Sprayable hydrogel sponge for neurovascular microenvironment reconstruction and inflammation modulation in diabetic wound healing

Abstract

Diabetic wounds are characterized by chronic inflammation, vascular insufficiency, and peripheral neuropathy, which collectively disrupt the neurovascular microenvironment essential for coordinated tissue regeneration. However, strategies targeting neurovascular regeneration remain limited. Here, we developed a sprayable hydrogel sponge based on gelatin methacryloyl and methacrylamide-modified ε-poly-L-lysine (S-GPL), co-functionalized with VEGF-mimetic peptide (KLT) and BDNF-mimetic peptide (RGI). The sprayable format conforms to irregular wound geometries, while the pneumatic spraying technique generates high-pressure microbubbles that create a porous structure, thereby enhancing exudate absorption and sustained peptide release as a sponge dressing. Additionally, the incorporation of KLT and RGI facilitates the reconstruction of the neurovascular microenvironment. In vitro, KLT promoted endothelial cell maturation and cytokine secretion, whereas RGI enhanced Schwann cell activity. Notably, S-GPL^KLT/RGI facilitated intercellular interactions between RSCs and HUVECs, highlighting the cellular mechanisms underlying neurovascular communication. In a full-thickness diabetic wound model in rats, the hydrogel accelerated wound closure, re-epithelialization, and matrix remodeling. These effects were accompanied by enhanced neovascularization and axonal regeneration, along with the formation of a spatially organized neurovascular niche, as evidenced by CD31⁺ capillaries closely aligned with PGP9.5⁺ nerve fibers. Building upon the intrinsic anti-inflammatory properties of S-GPL, transcriptomic and immunohistochemical analyses further revealed that S-GPL^KLT/RGI treatment suppressed the IL-17 signaling pathway. However, the relationship between immunomodulation and neurovascular reconstruction warrants further investigation. Collectively, this study presents a sprayable antibacterial hydrogel that not only reconstructs the neurovascular microenvironment but also mitigates chronic inflammation, offering a clinically translatable strategy for diabetic wound management.

Graphical abstract

Schematic illustration of the design and therapeutic mechanism of S-GPL^KLT/RGI for diabetic wound repair. The dual-network hydrogel is formed by combining GelMA and PLMA, functionalized with VEGF-mimetic peptide (KLT) and BDNF-mimetic peptide (RGI). Upon spraying and 405 nm UV-induced crosslinking, the hydrogel conforms to irregular wound surfaces and exhibits antibacterial activity against E. coli and S. aureus. In the diabetic wound microenvironment characterized by microangiopathy and peripheral neuropathy, S-GPL^KLT/RGI activates endothelial cells and Schwann cells through direct stimulation and paracrine signaling. These coordinated cellular responses facilitate neurovascular niche reconstruction and inflammation resolution, thereby promoting tissue regeneration and wound closure by day 14.

Highlights

• •Sprayable hydrogel sponge with mimetic peptides was developed for diabetic wound.
• •Spraying formed porous structure for exudate absorption and sustained peptide release.
• •Hydrogel enhanced neurovascular regeneration and IL-17 suppression in diabetic wounds.

1. Introduction

Diabetes mellitus is one of the most prevalent metabolic disorders worldwide. Its associated complications significantly compromise patients’ quality of life and place an increasing burden on healthcare systems [1]. Among these, diabetic wounds (DWs) represent one of the most debilitating complications, affecting nearly one-third of diabetic patients [2] and are associated with high rates of amputation and mortality [3]. DWs are typically located in high biomechanical stress areas, such as toes and joints, and manifest as irregular, full-thickness dermal ulcers with chronic non-healing profiles [4]. Traditional wound dressings often fail to conform to the complex shapes of these wounds, thereby delaying healing and causing secondary tissue damage. Therefore, there is an urgent clinical need for advanced wound dressings that combine morphological adaptability with controlled delivery of bioactive agents to effectively manage DWs (see Scheme 1).

Scheme 1.

Schematic illustration of the design and therapeutic mechanism of S-GPL^KLT/RGI for diabetic wound repair. The dual-network hydrogel is formed by combining GelMA and PLMA, functionalized with VEGF-mimetic peptide (KLT) and BDNF-mimetic peptide (RGI). Upon spraying and 405 nm UV-induced crosslinking, the hydrogel conforms to irregular wound surfaces and exhibits antibacterial activity against E. coli and S. aureus. In the diabetic wound microenvironment characterized by microangiopathy and peripheral neuropathy, S-GPL^KLT/RGI activates endothelial cells and Schwann cells through direct stimulation and paracrine signaling. These coordinated cellular responses facilitate neurovascular niche reconstruction and inflammation resolution, thereby promoting tissue regeneration and wound closure by day 14.

Delayed DW healing is driven by a complex interplay of sustained inflammation, microvascular dysfunction, and peripheral neuropathy [5]. Under persistent hyperglycemia, immune dysregulation and excessive release of pro-inflammatory cytokines (e.g., IL-1β and TNF-α) prolong the inflammatory phase, hinder extracellular matrix (ECM) remodeling [6], and increase the risk of infection, ultimately impeding tissue regeneration [7]. Moreover, diabetic peripheral neuropathy and microvascular rarefaction synergistically disrupt the neurovascular architecture at the wound site [5], further destabilizing local homeostatic regulation [8]. Accumulating evidence indicates that the nervous and vascular systems engage in bidirectional crosstalk, orchestrating wound healing through paracrine signaling and spatial scaffolding cues [9]. Therefore, targeted reconstruction of the neurovascular microenvironment represents a promising therapeutic strategy to overcome the regenerative deficits in DWs.

Diabetic wound healing faces the dual challenges of impaired angiogenesis and dysregulated neural regeneration. Although prior studies have highlighted the therapeutic potential of growth factors such as vascular endothelial growth factor (VEGF) and brain-derived neurotrophic factor (BDNF) in promoting tissue repair [[10], [11], [12]], their clinical application is hindered by their proteolytic instability and inactivation within the hostile inflammatory microenvironment [13]. Bioactive mimetic peptides derived from the functional domains of VEGF and BDNF have emerged as promising alternatives to native growth factors, offering enhanced proteolytic stability, greater synthetic accessibility, and high receptor-binding specificity [14,15]. Their stability and functionality have been further improved through advances in structural biology and synthetic chemistry [16]. Moreover, the hydrogel matrix provides an additional layer of protection and serves as a sustained-release reservoir, enabling prolonged bioactivity of these peptides in the wound microenvironment. However, most current hydrogel delivery systems, although capable of sustained peptide release, lack the ability to simultaneously recapitulate neurovascular signaling dynamics. This critical gap underscores the need for multifunctional biomaterials that integrate coordinated delivery of pro-angiogenic and neurogenic peptides with microenvironmental immunomodulation, thereby promoting neurovascular coupling and restoring regenerative homeostasis in chronic wounds.

Sprayable hydrogels, owing to their unique mode of administration, represent a promising alternative to conventional hydrogel bandages. As in situ-forming biomaterials, sprayable hydrogels exhibit strong substrate adhesion, conformal adaptability, and uniform coating capability, making them particularly well-suited for treating irregularly shaped or mechanically dynamic wounds [17,18]. The integration of pneumatic spray systems further broadens the applicability of these materials. Unlike hydrogel bandages, pneumatic spray technology allows for rapid and large-area application, while concurrently inducing microbubble formation under high-pressure conditions. This process generates a porous, sponge-like internal architecture with interconnected pore networks. Such structures endow the hydrogel with favorable swelling capacity, thereby maintaining a moist wound environment, facilitating exudate management, and alleviating patient discomfort [19]. In addition, the increased specific surface area enhances controlled drug release [20] and facilitates cellular adhesion, proliferation, and tissue integration [21]. Importantly, key physical parameters of the hydrogel—including pore size, degradation rate, and drug release kinetics—can be finely tuned by modulating spray conditions to generate pores of varying diameters. Collectively, these features confer sprayable hydrogels with superior functionality, operational convenience, and structural adaptability, establishing them as a versatile platform for enhancing therapeutic efficacy in the management of chronic and complex wounds.

ε-Poly-L-lysine (ε-PL) is a naturally occurring cationic polypeptide produced by Streptomyces albulus. It was initially identified for its potent broad-spectrum antimicrobial activity [22]. It exerts bactericidal effects by electrostatically disrupting microbial membranes via its lysine residues [23], showing broad-spectrum activity against common DW pathogens with low risk of resistance development [24]. The beneficial role of the naturally predominant bacteria capable of producing antimicrobial peptides in promoting the healing of infected diabetic wounds has been well-documented [25], highlighting the therapeutic potential of ε-PL in wound infection control. Leveraging recent advances in synthetic biology and polymer chemistry, we modified ε-PL with methacrylamide (yielding PLMA) to enable its stable integration into gelatin methacryloyl (GelMA)-based hydrogel network for biomedical applications [26,27].

In this study, we developed a sprayable, in situ-forming functional hydrogel sponge designed to reconstruct the neurovascular microenvironment and promote tissue regeneration in DWs. The hydrogel, composed of GelMA and PLMA, was applied using a pneumatic spray gun, where high-velocity airflow facilitated the formation of a porous architecture. Methacrylate modification enabled rapid UV-triggered crosslinking, ensuring precise adaptation to anatomically complex wound regions. VEGF-mimetic peptide KLT (GGGKLTWQELYQLKYKGIGG) and BDNF-mimetic peptide RGI (GGGRGIDKRHWNSQGG) were grafted into the network via a methacrylate linkage, thereby enabling controlled release during hydrogel degradation [28] and synergistically enhancing angiogenesis and neurogenesis. Our study aims to systematically evaluate the biofunctional performance of GelMA–PLMA sprayable hydrogel sponge with KLT and RGI (S-GPL^KLT/RGI), both in vitro and in vivo, focusing on its capacity to reconstruct the neurovascular microenvironment, modulate inflammation, and accelerate wound closure. This work proposes a novel biomaterial-based and precision-targeted strategy for the clinical management of DWs.

