Development of a gelatin methacryloyl double-layer membrane incorporated with nano-hydroxyapatite for guided bone regeneration

Abstract

Guided bone regeneration (GBR) is an effective technique for treating bone defects, with barrier membranes playing a critical role in preventing soft tissue invasion while supporting bone formation. However, conventional collagen GBR membranes have limitations, including poor mechanical strength, high swelling ratio, rapid biodegradation, and fragile structures. In this study, we developed a heterogeneous double-layer membrane with tunable physical, chemical, and biological properties, fabricated through simple photopolymerization and lyophilization of gelatin methacryloyl (GelMA) and nanohydroxyapatite (nHA). By adjusting the crosslinking time, methacrylation degree, and nHA concentration, the cryogels showed porous microstructures with different pore sizes ranging from 93 to 360 μm. Compressive mechanical testing, swelling measurements, and in vitro/in vivo biodegradation assays confirmed that the methacrylation of gelatin increased the compressive modulus to 29.02 MPa (p = 0.0002), reduced the swelling ratio to 714% (p = 0.002), and slowed the degradation rate to 41.2% after 48 hours (p = 0.002). Incorporating nHA further enhanced the mechanical properties and extended the degradation time. GelMA and nHA–GelMA cryogels exhibited excellent biocompatibility and promoted osteogenic differentiation of bone marrow mesenchymal stem cells (BMSCs), particularly in the nHA–GelMA cryogel with large pore sizes. We selected a GelMA cryogel with the smallest pore size for optimal barrier function and an nHA–GelMA cryogel with the highest osteogenic potential to construct the double-layer GBR membrane. In a rat calvarial defect model, this novel membrane significantly enhanced bone regeneration, demonstrating markedly improved bone volume/tissue volume (BV/TV) and bone mineral density (BMD) compared to the control group (p = 0.0042 and p = 0.0088, respectively), with efficacy comparable to that of a commercial GBR membrane. These findings demonstrate the promising potential of this simple, tunable double-layer GelMA/nHA cryogel membrane as a superior alternative for GBR applications.

---

1. Introduction

Bone is crucial for many vital functions, such as movement and organ protection. Although bone has the ability to regenerate, large defects that exceed its self-repair capacity may require materials that promote bone regeneration. In the oral and maxillofacial region, many patients still face the dilemma of bone loss or insufficiency, resulting from periodontal diseases, tumors, and trauma, which increases the difficulty and the risk of implant failure. To overcome the lack of bone mass, guided bone regeneration (GBR) has become a clinically standard modality for reconstructing alveolar bone defects, often combined with bone graft substitutes.^1–3

The concept of GBR, derived from guided tissue regeneration, involves using a barrier membrane to cover the bone defect area, excluding the rapidly proliferating epithelial and connective tissue while promoting the growth of slower-proliferating osteogenic cells.⁴ The classical GBR principle prioritizes space maintenance and osteoconduction over osteoinductivity, as the membrane's primary function is to block non-osteogenic tissues. An ideal GBR membrane should exhibit good biocompatibility, low permeability, an effective barrier function to prevent epithelial and connective tissue ingrowth, no cytotoxic effects, and structural stability to maintain space for osteogenesis. GBR membranes are divided into several types according to different classification standards, of which the most popular one is to divide them into resorbable and non-resorbable types. Non-resorbable membranes, such as polytetrafluoroethylene (PTFE) material, offer good biocompatibility but require a secondary surgery for removal, increasing the treatment complexity.⁵ Resorbable membranes, with their good in vivo biodegradation ability, provide an alternative. Collagen, chitosan, and alginate are common materials, with collagen from bovine or porcine sources being the most widely used.^6–8 BioGide®, a commercially available resorbable GBR membrane, is derived from porcine skin type I and type III collagen with a double-layer structure.⁹ One dense layer blocks rapidly growing fibroblasts, while the loose layer facilitates osteoblast growth and promotes osteogenesis.¹⁰ However, collagen-derived GBR membranes are limited by poor mechanical strength, high swelling ratios, rapid biodegradation, and fragile structures.^11–13

As a hydrolysis product of collagen, gelatin offers excellent biocompatibility and biodegradability, and low antigenicity.¹⁴ It retains arginine–glycine–aspartic acid (RGD) sequences for cell–material attachment and a matrix metalloproteinase (MMP) target sequence for degradation.^15,16 Compared to collagen, gelatin promotes cell migration, proliferation, and differentiation, with lower cytotoxicity and an easier andsimpler production process. However, its instability at body temperature restricts its wide application in tissue engineering.¹⁷ Introducing methacryloyl substituent groups enables gelatin to photo-crosslink at room temperature via photopolymerization, enhancing its chemical and physical properties. Gelatin methacryloyl (GelMA) preserves cell-promoting capabilities, as less than 5% of residual amino acids react with methacrylic anhydride.¹⁸ Most functional RGD sequences remain intact post-crosslinking, ensuring excellent cell adhesion and proliferation. Furthermore, properties such as porosity, swelling ratio, mechanical modulus, and cell response can be tuned by adjusting the methacrylation degree, crosslinking time, or other synthesis parameters.^15,19–21 Given its safe and straightforward synthesis, GelMA is an excellent candidate for GBR membrane fabrication.²²

Osteogenesis potential is another critical factor for an ideal GBR membrane. Hydroxyapatite (HA), with a chemical structure similar to the primary inorganic component of bone matrix, Ca₁₀(OH)₂(PO₄)₆, is a widely used bone substitute.²³ Its unique osteoconductive and osteoinductive properties make it highly effective for bone regeneration.²⁴ Nano-hydroxyapatite (nHA) further enhances cell attachment, proliferation and osteogenic differentiation due to its nanostructure.^25,26

The study aims to design and fabricate a novel double-layer GBR membrane composed of GelMA and nHA, featuring a dense-and-loose porous structure to promote effective bone regeneration. We evaluate the membrane's physical and chemical properties, as well as its in vivo and in vitro degradation behavior. In vitro cell adhesion, proliferation, and biocompatibility were assessed by seeding L929 fibroblasts and bone marrow mesenchymal stem cells (BMSCs) on the materials. Osteogenic potential was evaluated through ALP staining and quantitation. Finally, the in vivo bone regeneration capacity of the double-layer membrane was tested in a critical-size rat cranial bone defect model.

2. Materials and methods

2.1. Materials

Gelatin derived from bovine skin was purchased from Solarbio (Beijing, China). Methacrylic anhydride (MA), 2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone (Irgacure 2959), and hydroxyapatite (HA) nanopowder (<200 nm) were purchased from Sigma-Aldrich (Missouri, United States). Collagenase I was purchased from Biofroxx (Einhausen, Germany). All reagents were of analytical grade. RPMI 1640 Medium, Minimum Essential Medium Alpha (α-MEM), phosphate-buffered saline (PBS), fetal bovine serum and penicillin–streptomycin solution were purchased from Hyclone and Gibco (Massachusetts, United States). The BCA Protein Assay Kit, cell lysis buffer without inhibitors, BCIP/NBT alkaline phosphatase color development kit and alkaline phosphatase assay kit were purchased from Beyotime (Shanghai, China). The cell Counting Kit-8 (CCK-8) was purchased from APExBIO (Massachusetts, United States). TRITC phalloidin and antifade mounting medium with DAPI were purchased from Solarbio (Beijing, China).

