Injectable deferoxamine-loaded microsphere hydrogels for inhibition of ferroptosis and promotion of third-degree burn wound healing

Abstract

In chronic third-degree burn wounds, high levels of inflammatory factors such as IL-1β lead to ferroptosis in surrounding cells, thereby delaying the healing process of the burn wounds. To promote the healing of such wounds, an injectable DFO@GM-H hydrogel was developed. This hydrogel was prepared by loading deferoxamine (DFO) into gelatin microspheres (GMs), and further crosslinking in a mild reaction. The resulting DFO@GM-H hydrogels demonstrated effective self-healing and injectability, rendering them well-suited for application in irregular burn wounds. DFO@GM-H hydrogels not only maintained a moist healing environment for the burn wound healing, but also exhibited the ability to slowly release DFO, therefore overcoming the limitation of the short half-life of DFO in vivo. With sustained release and enhanced stability of DFO from the injectable hydrogels, DFO@GM-H hydrogels demonstrated excellent biocompatibility, effectively promoted cell proliferation, migration and angiogenesis, showed antibacterial activity, and effectively inhibited cellular ferroptosis both in vitro and in vivo. Consequently, DFO@GM-H hydrogels accelerated the healing of burn wounds in rats via ferroptosis inhibition and enhanced cellular response. Therefore, DFO@GM-H hydrogel is anticipated to emerge as a novel bioactive dressing for the management of burn wounds.

Graphical abstract

Highlights

• •Bioactive hydrogel was prepared by loading deferoxamine into gelatin microspheres and crosslinked in mild reaction.
• •Hydrogels showed good self-healing and injectability, and continuously released deferoxamine during 7 days.
• •Hydrogels effectively promoted cell proliferation, migration and angiogenesis via inhibition of ferroptosis.
• •Hydrogels accelerated the third-degree burn wound healing in rats in vivo.

1. Introduction

Burns represent a significant a global public health concern, characterized by a persistent upward trend in the incidence of new cases, particularly among populations with lower socioeconomic status and in underdeveloped regions [1,2]. It is estimated that approximately 5 billion people lack access to proper acute burn management due to a lack of safe and affordable surgical treatments [3]. Early debridement and skin transplantation are the most common surgical interventions used in the treatment of burn wounds [4]. A third-degree burn is defined as a full-thickness burn, destroying all three layers of skin. Third-degree burns are a serious type and often require coverage of the burn with a bandage or skin grafts to close the wound [5]. The microenvironment of burn wounds often exhibits a disturbance in reactive oxygen species (ROS) and cytokine production, leading to persistent inflammation and increased cell death [[6], [7], [8]]. Furthermore, the prolonged and excessive accumulation of ROS results in an increased secretion of pro-inflammatory cytokines, thereby triggering a vicious cycle of inflammation that prolongs the healing process [9,10]. Moreover, the loss of natural barriers significantly increases the risk of microbial invasion, with both Staphylococcus aureus (S. aureus) and Pseudomonas aeruginosa (P. aeruginosa) being the most common species in burns [11]. Consequently, the improvement of delayed healing in third-degree burns is in urgent need in this field. However, the efficacy and functionality of current clinical treatment methods are deficient, resulting in suboptimal therapeutic outcomes.

Hydrogels are regarded as an optimal three-dimensional structural biological dressing due to their considerable drug-loading capacity, water absorption, and moisturizing properties [12,13]. However, their use in certain applications as burn wound dressing is limited by their inability to accommodate bioactive molecules, particularly in cases requiring injection or irregular sizes. To address this limitation, researchers have developed porous microspheres (with a diameter of about 1–1000 μm) to deliver various bioactive substances to enhance wound healing effects [[14], [15], [16]]. Specifically, the combination of anti-inflammatory agents, antioxidants, and proliferative agents has been considered as highly effective approach for the treatment of chronic wounds, and this can be achieved by incorporating these bioactive substances into the porous microspheres.

Emerging evidence indicates that ferroptosis exerts a pivotal function in the process of healing of chronic wounds [[17], [18], [19]]. However, research on the relationship between ferroptosis and the healing of burn wounds remains limited. Deferoxamine (DFO) is a small molecule iron chelator that has been FDA-approved for the treatment of iron overload diseases. In addition to its iron-chelating properties, DFO has been shown to promote angiogenesis [20] and enhance collagen deposition [21,22]. In addition, iron chelator can impede biofilm formation by competing for available iron and modulating the gene expression [23]. However, the half-life of DFO in the body is approximately 20–30 min, as it rapidly degrades and is cleared by the body [24]. Therefore, multiple doses are generally required to achieve effective DFO concentrations.

Based on these considerations, herein, we aim to explore the functional effect of DFO-loaded hydrogels on the healing of third-degree burn wounds by regulating ferroptosis. In this study, an injectable DFO@GM-H hydrogel has been developed for acceleration of burn wound healing. DFO@GM-H hydrogel scaffold has been formed by loading DFO into gelatin microspheres (GMs), and further crosslinking with glutaraldehyde (GA) (Fig. 1a). The hydrogel system shows the advantages, including ease of preparation, adequate injectability and favorable biocompatibility, thus rendering it a suitable delivery platform for DFO. The hydrogel remodels the mechanical properties of third-degree burn wounds, providing mechanical support for cells and further guiding cell behavior as a temporary substitute. In addition, the hydrogel sustained releases the DFO loaded within the scaffold to induce angiogenesis, exhibit antibacterial activity, and inhibit the ferroptosis of cells, thereby significantly accelerated the healing process of third-degree burn wounds.

Fig. 1.

Schematic diagram of the preparation and application of DFO@GM-H hydrogel dressing. (a) Schematic illustration of the preparation process of DFO@GM-H hydrogel. The right panel indicates the chemical structure of DFO@GM-H hydrogel. (b) Schematic illustration of changes in the microenvironment of third-degree burn wound. (c) Schematic illustration of the application of DFO@GM-H hydrogel dressing on third-degree burn wound to promote wound healing. The right panel illustrates the mechanism of multiple effects of DFO@GM-H hydrogel on promotion of wound healing, including regulation of cell behavior, inhibition of bacteria, and inhibition of ferroptosis.

2. Materials and methods

2.1. Materials

Gelatin, isopropyl palmitate, Tween 80, isopropanol, pentanal, acetic acid, deferoxamine (DFO) and glutaraldehyde (GA) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. All other chemical reagents are of analytical grade.

2.2. Synthesis of gelatin microspheres (GMs)

First, 5 % (w/v) gelatin was subjected to heating until complete dissolution in DI water. Then, the gelatin solution was added dropwise into a beaker containing isopropyl palmitate and Tween 80. The content of the beaker was stirred thoroughly for 20 min for effective emulsification at 50 °C. Afterwards, the solution was cooled to room temperature. The precipitation was collected and added to isopropanol for dehydration during a period of 30 min. Thereafter, the mixture was filtered and vacuum dried, which resulted in the powder of gelatin microspheres (GMs).

2.3. Preparation of DFO-loaded GMs (DFO@GM)

10 % GMs (w/v) were dissolved in phosphate-buffered saline (PBS) and left to swell at room temperature for 10 min. Then DFO was added at a concentration of 5 μg/ml and the mixture was stirred thoroughly for 10 min, to encapsulate DFO into the microspheres of GMs (Fig. 1a). Subsequently, the mixture was subjected to centrifugation. Thereafter, the DFO-loaded GMs were thoroughly washed and centrifugated with DI water for three times, and the resulting DFO-loaded GMs was labeled as DFO@GM.

The encapsulation efficiency of DFO into DFO@GM was calculated using the following equation [15]:

$$EE(\%)=\frac{W_0-W_{Sol}}{W_0}\times 100\%$$

where W₀ and W_Sol represent the mass of original DFO and the mass of the un-loaded DFO in the suspension, respectively.