2. Materials and methods

2.1. Materials

Gelatin (G108394) and methacrylic anhydride (M102519) were purchased from Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). Lithium phenyl (2,4,6-trimethylbenzoyl) phosphonate (LAP, EFL-LAP) and poly-L-lysine methacryloyl (EFL-PLMA) were obtained from EFL-Tech Co., Ltd. (Suzhou, China). The VEGF-mimetic peptide KLT with sequence (MA)-GGGKLTWQELYQLKYKGIGG (purity >95 %), and the BDNF -mimetic peptide RGI with sequence (MA)-GGGRGIDKRHWNSQGG (purity >95 %) was synthesized and sterilized by BioAct Peptide Biotechnology LLC. (Beijing, China).

2.2. Preparation of GelMA

GelMA was synthesized according to the following protocol. Briefly, 10 g of gelatin was dissolved in 100 mL of deionized water with stirring at 50 °C until fully dissolved. Subsequently, 8 mL of methacrylic anhydride was added to the gelatin solution, and the reaction was carried out at 50 °C for 3 h. The resulting mixture was dialyzed against distilled water at 40 °C for one week to remove unreacted reagents, followed by freeze-drying to obtain the solid product.

2.3. Fabrication of hydrogels

The hydrogel precursor solution was prepared by dissolving GelMA, LAP, and EFL-PLMA in PBS at desired concentrations (GelMA: 5 %, LAP: 0.25 %, EFL-PLMA: 1 %, 3 %, 5 %). For peptide-functionalized hydrogel, mimetic peptides were added at the intended concentration to the precursor solution (300 μg/ml).

During the preparation of the GelMA-PLMA sprayable hydrogel (S-GPL), the precursor solution consisting of GelMA, LAP, and EFL-PLMA was applied using a pneumatic spray gun (TAMIYA 74537, Japan) equipped with a 0.5 mm diameter nozzle. Compressed air supply was maintained at 0.1 MPa through a Vogue AI series air compressor (Hosheng Pneumatic Machinery Co., China). The spraying process was conducted at an optimal working distance of 5–10 cm from the target surface, ensuring uniform deposition while preserving the structural integrity of the hydrogel matrix and cross-linked using a portable ultraviolet (UV) lamp (405 nm, EFL-LS-1600-405) for rapid gelation. Specifically, GelMA hydrogel was cross-linked under 405 nm UV irradiation for 20 s. For hydrogels incorporating peptides, the peptides were mixed thoroughly into the precursor solution before gelation under identical conditions. The GelMA bulk hydrogel and GelMA-PLMA bulk hydrogel (B-GPL) were prepared following the same procedure described above, except that the precursor solutions were not sprayed.

2.4. Characterization of hydrogels

Fourier transform infrared spectroscopy (FT-IR, Thermo Fisher, USA) was performed to analyze the chemical bonds present in samples. The FTIR spectra were recorded in the wavenumber range of 400–4000 cm⁻¹.

¹H nuclear magnetic resonance spectroscopy (¹H NMR, JNM-ECA600, JEOL Ltd., Japan) was used to confirm the substitution of free amino groups in gelatin with methacrylate groups. Samples were dissolved in deuterium oxide (D₂O, Sigma-Aldrich, USA), and the spectra were recorded at ambient temperature.

To determine the gelation time of hydrogels under illumination, rheological analysis was conducted using a rheometer (MCR302e, Anton Paar GmbH, Austria) equipped with a 10-mm diameter parallel plate at a gap of 1 mm and a temperature of 37 °C. Hydrogel precursor solutions were placed under the probe, and oscillatory tests were performed at 10 % strain and 1 Hz frequency for 100 s, recording data every 0.5 s. Irradiation started at 10 s using a 405 nm LED (25 mW). The gelation point was identified when the storage modulus (G′) exceeded the loss modulus (G″). The final storage modulus was recorded as the limit modulus of the hydrogel. Additionally, the viscoelastic properties of three prepared hydrogel samples were evaluated through frequency sweeps (0.1–10 Hz at 1 % strain) and strain sweeps (0.1–1000 % at 1 Hz frequency), and the corresponding loss factors and fracture strains were calculated. The loss factors of hydrogels were with the following equation:

$$Lossfactor\left(\mathit{tan}\delta \right)=lossmodulus/storagemodulus$$

The morphology of hydrogels (GelMA bulk gel, B-GPL, S-GPL) was examined using scanning electron microscopy (SEM, GEMINISEM 500, Carl Zeiss, Germany). Hydrogels were dried, sputter-coated with gold, and imaged. Pore sizes were quantified using ImageJ 1.51k (Wayne Rasband, NIH, USA). The morphology of microbubbles architecture of S-GPL was examined using SEM. After critical-point drying with CO₂, samples were rapidly immersed in liquid nitrogen (−196 °C) for 5 min to achieve complete vitrification, followed by precise mechanical fracture to expose the cross-section.

The mechanical strength of hydrogels was assessed by compression testing. Cylindrical hydrogel samples (diameter: 10 mm, height: 10 mm; GelMA bulk gel, B-GPL, S-GPL) were compressed using a universal testing machine (E43.104, MTS Industrial Systems) at a rate of 5 mm/min.

To evaluate swelling properties, hydrogels (GelMA bulk gel, B-GPL, S-GPL) were tested in PBS at 37 °C (n = 5). Briefly, pre-weighed hydrogels (500 μL, initial weight: m₀) were placed in nylon bags (initial wet bag weight: m₁) and submerged in PBS. At predetermined time intervals (0, 0.5, 1, 2, 4, 6, 8, 12, and 24 h), excess surface water was gently removed, and total weight (m₂) was measured. The swelling ratio (%) was calculated using the following equation:

$$Swellingratio(\%)=\frac{m_2-m_1}{m_0}\times 100\%$$

The fluorescent quantification assay was used to detect the release profile of hydrogels. In detail, 200 μL of precursor hydrogel solution (containing 300 μg/mL FITC-labeled MA-KLT and MA-RGI) was either directly dropped or pneumatically sprayed into 48-well plates and crosslinked in situ using blue light. After gelation, 1 mL of PBS was added to each well and incubated for 6 h to extract unbound peptides. The supernatant was then collected, and absorbance at 205 nm (A₂₀₅) was measured using a NanoDrop spectrophotometer. Peptide concentrations were calculated using a standard curve to estimate the fraction of unbound peptides.

The degradation behaviors of hydrogels (GelMA bulk gel, B-GPL, S-GPL) were assessed by placing equal masses of swollen hydrogels at equilibrium into vials containing PBS at 37 °C (n = 5). At predetermined intervals (every two days up to 24 days), samples were removed, lyophilized, and weighed. The remaining weight percentage was calculated according to the following equation:

$$Remainingweight(\%)=\frac{W_t-W_0}{W_0}\times 100\%$$

2.5. Antibacterial activity evaluation

The antibacterial activity of the hydrogel was evaluated using Escherichia coli (E. coli, ATCC 8739) and Staphylococcus aureus (S. aureus, ATCC 29213). Briefly, 50 μL of hydrogel precursor solution was added into a 96-well microplate and cross-linked under ultraviolet (UV) irradiation for 18–20 s for rapid gelation. Subsequently, 125 μL of bacterial suspension (10⁶ CFU/mL) was added onto the hydrogel surface. Wells containing bacterial suspension without hydrogel served as negative controls. The inoculated plates were incubated at 37 °C for 48 h, and optical density (OD) values were measured at 3, 6, 9, 12, 24, 36, and 48 h (n = 3). The outer wells of the plate were filled with water to prevent evaporation during incubation. Based on OD values, the 12-h time point was selected for further analysis. At this time point, the bacterial suspensions were diluted 10⁴-fold, and 100 μL of each diluted solution was spread onto solid agar plates, followed by incubation at 37 °C for 24 h. A bacterial suspension in PBS was utilized as a negative control. Bacterial colonies were photographed, and the killing ratio was calculated using the following equation:

$$Kill(\%)=\frac{Cellcountofcontrol-Survivorcountonhydrogel}{Cellcountofcontrol}\times 100\%$$

2.6. Cell culture

The human umbilical vein endothelial cells (HUVECs), rat Schwann cells-96 (RSCs) were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). HUVECs was conserved in an endothelial culture medium (Pricella) with 5 % FBS, 1 % PS, and 1 % endothelial cell growth supplement and then cultured in an incubator containing 5 % CO2 at 37 °C. RSCs was conserved in High-Glucose Dulbecco's Modified Eagle's Medium (high glucose, DMEM, Pricella) with 10 % FBS and 1 % PS and were also cultured in an incubator containing 5 % CO2 at 37 °C.

To evaluate the biological functions of different hydrogels, hydrogel extraction solutions were prepared according to ISO 10993-5. Briefly, 200 μL hydrogels were immersed in 1 mL of the corresponding culture medium and incubated at 37 °C for 72 h to obtain hydrogel extraction solutions. HUVECs and RSCs were cultured with these extraction solutions in 12-well culture plates, and a blank control group was established by culturing cells directly in the medium. HUVECs and RSCs were seeded at a density of 2 × 10⁴ cells/well and cultured with the respective hydrogel extraction solutions for predetermined durations prior to subsequent in vitro analyses.