2.2. Synthesis of GelMA with different substitution degrees

10 g of gelatin was fully dissolved in 100 ml of PBS by constant stirring at 60 °C. Then, 10 ml and 30 ml of MA were separately added to the gelatin solution drop by drop at a temperature of 50 °C. After a 3-hour reaction, 200 ml of preheated PBS was added to the solution to terminate the reaction between gelatin and MA. Then all the solution was dialyzed in a 12–14 kDa dialysis bag in ultrapure water at 37 °C for 7 days. The dialysis water was changed every 6 hours to remove unreacted MA. After dialysis, the solution was frozen at −20 °C for 24 hours and lyophilized for 72 hours to obtain GelMA cryogels with different substitution degrees.

2.3. Nuclear magnetic resonance (NMR) spectra analysis of GelMA

NMR spectra were used to determine the substitution degrees of the synthetic GelMA. 5 mg of gelatin and 5 mg of GelMA with different substitution degrees were fully dissolved in 0.5 ml of deuterium oxide (D₂O) at 50 °C. The solutions were added to the NMR tubes and cooled. The ¹H NMR spectra were recorded using a 400 Hz Bruker AV spectrometer (Switzerland). The methacrylation substitution can be calculated using the following formula:

2.4. Fabrication of GelMA and nHA–GelMA cryogels

GelMA with two different substitution degrees was dissolved in PBS containing 0.5 wt% I2959 at 65 °C to prepare 10% wt% GelMA solutions. The solutions were poured into cylindrical molds under light-proof conditions and irradiated with 450 nm UV light for 10 or 30 seconds. The four resulting GelMA hydrogel groups were named GM60-10s, GM60-30s, GM90-10s, and GM90-30s. GelMA with 60% methacrylation substitution was mixed with 0, 5, 10, or 20 mg ml⁻¹ of nHA and 0.5 wt% I2959 at 65 °C. The solutions were added to cylindrical molds and crosslinked under 450 nm UV light for 10 seconds. These were labeled 0HA, 5HA, 10HA, and 20HA. All hydrogels were frozen at −20 °C for 24 hours and lyophilized under vacuum for 48 hours to form porous cryogels.

2.5. Fourier transform infrared spectrometry (FT-IR) analysis

FT-IR analysis was performed to analyze the chemical bonds in gelatin, GelMA and nHA–GelMA cryogels. All cryogels were ground into powder and analyzed using a Fourier transform infrared spectrometer (Nicolet iS10, Thermo Scientific, MA, USA).

2.6. Scanning electron microscope (SEM) analysis

The microstructure of gelatin, GelMA, and nHA–GelMA cryogels was imaged and analyzed using a scanning electron microscope (Inspect F50, FEI, Hillsboro, OR, USA). The cryogels were trimmed into cylinders (8 mm diameter × 10 mm height). Cross-sections of the cryogels were gold sputtered prior to SEM analysis. The average pore size was calculated by measuring the pore diameters of three randomly selected sections per cryogel.

2.7. Compressive mechanical testing

Gelatin, GelMA, and nHA–GelMA cryogels were trimmed into cylinders (5 mm diameter × 2 mm height, measured using a micrometer). Compressive testing was performed using a universal material testing machine (Instron 5967). A 100 mm diameter probe was used to compress the cryogel surface at 1 N s⁻¹ until the height was reduced to 80% of the original. Stress–strain curves were plotted from three replicates per group.

2.8. Swelling measurement

The gelatin, GelMA, and nHA–GelMA cryogels were trimmed into cylinders (8 mm diameter × 10 mm height). Their initial weights were measured and recorded as W₀. Then all the cryogels were immersed in PBS solution at 37 °C. After 6, 12, and 24 hours, excess surface water was removed using filter paper, and samples were weighed again (W_t). The swelling ratio at each time point can be calculated using the following formula:

Each group contained three duplicate samples.

2.9. In vitro degradation testing

Cryogels were pre-swollen in PBS for 24 hours to reach equilibrium. After excess surface water was removed, the samples were weighed (W₀). Then, the samples were immersed in a 0.5 g ml⁻¹ collagenase type I solution at 37 °C. The collagenase solution was refreshed daily. Sample weights (W_t) were recorded at 6, 9, 12, 24 hours, and then daily. The degradation ratio can be calculated using the following formula:

Each group contained three duplicate samples.

2.10. In vivo biodegradation testing and toxicity assessment

The gelatin, GelMA, and nHA–GelMA cryogels were trimmed into cylinders with a diameter of 10 mm and height of 2 mm and sterilized. Their weights were measured and recorded as W₀. Twenty-seven 8-week-old male Sprague–Dawley rats, weighing around 250 g, were randomly divided into 9 groups corresponding to the different materials. The animal experiments were approved by the Ethics Committee of West China Hospital of Stomatology (WCHSIRB-D-2020-492). The surgery was conducted under the guidelines for care and use of laboratory animals of Sichuan University. As described before,²⁷ the rats were anesthetized using 10% chloral hydrate at a dose of 0.33 mg ml⁻¹. A 1.5 cm length sagittal incision was made on the back skin, and three separate subcutaneous pockets were made by blunt dissection along the incision. Each pocket received a sterilized sample, and the incision was closed with a suture. Each group included three rats. After surgery, the rats were monitored continuously for health and diet. Fourteen days post-operation, all the rats were sacrificed under general anesthesia, and the embedded samples were retrieved. All soft tissue around the samples was removed. The samples were frozen at −20 °C for 24 hours and lyophilized under vacuum for 48 hours. The weights of the retrieved samples were recorded as W_t. The in vivo degradation ratio can be calculated as follows:

Organs including the heart, liver, spleen, lungs, and kidneys were harvested, fixed, dehydrated in graded ethanol, embedded in paraffin, sectioned at 4 μm thickness, and stained with hematoxylin and eosin (H&E).

2.11. Cell proliferative activity evaluation

The mouse-derived L929 fibroblast cell line and rat bone marrow mesenchymal stem cells (BMSCs) isolated from two-week-old rats’ femurs and tibias, were used to study the cytocompatibility of GelMA and nHA–GelMA cryogels. After UV sterilization for 30 minutes, all cryogels were immersed in PBS solution containing 10% penicillin–streptomycin for 24 hours. Then GelMA and nHA–GelMA cryogels were separately immersed in α-MEM and RPMI 1640 medium for 24 hours to obtain extraction media. BMSCs and L929 cells were seeded in 96-well plates at a density of 3000 cells per well, and the corresponding extracted media were added. The CCK-8 assay was used to assess the proliferative activity of the two kinds of cells. At 1, 3, 5, and 7 days after seeding, the CCK-8 test solution was added to the plate and incubated at 37 °C under light-proof conditions for 30 minutes. Then the spectrophotometry at 450 nm was read with a microplate reader.

2.12. Live/dead cell staining

The sterilization of GelMA and nHA–GelMA cryogels and the preparation of extraction media are described in section 2.11. The L929 and BMSCs were seeded in 48-well plates at a density of 9000 cells per well. After 1, 3, and 5 days of culture, calcein-AM and PI live/dead staining were added and incubated for 30 minutes under light-proofconditions to detect the cell viability. An inverted fluorescence microscope was used to record the staining results.