2.4. Synthesis of DFO@GM-H hydrogels

The obtained DFO@GM solution was then mixed with 0.1 % GA (v/v), and allowed for crosslinking at room temperature for 24 h (Fig. 1a). The hydrogel was formed based on Schiff reaction between GA and gelatin [25,26]. The hydrogel was washed three times with ethanol and DI water, respectively. The obtained hydrogel was named as DFO@GM-H hydrogel.

2.5. Characterization

The condition and appearance of the microspheres and hydrogels were monitored using a high-resolution cell phone (Apple iPhone 15) and an inverted microscope (IX73, Olympus). The microstructure of the microspheres and the hydrogels was investigated using field emission scanning electron microscopy (SEM, Zeiss SIGMA). The average pore size was calculated by analyzing SEM images using ImageJ software (NIH, USA).

The synthesis of DFO@GM-H hydrogel was evaluated using Fourier transform infrared spectroscopy (FTIR, Thermo Fisher NICOLET 6700) to track changes in chemical structure. According to the various states of the samples, two FTIR characterization modes were used. The KBr pellet method was used for powder samples of GM and DFO, and the attenuated total reflectance (ATR) mode was used for liquid specimens of GA and DFO@GM, as well as hydrogel samples of GM-H and DFO@GM-H. All spectra were recorded across a spectral range of 4000-400 cm⁻¹ with 2 cm⁻¹ resolution.

2.6. Rheological property of the hydrogels

The rheological properties of the hydrogels were evaluated on a rheometer (AR2000ex, TA Instruments). Hydrogel discs measuring 25 mm in diameter and 1 mm in thickness were positioned between a pair of 25 mm round plates. Frequency sweep test was conducted to determine the storage modulus (G′) and loss modulus (G″) of hydrogels across a frequency ranging of 0.1–100 rad/s at a strain of 1 % in the linear viscoelastic region. Moreover, strain amplitude sweep tests were performed at 37 °C with strain ranging of 0.1 %–1000 %, to detect the critical strain point of hydrogels. Then, G′ and G″ of hydrogels during three cycles of oscillation between 1 % and 1000 % strain were measured at a frequency of 1 rad/s at 37 °C. To analyze the shear thinning properties of hydrogels, the viscosity of the hydrogels was measured by loading the hydrogels with a steady rate sweep with a shear rate ranging of 0.1–600 s⁻¹ at 37 °C.

2.7. Self-healing and injectable property of the hydrogels

The self-healing capability of the hydrogels was evaluated through macroscopic testing and observation. Two pieces of hydrogel were cut out and then allowed to come into contact at room temperature. Afterwards, the healed hydrogels were subjected to various external mechanical stimuli, and a high-resolution cell phone was used to monitor the various conditions.

In addition, the injectable capability of the hydrogels was observed through macroscopic testing. The hydrogel was loaded into a syringe, and then injected into a 3D-printed mold to recover their shapes. The injection processes and formation of hydrogels in the mold were video recorded and analyzed.

2.8. Drug-release capacity of DFO@GM-H hydrogels

Firstly, a standard curve was established with different DFO concentrations. Briefly, 0.1 mL of DFO solution was added to a tube at concentrations of 0, 0.1, 0.5, 1.5, 2, 2.5 and 3 mg/mL. Subsequently, 0.1 mL of a 0.75 mg/mL ferric chloride (FeCl₃) solution were added and thoroughly mixed in the dark for 10 min. The absorbance of FeCl₃ solutions was measured at 485 nm using a microplate reader (VICTOR Nivo Multi Microplate Reader, PerkinElmer), to obtain a standard curve for the DFO concentration.

To evaluate the drug release profile of DFO@GM-H hydrogels, the samples were immersed in PBS with pH values of 5, 7, and 9, respectively. The solution was placed in a dialysis bag, which was then immersed in PBS of the corresponding pH values. At regular intervals, a suitable volume of PBS solution was withdrawn to determine the drug content. 0.1 mL of the solution was mixed with 0.1 mL of FeCl₃, and the absorbance was measured as above described. By comparing the solution to the standard curve above, the DFO concentration was ascertained.

The accumulative DFO release from the hydrogels were plotted over incubation time. In drug release experiment, three samples were analyzed per group, with each experiment repeated three times.

2.9. In vitro degradation test

To assess in vitro degradation, the initial weight of dried DFO@GM-H hydrogels was measured, and then the hydrogels were immersed into PBS with different pH values of 5, 7, and 9, respectively. The samples were placed into a shaker (37 °C, 150 rpm) for 7 days. At predetermined time points, the residual hydrogels were collected and freeze-dried. Then the weight of the residual hydrogels was measured. The remaining hydrogel (%) was calculated using the following equation:

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

where W₀ and W_t indicated the weight of initial hydrogels and the residual hydrogels at specific time points, respectively.

2.10. Bioactivity test and cytocompatibility evaluation

Human skin fibroblasts (HSFs, Hunan Fenghui Biotechnology Co., Ltd) were cultured in DMEM medium (Thermo Fisher Scientific) supplemented with 10 % fetal bovine serum (FBS) and 1 % penicillin-streptomycin (PS) in a 5 % CO₂ incubator at 37 °C. Human umbilical vein endothelial cells (HUVECs, Hunan Fenghui Biotechnology Co., Ltd) were cultured in 1640 medium (Thermo Fisher Scientific) supplemented with 10 % FBS and 1 % PS. The cells were amplified routinely and used for experiments.

The hydrogel extract solution was used for cell experiments. In brief, GM-H and DFO@GM-H hydrogels were sterilized by three rounds of washing in 75 % ethanol and PBS. They were then incubated for 24 h at 37 °C with 10 mL of complete cell culture medium. The solution was then filtered by a 0.22 μm filter (Yeasen Biotechnology) to obtain the hydrogel extract solution.

2.10.1. Cytotoxicity and cell proliferation test

Different DFO concentrations were used to detect the toxicity of DFO to HSFs. Briefly, HSFs were seeded at a density of 2000 cells/well in 96-well plate, and incubated with cell culture medium containing varying DFO concentrations. After incubation for 0, 1, 3 and 5 days, cell viability was evaluated using CCK-8 kit (Beyotime Biotechnology). In brief, 10 μL of CCK-8 solution was added to each well and incubated for 2 h. The absorbance was recorded at 450 nm using the microplate reader. Appropriate DFO concentration was screened and employed for subsequent experiments.

Cytotoxicity of hydrogels on HSFs was then investigated using the CCK-8 assay. HSFs were seeded at a density of 2000 cells/well in 96-well plate, and adhered to the plate for 24 h, then the culture medium was replaced with 100 μL of hydrogel extract solution. The cell culture medium served as a control medium. The medium was refreshed every 24 h. After 24 h or 48 h incubation, cell viability of HSFs was evaluated using CCK-8 kit as above described. The relative cell viability was expressed as:

$$Cellviability(\%)=\frac{A_t-A_0}{A_c-A_0}\times 100\%$$

where A_t, A_c and A₀ represent the absorbance of the test sample, the control group and the blank group, respectively.

2.10.2. Hemolysis test

Firstly, 100 mg DFO@GM-H hydrogel was incubated in 10 mL saline (0.01 M, pH 7.4) at 37 °C with agitation at 100 rpm for 24 h. Then the supernatant was collected and filtered by a 0.22 μm filter to obtain sterile hydrogel extract solution. Fresh rat citrated blood was diluted with saline (saline: blood = 5 : 4). 100 μL of the diluted blood was added to 3 mL of the hydrogel extract solution and incubated for 1 h at 37 °C. Saline-treated blood was used as negative control. Triton-X-treated blood were used as positive control. Following 5 min centrifugation at 1000 rpm, 100 μL supernatant was added into 96-well plate and the absorbance was measured at 540 nm using a microplate reader. The following formula was used to determine the hemolysis (%):

$$Hemolysis(\%)=\frac{A_{\mathrm{S}}-A_N}{A_P-A_N}\times 100\%$$

where A_S, A_P and A_N represent the absorbance of the sample, the positive control and the negative control, respectively.