2.7. Cell viability, proliferation and morphology assessment

The CCK-8 cell proliferation assay was performed as per the manufacturer's instructions. Specifically, HUVECs were planted in the 48-well plates alone or different hydrogel solutions at a density of 1 × 10⁴ cells/well. The medium was removed after 1, 4 and 7 days of incubation, and a fresh culture medium containing 10 % v/v Cell Counting Kit-8 (CCK-8, Dojindo Molecular Technologies Inc., Japan) was added to the 48-well plates, followed by incubation for 1 h at 37 C in a 5 % CO₂ humidified cell culture incubator in the absence of light. Subsequently, 100 μ L supernatant medium was transferred to the 96 well plates, and the absorbance at a wavelength of 450 nm was measured using a microplate spectrophotometer (EnSpire, PerkinElmer, USA). To evaluate cell compatibility of the hydrogels, HUVECs were planted in the 48-well plates alone or with different hydrogel solutions at a density of 2 × 10⁴ cells/well. Cells were then incubated with Calcein AM (2 μM) for 20 min, rinsed with PBS three times, and incubated with propidium iodide (PI) (2 μM) for 10 min further. And washed with PBS three times once more. After rinsing with PBS three times, the Live/Dead staining images were photographed with scanning confocal laser microscopy (Zeiss, Airyscan2-LSM980, Germany).

2.8. Hemolysis assay

The hemolysis ratios of different hydrogels were investigated based on a previously documented method [29]. Whole blood samples were collected from male New Zealand white rabbits weighing between 2.0 and 2.5 kg. Rabbit blood was centrifuged at 3500 rpm for 10 min, after which the erythrocytes were washed three times with phosphate-buffered saline (PBS). Subsequently, an erythrocyte suspension was prepared by diluting 5 mL of purified erythrocytes in 95 mL of PBS. Next, 500 μL of hydrogel solutions were thoroughly mixed with 500 μL of erythrocyte suspension in a 2 mL centrifuge tube. The tubes were incubated at 37 °C for 60 min, followed by centrifugation at 3500 rpm for 10 min. The absorbance of the supernatant was measured at 540 nm using a Tecan Spark 10 M automatic microplate reader (EnSpire, PerkinElmer, USA). All experiments were performed in triplicate for each sample. The hemolysis ratio (%) of each hydrogel was calculated using the following equation:

$$Hemolysisratio(\%)=\frac{As-An}{Ap-An}\times 100\%$$

where As, An, and Ap represent the absorbance values of the supernatant, negative control (PBS), and positive control (0.1 % Triton X-100) groups, respectively.

2.9. Antioxidant effects evaluation

The reactive oxygen species (ROS) scavenging abilities of GelMA, S-GPL, and S-GPL^KLT/RGI hydrogels were determined using the DPPH scavenging assay. Briefly, hydrogel samples were immersed in 1 mL of DPPH solution (0.1 mM) and incubated at 37 °C for 1 h ddH₂O was used as the blank control, and vitamin C was employed as the positive control. The absorbance at 517 nm was measured after incubation, and the DPPH scavenging efficiency was calculated according to the following equation:

$$Clearancerate(\%)=\frac{A_{blank}+A_{sample}}{A_{blank}}\times 100\%$$

where A_blank represents the absorbance of the blank control, and A_sample represents the absorbance of the sample solution after incubation for 1 h.

2.10. Anti-inflammatory effects evaluation

The murine RAW 264.7 macrophages were stimulated with 100 ng/mL lipopolysaccharide (LPS, Invitrogen, UK) and 20 ng/mL interferon-γ (IFN-γ, Abcam, UK) and divided into four groups: (1) Negative control (Control): macrophages cultured in normal medium without LPS stimulation; (2) GelMA group: LPS-stimulated macrophages treated with GelMA hydrogel; (3) S-GPL group: LPS-stimulated macrophages treated with S-GPL hydrogel; (4) S-GPL^KLT/RGI group: LPS-stimulated macrophages treated with S-GPL^KLT/RGI hydrogel. After 24 h of incubation, the supernatant was collected for ELISA (BYGR500009 and BYHS101447, Byabscience, China) to quantify pro-inflammatory cytokines IL-6 and TNF-α. Total RNA was extracted for qRT-PCR analysis of M1 polarization markers (CD86, TNF-α) and M2 markers (CD206, Arg-1).

2.11. Transwell migration and scratch wound healing assessment in vitro

HUVECs suspension was seeded into the upper 8.0 μm pore inserts (Labselect, China) containing serum-free medium. S-GPL, S-GPL^KLT, S-GPL^RGI and S-GPL^KLT/RGI hydrogels (n = 3) were prepared on the lower chamber of 24-well culture plates. After 24 h of incubation, the cells on the bottom side were stained with 0.5 % crystal violet to be quantified under an optical microscope.

The effect of different hydrogels on cell migration profiles was investigated with a scratch wound healing assessment. HUVECs were seeded on the 24-well culture plate at a density of 3 × 10⁴ cells/well until the cell confluence reached approximately 90 %. The monolayer was scratched with a 200 μL pipette tip and washed with PBS to remove floating cells. The remaining cells then received different interventions and were imaged at 0 and 24 h post-wounding. The proportion of the closure area to the initial scratch area was measured and analyzed with ImageJ 1.51k (Wayne Rasband, NIH, USA) to record the migration rate in the different groups.

2.12. Tube formation ability assessments in vitro

To assess the effect of hydrogels on tube formation, harvested HUVECs were seeded onto growth factor-reduced Matrigel (BD Biosciences, USA). Briefly, thawed Matrigel (10 μL/well) was added into a μ-Slide (IBIDI, Germany) and incubated at 37 °C for 30 min to solidify. After being cultured with S-GPL, S-GPL^KLT, S-GPL^RGI and S-GPL^KLT/RGI hydrogel solutions for 48 h, 5000 cells were transferred onto the solidified Matrigel in the μ-Slide and incubated for an additional 4 h. Tube formation was observed and photographed in five independent fields using the same microscope described above.

2.13. Indirect co-culture assessment in vitro

Indirect co-culture models were established to evaluate the paracrine signaling effects of HUVECs on RSCs. Briefly, HUVECs were seeded into upper chamber inserts (8.0 μm pore size, Labselect, China) and cultured respectively in fresh medium, S-GPL hydrogel solution, or S-GPL^KLT (or S-GPL^RGI) hydrogel solutions (n = 3) for pre-stimulation. After 3 days, the culture medium was removed, and cells were washed with PBS. Subsequently, RSCs were seeded at a density of 3 × 10⁴ cells/well into the lower chambers of 24-well culture plates (n = 4). Cells in both upper and lower chambers were then cultured in fresh corresponding medium for an additional 48 h. Cells from the lower chambers were harvested, and further analyses including qRT-PCR and immunofluorescence staining were performed as described.

2.14. Quantitative reverse transcription polymerase chain reaction (qRT- PCR) in vitro

Firstly, total RNA was extracted using the RNAsimple Total RNA Kit (Tiangen, China), and then the complementary DNA was synthesized using FastKing RT kit (Tiangen, China). The primers were synthesized by SeqHealth (Wuhan, China). SYBR Green supermix (Bio-Rad, USA) was utilized for amplification and detection of cDNA targets on a CFX96 real-time PCR detection system (Bio- Rad, USA). The relative gene expressions were normalized with the housekeeping gene, GAPDH, followed by the analysis carried out in 2 ΔΔCt method. The primer sequences of target genes used in qRT-PCR are presented in Table S1.

2.15. Immunofluorescence staining in vitro

For in vitro immunofluorescence staining, HUVECs and RSCs were performed with the fixation in 4 % paraformaldehyde (PFA) for 1 h at 4 °C after the preset culture period. Then, the samples were per meabilized in 0.1 % Triton X-100 (Sigma, USA) for 5 min, and blocked with 10 % normal goat serum (Solarbio, China) for 30 min. After incubation in primary antibody solutions overnight at 4 °C, the samples were rinsed with PBS three times and immersed in secondary antibodies and DAPI (sc-74421, Santa Cruz, USA) for 1 h at room temperature in the dark. The immunofluorescence staining images were photographed with scanning confocal laser microscopy, and quantification analysis was performed with ImageJ 1.51k (Wayne Rasband, NIH, USA).

2.16. Animal surgery

All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Peking Union Medical College Hospital (Beijing, China, approval number: MDKN-2024-033). Eight-week-old male Sprague−Dawley (SD) rats (200–250 g, n = 30), were used in this study. A diabetic rat model was introduced using intraperitoneal injection of premade streptozotocin (STZ, Sigma-Aldrich, USA) solution at a dose of 60 mg/kg, and considered as successfully established if the blood glucose level of the rat was above 16.7 mmol/L one week after the injection. The rats were anesthetized with tribromoethanol vapor inhalation. Then, two full-thickness round-shaped skin wound (1 cm diameter) was created on the bilateral lower back of the rats, symmetrically aligned along the spine. The rats were randomly divided into five groups, while the groups were treated with: (1) saline solution (Control group); (2) S-GPL alone (S-GPL group); (3) S-GPL with KLT (S-GPL^KLT group); (4) S-GPL with RGI (S-GPL^RGI group); and (5) S-GPL with KLT and RGI (S-GPL^KLT + RGI group). A standardized spraying protocol was employed to control the thickness of the sprayed hydrogel layer. On Day 0, 200 μL of the precursor solution was sprayed from 10 cm onto the defect area and subsequently crosslinked using a 405 nm portable UV light source (EFL-LS-1600-405) for 18 s, resulting in a hydrogel layer approximately 4 mm in thickness. Following application, the hydrogel was not removed and was allowed to degrade spontaneously in situ. The control group received an equivalent volume of saline solution. The wounds were photographed 0,3,7,10 and 14 days post-surgery, and wound areas were measured by Image Pro Plus 6.0 (Media Cybernetics, Silver Spring, USA) software. The wound healing rate was determined using the formula:

$$Woundhealing(\%)=\frac{S_0-S_t}{St}\times 100\%$$

where S_t and S₀ denote the wound areas on the observation day (t) and immediately after the operation, respectively.