2.13. Cell adhesion observation

Following the sterilization protocol in section 2.11, cryogels were placed in 48-well plates and separately immersed in RPMI 1640 and α-MEM containing 10% FBS and 1% penicillin–streptomycin for 24 hours to allow full infiltration. L929 and BMSCs were separately seeded on GelMA cryogels and nHA–GelMA cryogels at a density of 3 × 10⁴ cells per well and cultured for 3 days. Cells were then fixed with 4% glutaraldehyde and dehydrated using graded ethanol. Gold-sputtered samples were observed under SEM to assess cell adhesion.

2.14. Cell morphology observation

The sterilization of GelMA and nHA–GelMA cryogels is described in section 2.11. All the sterilized cryogels were placed in 48-well plates and separately immersed in RPMI 1640 and α- MEM containing 10% FBS and 1% penicillin–streptomycin for 24 hours to allow full infiltration. L929 and BMSCs were separately seeded onto GelMA cryogels and nHA–GelMA cryogels at a density of 3 × 10⁴ per well. After 3 days of culture, the cells were fixed with 4% glutaraldehyde and then treated with Triton X-100 for 15 minutes and 10% bovine serum albumin (BSA) for 30 minutes. Rhodamine-labeled phalloidin and DAPI were used for fluorescent staining of the cells separately for 30 minutes and 5 minutes. The cell morphology was observed by confocal microscopy at 488 nm and 562 nm wavelengths.

2.15. Barrier function study of GelMA cryogels

The sterilization and infiltration of GelMA cryogels are described in section 2.11. L929 cells were seeded onto GelMA cryogels at a density of 3 × 10⁴ per well. After 3 days of culture, the cells were fixed, permeabilized, and blocked as above. Rhodamine-labeled phalloidin and DAPI were used for the fluorescent staining of the cells separately for 30 minutes and 5 minutes. Longitudinal sections along the cryogel diameter were observed by confocal microscopy at 488 nm and 562 nm wavelengths to assess the barrier function of different GelMA cryogels.

2.16. In vitro osteogenic differentiation study

The sterilization and infiltration of GelMA cryogels are described in section 2.11 BMSCs were seeded onto nHA–GelMA cryogels at a density of 3 × 10⁴ per well. Once the cells reached 80% confluence, osteogenic-induction media were added and changed every two days. After 7 days of osteogenic differentiation, cells were fixed with 4% glutaraldehyde. ALP staining was conducted according to the instructions of the BCIP/NBT alkaline phosphatase detection kit (Beyotime, China) at 37 °C for 2 hours. For ALP quantification, RIPA lysis buffer (Beyotime, China) was added to obtain the gross protein. The ALP concentration was measured using an ALP assay kit (Beyotime, China) at 405 nm wavelength.

2.17. Fabrication of the double-layer membrane

GelMA60 and GelMA90 were separately dissolved in PBS containing 0.5 wt% I2959 at 65 °C to obtain 10 wt% GelMA solutions. The GelMA90 solution was poured into cylindrical molds and UV crosslinked (450 nm, 30 s). The GelMA60 solution was mixed with 10 mg ml⁻¹ nHA, then added to cylindrical molds containing the crosslinked GelMA90 cryogel and crosslinked under UV for another 30 seconds. The double-layer cryogels were frozen at −20 °C for 24 hours and lyophilized under vacuum for 48 hours.

2.18. Animal study

The animal experiments were approved by the Ethics Committee of West China Hospital of Stomatology (WCHSIRB-D-2020-492). All the experimental procedures were conducted following the Guidelines for the Care and Use of Laboratory Animals of Sichuan University. Thirty male SD rats of SPF grade, 8 weeks old, and weighing 250–300 g, were purchased from Chengdu Dashuo. Rats were randomly assigned to 1 of 5 groups (n = 6):

(A) 10HA-GM90-30s double-layer GelMA membrane;

(B) GM90-30s single-layer GelMA membrane;

(C) 10HA single-layer GelMA membrane;

(D) Bio-Gide® GBR barrier membrane as positive control; and

(E) blankcontrol

As described previously,²⁸ rats were anesthetized with 10% chloral hydrate (0.33 mL per 100 g), shaved, and disinfected. A 1.5 cm sagittal incision was made along the midline from the eyes to the ears to expose the calvarium. A 5 mm-diameter trephine was used to create a 5 mm cranial defect under saline irrigation, avoiding the midsagittal suture. Different membrane materials, as described above, were used to cover the defect area, and no membrane was used in the blank control group. Postoperatively, erythromycin and an intramuscular injection of 200 000 units of penicillin solution were administered. Eight weeks after the operation, all the groups were euthanized with excess anesthesia. The cranial bones containing the bone defect areas were harvested, fixed in 4% paraformaldehyde for 48 h and preserved in 0.5% paraformaldehyde.

2.19. Micro-computed tomography (micro-CT) analysis

Samples were scanned and reconstructed using a Micro-CT scanner (VivaCT80, Scanco Medical, Switzerland) at 70 kV voltage, 114 μA current, and 12 μm voxel size. The bone mineral density (BMD) and bone volume/total bone volume (BV/TV) of the defects were analyzed using Scanco software. Each group had at least three samples for statistical analysis.

2.20. Histological staining

The samples were fully decalcified in 10% EDTA buffer solution for 4 weeks, dehydrated with graded alcohol, paraffin-embedded, and sectioned at 4 μm. Finally, hematoxylin and eosin (H&E) staining and Masson's trichrome staining were performed.

2.21. Immunohistochemical (IHC) staining

All the groups were examined by IHC staining. Osteocalcin (OCN; rabbit anti-rat, Servicebio, GB11233, 1 : 100), collagen type I (Col I; rabbit anti-rat, Servicebio, GB11022-3, rabbit, 1 : 100), and CD31 (rabbit anti-rat, Abcam, ab182981, 1 : 1000) were used to analyze vascularization and osteogenesis in the defect area. Sections were incubated overnight at 4 °C, followed by incubation with a biotinylated secondary antibody (Servicebio, Wuhan, China) for 30 min at room temperature. Finally, the slides were observed and captured under a microscope (Leica, Germany).

2.22. Statistical analysis

GraphPad Prism 8 was used for statistical analysis. At least 3 parallel samples were set for each group, and the results were expressed in the form of mean ± standard deviation. A t-Test was performed between two groups, and one-way ANOVA was performed among multiple groups. When P < 0.05, it was recorded as a significant difference.

3. Results and discussion

3.1. Synthesis and identification of GelMA

The ¹H NMR spectra of gelatin and GelMA with different methacrylation degrees are shown in Fig. 1a. Compared with gelatin, the GelMA spectra exhibited new signals at δ = 5.3 ppm and δ = 5.5 ppm, corresponding to the acrylic protons of the methacrylate groups. Notably, the intensity of these signals increased with higher methacrylic anhydride (MA) additions. The disappearance of the signal at δ = 2.8 ppm was associated with grafted methacrylate groups, while the signal at δ = 7.3 ppm represented unreacted amino acid residues in gelatin. To determine the methacrylation degrees, the ratio of grafted methacrylate groups and unreacted amine groups was calculated and is presented in Fig. 1b. Increasing the amount of methacrylic anhydride resulted in a higher methacrylation degree. When 10 ml and 30 ml of methacrylic anhydride were added to 100 ml of a 10% gelatin solution, the corresponding methacrylation degrees reached 60% and 90%, respectively. Accordingly, the synthesized GelMAs were designated as GelMA 60 and GelMA 90. The synthesis process of GelMA was straightforward, involving only a temperature-controlled methacrylation reaction followed by dialysis.