2.10.3. Scratch assay

First, cell migration of HSFs was investigated using scratch assay. HSFs at logarithmic growth stage were seeded at a density of 2 × 10⁵ cells/well in 6-well plate and incubated for 24 h. The monolayer was then scraped with a 200 μL pipette tip, rinsed, and imaged with an inverted microscope. Next, the medium was replaced with the hydrogel extract solution. HSFs incubated with normal cell culture medium served as control. The cells were incubated for another 24 h, and optical images were taken using an inverted microscope. The distances between two sides of the scratch were measured using ImageJ software.

2.10.4. Transwell assay

Meanwhile, cell migration and invasion of HUVECs was assessed using Transwell assay. The test was performed in 24-well transwell chambers with 8-μm-pore filter inserts (Corning). Briefly, HUVECs were seeded in the inserts at a density of 2 × 10⁴ cells/well in serum free culture medium. The lower chamber was filled with hydrogel extract solution with 20 % FBS. The cells were incubated for 24 h, fixed with formaldehyde, stained with crystal violet (Beyotime Biotechnology). The cells in the lower surface of the insert were observed and imaged with an inverted microscope. The number of cell migration and invasion was measured by ImageJ software.

2.10.5. Tube formation assay

Tube formation assay was performed to evaluate the angiogenesis potential of hydrogels. In brief, 50 μL of Matrigel (Corning) was added to 96-well plate, followed by gel at 37 °C for 1 h. Then, HUVECs at a density of 2 × 10⁴ cells/well were seeded onto the Matrigel, and incubated in 50 μL of the fresh hydrogel extract solution. HUVECs incubated in normal cell culture medium served as control. After 6 h incubation, the formation of capillary-like structures of HUVECS was imaged with an inverted microscope. Quantitative analysis of the tubular networks was conducted using angiogenesis analyzer in ImageJ software, including the number of branches and the total segment length.

2.10.6. Intracellular ferrous iron staining

The intracellular levels of ferrous iron (Fe²⁺) in HSFs were investigated using fluorescent staining of FerroOrange (Fe²⁺ indicator), to assess the extent of ferroptosis in the cells. In brief, FerroOrange (MKbio) was dissolved in dimethyl sulfoxide (DMSO) to form a 1 mM solution, which was further diluted in cell culture medium to a working concentration of 5 μM. HSFs were seeded in 6-well plate, and incubated with different treatments, including Erastin (a ferroptosis-inducer), IL-1β (a pro-inflammatory cytokine), IL-1β and hydrogel extract solution. Cells incubated in normal culture medium served as control. Afterwards, the cells were washed twice in PBS and incubated with 5 μM FerroOrange solution for 30 min at 37 °C in dark. Finally, the cells were immediately observed and imaged with fluorescent microscope (Aperio Versa 8, Leica), and quantitative analysis of the images was conducted with ImageJ software.

2.10.7. MDA assay

Malondialdehyde (MDA) content in HSFs was measured using MDA assay kit (Beyotime Biotechnology), to evaluate the lipid peroxidation of cells. In brief, HSFs were seeded in 6-well plate with different treatments similarly as above described. The cells were then collected and centrifuged at 11000 g at 4 °C for 10 min to harvest the supernatant. Then 200 μL MDA solution was added to each supernatant or standard solution (200 μL), the mixture was heated at 100 °C for 15 min and centrifuged at 1000 g for 10 min to harvest the supernatant. The supernatant (200 μL) was colorimetrically quantified at a wavelength of 532 nm (Thermo Scientific Multiskan FC). MDA levels were expressed as μmol/mg.

2.11. Antibacterial test

The antibacterial activity of the hydrogels was evaluated using S. aureus and Escherichia coli (E. coli). First, the antibacterial activity of hydrogels was evaluated using the agar diffusion assay. 100 μL of bacteria suspension (either S. aureus or E. coli) with a density of 1 × 10⁴ CFU/mL was spread on the culture dish with a triangular glass rod. Then the samples were put on the bacterial culture dish, including control (medical gauze), GM-H and DFO@GM-H hydrogels. The inhibitory zones were evaluated after 24 h of incubation at 37 °C.

Furthermore, the antibacterial activity of hydrogels was investigated by co-culture the samples with bacterial solutions. S. aureus and E. coli were co-incubated with different samples in a Luria-Bertani (LB) broth in different tubes at 37 °C overnight. Next, 100 μL of bacterial suspension from different groups was transferred to 96-well plate, and the viability of bacteria in each well was evaluate by CCK-8 assay as above described. The killing ratio of bacteria was calculated using the following equation:

$$Killingratio(\%)=\frac{A_c-A_t}{A_c}\times 100\%$$

where A_t and A_c represent the absorbance of the test sample and the control, respectively.

Moreover, P. aeruginosa and different samples was co-incubated in LB broth in 6-well plate at 37 °C for 5 days to form bacterial biofilms. Afterwards, the bacterial biofilm was fixed with methanol, rinsed, and stained with 1 % crystal violet. The stained bacterial biofilms were then dissolved in 33 % acetic acid, and the absorbance was measured at 570 nm using a microplate reader.

2.12. Western blotting

RIPA buffer (Beyotime Biotechnology, China) and protease suppressor (Biosharp, China) were used to lyse the cultured cells. A 10 % SDS-PAGE gel was employed to separate 20 μg protein from each sample, which was then electroblotted onto polyvinylidene difluoride membranes (MilliporeSigma). The membranes were blocked for 1 h with 5 % non-fat milk at room temperature. The separated proteins were probed overnight at 4 °C with the following primary antibodies, including ACSL4 mouse mAb (Proteintech, China, 66617-1-Ig, 1:2000), SLC7A11 rabbit mAb (Cell Signaling Technology, China, 12691, 1:1000), GPX4 mouse mAb (Proteintech, China, 67763-1-Ig, 1:2000), and GAPDH mouse mAb (Proteintech, China, 60004-1-Ig, 1:10000). Next, the membranes were rinsed for 10 min with Tris-buffered saline containing Tween 20, followed by 1 h incubation at room temperature with the appropriate horseradish peroxidase (HRP)-conjugated secondary antibodies, including HRP-goat anti-rabbit (Bioss, bs-0295G, 1:5000) and HRP-goat anti-mouse (Bioss, bs-0296G, 1:5000). Protein bands were visualized using enhanced chemiluminescence detection. Quantitative analysis of the immunoreactive bands was performed using ImageJ software. Three samples per group and three experimental replicates were conducted.

2.13. Animal studies

All animal experiments in this study were conducted in accordance with the guidelines approved by the Institutional Animal Care and Use Committee of Hubei Provincial Center for Disease Control and Prevention (IACUC Number: 202320225).

2.13.1. Preparation of third-degree burn wound model in rats

Third-degree burn model was established in rats according to previously reported protocols [27]. In details, male Sprague-Dawley (SD) rats weighing approximately 200 ± 10 g were anesthetized using chloral hydrate, and the skin on the dorsal were shaved. A curved metal punch (2 cm²) was heated to 80 °C by a tabletop thermoregulated burn machine (YLS-5Q, China). The heated punch was then pressed against the shaved skin of the rats at a pressure of 10 kPa for 18 s, resulting in a third-degree burn wound.

2.13.2. Application of DFO@GM-H hydrogels for burn wound healing

After 24 h of creation of burn wound model, necrotic tissue was surgically excised using scissors, and a biological adhesive (China Golden Elephant) was applied to secure a silicone ring around the burn wound to prevent premature contraction. The rats were randomly divided into four treatment groups: control group without treatment (negative control), GM-H hydrogel without DFO loading, free spraying of 5 μg/mL DFO solution, and DFO@GM-H hydrogel group. Four different materials were applied topically to cover the burn wounds on each rat, followed by the application of Tegaderm™ (3M, USA) over the wound area. The hydrogel dressings were changed every 3 days.