2.17. Laser speckle test

Seven days post-wounding, blood perfusion at the wound site was assessed using a laser speckle contrast imaging system (moorFLPI-2, Moor Instruments, UK), and images were captured with the moorFLPI software (V6.0). Blood perfusion was analyzed using an invisible near-infrared laser at a wavelength of 785 nm.

2.18. Histological evaluation

On days 7 and 14 post-operation, half skin samples from each of the two wound sites were collected, combined and fixed with 4 % (w/v) paraformaldehyde for 48 h, embedded in optimal cutting temperature compound (OCT, Tissue-Tek OCT Compound, USA), and sliced into 10 μ m-thick histologic sections using a cryostat microtome (CM 1950, Leica, Germany). The tissue sections were further prepared for standard hematoxylin and eosin (H&E) staining, Masson staining, immunohistochemistry and immunofluorescence staining. With immunohistochemistry, the inflammatory factors IL-6 and IL-17 were evaluated. With immunofluorescence staining, CD31 (Platelet endothelial cell adhesion molecule-1, PECAM-1/CD31) and α-SMA (α-Smooth Muscle Actin, α-SMA) antibodies were used for colocalization of newly generated vessels, and MBP (myelin basic protein, MBP) and PGP9.5 (Ubiquitin carboxyl-terminal hydrolase isozyme L1, UCHL1/PGP 9.5) were utilized to evaluate peripheral nerve regeneration. Additionally, CD31 and PGP9.5 were co-stained to observe the relative positions of nerves and vessels. Microscopy imaging was conducted using a Pannoramic SCAN scanner (3DHIESTECH, Hungary) and processed using CaseCenter 2.9SP1 software (3DHIESTECH, Hungary). To assess the in vivo toxicity of the hydrogel, the main organs (heart, liver, spleen, lung, and kidney) were harvested for H&E staining at 14 days. Inflammatory cells were observed following H&E staining, and the degree of staining was analyzed with ImageJ 1.51k (Wayne Rasband, NIH, USA).

The primary and secondary antibodies used in histological staining are listed in Table S2.

2.19. Transcriptome sequencing

The other half of skin samples collected on day7 and day14 were frozen with nitrogen and conducted RNA extraction and quality assessment. RNA-sequencing (RNA-seq) was performed using a NovaSeq Xplus sequencer (Illumina, San Diego, CA, USA). Differential expression genes (DEGs) were compared between the control and S-GPL^KLT/RGI groups. Downstream enrichment analysis including Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment and Gene Set Enrichment Analysis (GSEA) were conducted. Protein–protein interaction (PPI) analysis was performed using the STRING database (Search Tool for the Retrieval of Interacting Genes/Proteins; https://string-db.org).

2.20. Statistical analysis

Statistical comparisons were performed using GraphPad Prism software (10.2.2, San Diego, USA). One-way or two-way ANOVA was applied where appropriate, followed by Tukey's post hoc test for pairwise comparisons. Data are presented as mean ± SD, with significance indicated as follows: ∗∗∗∗p < 0.0001, ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, and ns denoting no statistical significance (p > 0.05)

3. Results and discussion

3.1. Fabrication and characterization of S-GPL

To address the unique challenges of diabetic wounds, we developed a sprayable, shape-adaptive hydrogel based on a GelMA-PLMA network system. GelMA synthesis was confirmed by ¹H nuclear magnetic resonance (¹H NMR) analysis (Fig. S1A), which showed the appearance of characteristic vinyl proton signals (5.2–5.6 ppm) absent in unmodified gelatin, verifying successful methacrylate grafting. Fourier transform infrared (FTIR) spectroscopy further verified the chemical structures of GelMA, PLMA, and the composite GelMA-PLMA hydrogel (Fig. S1B). GelMA displayed typical amide I (1630 cm⁻¹, C=O stretching) and II (1535 cm⁻¹, N–H bending) peaks, along with a 1720 cm⁻¹ methacryloyl band. PLMA exhibited similar amide peaks plus a distinct band at 1725 cm⁻¹ (methacrylamide C=O stretch), confirming successful ε-PL methacrylation. In the composite GelMA-PLMA hydrogel, the amide I/II bands of GelMA (1630 cm⁻¹ and 1535 cm⁻¹) and the corresponding PLMA bands (1640 cm⁻¹ and 1540 cm⁻¹) remained distinct. Meanwhile, their methacryloyl-related peaks (1720–1725 cm⁻¹) merged into a single broadened peak, suggesting physical entanglement between the two networks. Notably, the GelMA's amide I band shifted slightly (1630 → 1625 cm⁻¹), indicating potential hydrogen bonding between GelMA carboxyl groups and PLMA amino groups. Overall, these results confirm that each component retained its functional groups and interacted mainly via non-covalent forces (e.g., hydrogen bonding, electrostatic attraction) in the composite. The lack of new covalent bonds indicates that the hybrid network relies primarily on physical blending—an important feature for preserving the bioactivity of both components while ensuring structural stability.

Sprayable hydrogels provide a convenient, non-invasive method for delivering therapeutics with uniform coverage and rapid in situ gelation [30]. Their ease of use and multifunctionality make them promising for improving treatment outcomes and reducing the burden of chronic wound care. While previous studies have largely emphasized the ease of application via spraying, a critical research gap remains regarding how the spray process influences key material properties—such as rheology, swelling capacity, and structural stability—that are essential for clinical efficacy [31]. To investigate these effects, we compared the microstructure and physical properties of three hydrogels: GelMA, B-GPL (GelMA–PLMA bulk hydrogel), and S-GPL (GelMA–PLMA sprayable hydrogel). Cross-sectional SEM images confirmed that high-pressure spraying introduced microbubbles into S-GPL, yielding a layered porous architecture, while B-GPL exhibited an almost smooth microstructure (Fig. 1A and B). The bubbles primarily originate from the shearing of air by the liquid during spraying. Due to rapid photopolymerization, the bubbles are trapped before they can escape, forming a porous architecture within the gel [32]. Specifically, S-GPL exhibited a reproducible pore size distribution, with a prominent peak in the 10–30 μm range. (Fig. S2). Compared with in situ–formed GelMA and B-GPL, spray-fabricated S-GPL exhibits markedly higher porosity (67.00 ± 6.78 %) and specific surface area (21.72 ± 3.01 m²/g), outperforming GelMA (17.00 ± 3.16 %, 2.22 ± 0.80 m²/g) and B-GPL (30.40 ± 3.44 %, 5.36 ± 1.55 m²/g) (Fig. S3). In addition, gelation of 200 μL precursor solution resulted in final volumes of 295.7 ± 18.8 mm³ for S-GPL and 189.3 ± 2.5 mm³ for B-GPL (Fig. S4), indicating a volume increase due to the porous structure. Such a structure resembles native ECM, promoting cell adhesion, nutrient exchange, and tissue integration [33,34]. In line with this, a 3D-printed hydrogel with optimized porosity significantly accelerated wound healing [35,36], underscoring the importance of structural optimization for tissue regeneration.

Fig. 1.

Characterization of sprayable hydrogel. (A, B) Cross-sectional SEM images showing the porous microstructures of (A) B-GPL and (B) S-GPL hydrogels (scale bars = 20 μm). (C, D) Rheological performance of the hydrogels: (C) strain sweep and (D) frequency sweep curves (n = 3). (E) Mechanical properties, including compressive strength and fracture strain of the hydrogels (n = 3). (F) Swelling ratios in PBS at 37 °C for 24 h (n = 3). (G) Release ratio of KLT and RGI peptides from B-GPL^KLT/RGI, S-GPL^KLT/RGI, and their counterparts lacking MA conjugation (+KLT/RGI) (n = 3). (H, I) Joint adaptation and shape-conformability of S-GPL hydrogel. (J) OD600 values of E. coli and S. aureus cultured on GelMA hydrogels with varying PLMA concentrations over 0–48 h (n = 5). (K) Representative images of bacterial colonies on agar plates after 12 h of co-culture with different hydrogels. Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001.

The viscoelasticity properties of hydrogels were then evaluated by rheological measurements. In the linear viscoelastic region (γ = 0.1–50 %), S-GPL's storage modulus remained nearly constant (Fig. 1C), indicating it can withstand significant strain without structural failure. As shown in Fig. 1D, all hydrogels exhibited typical gel-like behavior (G′ > G″) over 0.01–5 Hz, reflecting elastic dominance and structural stability in this frequency range. Notably, the dual-network hydrogels (B-GPL and S-GPL) exhibited higher G′ values than GelMA (Fig. 1D), indicating that incorporating PLMA increases network density and stiffness. This enhanced mechanical stability is beneficial for maintaining hydrogel integrity under physiological conditions, especially in dynamic wound sites (e.g., joints). Moreover, the sprayed S-GPL showed slightly lower compressive strength and fracture strain than B-GPL, but a higher strain-at-break (Fig. 1E and S5), indicating greater deformability. Additionally, S-GPL had a very low loss factor (tan δ ≈ 0.05), signifying a predominantly elastic response.