Fig. 1 (a) ¹H-NMR spectra of gelatin and GelMA. (b) Calculated methacrylation degrees. (c) FT-IR spectra of GelMA with varying crosslinking times and methacrylation degrees. (d) FT-IR spectra of nHA–GelMA with different nHA concentrations.

3.2. Preparation of GelMA and nHA–GelMA cryogels

The GelMA and nHA–GelMA cryogels were fabricated through simple photopolymerization and lyophilization of GelMA, with or without nHA. By adjusting the crosslinking time and methacrylation degrees, these cryogels were suitable for constructing the GBR membranes. Compared to traditional collagen GBR membranes, which require complex synthesis involving toxic organic reagents like acetone and acetic acid,²⁹ GelMA synthesis is process-friendly and less technically demanding. Moreover, the GelMA photopolymerization enables fast and flexible crosslinking under mild conditions (neutral pH and moderate temperature). The temporal and spatial control of crosslinking facilitates consistent quality and scalability for mass production.

3.3. Chemical properties of GelMA and nHA–GelMA cryogels

The FT-IR spectra of gelatin, nHA, GelMA with varying crosslinking times and methacrylation degrees, and nHA–GelMA with different nHA concentrations are shown in Fig. 1c and d. The spectra of GM60-10s, GM60-30s, GM90-10s, and GM90-30s all showed characteristic absorption bands of functional groups, with peaks at 1240 cm⁻¹ (C–N stretching), 1541 cm⁻¹ (N–H bending), 1640 cm⁻¹ (C=O stretching), 2934 cm⁻¹ (C–H stretching), and 3303 cm⁻¹ (N–H stretching), confirming the successful introduction of methacrylate groups. The nHA spectrum showed a characteristic peak at 1061 cm⁻¹. All nHA–GelMA with different nHA concentrations showed this peak, with its intensity increasing proportionally to the nHA concentration.

3.4. Morphology of the cryogels

One of the key advantages of GelMA is its tunable microstructure, mechanical properties, and swelling behavior, which can be achieved by adjusting synthesis parameters such as concentration, methacryloyl degree, and crosslinking time.¹⁷ Cross-sectional microstructure images of gelatin, GelMA, and nHA–GelMA cryogels, captured by SEM, are shown in Fig. 2a and b. All cryogels had an interconnected microporous structure under SEM. In contrast to the irregular and disordered microstructure of gelatin, crosslinked GelMA and nHA–GelMA cryogels displayed an orderly interconnected micro-porous structure with different pore sizes. Increased crosslinking time resulted in smaller pore sizes, and the pore size further decreased with higher methacrylation degrees. Gelatin had the largest average pore size at 383 μm. As crosslinking time and methacrylation degree increased, the average pore size decreased from 342 μm in the GM60-10s group to 93 μm in the GM90-30s group. However, adding nHA to GM60-10s cryogels did not significantly affect the pore size, with 0, 5, 10, and 20 nHA cryogels showing similar pore sizes ranging from 340 μm to 360 μm. Cryogenic treatment of the GelMA hydrogels facilitated the formation of a porous structure.³⁰ Previous studies indicate that the average pore size was inversely related to GelMA solution concentration, cooling rate, and methacrylation degree.^21,31,32 Our study explored the relationship among crosslinking time, methacrylation time, and average pore size. By varying the crosslinking time (10 s and 30 s) and methacrylation degree (60% and 90%), we successfully achieved average pore sizes ranging from 342 μm to 93 μm. The double-layer GBR membrane was constructed with a 359 μm 10HA cryogel as the loose layer and a 93 μm GM90-30s cryogel as the dense layer, leveraging the pore size difference to support distinct cell barrier and osteogenic functions.

Fig. 2 The porous microstructure of gelatin, GelMA (a) and nHA–GelMA (b) cryogels observed under SEM, with quantitative analysis of pore size. All the samples showed an interconnected porous structure. Increasing crosslinking time and methacrylation degree led to a gradual decrease in pore size. However, nHA concentration did not affect pore size.

3.5. Mechanical properties of the cryogels

Excellent compressive mechanical properties are crucial for maintaining the barrier membrane's integrity and resisting pressure from outside during guided bone regeneration.³³ The compressive stress–strain curves of the gelatin, GelMA and nHA–GelMA groups are shown in Fig. 3a. The gelatin group exhibited the lowest compressive strength among all crosslinked cryogels, with a mean compressive modulus of 7.039 ± 0.2450 MPa. The curve trends indicated that the compressive strength was highly dependent on the crosslinking time, methacrylation degree, and nHA concentration. Increasing the crosslinking time, methacrylation degree, and nHA concentration progressively enhanced the compressive strength. Notably, adding 20 mg ml⁻¹ of nHA to GelMA60 crosslinked for 10 s resulted in a compressive modulus of 20HA (35.18 ± 5.701) that surpassed that of the GM90-30s group (29.02 ± 2.950), suggesting that nHA significantly reinforced the compressive mechanical properties and could play a critical role in maintaining the micro-space for osteogenic differentiation and new bone formation of BMSCs. These results showed that the compressive strength of the GelMA and nHA–GelMA groups was significantly enhanced compared to that of the gelatin group. When compared to collagen membranes in a previous study, the mechanical stress of our cryogels was two to three times higher than that of the oxidized sodium alginate–collagen membranes at the same strain level.²⁹

Fig. 3 (a) Compressive stress–strain curves of gelatin, GelMA and HA-GelMA. (b) The swelling ratio of gelatin, GelMA and HA-GelMA cryogels at 6 h, 12 h and 24 h. (c) The shape of gelatin and GelMA after swelling equilibrium. Compared to GelMA, gelatin cryogel lost its integrity and cylindrical shape. (d) The in vitro degradation rates of gelatin, GelMA and HA-GelMA cryogels.

3.6. Swelling ratio of the cryogels

The swelling ratios of the gelatin, GelMA, and nHA–GelMA cryogels at 6, 12, and 24 hours are presented in Fig. 3b. At each time point, the gelatin group exhibited the highest swelling ratio, reaching 1500% at 6 hours. Crosslinking effectively reduced the swelling properties. With the increase in crosslinking time and methacrylation degree, the swelling ratio gradually decreased at each time point, with the GM90-30s group demonstrating the lowest swelling ratio. However, the presence of nHA had no significant effect on the swelling ratio, as all four nHA–GelMA groups showed no notable differences at each time point. Moreover, after reaching swelling equilibrium in 24 hours, the gelatin group failed to maintain its original cylindrical shape and began to lose material integrity compared to the GelMA cryogels (Fig. 3c). This suggests that gelatin may not be suitable for providing a stable bone repair space, whereas all crosslinked GelMA cryogels maintained their stability for a longer duration.

3.7. Biodegradation of the cryogels

The in vitro degradation ratios of gelatin, GelMA, and nHA–GelMA cryogels in collagenase type I solution are shown in Fig. 3d. All materials eventually degraded completely, but at different rates. The gelatin group exhibited the shortest degradation time, lasting only one day. The GM60-10s and GM60-30s groups degraded in six and seven days, respectively, in collagenase type I solution, while the GM90-10s and GM90-30s groups lasted nine days. These results suggest that the methacrylation degree is the primary factor influencing the degradation, with an inverse relationship. The 5, 10, and 20 HA groups took nine days to fully degrade, compared to six days for the 0 HA group. However, no significant differences were observed among the degradation ratios of the 5, 10, and 20 HA groups, indicating that nHA effectively extended the degradation time of GelMA cryogels, independent of nHA concentration.