2.13.3. Wound closure rate analysis

The wound healing processes of each group were daily observed and photographed. The wound areas were analyzed using ImageJ software. The wound closure rate was calculated according to the following formula, which represents the change in wound size relative to the original wound size.

$$Woundclosurerate(\%)=\frac{S_0-S_n}{S_0}\times 100\%$$

where S₀ and Sn represent the wound area on day 0 and on day n, respectively.

2.13.4. Histological analysis and immunofluorescence staining

Wound specimens were collected on the 8th and 29th days of treatment. Hematoxylin and eosin (H&E) staining was conducted to evaluate the epidermal regeneration and the formation of granulation tissue within the burn wounds. Masson's trichrome staining was carried out to assess collagen deposition in the wound bed.

Furthermore, immunofluorescence staining of the regenerated skin was conducted using various antibodies to analyze the tissue inflammatory response (IL-1β), angiogenesis (CD31 and VEGF), ferrous iron content (ferritin), and markers associated with ferroptosis (ACLS4, SLC7A11, and GPX4). Tissue sections were incubated with primary antibodies overnight at 4 °C, including IL-1β mouse mAb (Proteintech, China, 66737-1-Ig, 1:300), Anti-CD31 rabbit mAb (Abcam, England, ab182981, 1:300), VEGFA rabbit pAb (Proteintech, China, 19003-1-ap, 1:500), ferritin light chain rabbit pAb ((Proteintech, China, 10727-1-AP, 1:500), ACSL4 rabbit mAb (Abcam, England, ab155282, 1:500), SLC7A11 rabbit mAb (Abcam, England, ab307601, 1:500), GPX4 mouse mAb (BOSTER, China, BA3802-1, 1:500). After rinsing, the samples were incubated with corresponding secondary antibodies conjugated to different fluorophores (Alexa Fluor 488 and 532 nm) at 37 °C for 1 h. Cell nuclei were stained with 4,6-diamidino-2-phenylindole (DAPI) at room temperature for 10 min. The slices were observed and imaged with fluorescence microscope, and quantitative analysis of the images was conducted with ImageJ software.

2.13.5. In vivo degradation test

To assess in vivo degradation, DFO@GM-H hydrogels were injected under the dorsal skin of C57BL/6 mice (∼20 g). And then the hydrogels on the dorsal skin was observed during 5 days. Skin samples with implanted hydrogels were collected on day 0 and day 5, and H&E staining was used to evaluate the degradation of the hydrogel in vivo.

2.13.6. In vivo biocompatibility test

On the 29th day of different treatments where the burn wounds completely healed, the vital organs (heart, liver, spleen, lung, and kidney) from different groups were excised and subjected to H&E staining. H&E staining images were analyzed to assess the potential toxicity of different samples compared to the control group.

2.14. Statistical analysis

All experimental data were subjected to statistical analysis, and the results were presented as mean ± standard deviation (S.D.). Statistical differences were determined using one-way ANOVA and Student's t-test. A significant difference was considered when p < 0.05.

3. Results

3.1. Characterization of DFO@GM-H hydrogels

DFO@GM-H hydrogel was synthesized based on the Schiff base reaction involving in aldehydes of GA and amines of gelatin (Fig. 1a). Before crosslinking, DFO-encapsulated GMs (DFO@GM) were well dispersed in solution (Fig. 2a). After crosslinking, the sol state of DFO@GM solution transitioned into gel state of DFO@GM-H hydrogels (Fig. 2a). Optical images further confirmed the formation of both GM-H and DFO@GM-H hydrogels (Fig. S1). The microscopic morphology of the microspheres and hydrogels was analyzed using SEM. The diameter of the DFO@GM microspheres was approximately 334.8 ± 38.2 μm, and their surface appeared relatively smooth (Fig. 2b). After the crosslinking process with GA, the resulting DFO@GM-H hydrogel demonstrated a connected and looser porous structure, with pore sizes ranging from 30 to 50 μm (Fig. 2b). This structural configuration is advantageous for the release of the encapsulated DFO. Furthermore, the increased specific surface area of DFO@GM-H hydrogels is beneficial for cell adhesion, thereby facilitating cell migration and angiogenesis, which are critical processes in wound healing.

Fig. 2.

Characterization and mechanical properties of DFO@GM-H hydrogels. (a) Optical images of sol-gel transition from DFO@GM solution to DFO@GM-H hydrogel. (b) SEM images of DFO@GM microspheres and DFO@GM-H hydrogels. (c) FTIR spectra during synthesis of DFO@GM-H hydrogel. (d) G′ and G″ of DFO@GM-H hydrogel at a strain of 1 %. (e) Strain amplitude sweep test to detect the critical stain point of DFO@GM-H hydrogel. (f) Step-strain measurement to confirm the self-healing capacity of DFO@GM-H hydrogel. (g) Self-healing process of two individual DFO@GM-H hydrogels. (h) Shear-thinning behavior of DFO@GM-H hydrogel. (i) Optical images of injection from syringe. (j) Injection and formation of hydrogels in different shapes.

Subsequently, FTIR spectroscopy was used to analysis the chemical structure during the synthesis of DFO@GM-H hydrogels. After emulsion of DFO into GMs (Fig. 1a), the obtained DFO@GM showed a broad band in the range of 3400–3200 cm⁻¹ (Fig. S2, Fig. 2c), which included the characteristic bands of the O-H stretching, the C-H stretching, and the N-H stretching of amide group of gelatin. And a strong band at 1638 cm⁻¹ was assigned to the C=O stretching of amide I of gelatin (Fig. S2, Fig. 2c). For GA, the characteristic bands at 2985 cm⁻¹ and 2913 cm⁻¹ were assigned to the C-H stretching of alkyl and C-H stretching of aldehyde, respectively (Fig. 2c) [28]. The band at 1638 cm⁻¹ represented the C=O stretching of aldehyde (Fig. 2c). And the bands at 1431 cm⁻¹, 1109 cm⁻¹, and 1004 cm⁻¹ were assigned to the bending of the CH₂, the C–C stretching, and the C-H bending, respectively (Fig. 2c). After the crosslinking between DFO@GM and GA, the obtained DFO@GM-H hydrogel exhibited an enhanced intensity of the bands at 3400–3200 cm⁻¹ and at 1638 cm⁻¹ (C=O stretching), however, with a decreased intensity in the characteristic bands of GA (Fig. 2c). These results indicate the successful synthesis of DFO@GM-H hydrogels.

3.2. Rheological property, self-healing and injectable property of DFO@GM-H hydrogels

The rheological properties of the hydrogels were analyzed using a rheometer. Generally, GM-H hydrogels without drug loading and DFO@GM-H hydrogels exhibited similar rheological characteristics (Fig. 2d–h, Fig. S3), indicating that the incorporation of DFO into the hydrogels did not significantly affect their mechanical properties. For both GM-H and DFO@GM-H hydrogels, the storage modulus (G′) was greater than the loss modulus (G″) at a strain of 1 % (Fig. 2d, Fig. S3a), indicating the gel-like character of both hydrogels [29].

To characterize the self-healing properties of hydrogels, a strain amplitude sweep test was first conducted to determine the critical strain point at which the gel network is disrupted, which was found to be ∼20 % (Fig. 2e, Fig. S3b). Based on this critical strain point, a step-strain oscillation test was performed, by varying the strain between 1 % and 100 % strain. Under small strain (1 %), both hydrogels exhibited a higher G′ than G′′ (Fig. 2f, Fig. S3c). As the strain switched to large strain (100 %), G″ was higher than G′ (Fig. 2f, Fig. S3c), indicating the collapse of the hydrogel network. Upon returning to small strain conditions, both G′ and G″ of the hydrogels recovered (Fig. 2f, Fig. S3c), demonstrating that the network of both hydrogels was restored even after three cycles of strain oscillation. This result illustrates the repeatable self-healing properties of both hydrogels. Moreover, the self-healing ability of DFO@GM-H hydrogel was macroscopically evaluated. After being cut and contacted, two pieces of DFO@GM-H hydrogel showed fast healing within 30 s at the interfaces of damage (Fig. 2g). Moreover, the healed DFO@GM-H hydrogel can resist gravity and maintain an integrated shape under stretching (Fig. 2g). The above results indicate the fast and excellent self-healing properties of DFO@GM-H hydrogels, which can be contributed to the abundant hydrogen bonds within the hydrogel network (Fig. 1a). DFO@GM-H hydrogel with excellent self-healing ability can maintain the stability when subjected to external mechanical forces during dressing application.