To further assess functional performance, swelling and degradation behaviors were analyzed. As illustrated in Fig. 1F, S-GPL had the highest swelling ratio (193.43 % ± 13.97 %), significantly outperforming GelMA and B-GPL. This high swelling capacity is likely due to the spray-induced porous architecture, which promotes fluid uptake—a desirable feature for exudate management in DWs [37]. In addition, the S-GPL^KLT/RGI hydrogel achieved sustained peptide release over ∼23 days (Fig. 1G), exhibiting a “slow–fast–slow” pattern: an initial lag phase due to the dense MA-crosslinked network, followed by accelerated release during degradation and pore expansion, and a final plateau as peptides were depleted. Only 1.16 ± 0.54 % of the peptide was detected in the initial elution from S-GPL^KLT/RGI, indicating that over 98 % of KLT and RGI were effectively grafted or retained within the hydrogel (Fig. S6). In contrast, in the absence of MA conjugation, S-GPL + KLT/RGI released nearly all peptides within 5 days, confirming that MA crosslinking is essential for sustained delivery. The release profile of S-GPL^KLT/RGI aligned well with its ∼20-day degradation period (Fig. S7) and the typical timeline of chronic wound healing [38]. This controlled degradation results from hydrolysis of ester and amide bonds and enzymatic cleavage of the GelMA-PLMA backbone, enabling prolonged therapeutic release. Together, these mechanisms facilitate gradual matrix degradation and enable prolonged release of therapeutic peptides.

To evaluate the practical applicability of the S-GPL hydrogel, we tested its performance in adapting to both dynamic and irregular wound environments. In a flexed joint model (Fig. 1H), S-GPL was evenly sprayed onto a bent finger and photo-crosslinked under 405 nm light for 18 s (based on gelation kinetics; Fig. S8). The hydrogel conformed seamlessly to the joint curvature and retained adhesion after movement, indicating excellent morpho-adaptability and mechanical compliance suitable for dynamic anatomical sites. Similarly, on an irregularly shaped wound model (Fig. 1I), the sprayed hydrogel uniformly filled the wound cavity and photocured into a conformal dressing, highlighting its precise shape adaptability for sealing complex wound geometries.

In conclusion, combining a dual-network architecture (which increases crosslink density and mechanical robustness) with spray-induced porous architecture (which creates a sponge-like structure) endows the hydrogel with enhanced elasticity, exudate absorption capacity, and adaptability to complex surface geometries. Consistent with previous studies emphasizing the role of porous elasticity in wound healing [39], S-GPL achieves a favorable balance between structural integrity, adaptability, and bioactivity, making it a promising candidate for chronic wound management. Clinically, we recommend supplementary spraying as needed based on exudate, infection risk, and hydrogel degradation. If significant degradation or infection occurs, timely replenishment or replacement is recommended to maintain efficacy and safety.

3.2. Antibacterial properties and biocompatibility of S-GPL

The antimicrobial efficacy of ε-PL can be influenced by factors such as molecular weight and electrostatic masking resulting from methacrylate. Tu et al. reported that ε-PL exhibits a minimum inhibitory concentration of 0.064 mg/mL against Staphylococcus aureus (S. aureus) and 16.384 mg/mL against Escherichia coli (E. coli) [40]. However, ε-PL has been shown to be immunogenic, and high ε-PL levels can be cytotoxic in vitro [41], making it important to optimize its dosage. Currently, no guidelines exist for the optimal ε-PL content in dual-network hydrogels. Therefore, we formulated hydrogels with 5 % GelMA and varying PLMA concentrations (1 %, 3 %, 5 %; denoted S-GPL1 %, S-GPL3 %, S-GPL5 %) to identify the composition that best balances antibacterial activity and biocompatibility.

To evaluate the antibacterial properties, bacterial contact-killing assays were performed using S. aureus and E. coli, which are the two most common pathogen in DWs [42]. Experimental groups included S-GPL1 %, S-GPL %, S-GPL5 %, with GelMA and PBS as controls. The optical density at 600 nm (OD600) values at different times were measured over a 48 h period. As shown in Fig. 1J–S aureus proliferated rapidly in the control and GelMA groups (OD600 peaking at ∼24 h). In contrast, S-GPL1 % exhibited delayed and significantly suppressed bacterial growth, while S-GPL3 % and S-GPL5 % maintained OD600 values close to baseline throughout the 48 h period, indicating complete bacterial inhibition. A slightly weaker antibacterial effect was observed with E. coli: control and GelMA showed a sharp OD600 increase (∼12 h peak), whereas S-GPL3 % and S-GPL5 % almost completely suppressed bacterial growth (minimal OD increase), and S-GPL1 % achieved only moderate suppression. These results were corroborated by agar plate cultures with bacterial solutions after 12 h (Fig. 1K). S-GPL3 % exhibited antibacterial rates of 97.79 % against E. coli and 98.96 % against S. aureus, while S-GPL5 % achieved 99.67 % and 100 % inhibition, respectively. Overall, both S-GPL3 % and S-GPL5 % hydrogels exhibited potent and sustained antibacterial activity against both S. aureus and E. coli, indicating their strong potential for use in infection-resistant wound dressings. Given the potential confounding effects of infection on the assessment of neurovascular interaction, we did not incorporate infected wound models. Nonetheless, establishing such models will be an important next step to evaluate the therapeutic efficacy and translational relevance of the hydrogel system in clinically representative settings.

To assess cytocompatibility, HUVECs were cultured with extracts from S-GPL1 %, S-GPL3 % and S-GPL5 %, with GelMA and complete medium as controls. Cell viability (CCK-8 assay) increased over time in all groups (days 1, 4, 7), but S-GPL3 % led to significantly higher OD450 values than the other formulations on days 4 and 7 (Fig. S9), indicating that 3 % PLMA most effectively supported HUVEC proliferation. In contrast, the GP5 % group demonstrated a slight decrease in OD450 compared to the control on both days, indicating that excessive PLMA content may exert an inhibitory effect on cell growth. Live/Dead staining confirmed these trends (Fig. S10), affirming that 3 % PLMA incorporation improves hydrogel biocompatibility. Cytoskeletal staining showed that HUVECs cultured on the S-GPL3 % hydrogel exhibited more pronounced cellular spreading and cytoskeletal extension (Fig. S11). Furthermore, S-GPL3 % exhibited excellent blood compatibility with a hemolysis rate of ∼0.5 % (Fig. S12), well below the 2 % threshold (ASTM F756-00) for non-hemolytic materials. Studies have shown that ε-PL can improve goat sperm quality by enhancing antioxidant capacity [43]. As shown in Fig. S13, S-GPL3 % exhibited a DPPH radical scavenging rate of ∼35 %, suggesting that mitigating oxidative stress may contribute to the promotion of cell proliferation [44]. However, this effect may be compromised at higher concentrations, as the increased cationic charge density of PLMA can disrupt cell membranes or nutrient transport, thereby reducing biocompatibility [45].

Together, these findings demonstrate that while PLMA augments the antibacterial activity of S-GPL hydrogels, its concentration must be optimized to ensure cytocompatibility. Among the variants tested, S-GPL3 % achieved the best balance of potent antimicrobial efficacy and favorable biocompatibility, making it the most promising formulation. Therefore, we selected 3 % PLMA as the standard composition (S-GPL refers to this formulation for all other experiments).

3.3. GPL^KLT/RGI promotes endothelial cell recruitment and vascularization in vitro

Angiogenesis is a critical prerequisite for DW healing, as newly formed blood vessels provide essential oxygen and nutrients to regenerating tissues [46]. However, hyperglycemia-induced protease activity leads to endothelial progenitor cell damage and VEGF degradation [[47], [48], [49]], severely impairing VEGF-driven angiogenesis at the wound site [50]. To overcome this limitation, we incorporated the VEGF-mimetic peptide KLT (analogous to VEGF helix 17–25 sequence) and evaluated the effects of hydrogel-released factors on endothelial cell behavior. Transwell migration assays showed that significantly more HUVECs migrated in the presence of GPL^KLT and GPL^KLT/RGI extracts compared to GPL alone (Fig. 2A and D), indicating that KLT enhances endothelial recruitment. Notably, even GPL (without KLT) induced more HUVEC migration than blank medium (227 ± 30 cells/field), likely due to the proliferative influence of ε-PL. Scratch wound assays further corroborated these results (Fig. 2B and E).

Fig. 2.

GPL^KLT/RGI promotes endothelial cell recruitment and angiogenic differentiation in vitro. (A) Transwell assay showing HUVECs migration after treatment with different hydrogel extracts (scale bar = 100 μm). (B) Representative images of the scratch assay at 0 and 24 h (scale bar = 200 μm). (C) Tube formation assay demonstrating capillary-like network formation by HUVECs cultured with hydrogel extracts (scale bar = 100 μm). (D) Quantification of migrated cells in Transwell assay (n = 3). (E) Quantitative analysis of scratch closure (n = 3). (F, G) Semi-quantitative analysis of tube formation: (F) number of junctions and (G) segment length (n = 3). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n. s.: not significant.

