front. Bioeng. Biotechnol., 2023, 11, 1304835.
62. X. Liu, Y. Sun, J. Wang, Y. Kang, Z. Wang, W. Cao, J. Ye and C. Gao, Bioact. Mater., 2024, 34, 269–281.
63. C. Liang, A. Y. Park and J. Guan, Nat. Protoc., 2007, 2, 329–333.
64. S. Alizadeh, B. Seyedalipour, S. Shafieyan, A. Kheime, P. Mohammadi and N. Aghdami, Biochem. Biophys. Res. Commun., 2019, 517, 684–690.
65. J. Xiao, S. Chen, J. Yi, H. F. Zhang and G. A. Ameer, Adv. Funct. Mater., 2017, 27, 1604872.
66. C. K. Sen, S. Khanna, M. Venojarvi, P. Trikha, E. C. Ellison, T. K. Hunt and S. Roy, Am. J. Physiol. Heart Circ. Physiol., 2002, 282, H1821–H1827.
67. P. Wang, Y. Yuan, K. Xu, H. Zhong, Y. Yang, S. Jin, K. Yang and X. Qi, Bioact. Mater., 2021, 6, 916–927.
68. M. Rahimnejad, S. Derakhshanfar and W. Zhong, Burns Trauma, 2017, 5, 4.
69. N. Yamamoto, T. Oyaizu, M. Enomoto, M. Horie, M. Yuasa, A. Okawa and K. Yagishita, Sci. Rep., 2020, 10, 2744.
70. K. K. Kim, D. Sheppard and H. A. Chapman, Cold Spring Harb. Perspect. Biol., 2018, 10, a022293.
71. J. Zeng, T. Y. Huang, Z. Z. Wang, Y. F. Gong, X. C. Liu, X. M. Zhang and X. Y. Huang, Int. Immunopharmacol., 2021, 90, 107200.
72. Q. Chen, G. Wang, W. Li, Y. Jiang, X. Lu and P. Zhou, J. Diabetes Res., 2016, 2016, 1–10.
73. X. Wu, C. Ding, X. Liu, Q. Ding, S. Zhang, Y. Wang, C. Xin, H. Wei, R. Mao, G. Zhang, T. Zhao and W. Liu, Int. J. Biol. Macromol., 2024, 279, 135324.
74. W. S. Kang, E. Kim, H. Choi, K. H. Lee, K. J. Kim, D. Lim, S. Choi, Y. Kim, S. ah Son, J. S. Kim and S. Kim, Antioxidants, 2023, 12, 1171.
75. S. Bhattacharyya, F. Fang, W. Tourtellotte and J. Varga, J. Pathol., 2013, 229, 286–297.
76. S. Bhattacharyya, M. Wu, F. Fang, W. Tourtellotte, C. Feghali-Bostwick and J. Varga, Matrix Biol., 2011, 30, 235–242.
77. V. Sharma, J. Singh, Y. Kumar, A. Kumar, K. Venkatesan, M. Mukherjee and A. K. Sharma, Life Sci., 2025, 368, 123508.
78. D. Zhang, M. Zhu, P. Xu, X. Wen, G. Liang, W. Zheng, Y. Zeng, T. Sun, R. Fan, Y. Lu, X. Tan, M. Gong, T. Wang, J. Chen and J. Guan, ACS Nano, 2024, 18, 23428–23444.
79. T. Wüstemann, E. Madzharova, M. S. Wietecha, N. B. Ghyselinck, M. Höring, G. Liebisch, N. Zamboni, U. Auf dem Keller and S. Werner, JCI Insight, 2025, 10, e194188.
80. J. Luo, J. Chen, Z. Deng, X. Luo, W. Song, K. A. Sharff, N. Tang, R. C. Haydon, H. H. Luu and T. He, Lab. Invest., 2007, 87, 97–103.
81. T. Liu, Y. Wu, C. Wang, T. Zhou, R. Yang, X. Zhang, W. Yan, Q. Fan and Z. Lu, Mater. Today Bio, 2025, 35, 102626.
82. M. Liu, G. Ye, R. Li, T. Gao, C. Zhuang, S. Huang, S. Wang, J. Hu, A. P. Xiang and M. H. Jiang, Cell Biol. Int., 2025, 49, 1425–1440.
83. X. Yin, M. Liu, Y. Xu, H. Zhong, J. Zhang, H. Xie, Y. Xiang, J. Wan and D. Chen, Free Radical Biol. Med., 2026, 242, 431–443.
84. X. Zhang, Y. Liang, S. Huang and B. Guo, Adv. Colloid Interface Sci., 2024, 332, 103267.
85. G. Dahlén, Periodontology 2000, 2009, 49, 13–38.
86. T. J. Kim, J. L. Silva, M. K. Kim and Y. S. Jung, Food Chem., 2010, 118, 740–746.
87. H. Jafari, P. Ghaffari-Bohlouli, S. V. Niknezhad, A. Abedi, Z. Izadifar, R. Mohammadinejad, R. S. Varma and A. Shavandi, J. Mater. Chem. B, 2022, 10, 5873–5912.
88. J. Yang, J. Qin, B. Xu, H. Liu and X. Wang, Small, 2025, e09183.
89. J. Xi, G. Wei, L. An, Z. Xu, Z. Xu, L. Fan and L. Gao, Nano Lett., 2019, 19, 7645–7654.
90. B. Kaczmarek, Materials, 2020, 13, 3224.
91. S. Song and T. K. Wood, Environ. Microbiol., 2021, 23, 2335–2338.
92. J. Liu, L. Yang, B. V. Kjellerup and Z. Xu, Trends Microbiol., 2023, 31, 1013–1023.

---

Footnote

† These authors contributed equally to this work.

---