Further, the injectable capacity of the hydrogels was evaluated by characterizing the effect of high shear rates on gel viscosity. As the shear rate increased, the viscosity of both GM-H and DFO@GM-H hydrogels gradually decreased (Fig. 2h, Fig. S3d), indicating that the hydrogels had the ability of shear thinning, where the shear disrupted the dynamic crosslinks within the hydrogel network. Moreover, the injectable capacity of the DFO@GM-H hydrogel was visualized by injecting the hydrogel through a syringe needle (Fig. 2i). DFO@GM-H hydrogel was able to continuously form different letters and shapes when extruded through the syringe, while maintaining a stable gel state after injection (Fig. 2j). The injectability makes the DFO@GM-H hydrogel particularly well-suited for applications on irregular burn wound surfaces.

3.3. Drug release profile of DFO@GM-H hydrogels

The microenvironment of burn wounds varies with different pH values. The pH of unhealed chronic wounds typically changed within the alkaline range of 7.15–8.90 [30], with some instances reaching as high as 9.25 [31]. Therefore, the drug release profile of DFO@GM-H hydrogel was investigated under different pH values of 5, 7 and 9, respectively. When DFO@GM-H hydrogel was subjected to PBS with different pH conditions, the overall trend of DFO release from the hydrogel remained consistent (Fig. 3a). However, notable differences in release rate and total release amount were observed across the different pH conditions. At the lower pH of 5, DFO@GM-H hydrogel exhibited a rapid initial release, reaching maximum drug release after 24 h (Fig. 3a), probably due to the accelerated degradation of the hydrogel in the acidic microenvironment. Specifically, as the pH increased to 7 and 9, DFO release continued even after 24 h (Fig. 3a). Following 144 h, the DFO release in the pH 7 and 9 groups stabilized, exhibiting no significant fluctuations in drug concentration thereafter (Fig. 3a). These results indicate that the DFO@GM-H hydrogel facilitates a sustained release of DFO at pH levels exceeding 7, thereby extending the duration of DFO's therapeutic action.

Fig. 3.

Characterization of biocompatibility of DFO@GM-H hydrogels. (a) Cumulative release of DFO from DFO@GM-H hydrogel at different pH values. (b) Cell viability of HSFs incubated with extract of different hydrogels for 24 h and 48 h. (c) Hemolytic rate of red blood cells treated with extract of different hydrogels. The insert is the optical image of hemolysis test with different groups. (d) Representative images of migration of HSFs in scratch assay treated with different hydrogel extract at 0 h, 12 h and 24 h. The red lines indicate the migration boundary of HSFs. (e) Representative images of HUVECs treated with different hydrogel extract in Transwell assay. (f) Representative images of the tubular structures formed by HUVECs incubated in extract of different hydrogels for 6 h. (g) Quantification of migration rates of HSFs in scratch assay. (h) Quantification of migration and invasion of HUVECs in Transwell assay. (i) Quantification of angiogenesis of HUVECs, including the number of branches and the total segment length of HUVECs in tube formation assay. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns means non-significant (p > 0.05).

3.4. Degradation behavior of hydrogels in vitro

The degradation behavior of DFO@GM-H hydrogel was investigated in vitro in PBS solutions with different pH levels, which simulated the characteristics of different wound microenvironment. After incubation in PBS with different pH levels for 7 days, DFO@GM-H hydrogel gradually degraded (Fig. S5), and the acidic environment (pH = 5) induced a higher and faster degradation of hydrogels as compared to the neutral (pH = 7) and basic environment (pH = 9) (Fig. S5). After incubation of 7 days, the remaining weight of the hydrogels was at 11.7 %, 38.3 %, and 46.7 % for incubation at pH values of 5, 7, and 9, respectively (Fig. S5). This degradation behavior of hydrogels was consistent with the drug release profile (Fig. 3a). The higher the pH (7 and 9), the slower the hydrogel degradation (Fig. S5), the slower the drug release (Fig. 3a).

3.5. In vitro biocompatibility of DFO@GM-H hydrogels

Initially, we assessed the toxicity of various concentrations of DFO on HSFs. A concentration of 5 μg/mL of DFO demonstrated a higher cell survival rate (Fig. S4). Consequently, this concentration was selected for drug encapsulation in DFO@GM-H hydrogels. And the encapsulation efficiency of DFO into DFO@GM was calculated to be 83.9 % ± 0.6 % (Table S1).

Subsequently, we evaluated the cytotoxicity of both GM-H and DFO@GM-H hydrogels on HSFs using hydrogel extract solutions. In comparison to the control group, HSFs exhibited comparable cell survival rates after co-culturing with the GM-H hydrogel extract solution for both 24 and 48 h (Fig. 3b). Interestingly, co-culturing with the DFO@GM-H hydrogel extract solution resulted in an increased cell survival rate at both 24 and 48 h (Fig. 3b). These results suggest that pure GM-H hydrogel possesses favorable biocompatibility with HSFs, while DFO@GM-H hydrogel exhibits the capacity to enhance cell proliferation.

Furthermore, hemolysis tests revealed that the hemolysis rates of both GM-H and DFO@GM-H hydrogel extract groups were below 5 % (Fig. 3c), indicating the excellent blood compatibility of both hydrogels. The observed low cytotoxicity, effective promotion of cell proliferation, and excellent blood compatibility of DFO@GM-H hydrogels provided a solid foundation for their potential applications in wound dressings.

3.6. In vitro cell migration of DFO@GM-H hydrogels

As known, burn wound surfaces lead to a decrease in the migratory ability of cells within the wound [32]. Therefore, promoted cell migration is a prerequisite for a desired wound dressing. First, we used scratch assay to assess the cell migration ability of HSFs treated with different hydrogel extract solutions. Compared to the control group, both GM-H and DFO@GM-H hydrogel extract groups exhibited similar cell migration rates, with no significant statistical differences observed after 12 h of incubation (Fig. 3d and g). After 24 h of incubation, significant cell migration was observed in the scratch areas treated with GM-H and DFO@GM-H groups (Fig. 3d). The healing areas were 54.4 % ± 8.3 % and 77.7 % ± 8.8 % for GM-H and DFO@GM-H groups, respectively, which were significantly higher than the healing area of 18.1 % ± 0.9 % in the control group (Fig. 3g).

Additionally, we used Transwell assay to evaluate the migration and invasion capabilities of HUVECs treated with different hydrogel extract solutions. Consistent with the results of the scratch assay, after 24 h of incubation, HUVECs treated with GM-H and DFO@GM-H groups showed enhanced cell migration and invasion ability (Fig. 3e), with DFO@GM-H treatment showed the most significant effects on HUVECs migration and invasion (Fig. 3e). The average migration rates of HUVECs were 346.8 % ± 42.5 % for GM-H group and 510.0 % ± 111.9 % for DFO@GM-H group, both of which represented a significant increase compared to the control group (Fig. 3h).

3.7. In vitro angiogenesis activity of DFO@GM-H hydrogels

Angiogenesis is highly desired for accelerating the process of chronic wound healing [20]. We conducted tube formation assay on HUVECs to evaluate the angiogenesis activity of the hydrogels. Compared to the control group, the angiogenesis activity was significantly increased in HUVECs after 6 h of incubation with GM-H and DFO@GM-H hydrogel extract groups (Fig. 3f). The number of branches and total segments significantly increased after treatment with GM-H and DFO@GM-H groups (Fig. 3i), with DFO@GM-H treatment exhibiting the most significant tubular length and number of branches (Fig. 3i). These results demonstrate that DFO@GM-H hydrogels effectively promote cell migration and angiogenesis, which play a major role in burn wound healing, suggesting good potential of DFO@GM-H hydrogels for promotion of burn wound healing.