We also assessed angiogenic capacity using a Matrigel-based tube formation assay. HUVECs treated with GPL^KLT and GPL^KLT/RGI extracts formed a greater number of junctions and longer capillary-like networks than those treated with GPL extracts (Fig. 2C–F and G), while GPL worked modestly compared to blank medium. These findings indicate that under the antioxidative microenvironment provided by GPL, the inclusion of KLT markedly promotes endothelial recruitment and neovessel formation, potentially contributing to improved wound healing.

3.4. GPL^KLT/RGI promotes the maturation and paracrine functions of HUVECs and RSCs in vitro

We next investigated the functional activation of HUVECs. CD31 is a marker of endothelial maturity, and its elevated expression reflects a more angiogenic phenotype [51]. Immunofluorescence staining revealed markedly higher CD31 expression in HUVECs treated with GPL^KLT or GPL^KLT/RGI, with cells organizing into vessel-like structures (Fig. 3A), suggesting that KLT promotes endothelial maturation and differentiation. This observation was corroborated by increased CD31 mRNA levels in these groups (Fig. 3B). In addition to promoting phenotypic maturation, KLT enhanced the paracrine function of HUVECs. As a VEGF-mimetic, KLT not only directly stimulates VEGF receptors but also upregulated the HUVECs’ own VEGF expression (Fig. 3B), amplifying angiogenic signaling. Importantly, KLT treatment triggered the release of downstream factors including PDGF and FGF-2 (Fig. 3B). PDGF recruits fibroblasts and promotes collagen deposition and granulation tissue formation [52], while FGF-2 facilitates angiogenesis and stimulates keratinocyte proliferation and migration for re-epithelialization [53]; together, these factors act synergistically in DW healing [54].

Fig. 3.

GPL^KLT/RGI promotes endothelial and Schwann cell activation in vitro. (A) Representative immunofluorescence images of HUVECs stained for Actin (red), DAPI (blue), and CD31 (green) after 72 h incubation with extracts from GelMA, GPL^KLT, GPL^RGI, or GPL^KLT/RGI hydrogels (scale bar = 100 μm). (B) Relative gene expression of HUVECs after 72 h treatment with different hydrogel extracts (n = 5). (C) Relative gene expression of RSCs following 72 h treatment with hydrogel extracts (n = 5). Data are presented as mean ± SD. ∗P < 0.05, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n. s.: not significant.

Recent studies have shown that brain-derived neurotrophic factor (BDNF) plays a broad reparative role in peripheral tissues, not just the central nervous system [55]. In this study, we employed the functional peptide RGI, which mimics BDNF's key active sequence. RSCs were cultured with hydrogel extracts for 3 days and analyzed by qRT- PCR. RGI exposure (in GPL^RGI and GPL^KLT/RGI) downregulated the immature Schwann cell marker NCAM and upregulated the activation marker S100 and the myelination-associated protein PMP22 (Fig. 3C), indicating that RGI drives Schwann cells toward a myelinating phenotype [50]. RGI also induced higher expression of endogenous neurotrophic factors BDNF and nerve growth factor (NGF), further enhancing the regenerative potential. Interestingly, combining KLT with RGI yielded even greater effects: the GPL^KLT/RGI group showed significantly higher expression of these genes (except BDNF) compared to GPL^RGI alone. Schwann cells have been reported to respond to VEGF signals from various cellular sources, including endothelial cells and macrophages, thereby enhancing their proliferation, migration, and differentiation [56,57]. These findings imply that KLT may further enhance Schwann cell activation through VEGF receptor-mediated signaling, thereby revealing a potential cellular basis for neurovascular interaction.

Collectively, these results demonstrate that GPL^KLT/RGI orchestrates both endothelial and Schwann cell activation, thereby initiating a multifactorial network of regenerative signals for coordinated tissue regeneration in DWs. Notably, the overlapping effects of KLT on Schwann cells suggest potential cross-talk between neurovascular compartments. These findings highlight the capacity of GPL^KLT/RGI to activate parallel, interconnected pathways and reconstruct a functional neurovascular microenvironment critical for DW repair [58,59].

3.5. GPL^KLT induced HUVECs enhance RSCs activation through paracrine signaling

The close interplay between angiogenesis and neurogenesis is well recognized, as blood vessels often precede Schwann cell migration and axonal extension during nerve repair [60]. Building on our findings that GPL^KLT amplifies endothelial endogenous VEGF expression, and that Schwann cells respond to KLT-induced VEGF signaling pathway,we hypothesized that GPL^KLT-activated HUVECs might enhance RSCs function via soluble factors. To test this, we established an indirect transwell co-culture system (Fig. 4A) wherein HUVECs pretreated with GPL or GPL^KLT for 72 h were placed in the upper chamber and RSCs in the lower chamber (allowing only cytokines to diffuse between layers). This model enabled us to isolate the effect of HUVEC-derived factors on RSCs and examine whether hydrogel-induced endothelial activation can promote neurovascular microenvironment formation.

Fig. 4.

GPL^KLT induced HUVECs enhance RSCs activation. (A) Schematic illustration of the transwell co-culture system. Pre-induced HUVECs were seeded in the upper chamber, and RSCs were cultured in the lower chamber. (B, C) Relative gene expression of RSCs: (B) neurotrophic secretion-related genes and (C) maturation-related genes after monoculture or co-culture with HUVECs, HUVECs pre-induced by GPL or HUVECs pre-induced by GPL^KLT for 72 h (n = 3). (D) Representative immunofluorescence images of RSCs stained for MBP (green), Actin (red), and DAPI (blue). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗∗P < 0.0001, n. s.: not significant.

As shown in Fig. 4B and C, HUVECs activated by GPL^KLT significantly upregulated NGF expression in co-cultured RSCs, along with Schwann cell maturation markers (increased S100, decreased NCAM). Immunofluorescence staining further confirmed that RSCs co-cultured with KLT-pretreated HUVECs exhibited intense Myelin Basic Protein (MBP) signals, in contrast to the faint myelin formation in controls (Fig. 4D). Morphologically, these RSCs appeared more elongated and interconnected, suggesting that endothelial paracrine cues not only induce pro-myelination gene expression but also promote cytoskeletal arrangements conducive to Schwann cell migration and interaction. These observations reinforce that HUVECs – especially when primed by KLT – can, through secreted factors alone, substantially enhance Schwann cell maturation and readiness for nerve regeneration [61].

Overall, GPL^KLT-preconditioned HUVECs powerfully enhance Schwann cell activation via paracrine signaling. This finding aligns with emerging evidence that endothelial–Schwann cell cross-talk is pivotal in nerve repair. For instance, endothelial cell-derived exosomes have been shown to boost and maintain repair-supportive phenotypes in Schwann cells, thereby accelerating axonal regeneration and remyelination [62]. In our system, GPL^KLT provides a sustained VEGF-like stimulus, greatly amplifying endothelial secretion of beneficial factors. This synergistic effect of KLT underscores the biointeractive design of our dual-network hydrogel, which simultaneously promotes vascular activation and glial support.

3.6. S-GPL^KLT/RGI hydrogel accelerates DW healing in vivo

To evaluate therapeutic efficacy, we applied the biofunctional hydrogels in a diabetic rat skin wound model. Two full-thickness wounds (1 cm diameter) were created on each rat's dorsum and treated with S-GPL, S-GPL^KLT, S-GPL^RGI, or S-GPL^KLT/RGI hydrogels, with saline as a control (Fig. 5A). Wound closure was monitored on days 0, 3, 7, 10, and 14. S-GPL^KLT/RGI showed the most rapid closure at all time points (Fig. 5B and C), whereas control wounds exhibited Delayed healing, as expected for DWs. S-GPL^KLT and S-GPL^RGI also accelerated wound repair compared to S-GPL and control, with S-GPL^KLT more effective than S-GPL^RGI in the early phase. Notably, S-GPL alone yielded moderate improvement over control, indicating that the dual-network matrix itself provides a favorable healing environment. By day 14, S-GPL^KLT/RGI-treated wounds had the smallest residual area (8.9 % ± 2.6 % of original), followed by S-GPL^KLT (13.6 % ± 1.6 %), S-GPL^RGI (15.0 % ± 0.8 %), S-GPL (21.5 % ± 2.0 %), and control (31.3 % ± 4.7 %) (Fig. 5E). Consistently, laser speckle imaging on day 7 showed enhanced blood perfusion in the S-GPL^KLT and S-GPL^KLT/RGI groups relative to control and S-GPL, with S-GPL^KLT/RGI exhibiting the highest perfusion (Fig. 5D and F). S-GPL^KLT/RGI significantly improved perfusion (P < 0.05), indicating a synergistic pro-angiogenic effect of the two peptides. Importantly, H&E staining of major organs in the S-GPL^KLT/RGI group revealed normal tissue structure (Fig. S14), confirming the hydrogel's good systemic biocompatibility.

Fig. 5.

S-GPL^KLT/RGI accelerates wound closure in a diabetic rat model. (A) Schematic illustration of the sprayable hydrogel treatment applied to full-thickness wounds in streptozotocin-induced diabetic rats. (B) Representative wound images captured on days 0, 3, 7, 10, and 14 post-treatments. (C) Schematic diagram summarizing the wound healing process under different treatment conditions. (D) Representative laser speckle contrast images of blood perfusion in wound areas on day 7. (E) Quantification of wound closure rate over time (n ≥ 3), with statistical comparison performed between the control group and the S-GPL^KLT/RGI group. (F) Quantitative analysis of perfusion intensity (n = 3). Data are presented as mean ± SD. ∗P < 0.05, ∗∗∗∗P < 0.0001, n. s.: not significant.