Given the complexity of the in vivo enzymatic environment, we investigated the in vivo degradation and biocompatibility through subcutaneous implantation in rats. After 14 days, all cryogels were collected, lyophilized, and weighed, and the viscera from each group were examined by H&E staining. As shown in Fig. 4a, all cryogels were encapsulated by a fibrous capsule with some tissue and blood vessel infiltration after 14 days of implantation. The residual mass ratios for each group are presented in Fig. 4b. The gelatin group was fully degraded, consistent with reports of gelatin's instability at body temperature. Despite gelatin's advantages, such as low antigenicity and good compatibility, its rapid biodegradation limits its application in tissue regeneration.^34,35 Consequently, gelatin lacks the resistance and integrity required for a GBR membrane and was excluded from further studies. However, all the other cryogels, including GelMA and nHA–GelMA groups, presented relatively high residual ratios exceeding 90%. Although no statistically significant differences were observed among the groups, a trend was noticed: increasing crosslinking time, methacrylation degree, and nHA concentration reduced the in vivo degradation rate. In our study, compared to gelatin, all crosslinked cryogels demonstrated significantly improved degradation profiles, extending in vitro degradation to 6–9 days and maintaining stability for 14 days in vivo, highlighting their potential as GBR membranes. Previous studies reported that GelMA showed an 18.7% mass loss after incubation for 14 days in PBS and a slightly higher degradation rate in 1.75 μg ml⁻¹ collagenase.³⁶ Additionally, GelMA hydrogels remained stable for at least four weeks in a rat arteriovenous loop model, where a PTFE chamber partially isolated some of the collagenases in vivo.³⁷ In our study, both GelMA and nHA–GelMA showed prolonged degradation times compared to uncrosslinked gelatin. In the subcutaneous implantation model, these cryogels showed less than 10% mass loss over two weeks and retained their original cylindrical shape, demonstrating slow biodegradation.

Fig. 4 Morphology of the GelMA and HA-GelMA cryogels after 14 days of subcutaneous implantation. (a) Cryogels encapsulated by a fibrous capsule with some tissue and blood vessel infiltration. (b) Quantitative analysis of residual weight of implanted cryogels, showing that all the GelMA and nHA–GelMA cryogels retained approximately 90% of their original weight.

3.8. Biocompatibility of the cryogels

H&E staining was performed to evaluate in vivo biocompatibility after 14 days of subcutaneous implantation in rats. As shown in Fig. 5, compared to the control group, the histological structures of the heart, liver, spleen, lungs, and kidneys in the gelatin, GelMA, and nHA–GelMA groups remained regular and organized. These findings indicated that all the cryogels were non-toxic and exhibited excellent biocompatibility.

Fig. 5 H&E staining of heart, liver, spleen, lungs, and kidneys in control and experimental groups. All organs exhibited typical histological structures (scale bar = 50 μm, magnification = 20×).

Cell viability of L929 fibroblasts in GelMA cryogel extract and BMSCs in nHA–GelMA cryogel extract liquid was assessed using the CCK-8 assay, with cells cultured in the untreated medium as the control group. Fig. 6a shows cell viability on days 1, 3, 5, and 7. L929 cells maintained excellent viability, exceeding 75% after 7 days, indicating that GM60-10s, GM60-30s, GM90-10s, and GM90-30s cryogels were non-toxic to L929 cells. BMSCs also presented good viability, over 75%, except in the 20HA group. This reduced viability may be attributed to the high concentration of nHA, which alters the pH of the extract, creating a less favorable environment for cell proliferation. At the same time, live/dead cell staining was performed and imaged on days 1, 3, and 5, as shown in Fig. 6b and c. At each time point, live cells (green) predominated in both L929 and BMSC cell populations. Over time, the density of L929 and BMSC live cells increased, and the cells maintained normal morphology. The results from the CCK-8 assay and live/dead cell staining collectively demonstrate that GelMA and nHA–GelMA cryogels exhibited excellent cell compatibility, promoting cell growth and proliferation.

Fig. 6 (a) CCK-8 assay results for L929 fibroblasts and BMSCs cultured in cryogel on days 1, 3, 5, and 7 (p = 0.7182 and 0.5299, respectively). Live/dead staining of L929 (b) and BMSCs (c) on days 1, 3 and 5.

SEM images of L929 fibroblasts and BMSCs on the surface of GelMA and nHA–GelMA cryogels are shown in Fig. 7a and b. L929 cells adhered, spread and established cell-to-cell contact on GelMA cryogels with different crosslinking times and methacrylation degrees. These cells maintained a typical spindle-shaped or triangular morphology. As the pore size of the GelMA cryogels decreased, the L929 cell density also decreased. BMSCs showed good adhesion and spreading on the surface of nHA–GelMA cryogels across different nHA concentrations. However, due to high cell density, only a portion of the BMSCs exhibited the characteristic spindle shape.

Fig. 7 Cell adhesion of L929 (a) and BMSCs (b) on the surface of GelMA and nHA–GelMA cryogels with different crosslinking times, methacrylation degrees and nHA concentrations under observation with SEM on day 3.

The morphology of L929 and BMSCs on the surface of GelMA and nHA–GelMA cryogels after 3 days of culture was observed under a confocal microscope, as shown in Fig. 8a and b. Blue fluorescence (DAPI) indicated cell nuclei, while red fluorescence (rhodamine-labeled phalloidin) highlighted the cell membrane and cytoplasm. L929 cells presented a typical small spindle or triangular phenotype with extended pseudopodia on the surface of GelMA cryogels. Notably, as crosslinking time and methacrylation degree increased, the density of L929 cells decreased gradually, indicating that smaller pore sizes may hinder cell growth. BMSCs also presented a typical spindle shape on the surface of nHA–GelMA cryogels, with extended pseudopodia observed in some cells. However, higher nHA concentrations appeared to limit the extent of BMSC spreading.

Fig. 8 Cell morphology of L929 fibroblasts (a) and BMSCd (b) seeded on the surface of GelMA and nHA–GelMA cryogels with different crosslinking times, methacrylation degrees and nHA concentrations observed under a fluorescence microscope on day 3.

3.9. Barrier function of the cryogels

The average infiltration depth of L929 cells was measured using a confocal microscope to evaluate the barrier function of GM60-10s, GM60-30s, GM90-10s, and GM90-30s cryogels. After staining with rhodamine-labeled phalloidin and DAPI, cross-sectional images of the GelMA cryogels and the average infiltration depths for each group are shown in Fig. 9, with the upper side of the image representing the surface of the cryogel. L929 cells on GM60-10s cryogels exhibited the greatest infiltration depth, averaging 300.0 μm, while GM90-30s cryogels exhibited the shortest infiltration depth at 45.3 μm. L929 cells on GM90-30s cryogels displayed the highest density and intensive cell-to-cell contact. The differences in infiltration depth were statistically significant, indicating that increasing crosslinking time and methacrylation degree improved the barrier function of GelMA cryogels. In other words, as pore size decreased, the barrier function of GelMA cryogels improved, with the GM90-30s cryogel exhibiting the best barrier performance. A previous study on an oxidized sodium alginate–collagen heterogeneous membrane reported a similar trend, with infiltration depth decreasing as pore size was reduced from 240–310 μm to 30–60 μm.²⁹ Other studies have also shown that cell migration was restricted on scaffolds with smaller pore sizes.³⁸

Fig. 9 (a) Cross-sectional view of L929 cell infiltration depth into GelMA cryogels with varying crosslinking times and methacrylation degrees, with the upper side of the image representing the cryogel surface. (b) Quantitative analysis of the L929 cell infiltration depth for the four GelMA groups.

3.10. Osteogenesis-promoting function of the cryogels

ALP staining of osteogenically induced BMSCs on nHA–GelMA cryogels after 7 days, observed using a stereomicroscope, is shown in Fig. 10a. As the concentration of nHA increased, the cryogels gradually transitioned from translucent to ivory and opaque, with deeper blue ALP protein staining. However, the ALP staining peaked at an nHA concentration of 10 mg ml⁻¹ and became lighter at 20 mg ml⁻¹. This may be due to elevated phosphate ions in the culture medium lowering the pH, thereby affecting cell proliferation and differentiation. Quantitative ALP analysis (Fig. 10b) showed a similar trend, with the highest osteogenic differentiation at 10 mg ml⁻¹ nHA, slightly decreasing at 20 mg ml⁻¹. The differences were statistically significant.

Fig. 10 (a) ALP staining of BMSCs seeded on HA-GelMA cryogels after 14 days. (b) Quantitative ALP activity measured using a commercial kit at 405 nm, expressed as optical density (OD) values.

3.11. In vivo bone regeneration

The rat cranial bone defect model provides a uniform and reproducible platform for evaluating bone regeneration by radiographic and histological analyses.³⁹ A 5 mm cranial bone defect has been reported to yield 18.29% and 21.44% new bone formation at 1 and 3 months, respectively.⁴⁰ In this study, rat cranial bones containing the defect area were collected 8 weeks post-surgery and scanned using a micro-CT to examine and analyze the bone regeneration. The three-dimensional reconstruction images are shown in Fig. 11a. A scoring system was established to evaluate new bone formation based on bony bridging and union: complete bony bridging across the defect diameter scored 4; partial bridging scored 3; new bone at defect borders scored 2; bone spicules throughout the defect scored 1; and no bone formation scored 0.⁴¹ The blank control group exhibited sparse bone spicules scoring 1, whereas the 10HA-GM90-30s double-layer group had excellent bone regeneration ability, with new bone formation nearly filling the whole defect area, scoring 4. The GM90-30s and 10HA single-layer groups showed new bone extending from the defect borders to the center of the defect, scoring 3. Remarkably, bone formation in the BioGide® group was comparable to that in the 10HA-GM90-30s double-layer group, indicating that the double-layer membranes effectively enhanced bone reconstruction activity. Quantitative analysis (Fig. 11b) revealed a significant difference in bone volume/tissue volume (BV/TV) between the control group (0.04747 ± 0.009343) and the 10HA-GM90-30s double-layer group (0.3750 ± 0.1113) with p = 0.0042, which coordinated with the histologic scores. Similarly, BMD differed significantly between the control group (107.6 ± 7.087 mg cm⁻³) and the 10HA-GM90-30s double-layer group (402.6 ± 119.8 mg cm⁻³p = 0.0088). No statistical differences were observed in BV/TV and BMD between the 10HA-GM90-30s double-layer group and the BioGide® group (0.3093 ± 0.7593, p = 0.4230; 334.1 ± 96.13 mg cm⁻³, p = 0.4553). These results suggest that the 10HA-GM90-30s double-layer cryogels could significantly enhance new bone formation, matching the performance of BioGide®, a clinically used GBR membrane. The superior performance of the 10HA-GM90-30s double-layer membrane can be attributed to several factors. First, its dense-and-loose porous structure mimics that of BioGide®, facilitating bone regeneration through dual functionality: barrier maintenance and osteoconduction. The dense layer acts as a mechanical barrier, preventing the rapid infiltration of epithelial and connective tissue cells that compete with osteogenic cells for defect space. The loose, porous layer provides an optimal microenvironment for cell migration, vascularization, and osteogenic differentiation due to its high permeability and large surface area. Second, the GM90-30s dense layer had the highest compressive strength and lowest swelling ratio among the GelMA cryogels, maintaining space for osteoblast aggregation and differentiation. Third, while classical GBR emphasizes space maintenance and osteoconduction, the inclusion of osteoinductive nHA enhances bone formation by promoting mineral deposition and osteogenic differentiation.

Fig. 11 (a) 3D reconstruction of five different GBR membranes, including the control group. The 10HA-GM90-30s double-layer group with a dense-and-loose structure exhibited the highest bone regeneration capacity. The dotted circle demonstrates the 5 mm bone defect, and the CT images show the bone regeneration in cross-sectional views. (b) Morphometric analysis of bone volume/total volume (BV/TV) and bone mineral density (BMD) calculated at the defect area.

Histological analyses, including H&E staining and Masson's trichrome staining, further illustrate the tissue regeneration situation in the defect area, as shown in Fig. 12a and b. The control group was dominated by fibrous tissue, confirming the defect's critical size. In the 10HA-GM90-30s group, thick new bone formed, comparable to the BioGide® group, both scoring 4 qualitatively. In contrast, the GM90-30s and 10HA single-layer groups showed a combination of thin, incomplete bone layers and fibrous tissue. In Masson's staining, the newly formed bone appeared as red-and-blue tissue, with the red component darkening as the bone matured, while the fibrous tissue was stained red. The control group showed only fibrous tissue, whereas the 10HA-GM90-30s group exhibited thick red-and-blue newly formed bone tissue, with minimal fibrous tissue formed at the defect area. Biodegradable materials in aqueous environments undergo four stages: hydration, strength loss, mass integrity loss and solubilization—potentially affecting the membrane's barrier function during resorption or in vivo wound healing. The in vivo barrier and osteogenic abilities of the selected cryogels could also be analyzed in the rat cranial bone defects. The results of histological staining revealed that massive fibrous tissue was only observed in the defect area of the control group, while the other groups, with double- or single-layer cryogel membranes, predominantly formed bone tissue, indicating that the GM90-30s cryogel effectively resisted fibrous tissue intrusion in vivo.

Fig. 12 Images of H&E staining (a) and Masson's trichrome staining (b) from the sagittal view, with higher-magnification images of the red-boxed areas.

Specific osteogenic marker OCN, middle osteogenic marker Col I and early angiogenic marker CD31 were analyzed by IHC staining to investigate the promotion of osteogenesis and angiogenesis in different groups, as shown in Fig. 13a–c. The positive staining areas for OCN and Col I were mainly gathered at the edges of the new bone, especially obvious in the 10HA-GM90-30s double-layer group and the BioGide® group. As for CD31, positive-staining areas were mostly found in the fibrous tissue surrounding the new bone in the 10HA-GM90-30s double-layer group and single-layer group, while only a few could be found within the new bone tissue in the BioGide® group. Semi-quantification of positive-stained areas revealed that the control group had the lowest expression of OCN (2.597 ± 0.4402), Col I (1.930 ± 0.4263), and CD31 (3.003 ± 0.8740), while 1the 0HA-GM90-30s double-layer group exhibited the highest expression of OCN (8.012 ± 1.962), Col I (8.405 ± 1.077) and CD31(5.518 ± 1.556), with significant differences compared to the control group (p < 0.0001). No statistical differences were observed in OCN, Col I, and CD31 expression between the 10HA-GM90-30s double-layer and BioGide® groups (p = 0.2444, 0.0825, and 0.0572, respectively). These IHC results confirm that the 10HA-GM90-30s double-layer membrane significantly promoted osteogenesis and angiogenesis. Our results suggested that the 10HA-GM90-30s double-layer membrane with a dense-and-loose porous structure demonstrated splendid new bone formation ability in the cranial bone defect area.

Fig. 13 (a) The immunohistochemical staining for OCN, Col I and CD31 in the defect area tissues of each group at 8 weeks post-surgery with higher-magnification views. The 10HA-GM90-30s double-layer group showed the strongest positive brown staining as the BioGide® group. (b) The quantitative result of the immunohistochemical staining for OCN, Col I and CD31.

While promising, the molecular mechanisms behind the osteogenic effects of the nHA–GelMA membrane remain unclear and warrant further study using gene expression or pathway analyses. Additionally, the current double-layer membrane lacks intrinsic antibacterial properties, which may compromise its performance in clinical environments prone to infection. Future efforts may explore incorporating antibacterial agents or modifying surface chemistry to enhance the membrane's resistance to microbial contamination.

4. Conclusions

In this study, we developed a novel double-layer membrane composed of GelMA and nHA with tunable physical and chemical properties, showing significant osteogenesis-promoting function both in vitro and in vivo. Methacryloyl substitution enabled photo-crosslinking of gelatin, while nHA incorporation enhanced its osteogenic potential. By adjusting the crosslinking time, methacrylation degree, and nHA concentration, the cryogels showed a uniform porous microstructure with varying pore sizes. Our results confirmed that methacrylation of gelatin could significantly improve mechanical strength, swelling ratio, and degradation rate. The addition of nHA powder effectively extended the degradation time and enhanced the compressive mechanical properties. Furthermore, cell adhesion, proliferation, and morphology were also promoted when cells were seeded on the GelMA and nHA–GelMA cryogels. The GelMA cryogel with the smallest pore size showed the best barrier function and was selected as the dense layer for the GBR membrane. The nHA–GelMA with an nHA concentration of 10 mg ml⁻¹ had the best osteogenic ability and was chosen as the loose layer. Both selected layers offered excellent biocompatibility and a straightforward synthetic procedure. The double-layer nHA–GelMA cryogel membrane significantly enhanced bone regeneration, with efficacy comparable to the commercial GBR product BioGide®, underscoring its promising potential for translational applications.

Author contributions

JY Wang, WJ Zhang and Y Tian conceptualized the study and designed the experiments. JY Wang conducted all the cryogel synthesis, characterization test, and in vitro/in vivo experiments. XR Zhang assisted with cryogel synthesis. YR He supported the cryogel characteristic test. XH Wang contributed to the in vivo experiments. XL Xiao and S Cha aided in the in vitro cell experiments. D Bai provided guidance on experimental design. JY Wang drafted the manuscript, which was revised by WJ Zhang and Y Tian.

Conflicts of interest

The authors declare no competing financial interests. 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.

Data availability

The data supporting this article have been included in the figures and tables shown in the article.

Acknowledgements

This study was supported by the Sichuan Science and Technology Program, China (Grant No. 2025ZNSFSC0759).

References

1. F. Brugnami, A. Caiazzo, L. Signorelli and D. Capri, Regenerative orthodontics: GBR and corticotomy to stretch the limits of orthodontic treatment, Int. J. Periodontics Restor. Dent. , 2021, 41(1), 105–111.
2. M. Radenković, S. Alkildani, I. Stoewe, J. Bielenstein, B. Sundag, O. Bellmann, O. Jung, S. Najman, S. Stojanović and M. Barbeck, Comparative in vivo analysis of the integration behavior and immune response of collagen-based dental barrier membranes for guided bone regeneration (GBR), Membranes, 2021, 11(9), 712.
3. R. Pistilli, C. Barausse, M. Simion, L. Bonifazi, M. Karaban, A. Ferri and P. Felice, Simultaneous GBR and implant placement with resorbable membranes in the rehabilitation of partially edentulous and horizontally atrophic dental arches: A retrospective study on 97 implants with a 3- to 7-year follow-up, Int. J. Periodontics Restor. Dent. , 2022, 42(3), 371–379.
4. I. Elgali, O. Omar, C. Dahlin and P. Thomsen, Guided bone regeneration: Materials and biological mechanisms revisited, Eur. J. Oral Sci., 2017, 125(5), 315–337.
5. A. Cucchi, E. Vignudelli, A. Fiorino, G. Pellegrino and G. Corinaldesi, Vertical ridge augmentation (VRA) with Ti-reinforced d-PTFE membranes or Ti meshes and collagen membranes: 1-year results of a randomized clinical trial, Clin. Oral Implants Res., 2021, 32(1), 1–14.
6. P. Bunyaratavej and H. L. Wang, Collagen membranes: A review, J. Periodontol., 2001, 72(2), 215–229.
7. C. Xu, C. Lei, L. Meng, C. Wang and Y. Song, Chitosan as a barrier membrane material in periodontal tissue regeneration, J. Biomed. Mater. Res., Part B, 2012, 100(5), 1435–1443.
8. Y. Ueyama, K. Ishikawa, T. Mano, T. Koyama, H. Nagatsuka, K. Suzuki and K. Ryoke, Usefulness as guided bone regeneration membrane of the alginate membrane, Biomaterials, 2002, 23(9), 2027–2033.
9. Z. Sheikh, J. Qureshi, A. M. Alshahrani, H. Nassar, Y. Ikeda, M. Glogauer and B. Ganss, Collagen based barrier membranes for periodontal guided bone regeneration applications, Odontology, 2017, 105(1), 1–12.
10. S. A. Kay, L. Wisner-Lynch, M. Marxer and S. E. Lynch, Guided bone regeneration: Integration of a resorbable membrane and a bone graft material, Pract. Periodontics Aesthet. Dent., 1997, 9(2), 185–194 , quiz 196.
11. T. Wu, L. Huang, J. Sun, J. Sun, Q. Yan, B. Duan, L. Zhang and B. Shi, Multifunctional chitin-based barrier membrane with antibacterial and osteogenic activities for the treatment of periodontal disease, Carbohydr. Polym., 2021, 269, 118276.
12. R. M. Raftery, B. Woods, A. L. P. Marques, J. Moreira-Silva, T. H. Silva, S. A. Cryan, R. L. Reis and F. J. O'Brien, Multifunctional biomaterials from the sea: Assessing the effects of chitosan incorporation into collagen scaffolds on mechanical and biological functionality, Acta Biomater., 2016, 43, 160–169.
13. V. Charulatha and A. Rajaram, Influence of different crosslinking treatments on the physical properties of collagen membranes, Biomaterials, 2003, 24(5), 759–767.
14. J. Y. Lai and Y. T. Li, Functional assessment of cross-linked porous gelatin hydrogels for bioengineered cell sheet carriers, Biomacromolecules, 2010, 11(5), 1387–1397.
15. J. W. Nichol, S. T. Koshy, H. Bae, C. M. Hwang, S. Yamanlar and A. Khademhosseini, Cell-laden microengineered gelatin methacrylate hydrogels, Biomaterials, 2010, 31(21), 5536–5544.
16. Y. Liu and M. B. Chan-Park, A biomimetic hydrogel based on methacrylated dextran-graft-lysine and gelatin for 3D smooth muscle cell culture, Biomaterials, 2010, 31(6), 1158–1170.
17. K. Yue, G. Trujillo-de Santiago, M. M. Alvarez, A. Tamayol, N. Annabi and A. Khademhosseini, Synthesis, properties, and biomedical applications of gelatin methacryloyl (GelMA) hydrogels, Biomaterials, 2015, 73, 254–271.
18. A. I. Van Den Bulcke, B. Bogdanov, N. De Rooze, E. H. Schacht, M. Cornelissen and H. Berghmans, Structural and rheological properties of methacrylamide modified gelatin hydrogels, Biomacromolecules, 2000, 1(1), 31–38.
19. Y. Lee, J. M. Lee, P. K. Bae, I. Y. Chung, B. H. Chung and B. G. Chung, Photo-crosslinkable hydrogel-based 3D microfluidic culture device, Electrophoresis, 2015, 36(7–8), 994–1001.
20. S. V. Vlierberghe, V. Cnudde, P. Dubruel, B. Masschaele, A. Cosijns, I. D. Paepe, P. J. Jacobs, L. V. Hoorebeke, J. P. Remon and E. Schacht, Porous gelatin hydrogels: 1. Cryogenic formation and structure analysis, Biomacromolecules, 2007, 8(2), 331–337.
21. Y. C. Chen, R. Z. Lin, H. Qi, Y. Yang, H. Bae, J. M. Melero-Martin and A. Khademhosseini, Functional human vascular network generated in photocrosslinkable gelatin methacrylate hydrogels, Adv. Funct. Mater., 2012, 22(10), 2027–2039.
22. A. O. Elzoghby, Gelatin-based nanoparticles as drug and gene delivery systems: Reviewing three decades of research, J. Controlled Release, 2013, 172(3), 1075–1091.
23. S. Kargozar, R. K. Singh, H. W. Kim and F. Baino, “Hard” ceramics for “soft” tissue engineering: Paradox or opportunity?, Acta Biomater., 2020, 115, 1–28.
24. M. Mehrali, A. Thakur, C. P. Pennisi, S. Talebian, A. Arpanaei, M. Nikkhah and A. Dolatshahi-Pirouz, Nanoreinforced hydrogels for tissue engineering: biomaterials that are compatible with load-bearing and electroactive tissues, Adv. Mater., 2017, 29(8), 1603612.
25. F. Guo, C. Yuan, H. Huang, X. Deng, Z. Bian, D. Wang, K. Dou, L. Mei and Q. Zhou, Regulation of T cell responses by nano-hydroxyapatite to mediate the osteogenesis, Front. Bioeng. Biotechnol., 2022, 10, 884291.
26. P. Johansson, R. Jimbo, P. Kjellin, F. Currie, B. R. Chrcanovic and A. Wennerberg, Biomechanical evaluation and surface characterization of a nano-modified surface on PEEK implants: A study in the rabbit tibia, Int. J. Nanomed., 2014, 9, 3903–3911.
27. T. Gong, Z. Wang, Y. Zhang, Y. Zhang, M. Hou, X. Liu, Y. Wang, L. Zhao, N. D. Ruse, T. Troczynski and U. O. Häfeli, A comprehensive study of osteogenic calcium phosphate silicate cement: Material characterization and in vitro/in vivo testing, Adv. Healthc. Mater., 2016, 5(4), 457–466.
28. A. Jana, M. Das and V. K. Balla, In vitro and in vivo degradation assessment and preventive measures of biodegradable Mg alloys for biomedical applications, J. Biomed. Mater. Res., Part A, 2022, 110(2), 462–487.
29. Y. He, Y. Tian, W. Zhang, X. Wang, X. Yang, B. Li, L. Ge, D. Bai and D. Li, Fabrication of oxidized sodium alginate-collagen heterogeneous bilayer barrier membrane with osteogenesis-promoting ability, Int. J. Biol. Macromol., 2022, 202, 55–67.
30. S. T. Koshy, T. C. Ferrante, S. A. Lewin and D. J. Mooney, Injectable, porous, and cell-responsive gelatin cryogels, Biomaterials, 2014, 35(8), 2477–2487.
31. Z. Yuan, X. Yuan, Y. Zhao, Q. Cai, Y. Wang, R. Luo, S. Yu, Y. Wang, J. Han, L. Ge, J. Huang and C. Xiong, Injectable GelMA cryogel microspheres for modularized cell delivery and potential vascularized bone regeneration, Small, 2021, 17(11), e2006596.
32. F. Jonidi Shariatzadeh, A. Solouk, S. Bagheri Khoulenjani, S. Bonakdar and H. Mirzadeh, Injectable and reversible preformed cryogels based on chemically crosslinked gelatin methacrylate (GelMA) and physically crosslinked hyaluronic acid (HA) for soft tissue engineering, Colloids Surf., B, 2021, 203, 111725.
33. C. H. Hämmerle, J. Schmid, A. J. Olah and N. P. Lang, A novel model system for the study of experimental guided bone formation in humans, Clin. Oral Implants Res., 1996, 7(1), 38–47.
34. G. Yang, Z. Xiao, H. Long, K. Ma, J. Zhang, X. Ren and J. Zhang, Assessment of the characteristics and biocompatibility of gelatin sponge scaffolds prepared by various crosslinking methods, Sci. Rep., 2018, 8(1), 1616.
35. S. Liu, X. Sun, T. Wang, S. Chen, C. G. Zeng, G. Xie, Q. Zhu, X. Liu and D. Quan, Nano-fibrous and ladder-like multi-channel nerve conduits: Degradation and modification by gelatin, Mater. Sci. Eng., C, 2018, 83, 130–142.
36. C. S. Russell, A. Mostafavi, J. P. Quint, A. C. Panayi, K. Baldino, T. J. Williams, J. G. Daubendiek, V. Hugo Sánchez, Z. Bonick, M. Trujillo-Miranda, S. R. Shin, O. Pourquie, S. Salehi, I. Sinha and A. Tamayol, In situ printing of adhesive hydrogel scaffolds for the treatment of skeletal muscle injuries, ACS Appl. Bio Mater., 2020, 3(3), 1568–1579.
37. S. Heltmann-Meyer, D. Steiner, C. Müller, D. Schneidereit, O. Friedrich, S. Salehi, F. B. Engel, A. Arkudas and R. E. Horch, Gelatin methacryloyl is a slow degrading material allowing vascularization and long-term usein vivo, Biomed. Mater., 2021, 16(6), 065004.
38. C. Chu, Y. Wang, Y. Wang, R. Yang, L. Liu, S. Rung, L. Xiang, Y. Wu, S. Du, Y. Man and Y. Qu, Evaluation of epigallocatechin-3-gallate (EGCG) modified collagen in guided bone regeneration (GBR) surgery and modulation of macrophage phenotype, Mater. Sci. Eng., C, 2019, 99, 73–82.
39. P. S. Gomes and M. H. Fernandes, Rodent models in bone-related research: The relevance of calvarial defects in the assessment of bone regeneration strategies, Lab. Anim., 2011, 45(1), 14–24.
40. A. Vajgel, N. Mardas, B. C. Farias, A. Petrie, R. Cimões and N. Donos, A systematic review on the critical size defect model, Clin. Oral Implants Res., 2014, 25(8), 879–893.
41. P. P. Spicer, J. D. Kretlow, S. Young, J. A. Jansen, F. K. Kasper and A. G. Mikos, Evaluation of bone regeneration using the rat critical size calvarial defect, Nat. Protoc., 2012, 7(10), 1918–1929.

---