3.8. Antibacterial effect

Burn wounds are frequently infected by bacteria [33]. Consequently, wound dressings with potent antibacterial properties have attracted significant clinical attention [13]. To evaluate the antibacterial activity of hydrogels, first, we employed the agar diffusion assay. Compared to control of medical gauze, both GM-H and DFO@GM-H hydrogels formed distinct inhibition zones against the Gram-positive bacterium S. aureus and the Gram-negative bacterium E. coli (Fig. 4a). Especially, DFO@GM-H hydrogel exhibited larger inhibition zone compared to GM-H hydrogel (Fig. 4b).

Fig. 4.

Anti-bacterial activity of DFO@GM-H hydrogels. (a) Image of inhibition zone against S. aureus and E. coli after cultured with control (medical gauze) and hydrogels. (b) Quantification of inhibitory zone diameter against S. aureus and E. coli. (c–d) Images and killing ratio of bacterial colonies of S. aureus (c) and E. coli (d) after cultured with control and DFO@GM-H hydrogel. (e) Images of bacterial biofilms of P. aeruginosa after cultured with control and DFO@GM-H hydrogel. (f) Quantification of bacterial biofilms of P. aeruginosa. ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Subsequently, we investigated the antibacterial activity of DFO@GM-H hydrogels by co-culture the samples in bacterial solutions for 24 h. Then the supernatant was spread on LB agar plates and incubated for additional 24 h. A significant reduction in colony-forming units (CFUs) in the DFO@GM-H hydrogel group was observed compared to the control group of medical gauze (Fig. 4c and d). Quantification showed that 98.7 % ± 0.2 % of S. aureus and 98.7 % ± 0.1 % of E. coli were killed with treatment of DFO@GM-H hydrogels (Fig. 4c and d).

Moreover, given the propensity of P. aeruginosa to form biofilms and cause chronic infections [34], control (medical gauze) or DFO@GM-H hydrogel was co-cultured with the bacteria of P. aeruginosa for 5 days. Subsequent crystal violet staining demonstrated that minimal biofilm was formed in DFO@GM-H hydrogel groups compared to the control (Fig. 4e and f).

In conclusion, DFO@GM-H hydrogel effectively inhibits bacterial penetration and suppresses biofilm formation, therefore exhibits excellent antibacterial activity.

3.9. Inhibition of ferroptosis by DFO@GM-H hydrogels

In the microenvironment of third-degree burn wounds, there are persistently elevated levels of inflammatory factors, such as IL-1β [35], which contribute to ferroptosis in surrounding cells, thereby delaying the healing process of the burn wounds. The accumulation of intracellular ferrous iron (Fe²⁺) and lipid peroxidation are two major biochemical events that lead to ferroptosis [36]. Fe²⁺ is unstable and highly reactive, generating hydroxyl radicals through the Fenton's reaction [37]. These radicals can directly react with polyunsaturated fatty acids (PUFAs) in the cell membrane and plasma membrane, producing a large amount of lipid ROS, thereby triggering cellular ferroptosis [[38], [39], [40]].

Therefore, we simulated the burn wound microenvironment by culturing the HSFs with the addition of ferroptosis-inducer (Erastin) or with the intervention of IL-1β. We used the probe of FerroOrange (Fe²⁺ indicator) to measure intracellular Fe²⁺ levels under different intervention conditions [41]. As shown, the intracellular Fe²⁺ levels significantly increased with IL-1β intervention, with no significant difference compared to the results obtained using the ferroptosis inducer of Erastin (Fig. 5a and b). To assess the protective ability against inflammation provided by DFO@GM-H hydrogels, HSFs received IL-1β intervention followed by incubation with DFO@GM-H hydrogel extract. After treatment with DFO@GM-H hydrogel, the intracellular Fe²⁺ levels decreased to the levels comparable to those of the untreated control group (Fig. 5a and b).

Fig. 5.

Inhibition of ferroptosis by DFO@GM-H hydrogels. (a) Representative fluorescent images of intracellular Fe²⁺ accumulation in HSFs exposed to different groups for 24 h. Intracellular Fe²⁺ were stained with FerroOrange in red fluorescence. (b) Quantitative analysis of intracellular Fe²⁺ accumulation in HSFs. (c) Quantitative analysis of MDA content in HSFs exposed to different groups for 24 h. (d) Protein expression of ACSL4, SLC7A11, GPX4 and GAPDH in HSFs treated with different groups by western blot. (e–g) Quantification of fold-change of ACSL4, SLC7A11 and GPX4 in HSFs by western blot. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns means non-significant (p > 0.05).

Moreover, since the accumulation of lipid peroxidation is a significant biochemical event that leads to ferroptosis [36], we measured the levels of lipid peroxidation products MDA in the HSFs. The results indicated that the MDA levels in HSFs were consistent with the trend observed in the Fe²⁺ content within the cells (Fig. 5c).

Subsequently, we performed western blotting on HSFs under various intervention conditions, utilizing three commonly recognized indicators of ferroptosis: ACSL4, SLC7A11, and GPX4 [42]. As known, ACSL4 promotes ferroptosis by facilitating the esterification of PUFAs to acyl-CoA [43]. In contrast, the expression level of SLC7A11 is associated with ferroptosis resistance [44,45]. Meanwhile, GPX4 also played an important role in inhibiting ferroptosis by converting toxic lipid peroxides into non-toxic lipid alcohols [46]. The protein expression levels following IL-1β intervention were consistent with those observed when using the ferroptosis inducer Erastin (Fig. 5d–g), indicating that IL-1β can effectively induce cellular ferroptosis. After treating the HSFs with DFO@GM-H hydrogel extract, the expression levels of these ferroptosis-related proteins returned to levels similar to those of the control group (Fig. 5d–g).

The results indicate that DFO@GM-H hydrogels effectively reduced the intracellular Fe²⁺ levels and MDA levels in HSFs (Fig. 5a–c). Consequently, this treatment significantly inhibited the process of ferroptosis in HSFs by decreasing the expression of ACSL4 and increasing the expression of SLC7A11 and GPX4 in HSFs (Fig. 5d–g). The reduction of ferroptosis in burn wound cells through DFO@DM-H treatment has the potential to promote the healing of burn wounds.

3.10. In vivo burn wound healing performance of DFO@GM-H hydrogels

To investigate the burn wound healing performance of DFO@GM-H hydrogels, we established a rat model of chronic wounds with third-degree burns and applied the hydrogels to the burn wounds (Fig. 6a). Wounds healing in the different dressing groups were monitored from day 0 to day 29 post-wounding (Fig. 6b). On day 8, 17, and 22, the healing effects in the GM-H group and the free DFO group were similar, with no significant differences observed throughout the healing process (Fig. 6b and c). In contrast, wound healing in the DFO@GM-H hydrogel group was significantly improved compared to the other three groups (Fig. 6b and c). This enhancement may be attributed to a synergistic effect between the moisturizing action of the GM-H hydrogels and the therapeutic effect of the released DFO on cellular response. According to the analysis of wound area from the wound images, on day 8 post-wounding, the relative re-epithelialization areas for the control, GM-H hydrogel, free DFO group, and DFO@GM-H hydrogel group were 26.0 % ± 8.3 %, 44.0 % ± 7.7 %, 44.7 % ± 13.6 %, and 52.3 % ± 13.88 %, respectively (Fig. 6c and f). On day 17, the relative re-epithelialization areas for the four groups increased to 76.3 % ± 12.7 %, 87.4 % ± 1.3 %, 78.9 % ± 6.3 %, and 94.9 % ± 2.5 %, respectively (Fig. 6c and f). On day 29 post-wounding, all burn wounds had healed in the four treatment groups (Fig. 6f), with no significant difference in the relative re-epithelialization area (Fig. S6a).

Fig. 6.

In vivo burn wound healing performance of DFO@GM-H hydrogels in rat. (a) Schematic schedule of creation of burn wounds and application of DFO@GM-H hydrogels for wound healing. (b) Representative images of burn wound healing progress in different groups during 22 days post-wounding. (c) Wound traces during the healing process. (d) Representative H&E and (e) Masson's staining images of wound tissues from different groups on day 8 post-wounding. (f) Quantitative analysis of re-epithelialized area during healing process. (g) Quantitative analysis of granulation tissue thickness in wound tissues from different groups on day 8 post-wounding. (h) Quantitative analysis of collagen deposition in wound tissues from different groups on day 8 post-wounding. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns means non-significant (p > 0.05).

3.11. Histological evaluation

To further validate the burn wound healing effects of the DFO@GM-H hydrogels, histological analyses were assessed on day 8 and day 29 post-wounding. On day 8, granulation tissue thickness in the DFO@GM-H hydrogel group measured 237.9 μm ± 21.3 μm, significantly lower than that in the other groups (Fig. 6d and g). This observation can be ascribed to the earlier transition of granulation tissue to collagen in the DFO@GM-H hydrogel group (Fig. 6e). This transition resulted in a diminished quantity of residual granulation tissue, thereby contributing to an increase in the tensile strength of the wound. On day 8, collagen deposition in the DFO@GM-H hydrogel group reached 83.2 % ± 1.8 %, significantly higher than that in the control and in the GM-H hydrogel groups (Fig. 6h). However, on day 29, no significant differences in collagen deposition were observed among the four groups (Fig. S6b–c). These results indicate that DFO@GM-H hydrogels can significantly promote wound healing in rats with third-degree burns, particularly during the early stages of the healing process.

3.12. In vivo degradation and biocompatibility of DFO@GM-H hydrogels

The degradation behavior of DFO@GM-H hydrogel was investigated in vivo. Upon subcutaneous injection of the hydrogel into the dorsal region of mice at day 0, the hydrogel was visible at the dorsal skin (Fig. S7a). After 5 days, the hydrogel became smaller and invisible due to the degradation and absorption by the animals (Fig. S7a). H&E staining images also showed the surrounding skin with implanted hydrogels at day 0 (Fig. S7a), after 5 days, the implanted region became small and loose (Fig. S7b), which indicated the degradation of hydrogels in vivo.

Additionally, on day 29 post-wounding, H&E staining of major organs (heart, liver, spleen, lung and kidney) were analyzed. No obvious histological changes of major organs were observed for the rats applied with different groups (Fig. S8), revealed no apparent in vivo toxicity of DFO@GM-H hydrogel as compared to the control. These results indicated that DFO@GM-H hydrogels possess excellent biocompatibility in vivo.

3.13. Immunofluorescence staining

Wound healing is a complex and intricate biological process that involves anti-inflammatory responses, cell proliferation and migration, angiogenesis, epithelialization, and the deposition of extracellular matrix [47,48]. To further investigate the healing mechanism of DFO@GM-H hydrogel, we conducted immunofluorescence analysis for IL-1β, CD31, and VEGF. First, we assessed the inflammatory response of the burn wounds by staining of IL-1β on days 8 and 29. On day 8 post-wounding, the tissue treated with GM-H hydrogel did not show significant differences in IL-1β expression compared to the control group (Fig. 7a and c), indicating a persistent inflammatory microenvironment in the third-degree burn wound. In contrast, after treatment with DFO and DFO@GM-H hydrogels, the fluorescence intensity of IL-1β in the skin tissues significantly decreased (Fig. 7a and c), which may be attributed to the anti-inflammatory effects of DFO [49]. By day 29 post-wounding, compared to the control group, IL-1β staining in the three treatment groups all decreased, with no significant differences among them (Fig. S9a, Fig. S9c), likely due to the wound healing process entering its late stage, during which only the control group had not yet fully re-epithelialized.

Fig. 7.

Immunofluorescence staining and quantification of inflammation and angiogenesis at burn wound sites. (a) Representative fluorescent staining images of IL-1β of skin tissues on day 8 post-wounding. (b) Representative fluorescent staining images of CD31 and VEGF of skin tissues on day 8 post-wounding. (c) Quantitative analysis of relative coverage of IL-1β in skin tissues on day 8 post-wounding. (d–e) Quantitative of CD31 and VEFG levels in skin tissues on days 8 post-wounding. ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns means non-significant (p > 0.05).

Angiogenesis is one of the critical steps in the wound healing process, and its reduction can significantly hamper the healing of burn wounds [50]. CD31 plays an important role in angiogenesis, and CD31 staining is widely used to evaluate angiogenesis [51]. Vascular endothelial growth factor (VEGF) is a major pro-angiogenic factor that promotes the proliferation and migration of endothelial cells [52]. Angiogenesis in the healed skin tissues was evaluated using both CD31 and VEGF staining. On day 8 post-wounding, the fluorescence intensity of CD31 and VEGF in tissues treated with DFO@GM-H hydrogels was significantly enhanced and exhibited broader coverage (Fig. 7b and 7d–e). In contrast, the fluorescence intensity of CD31 and VEGF in the GM-H hydrogel group and the free DFO group was only slightly elevated compared to the control group (Fig. 7b and 7d–e). The poor healing of burn wounds in the control group was attributed to the low efficiency of endothelial cell proliferation and migration, as well as persistent inflammatory responses [53,54]. The enhanced angiogenesis by DFO@GM-H hydrogel treatment may be due to not only the moist healing environment provided by the hydrogels, which is more conducive to angiogenesis, but also the released DFO itself having the ability to promote angiogenesis, together resulting in synergistically enhance angiogenesis in DFO@GM-H hydrogel.

Furthermore, we explored the cellular ferroptosis during the healing process of third-degree burn wounds with different treatments. As known, ferritin can convert excess free Fe²⁺ into Fe³⁺ and store iron in the body, which can indirectly reflect the levels of free Fe²⁺ within the cells [55]. We detected the accumulation of intracellular Fe²⁺ levels in skin tissues using ferritin staining. There was no significant difference in intracellular Fe²⁺ content between the GM-H hydrogel group and the control group (Fig. 8a and c), while the intracellular Fe²⁺ levels in the DFO group and the DFO@GM-H hydrogel group were significantly decreased (Fig. 8a and c), which were only half of the Fe²⁺ levels of the control group (Fig. 8c). These results indicate that DFO@GM-H hydrogel treatment effectively inhibit ferroptosis in burn wounds.

Fig. 8.

Immunofluorescence staining and quantification of ferroptosis at burn wound sites. (a) Representative fluorescent images of intracellular Fe²⁺ accumulation in skin tissues on day 8 post-wounding. (b) Representative fluorescent staining images of ACLS4, SLC7A11 and GPX4 in skin tissues on day 29 post-wounding. (c) Quantitative analysis of intracellular Fe²⁺ level in skin tissues on day 8 post-wounding. (d–f) Quantitative analysis of ACLS4, SLC7A11 and GPX4 levels of skin tissues on days 29 post-wounding. ∗p < 0.05, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns means non-significant (p > 0.05).

Next, we conducted immunofluorescence staining to assess key indicators of ferroptosis in wound tissues. On days 8 and 29 post-wounding, the fluorescence intensity of ACLS4, a protein that promotes cellular ferroptosis, was lowest in the DFO@GM-H hydrogel group among the four groups (Fig. 8b and d, Fig. S9b, Fig. S9d). In contrast, the fluorescence intensity of SLC7A11 and GPX4, which are protective proteins involved in the ferroptosis process, was highest in the DFO@GM-H hydrogel group (Fig. 8b and 8e–f). Together, these results indicate that DFO@GM-H hydrogel effectively promotes the wound healing process by decreasing inflammation (Fig. 7a), enhancing angiogenesis (Fig. 7b) and inhibiting ferroptosis in surrounding cells (Fig. 8).

4. Discussion

Our study effectively developed an injectable DFO@GM-H hydrogel system as third-degree burn wound dressing. The hydrogels were prepared through the encapsulation of DFO within gelatin microspheres, and subsequently crosslinked using glutaraldehyde (Fig. 1a). The resulting DFO@GM-H hydrogels exhibited both injectability and self-healing characteristics (Fig. 2), making them suitable for various shapes of irregular burn wounds. The obtained DFO@GM-H hydrogels were particularly tailored for the treatment of third-degree burn injuries, capitalizing on their moisturizing properties and the controlled release of DFO (Fig. 3a). This approach is both simple and efficient, allowing for the integration of multiple functionalities within a single hydrogel system.

Third-degree burn wounds are considered difficult-to-heal chronic wounds due to long-term exposed to a high environment of inflammatory factors such as IL-1β [56]. Ferroptosis, a distinct form of non-apoptotic cell death, is characterized by intracellular iron overload and excessive production of lipid ROS [45,57]. Increasing evidence suggests that ferroptosis played a crucial role in the healing process of chronic wounds in diabetes, infected and aging skin [[17], [18], [19],58]. However, research on the relationship between ferroptosis and the healing of third-degree burn wounds was rarely reported. IL-1β can increase the levels of ROS, lipid ROS, and the lipid peroxidation product MDA in chondrocytes, and alter the expression of ferroptosis-related proteins. Herein, this study aims to explore the effect of DFO@GM-H hydrogels on the healing process of third-degree burn wounds by regulating ferroptosis through in vitro and in vivo models.

In the in vitro model, DFO@GM-H hydrogels exhibited sustained release of DFO during 7 days (Fig. 3a), overcoming the limitation of the short half-life of DFO in vivo. Moreover, DFO@GM-H hydrogels can slowly release the drug in the alkaline microenvironment of chronic wounds resulting from third-degree burns (Fig. 3a), which corresponds with the degradation behavior of hydrogel (Fig. S5). The sustained and slow drug release rate is beneficial for the treatment of chronic burn wounds. Furthermore, the DFO@GM-H hydrogel showed good biocompatibility both in vitro and in vivo (Fig. 3b, Fig. S8), and exhibited excellent blood compatibility (Fig. 3c). DFO@GM-H hydrogels promoted the migration of HSFs and HUVECs (Fig. 3d–e, Fig. 3g and h), as well as accelerated the tube formation efficiency of HUVECs (Fig. 3f and i). The enhanced cell migration and angiogenesis were beneficial for the burn wound repair.

In third-degree burn wounds, the elevated risk of infection plays a critical role in the healing process [59]. DFO@GM-H hydrogel demonstrated excellent antibacterial activity against S. aureus, E. coli, and P. aeruginosa (Fig. 4). The antibacterial efficacy can be attributed to the capacity of DFO of chelating iron ions in the microenvironment. Moreover, low concentrations of iron ions inhibited the regulation of bacterial quorum sensing, thereby augmented the antibacterial effect [60]. In addition to the antibacterial properties, DFO@GM-H hydrogel facilitates the re-epithelialization of the wound (Fig. 6), and establishes a physical barrier that helps to mitigate the risk of infection.

Subsequently, we investigated the inflammatory microenvironment and ferroptosis of third-degree burn wounds. The intervention of HSFs with ferroptosis inducer of Erastin and the addition of IL-1β to HSFs led to the accumulation of intracellular Fe²⁺ levels and the generation of lipid peroxidation products of MDA (Fig. 5a–c). Furthermore, IL-1β intervention of HSFs significantly regulated the expression of factors related to the ferroptosis signaling pathway (Fig. 5d–g), thereby promoting the process of cellular ferroptosis. However, after treatment with DFO@GM-H hydrogels, both intracellular Fe²⁺ levels and lipid peroxidation product of MDA returned to normal levels (Fig. 5a–c), and the ferroptosis pathway was effectively inhibited (Fig. 5d–g). Together, DFO@GM-H hydrogels played a protective role in the ferroptosis induced by IL-1β.

The elevation of IL-1β establishes a detrimental feedback loop characterized by reciprocal interactions with oxidative stress responses [61]. Overexpressed oxidative stress, along with exaggerated inflammation and compromised angiogenesis, substantially impede the chronic wound healing process [13,20,62]. In the in vivo study, DFO@GM-H hydrogels with sustained DFO release effectively accelerated the burn wound healing in rat models (Fig. 6a–c), with increased re-epithelialization area and collagen deposition in the wound bed (Fig. 6d–h). After treatment with DFO@GM-H hydrogels, the inflammatory response marker of IL-1β in the wound tissues significantly decreased (Fig. 7a and c), while the expression of angiogenesis markers of CD31 and VEGF significantly increased (Fig. 7b and 7d–e). Furthermore, following DFO@GM-H hydrogels treatment, the expression levels of the key proteins associated with ferroptosis in the wound area were effectively regulated (Fig. 8b and 8d–f), leading to the reduction in the incidence of ferroptosis in the surrounding cells (Fig. 8a and c), thereby promoting wound healing (Fig. 6). DFO@GM-H hydrogels promoted burn wound healing via inducing angiogenesis, anti-inflammation and mitigated ferroptosis. Our observations was in agreement with previous report where the controlled release of DFO promoted chronic wound healing via chelating Fe²⁺, and subsequent decreasing ROS levels and promoting angiogenesis [20].

Together, DFO@GM-H hydrogels exhibited sustained release of DFO, and effectively inhibited the ferroptosis (Fig. 5, Fig. 8). This resulted in enhanced cell migration and angiogenesis (Fig. 3), suppression of bacterial infection (Fig. 4), reduction of the inflammatory response (Fig. 7), and ultimately improved the healing outcomes for third-degree burn wounds in rats (Fig. 6). In animal study, although the use of free DFO solution alone resulted in regulation in ferroptosis indicators in rat wounds (Fig. 8), however, these changes were not significant when compared to the sustained release from DFO@GM-H hydrogels. Moreover, the rate of burn wound healing and the efficiency of collagen deposition in the free DFO group were inferior to those in the DFO@GM-H hydrogel group (Fig. 6). This is primarily due to the short half-life of DFO and the limitations of the dry healing environment. DFO@GM-H hydrogels provided both sustained release of DFO and moist environment for the burn wounds, therefore resulting in significantly more beneficial effects on burn wound healing.

5. Conclusion

In summary, we designed and successfully fabricated DFO@GM-H hydrogels with inhibition of ferroptosis to promote burn wound healing. DFO@GM-H hydrogels exhibited self-healing and injectability so that can be effectively filled to the irregular burn wounds. DFO@GM-H hydrogels exhibited sustained drug release over 7 days, and demonstrated a more prolonged release state in an alkaline microenvironment. The hydrogels had the ability to reduce inflammation, promote cell proliferation and migration, enhance angiogenesis, exhibit antibacterial activity, and continuously inhibit cellular ferroptosis both in vitro and in vivo, thereby accelerating the healing process of third-degree burn wounds in rats. DFO@GM-H hydrogels can significantly inhibit cellular ferroptosis in the healing of third-degree burn wounds, suggesting that ferroptosis may become a novel therapeutic target for chronic non-healing wounds.

CRediT authorship contribution statement

Langjie Chai: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Jianglong Huang: Funding acquisition, Formal analysis. Min Wang: Methodology. Yihui Huang: Formal analysis. Zhuo Huang: Software, Data curation. Ruiyu Zhang: Methodology, Investigation. Lu He: Data curation. Haijie Wang: Formal analysis. Danyang Chen: Validation, Methodology, Funding acquisition. Yifeng Lei: Writing – review & editing, Supervision, Project administration, Funding acquisition, Formal analysis. Liang Guo: Supervision, Resources, Funding acquisition, Conceptualization.

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 is the Supplementary data to this article:

Data availability

Data will be made available on request.