3.7. S-GPL^KLT/RGI hydrogel enhances the tissue reconstruction and modulates the angiogenic-neurogenic microenvironments in vivo

Re-epithelialization and collagen deposition are pivotal for rebuilding skin tissue during wound healing [63]. To further assess tissue regeneration, H&E and Masson's trichrome staining were performed on day 14 (Fig. 6A and B). All S-GPL-treated wounds showed greater epidermal thickness and collagen content than controls (Fig. 6F and G), indicating that the S-GPL matrix provided a favorable microenvironment for wound healing. Among the treatments, S-GPL^KLT showed superior effects over S-GPL^RGI, while S-GPL^KLT/RGI produced the greatest epithelial thickness (52.7 % ± 4.5 %) and highest collagen deposition (236.8 % ± 5.8 % of control). These findings suggest that KLT exerts a more direct impact on tissue regeneration, whereas RGI enhances the regenerative effect synergistically. This may be attributed to the fact that VEGF not only promotes angiogenesis but also exerts non-endothelial effects by directly stimulating keratinocyte survival, proliferation, and migration [[64], [65], [66]]. The synergistic role of RGI requires to be further investigated. Effective re-epithelialization is essential for restoring the epidermal barrier and preventing microbial invasion [67], while organized collagen deposition provides the mechanical scaffold necessary for dermal regeneration and tensile strength restoration [68]—both are key determinants of successful healing in chronic DWs. Collectively, the histological data indicate that S-GPL^KLT/RGI promotes both the structural and functional maturation of regenerated skin.

Fig. 6.

S-GPL^KLT/RGI promotes cutaneous tissue reconstruction and modulates the angiogenic–neurogenic microenvironment. (A, B) Representative H&E and Masson's trichrome staining of wound tissues on day 14 post-treatment. (C) Representative immunofluorescence images of CD31 and α-SMA expression on day 7 (scale bars = 1 mm). (D) Representative immunofluorescence images of MBP and PGP9.5 expression on day 14 (scale bars = 500 μm). (E) Dual immunofluorescence of CD31 and PGP9.5 showing spatial colocalization of blood vessels and nerve fibers on day 14 (scale bars = 500 μm). (F, G) Quantification of (F) epithelial thickness and (G) collagen deposition (n = 3). (H, I) Quantification of (H) CD31⁺ area and (I) the number of CD31⁺/α-SMA⁺ vessel (n = 3). (J) Quantification of regenerated nerve fibers based on PGP9.5 and MBP staining (n = 3). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n. s.: not significant.

To further investigate the angiogenesis during the proliferative phase, immunofluorescence staining was first performed. CD31 and α-SMA are classical markers of vascular development: CD31 marks newly formed endothelium in the early stages of angiogenesis, while α-SMA indicates the recruitment of perivascular cells (primarily smooth muscle cells or pericytes) and reflects vessel maturation [69]. At day 7, immunostaining revealed nascent blood vessels beneath the wound bed in all hydrogel-treated groups (Fig. 6C). S-GPL^KLT and S-GPL^KLT/RGI significantly increased the area of CD31⁺ vessels compared to controls (Fig. 6H), indicating a higher capillary density. Notably, the number of CD31⁺/α-SMA⁺ double-positive vessels was further elevated in the S-GPL^RGI and S-GPL^KLT/RGI group (Fig. 6I), suggesting that RGI facilitates the maturation of new vasculature beyond the capillary stage. This finding aligns with previous studies that neurotrophic cues can modulate the perivascular environment and support vessel stabilization, potentially by recruiting pericytes and smooth muscle cells [70].

Peripheral nerve regeneration was assessed at day 14 by co-staining for MBP and PGP9.5. Co-localization of MBP⁺ Schwann cells and PGP9.5⁺ axons confirmed the presence of newly myelinated nerve fibers, primarily located in the lower reticular dermis (Fig. 6D). Quantitatively, S-GPL^RGI significantly increased nerve fiber density, while S-GPL^KLT/RGI exhibited further enhanced effects (Fig. 6J). S-GPL^KLT also showed a statistically significant improvement over the S-GPL. These results suggest that RGI plays a predominant role in activating Schwann cells and promoting myelination, whereas KLT may further support neural regeneration indirectly through the paracrine effects we reported and by enhancing vascularization. Restoration of cutaneous innervation is crucial in chronic DW healing [71], and reinnervation with myelinated fibers provides more stable functional recovery and helps prevent chronic sensory disturbances [72]. Thus, the S-GPL^KLT/RGI hydrogel offers a promising strategy to achieve targeted reinnervation and functional repair in DWs complicated by peripheral neuropathy.

The reconstruction of the neurovascular microenvironment has emerged as a pivotal strategy in the regeneration of diverse tissues, including the central nervous system [73], dental pulp [74], and skeletal muscle [75]. As a specialized microenvironment, the neurovascular microenvironment provides localized trophic support and paracrine cues while also interfacing with the circulatory system to modulate nutrient delivery, immune homeostasis, and progenitor cell mobilization [76,77]. In DWs, where vascular insufficiency and peripheral neuropathy are prominent pathological features [78,79], restoring a functional neurovascular microenvironment is particularly important for enabling coordinated repair. To visualize this relationship, co-immunostaining of CD31 and PGP9.5 was performed on day 14 to investigate the spatial relationship between regenerated vessels and nerves (Fig. 6E). S-GPL^KLT, S-GPL^RGI and S-GPL^KLT/RGI groups all exhibited niche-like structures, characterized by CD31⁺ capillaries closely opposed to PGP9.5⁺ nerve fibers. However, differences in the ratio of vascular to neural components were observed among these groups, suggesting that the degree of neurovascular coupling may vary depending on the specific peptide incorporated. The observed organization reinforces the notion that restoring spatial neurovascular coordination is essential for orchestrated tissue regeneration. In particular, the tight association of vessels and nerves in the S-GPL^KLT/RGI group suggests the presence of a refined microenvironment that simultaneously provides nutritional perfusion, thereby facilitating effective tissue repair. Notably, the optimal balance of neural and vascular components in a regenerative niche remains undefined; future studies should explore the dose-dependent contributions of KLT and RGI to optimize neurovascular coupling in chronic wound healing.

To further investigate the mechanism of neurovascular coupling, we curated a neurovascular interaction gene set comprising 48 literature-validated genes involved in neurovascular development, vascular-guided Schwann cell migration, and reciprocal signaling [70,[80], [81], [82], [83], [84], [85]]. GSEA revealed significant enrichment of this gene set in the S-GPL^KLT/RGI group, highlighting active neurovascular reprogramming (Fig. S15). Representative genes included Shh, Sema3a, and Wnt family members (Wnt4/5a/5b/6), all known to regulate neurovascular patterning [82,86]. Protein–protein interaction (PPI) analysis (STRING database) based on this gene set identified major axes such as VEGF–NRP1, BDNF–MAPK3, and CXCL12–CXCR4 as central nodes in the network (Fig. S16), suggesting that the observed regeneration is driven by tightly coordinated signaling across cell types. Given that the other two axes are intrinsically activated by mimetic peptides, we selected the CXCL12–CXCR4 axis for further in vitro validation. Transwell assays demonstrated that either rhCXCL12 or conditioned medium from S-GPL^KLT/RGI–treated RSCs (S-GPL^KLT/RGI RSC_S-CM) significantly enhanced HUVEC migration, while the CXCR4 antagonist AMD3100 markedly suppressed this effect (by 73.4 % and 64.2 %, respectively), confirming a key paracrine role of Schwann cells in directing endothelial migration (Fig. S17). These results suggest that the CXCL12–CXCR4 axis plays an important, but not exclusive, role in mediating endothelial migration induced by S-GPL^KLT/RGI–stimulated Schwann cells. Further exploration of the underlying mechanisms governing neurovascular crosstalk should be a key direction for future research.

Taken together, S-GPL^KLT/RGI promotes not only vessel and neural regeneration but also their spatial integration into a functional neurovascular microenvironment. This dual-modulatory strategy provides a promising approach to overcome the impaired neurovascular coupling observed in DWs and contributes to orchestrated tissue repair and functional skin restoration. Notably, our approach exemplifies the emerging paradigm of simultaneously revitalizing angiogenesis and neurogenesis in chronic wound therapy.

3.8. Transcriptomic analysis reveals a temporal shift from angiogenesis to neurogenesis and IL-17 pathway suppression during wound healing

While S-GPL^KLT/RGI demonstrated significant efficacy in promoting wound repair, the underlying mechanisms — particularly those mediating neurovascular regeneration — remain incompletely understood. Moreover, it is unclear how a multifunctional material dynamically modulates key signaling pathways over time to orchestrate regeneration. To address these questions, we performed transcriptomic profiling of wound tissues from S-GPL^KLT/RGI-treated and control rats at days 7 and 14 post-injury (Fig. 7A). As illustrated by the volcano plots (Fig. 7B), the treatment induced marked transcriptional changes at both time points, with 330 and 289 differentially expressed genes (DEGs) identified on days 7 and 14, respectively. Unsupervised clustering of DEGs revealed high intra-group similarity and clear separation between S-GPL^KLT/RGI and controls (Fig. 7C), indicating that the hydrogel exerts a significant, time-dependent regulatory effect on the wound transcriptome.

Fig. 7.

Temporal coordination of neurovascular regeneration and immune modulation revealed by transcriptomic analysis. (A) Schematic of the RNA-seq workflow for control and S-GPL^KLT/RGI-treated wound tissues at days 7 and 14 (n = 3). (B) Volcano plots illustrating differentially expressed genes (DEGs) between S-GPL^KLT/RGI and control groups at each time point. (C) Heatmap showing unsupervised clustering of DEGs across all samples. (D) Gene Ontology enrichment analysis of DEGs. (E) Kyoto Encyclopedia of Genes and Genomes analysis of DEGs. (F) Gene Set Enrichment Analysis plot of the IL-17 signaling pathway. (G) Immunohistochemical staining of IL-17 in wound tissues on days 7 and 14. (H) Quantification of IL-17 expression based on integrated optical density (n = 3). Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, n. s.: not significant.

To elucidate the biological processes affected by S-GPL^KLT/RGI, we performed GO and KEGG enrichment analyses of the DEGs. On day 7, GO terms were predominantly enriched in cytokine responses and angiogenesis (Fig. 7D), indicating early activation of immune regulation and vascular regeneration in the treatment group. Consistently, KEGG analysis showed significant enrichment of the PI3K-Akt and VEGF signaling pathways (Fig. 7E), both well-established drivers of angiogenesis and endothelial cell activation [[87], [88], [89]]. This is consistent with previous reports that activating PI3K/Akt signaling in engineered hydrogels can promote neurovascular regeneration in diabetic wounds [90]. By day 14, the enrichment profile had shifted from angiogenesis to neurogenesis. GO terms were associated with neurogenic pathways (axon guidance, synapse assembly), re-epithelialization, and skin appendage development (Fig. 7D). Correspondingly, KEGG pathways related to neuroactive ligand–receptor interactions and neurotrophin signaling were upregulated (alongside continued enrichment of angiogenic pathways) (Fig. 7E). This temporal shift suggests that S-GPL^KLT/RGI initially promotes immune regulation and vascular reconstruction, then subsequently enhances neural network formation and tissue maturation at later stages of healing.

Among the significantly modulated pathways, the IL-17 signaling axis was consistently downregulated in S-GPL^KLT/RGI-treated wounds at both time points. IL-17 is a hallmark cytokine of chronic inflammation and is closely linked to delayed healing in diabetes [[91], [92], [93]]. It promotes matrix metalloproteinases (MMPs) that degrade the ECM and disrupt collagen–fibronectin organization, thereby impairing cell adhesion and migration [94]. To validate this observation, GSEA was performed on the KEGG IL-17 signaling pathway (rno04657). The results revealed overall suppression of IL-17 pathway in the treatment group at both time points (Fig. 7F). Specifically, downstream targets of the IL-17 receptor, including MAPK pathway genes (e.g., Fos, Fosb) and the pro-inflammatory chemokine Cxcl1, were significantly downregulated (Fig. S18). What's more, the immunohistochemical staining for IL-17 and its downstream effector IL-6 further validated the downregulation of inflammation. In the control group, IL-17 expression remained high at both days 7 and 14, indicative of sustained inflammation (Fig. 7G and H). In contrast, S-GPL^KLT/RGI markedly reduced IL-17 levels, while a moderate decrease was also observed in the S-GPL group, suggesting that the hydrogel base confers partial anti-inflammatory effects. A similar expression pattern was observed for IL-6 (Fig. S19), a key inflammatory marker in DW microenvironments [95]. Given that IL-17 and IL-6 are key cytokines secreted by Th17 cells and M1-type macrophages, respectively, and their sustained high expression is a major contributor to delayed healing and tissue damage in diabetic wounds [96,97]. These results collectively confirm the strong anti-inflammatory efficacy of S-GPL^KLT/RGI in vivo.

To further investigate the mechanism underlying inflammation suppression, we conducted in vitro experiments using an LPS-stimulated RAW264.7 macrophage inflammation model. The results demonstrated that S-GPL hydrogels containing ε-PL significantly suppressed the expression of pro-inflammatory cytokines IL-6 and TNF-α, while promoting macrophage polarization toward an anti-inflammatory M2 phenotype (CD206⁺, Arg-1⁺) (Fig. S20). Notably, S-GPL^KLT/RGI showed comparable performance to S-GPL in this model, suggesting that ε-PL may be the key contributor rather than the mimetic peptides in vitro. Mechanistically, ε-PL may exert its immunomodulatory effects through ROS scavenging, inhibition of macrophage overactivation, and suppression of pro-inflammatory cytokine production [[98], [99], [100]], collectively establishing a favorable microenvironment for subsequent neurovascular regeneration. However, the more pronounced anti-inflammatory effect observed in vivo in the S-GPL^KLT/RGI group. This enhanced effect may result from the peptide-induced neurovascular regeneration, which in turn contributes to resolving inflammation and restoring immune homeostasis.

In summary, the anti-inflammatory activity and neurovascular regeneration are interrelated and facilitate each other. On one hand, S-GPL exerts direct anti-inflammatory effects. Sustained IL-17 expression can impair both angiogenesis and neurogenesis. It damages endothelial progenitor cells, thereby hindering neovessel formation [101], and activates the chromatin-modifying enzyme JMJD3, which upregulates chemokine, enhances neutrophil infiltration, and exacerbates inflammatory damage to peripheral nerve [102]. The anti-inflammatory effect observed in the S-GPL group establish a favorable microenvironment for subsequent neurovascular regeneration. On the other hand, the further reduction of IL-17 in the S-GPL^KLT/RGI group comparing with S-GPL may be associated with the reestablishment of a functional neurovascular microenvironment. Neovascularization can modulates immune cell recruitment, with M2 macrophage polarization facilitating the release of anti-inflammatory cytokines (IL-10, TGF-β) that suppress IL-6 and IL-17 [103]. In parallel, the nervous system contributes to immunomodulation by promoting Treg differentiation and inhibiting Th17 cells, as demonstrated by human neural stem cell therapy, ultimately reducing IL-17 levels and supporting repair [104,105].

Taken together, the suppression of IL-17 signaling may function both as a prerequisite for and a result of neurovascular regeneration. By reconstructing the neurovascular microenvironment, S-GPL^KLT/RGI restores immune homeostasis and alleviates chronic inflammation, effectively disrupting the pathological cycle of “persistent inflammation–impaired repair” in DWs and enabling temporally coordinated, functionally integrated tissue regeneration.

4. Conclusion

In this study, we successfully developed a sprayable, bio functional hydrogel sponge functionalized with VEGF- and BDNF-mimetic peptides to address the complex pathological microenvironment of chronic DWs. The pneumatic spraying strategy enables seamless adaptation to complex wound geometries owing to its conformability and mechanical compliance. The high-velocity airflow during spraying generates a porous architecture that facilitates exudate management and promotes tissue integration. Moreover, this system accommodates a wide range of therapeutic payloads—including proteins, exosomes, small-molecule drugs, and microspheres—by adjusting the nozzle diameter. Nonetheless, the limited control over gas-induced porosity and the potential impact of UV crosslinking on the bioactivity of sensitive therapeutics may influence therapeutic efficacy. Mechanistically, S-GPL^KLT/RGI modulated endothelial and Schwann cell phenotypes and HUVEC's paracrine functions, thereby promoting neurovascular crosstalk and coordinated niche reconstruction. In vivo, the hydrogel significantly accelerated wound closure, enhanced tissue reconstruction, and facilitated the spatial integration of vasculature and peripheral nerves. Transcriptomic analyses revealed a temporally coordinated transition from angiogenesis to neurogenesis, accompanied by sustained suppression of IL-17 signaling. These findings highlight the ability of the hydrogel system to resolve chronic inflammation and re-establish a pro-regenerative microenvironment. Notably, IL-17 pathway suppression showed a strong correlation with enhanced neurovascular regeneration. Although the precise mechanisms underlying neurovascular interactions and their link to IL-17 modulation remain to be fully elucidated, our study demonstrates the therapeutic potential of targeting the neurovascular niche through multifunctional biomaterials. Future studies employing cell-type–specific strategies will be critical to further dissect these interactions and advance our understanding of tissue regeneration. This work provides a framework for designing next-generation hydrogel systems that leverage advanced materials to enable precision therapy tailored to the pathophysiological mechanisms of refractory wounds.

CRediT authorship contribution statement

Xiaozhen Zhou: Writing – review & editing, Writing – original draft, Validation, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization. Pengchao Ma: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis. Yihao Liu: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis. Zhaojian Wang: Writing – review & editing, Validation, Supervision, Resources, Project administration, Methodology, Conceptualization. Shida Chen: Validation, Resources, Methodology, Investigation. Zekun Cheng: Methodology, Investigation. Songlu Tseng: Supervision, Resources, Methodology, Investigation. Hui Wu: Writing – original draft, Supervision, Resources, Methodology, Investigation. Mengdi Zhang: Validation, Methodology, Investigation. Fengzhou Du: Validation, Formal analysis. Nanze Yu: Methodology, Investigation. Xiao Long: Writing – review & editing, Validation, Supervision, Resources, Project administration, Methodology, Conceptualization. Jiuzuo Huang: Validation, Resources, Methodology, Funding acquisition, Formal analysis. Xiumei Wang: Writing – review & editing, Validation, Supervision, Resources, Project administration, Methodology, Conceptualization.

Ethics approval and consent to participate

In the study “Sprayable hydrogel sponge for neurovascular microenvironment reconstruction and inflammation modulation in diabetic wound healing”, Sprague-Dawley (SD) rats were used to establish a diabetic wound model for in vivo experiments. All procedures were conducted in accordance with relevant laws and institutional guidelines, and were approved by the Beijing Medconnor Laboratory Animal Welfare Ethics Committee in September 2024 (Ethical Approval No. MDKN-2024-033).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: