3D-printed GelMA-Alginate microsphere scaffold with staged dual-growth factor release for enhanced bone regeneration

Abstract

The process of bone regeneration is inherently complex, relying on the precise regulation of multiple growth factors. However, when bone defects exceed a critical size, this delicate regulatory balance is disrupted, severely hindering new bone formation. In this study, we developed a bioengineered scaffold with staged growth factor release capabilities to promote efficient bone repair, constructed by integrating gelatin methacryloyl (GelMA) with embedded alginate microspheres (AM). Through microfluidic technologies and 3D printing, AM were fabricated and integrated into the GelMA matrix to fabricate the composite scaffold, which not only maintains structural integrity but also enables gradient-controlled release of vascular endothelial growth factor (VEGF) and bone morphogenetic protein-2 (BMP-2). This design ensures uniform bioactivity of the printed constructs, optimizing conditions for bone regeneration. In vitro experiments demonstrated that the scaffold exhibited an initial burst release of VEGF, followed by a sustained, prolonged release of BMP-2 mimicking the natural sequential processes of tissue vascularization and subsequent osteogenesis. Upon in vivo implantation, the scaffold has demonstrated the ability to effectively repair a critical-sized cranial defect in a rat model within a 12-week period. These findings highlight that the bioink-printed construct, through its spatiotemporally coordinated delivery of VEGF and BMP-2, effectively stimulates bone regeneration. The staged growth factor-releasing 3D printed scaffold offers a promising therapeutic strategy, particularly for treating large-scale bone defects, by significantly enhancing the efficiency of bone defect repair.

Graphical abstract

1. Introduction

Addressing bone defects larger than a critical threshold is a significant clinical challenge, as these often fail to heal naturally [1,2]. Traditional treatments, such as autografts, allografts, and xenografts, are hampered by issues like limited availability, insufficient functionality, and immune reactions [3]. In response, there's a growing focus on 3D bioprinting, whose ability to integrate both structure and function is pivotal for treating large bone defects, promising a solution to the limitations of current bone repair methods [4,5].

In the field of 3D bioprinting for bone repair, the bioink plays a pivotal role, necessitating not only excellent printability but also the ability to foster a conducive microenvironment for cell adhesion, proliferation, and differentiation [6]. Hydrogels have gained prominence as bioinks due to their inherent biocompatibility, minimal cytotoxicity, and their structural resemblance to the extracellular matrix (ECM), which is largely attributed to their high water content [7,8]. Alginates, gelatin methacrylate anhydride (GelMA), collagen, and similar hydrogels are currently under extensive investigation for their use as bioinks in 3D bioprinting for bone repair [4,6].

To augment the capabilities of hydrogel bioinks in bone repair, the integration of growth factors has become a favored strategy [9,10]. For instance, Chai and colleagues have developed a bioactive GelMA bioink enriched with BMP-2, which has been shown to enhance the adhesion, proliferation, and osteogenic differentiation of bone marrow stem cell (BMSC) [11]. Sun and colleagues have also introduced hydrogel bioinks that incorporate GelMA and BMP-4-loaded mesoporous silica nanoparticles, which have been demonstrated to significantly expedite osteogenic differentiation and bone regeneration in calvarial critical-size defect models [12].

However, bone defect repair is a multifaceted physiological process that requires the intricate coordination of various growth factors, such as VEGF and BMP [13]. These factors are instrumental in orchestrating the remodeling of blood vessels and the stimulation of osteogenic differentiation, thereby synergistically promoting bone regeneration. The simultaneous delivery of VEGF and BMP has been found to enhance bone formation in a synergistic manner[[14], [15], [16], [17]]. It is also noteworthy that in the natural process of bone defect repair, VEGF is expressed early on, facilitating vascular regeneration in the defect area, while BMP is expressed continuously and gradually to stimulate osteogenic differentiation [[18], [19], [20]].

The pursuit of bioinks that can emulate the physiological sequence of vascularization followed by osteogenesis during bone repair may maximizing the therapeutic potential of 3D printed constructs. Current methodologies predominantly rely on post-printing interventions such as absorption, coating, or multi-ink printing, to integrate growth factors[21,22]. These techniques, while effective, may not ensure a coherent distribution of bioactivity throughout the construct. The proposition of a unified bioink, embedded with a variety of biofactors, offers a promising alternative by potentially ensuring a more homogenous bioactivity across the 3D printed structure. However, as our literature review reveals, there is a notable absence of research on the precise, controlled release of multiple growth factors within a single bioink system.

Herein, we have developed a pioneering Gel/AM bioprinted scaffold, employing GelMA hydrogel and strategically dispersed alginate microspheres (AMs) as a bioink for bioprinting. Specifically, the AMs are well integrated within the GelMA hydrogel for Gel/AM bioprinted scaffolds to provide a gradient encapsulation of VEGF and BMP-2. In vitro studies demonstrated a rapid release of VEGF, succeeded by a gradual and sustained release of BMP-2 from the scaffolds. This sequential release strategy is meticulously tailored to align with the physiological sequence of vascularization and subsequent osteogenesis. The obvious migration and tube formation of human umbilical vein endothelial cells (HUVEC) to Gel/AM bioprinted scaffold were observed and significant osteogenic differentiation of BMSC to Gel/AM bioprinted scaffold was observed by immunofluorescence staining of Runx2 and Osterix in vitro. Finally, the efficient biological properties for vascular regeneration and bone regeneration of the Gel/AM bioprinted scaffold were tested in the rat critical-sized bone defect model (Fig. 1).

Fig. 1.

Schematic diagram illustrating the preparation process of the Gel/AM composite bioink and the Gel/AM 3D bioprinted scaffold, followed by the in vivo implantation of the Gel/AM 3D bioprinted scaffold for bone repair.

2. Materials and methods

2.1. Materials

Sodium alginate, span 80, β-glycerophosphate, dexamethasone and triton X-100 were purchased from Sigma-Aldrich (St Louis, MO, USA). CaCl₂, crystal violet and collagenase were purchased from Beijing Solarbio Science & Technology Co., Ltd (Beijing, China). Anhydrous alcohol, isoamyl acetate, citrate sodium and tribromoethanol were purchased from Shanghai Macklin Biochemical Co., Ltd (Shanghai, China). GelMA were purchased from EFL-Tech Co., Ltd (Suzhou, China). FITC-BSA and Cy3-BSA were purchased from Xi'an Qiyue Biology Co., Ltd (Xi'an, China). BMP-2, VEGF and enzyme-linked immunosorbent assay (ELISA) kit were purchased from Sino Biological Inc (Beijing, China), Human Umbilical Vein Endothelial Cells (HUVECs) was purchased from GuangZhou Jennio Biotech Co., Ltd (Guangzhou, China). Endothelial Cell Medium (ECM) was purchased from ScienCell Research Laboratories (San Diego, CA, USA). Paraformaldehyde fixing solution and 4′, 6 diamidino-2-phenylindole (DAPI) were purchased from Beijing Leagene Technology Co., Ltd (Beijing, China). Calcein-AM was purchased from US Everbright Inc (Beijing, China). α-MEM was purchased from Beijing ThermoFisher Scientific Biochemical Products Co., Ltd (Beijing, China). L-Ascorbic Acid Phosphate Magnesium was purchased from FUJIFILM Wako Pure Chemical Corporation (Japan). Antibodies against Runx2 (host: rabbit; Cat. No. ab192256; dilution, 1:1000) and antibodies against Osterix (host: rabbit; Cat. No. ab63856; dilution, 1:1000) were purchased from Abcam. Secondary antibodies (goat anti-rabbit Alexa Fluor 488-conjugated, Cat. No. SA00013-2) was purchased from Proteintech (Chicago, USA). rhodamine phalloidin solution was purchased from Cytoskeleton Inc (NJ, USA). Anti-CD31 (host: rabbit; Cat. GB11063-2; dilution, 1:500), anti-α-SMA (host: rabbit; Cat. GB111364; dilution, 1:500), goat anti-rabbit secondary antibody (CY3-conjugated, Cat. GB21303; dilution, 1:300) and goat anti-rabbit secondary antibody (Alexa Fluor 488-conjugated, Cat. GB25301; dilution, 1:400) were purchased from Wuhan Servicebio Technology Co., Ltd (Wuhan, China)

2.2. Preparation of alginate microspheres (AM)

Alginate microspheres (AM) were created using a droplet-based microfluidic device (Suzhou Wenhao Technology Co., Ltd., Suzhou, China) in conjunction with Ca²⁺ ion crosslinking. The dispersed phase consisted of a mixture of sodium alginate (Sigma-Aldrich, St. Louis, MO, USA) aqueous solution (1.5 % wt). The continuous phase comprised liquid paraffin (Shanghai Macklin Biochemical Co., Ltd., Shanghai, China) and 2 % wt Span 80 (Sigma-Aldrich, St. Louis, MO, USA) that stabilized the droplets formed in the microfluidic device. The flow rates of the dispersed and continuous phases were adjusted to 0.4 mL/h ∼0.45 mL/h and 4.0 mL/h ∼5.0 mL/h, respectively, to achieve uniform alginate droplets (50 μm–60 μm). The alginate droplets were collected in a bath containing 2.0 % wt CaCl₂ aqueous solution (Beijing Solarbio Science & Technology Co., Ltd., Beijing, China) and cross-linked by Ca²⁺ to fabricate AM. The collected AM were centrifuged and washed several times to remove liquid paraffin and CaCl2 aqueous solution, yielding AM for further analysis.

2.3. Morphology of AM

SEM was employed to examine the structures of AM. The obtained AM were dehydrated with 30 %, 50 %, 70 %, and 90 % gradient alcohol respectively. Subsequently, they underwent additional dehydration with anhydrous alcohol (Shanghai Macklin Biochemical Co., Ltd., Shanghai, China) for three 20-min cycles. Following this, the AM were immersed in isoamyl acetate (Shanghai Macklin Biochemical Co., Ltd., Shanghai, China) twice for 20 min each time to substitute ethanol and then processed using the supercritical CO₂ fluid drying method. The resulting dried microspheres were dispersed onto double-sided conductive tape, which was affixed to the opposing side of a copper plate. A gold spraying procedure with a thickness of 10 nm was then applied. The SEM (ZEISS SUPRA® 55, Carl Zeiss, Germany) was utilized to scrutinize the morphological characteristics of the AM.

2.4. Mechanical testing of Gel/AM composite bioinks

The Gel/AM composite bioinks' compressive strength was assessed using an electromechanical universal testing instrument (Z050, Zwick, Germany). The bioinks were categorized into four groups: Gel/AM-0, Gel/AM-5, Gel/AM-10, and Gel/AM-15 based on the AM content (v/v) in a 5 % w/v GelMA solution (EFL-Tech Co., Ltd, Suzhou, China), corresponding to concentrations of 0 %, 5 %, 10 %, and 15 %, respectively. All bioinks underwent photo-crosslinking with a UV device (LS1601, EFL-Tech Co., Ltd, Suzhou, China) for 30 s and were prepared with a 9 mm diameter and 9 mm height. Subsequently, each hydrogel group (n = 5) was compressed perpendicularly at a rate of 5 mm/min at room temperature until fracture. Compressive stress-strain curves were recorded for each hydrogel and averaged from five independent samples. The ultimate compressive strength was noted at the hydrogel's fracture point, and the compressive modulus was determined by the slope of a linear segment in the stress–strain curve.

2.5. Fabrication and characterization of 3D printed Gel/AM scaffold

In this study, a 3D bioprinter (Bio-x, Cellink, Sweden) was employed for printing Gel/AM scaffold. Initially, various volumes (v/v) of AM were added to a GelMA solution (5 % w/v) prepared as bioinks for 3D printing Gel/AM scaffold. The crosslinking agent used was 0.25 % lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP). The Gel/AM bioinks were categorized into four groups: Gel/AM-0, Gel/AM-5, Gel/AM-10, and Gel/AM-15. Subsequently, the Gel/AM bioinks were loaded into plastic sterile syringe barrels and placed in a refrigerator at 4 °C for 10 min to facilitate rapid thermo-crosslinking, ensuring stable filament deposition before 3D printing. The temperature of the syringe barrel controller was then set to about 10 °C, while the printed platform temperature was adjusted to be 15 °C. The extrusion nozzle had an inner diameter of 27G. The printing process was executed at a speed of 4∼6 mm/s with a pressure range of 50∼80 kPa. Post-printing, the Gel/AM scaffold underwent exposure to UV light for 30 s to establish a stable structure through photo-crosslinking for subsequent characterization.

For the characterization of the Gel/AM scaffold, freeze-drying was performed using a Freeze Dryer (CTFD-12P, Qingdao Creatrust Technology Co., Ltd., Qingdao, China). The lyophilized scaffold were halved to expose an interior cross-section for observation. Subsequently, all scaffold were affixed to conductive tape, coated with gold spray, and subjected to scanning electron microscopy to analyze the interior morphology, structure, and distribution of AM within the scaffold.

2.6. Imitate encapsulation and release of growth factors in AM and Gel/AM scaffold

To assess the encapsulation and distribution of growth factors in AM and Gel/AM scaffold, FITC-BSA (Xi'an Qiyue Biology Co., Ltd., Xi'an, China) and Cy3-BSA (Xi'an Qiyue Biology Co., Ltd., Xi'an, China) were employed to mimic the presence of growth factors. Specifically, 400 μg of FITC-BSA was dissolved in a 1 mL aqueous solution of sodium alginate (1.5 % wt), serving as the dispersed phase for fabricating FITC-BSA-loaded AM using microfluidic droplet technology. Subsequently, 100 μl of FITC-BSA/AM was added to 1 mL of GelMA solution (5 % wt) mixed with 100 μg of Cy3-BSA to produce Gel/AM scaffold. The encapsulation and distribution of BSA were monitored using a confocal fluorescence microscope (A1R, Nikon, Tokyo, Japan).

To simulate and visualize the release of growth factors in AM and Gel/AM scaffold, fluorescent BSA was loaded into both. Specifically, 100 μl of pure FITC-BSA/AM was immersed in 1 mL ddH₂O and incubated at 37 °C for 14 days. The remaining quantity of FITC-BSA preserved in AM was observed under a confocal fluorescence microscope at intervals of 0 days, 3 days, 7 days, and 14 days. Similarly, Gel/AM scaffold were immersed in 1 mL ddH₂O and incubated at 37 °C for 14 days, and the remaining quantity of FITC-BSA and Cy3-BSA preserved in Gel/AM scaffold was observed under a confocal fluorescence microscope at intervals of 0 days, 3 days, 7 days, and 14 days. All procedures in this section were conducted in the dark.

2.7. The release profile of VEGF and BMP-2 in Gel/AM scaffold

To precisely assess the release of growth factors in AM and Gel/AM scaffold, VEGF and BMP-2 were incorporated into Gel/AM scaffold. Initially, 400 μg of BMP-2 (400 μg/mL,Sino Biological Inc., Beijing, China) was dissolved in a 1000 μl aqueous solution of sodium alginate (1.5 % wt) as the dispersed phase for fabricating BMP-2-loaded AM using microfluidic droplet technology. Subsequently, 100 μl of BMP-2/AM was added to 1 mL of GelMA solution (5 % wt) mixed with 100 μg of VEGF (Sino Biological Inc., Beijing, China) as inks to produce Gel/AM scaffold. Next, 100 μl of BMP-2/AM and Gel/AM scaffold loaded with VEGF and BMP-2 were separately immersed in 1 mL ddH₂O and incubated at 37 °C for 35 days. The supernatant was collected and replaced with an equal volume at intervals of 3, 5, 7, 10, 14, 21, 28, and 35 days. After 35 days, AM and Gel/AM were completely degraded with citrate sodium (2 % wt, Shanghai Macklin Biochemical Co., Ltd., Shanghai, China) and collagenase (2 U/mL, Beijing Solarbio Science & Technology Co., Ltd., Beijing, China) to assess the remaining quantity of BMP-2 and VEGF preserved in AM and Gel/AM scaffold. Finally, the released VEGF or BMP-2 in all collected supernatants were quantified using an enzyme-linked immunosorbent assay (ELISA) kit (Sino Biological Inc., Beijing, China) following the manufacturer's instructions.

2.8. Effects of Gel/AM scaffold loaded with VEGF on HUVECs angiogenesis

2.8.1. Detection of cell migration by scratch assay

The scratch assay was conducted to assess the impact of Gel/AM scaffold loaded with VEGF on endothelial cell functionality. Human Umbilical Vein Endothelial Cells (HUVECs, GuangZhou Jennio Biotech Co., Ltd., Guangzhou, China) were cultured up to Passage 4 (P4) by detaching the cells using Trypsin-EDTA and resuspending them in Endothelial Cell Medium (ECM, ScienCell Research Laboratories, San Diego, CA, USA). Subsequently, HUVECs were seeded and cultured in 6-well plates at a concentration of 1 × 10⁵ cells/well until they reached confluency. Following this, a 200 μL pipette tip was horizontally used to scratch the confluent HUVECs layer, and the cells were gently washed with PBS. Post-scratch, HUVECs were co-cultured with Gel/AM scaffold in ECM medium without FBS for 24 h at 37 °C under different treatment conditions, including the VEGF + group (the Gel/AM scaffold loaded with VEGF) and the VEGF- group (the pure Gel/AM without VEGF loading). Finally, cell migration was observed, images were captured using an optical microscope (DMi8, Leica, Germany), and the number of migrating cells for each group was analyzed with Image J software (National Institutes of Health, Bethesda, MD, USA).

2.8.2. Detection of cell migration by transwell assay

HUVECs were seeded and cultured in transwell inserts with an 8 μm pore-size at a concentration of 5 × 10⁴ cells/well in 24-well plates. Gel/AM scaffold, including the VEGF + group and the VEGF- group, were placed in the bottom chambers with ECM medium. After 24 h of incubation at 37 °C, non-migrating cells in the upper chambers were removed using cotton swabs. Cells that had migrated to the lower surface of the upper chambers were fixed in a 4 % wt paraformaldehyde fixing solution (Beijing Leagene Technology Co., Ltd., Beijing, China) and stained with 0.5 % wt crystal violet (Beijing Solarbio Science & Technology Co., Ltd., Beijing, China) for 30 min. To quantify the cells that migrated through the upper chamber, images were captured using an inverted microscope, and the number of migrating cells for each group was analyzed with Image J software.

2.8.3. HUVECs tube formation assay on Gel/AM scaffold

Additionally, a tube formation assay was conducted to assess the impact of Gel/AM scaffold loaded with VEGF on endothelial cell functionality. Gel/AM bioinks, loaded with or without VEGF, were mixed with Human Umbilical Vein Endothelial Cells (HUVECs) at a concentration of 1 × 10⁶ cells/mL to fabricate 3D bioprinted scaffold. Subsequently, the 3D bioprinted scaffold loaded with HUVECs were cultured with ECM medium for 14 days and 21 days, which was changed every second day. Finally, HUVECs within the Gel/AM scaffold were stained with Calcein-AM (US Everbright Inc., Beijing, China), and HUVECs within the Gel/AM scaffold were also immunostained with the CD31 primary antibody followed by the fluorescently conjugated secondary antibody (Wuhan Servicebio Technology Co., Ltd., Wuhan, China). Tube formation of HUVECs within the pores of Gel/AM scaffold was then observed under an inverted fluorescent microscope.

2.9. Osteogenic bioactivity of Gel/AM bioprinted scaffold loaded BMP-2 and VEGF in vitro

To assess the osteogenic bioactivity of Gel/AM bioprinted scaffold loaded with BMP-2 and VEGF in vitro, BMSCs were mixed with Gel/AM bioinks at a concentration of 1 × 10⁶ cells/mL for Gel/AM bioprinted scaffold. The Gel/AM scaffold were categorized into four groups: BMSC scaffold (Gel/AM bioinks without growth factors). VEGF/BMSC scaffold (Gel/AM bioinks loaded with VEGF at a concentration of 100 μg/mL). BMP-2/BMSC scaffold (Gel/AM bioinks loaded with BMP-2 at a concentration of 40 μg/mL). BMP-2/VEGF/BMSC scaffold (Gel/AM bioinks loaded with BMP-2 and VEGF at a concentration of 40 μg/mL and 100 μg/mL respectively). All scaffold groups were cultured in α-MEM (ThermoFisher Scientific Biochemical Products Co., Ltd., China) medium for three days. After this initial culture period, the medium was replaced with osteogenic medium comprising 50 μM L-Ascorbic Acid Phosphate Magnesium (FUJIFILM Wako Pure Chemical Corporation., Japan), 10 mM β-glycerophosphate (Sigma-Aldrich, USA), and 10 nM Dexamethasone (Sigma-Aldrich, USA). The osteogenic induction assay was conducted for 7, 14, and 21 days, respectively. Following 7, 14, and 21 days of culture, BMSC in the Gel/AM scaffold were fixed, permeabilized, and blocked with 4 % paraformaldehyde, permeabilized, 0.5 % (v/v) Triton X-100 (Sigma-Aldrich, USA) and 1 % bovine serum albumin, respectively. The scaffold were then respectively incubated with antibodies against Runx2 and Osterix at 4 °C overnight. Subsequently, after washing with PBS, the scaffold underwent further incubation with goat anti-rabbit secondary antibodies (Alexa Fluor 488-conjugated, Cat. No. SA00013-2; Proteintech, Chicago, USA) at room temperature for 2 h. And then,the scaffold were stained with 100 nM rhodamine phalloidin solution (Cytoskeleton, USA) and 4′, 6 diamidino-2-phenylindole (DAPI, Beijing Leagene Technology Co., Ltd., Beijing, China) to visualize actin and nuclei, respectively. The BMSC in the scaffold were observed under a confocal microscope (CLSM, Japan). Following osteogenic induction, the cells were washed with PBS and fixed in 4 % paraformaldehyde for 20 min. They were then stained using an ALP staining solution (Solarbio, China) or an ARS staining solution (Cyagen, China) according to the manufacturer's protocols. Total RNA was isolated from BMSC using trizol reagent (TransGen Biotech, China) according to the manufacturer's protocol. First-strand complementary DNA (cDNA) was synthesized from total RNA using the TransScript All-in-One First Strand cDNA Synthesis SuperMix (TransGen Biotech, China). Quantitative real-time polymerase chain reaction (qPCR) was subsequently performed with PerfectStart Green qPCR SuperMix (TransGen Biotech, China) to quantify target mRNA expression.

2.10. Animal model of cranial bone defects and implantation of Gel/AM 3D-bioprinted scaffold

The animal assay received approval from the Laboratory Animal Ethics Committee of Shenzhen Hospital of Southern Medical University (Shenzhen, China; Approval No. 2023-0128). Sixty male Sprague Dawley rats, aged 8 weeks and weighing 200∼250g, were utilized in the study. The rats were anesthetized with 2.0 % wt tribromoethanol (0.3 mL/20g weight, Shanghai Macklin Biochemical Co., Ltd., Shanghai, China). Following skin disinfection, a 1.5 cm sagittal incision was made longitudinally on the rat's scalp to expose the cranium. A circular full-thickness bone defect with a diameter of 6 mm was created in the right parietal region using a trephine drill (Henan Province, China). The animals were then randomized into five groups: Empty group (cranial bone defect only). BMSC scaffold (Gel/AM bioinks without growth factors). VEGF/BMSC scaffold (Gel/AM bioinks loaded with VEGF at a concentration of 100 μg/mL). BMP-2/BMSC scaffold (Gel/AM bioinks loaded with BMP-2 at a concentration of 40 μg/mL). BMP-2/VEGF/BMSC scaffold (Gel/AM bioinks loaded with BMP-2 and VEGF at a concentration of 40 μg/mL and 100 μg/mL respectively). After implantation of Gel/AM scaffold, the fascia and skin were stitched sequentially. At 4 weeks, 8 weeks, and 12 weeks after the operation, the entire cranium was collected and fixed with a 4 % paraformaldehyde solution for further analysis.

2.11. Radiographic evaluation, histological tests and immunofluorescence staining

The collected cranium samples underwent micro-CT scanning using a micro-CT device (SkyScan1176; Bruker Corp., Billerica, MA, USA). The acquired data were reconstructed and analyzed using NRecon, Dataviewer, CTvox, and CTAn software (Ver. 1.1; Bruker Corp).

After decalcification with 10 % ethylenediaminetetraacetic acid (EDTA, Beijing Leagene Technology Co., Ltd., Beijing, China), the cranium samples were dehydrated, embedded in paraffin wax, and sectioned. All sections were then deparaffinized and stained with Hematoxylin and Eosin (H&E), Masson's trichrome stain and immunofluorescence staining of OCN for microscopic observation.

Furthermore, immunofluorescence staining of CD31 and α-SMA was conducted to evaluate the formation of new blood vessels at 4 weeks after the operation. Briefly, following deparaffinization and antigen retrieval, the tissue sections were pre-incubated and blocked with 3 % bovine serum albumin (BSA) at room temperature for 30 min. Subsequently, the sections were respectively incubated with anti-CD31 (host: rabbit; Cat. GB11063-2; dilution, 1:500; Wuhan Servicebio Technology Co., Ltd., Wuhan, China) and anti-α-SMA (host: rabbit; Cat. GB111364; dilution, 1:500; Wuhan Servicebio Technology Co., Ltd., Wuhan, China) antibodies at 4 °C overnight. After washing three times with PBS, the sections were respectively incubated with a goat anti-rabbit secondary antibody (CY3-conjugated, Cat. GB21303; dilution, 1:300,Wuhan Servicebio Technology Co., Ltd.) and goat anti-rabbit secondary antibody (Alexa Fluor 488-conjugated, Cat. GB25301; dilution, 1:400,Wuhan Servicebio Technology Co., Ltd., Wuhan, China) for 50 min. Finally, the sections were stained with DAPI nuclear staining for 10 min. Fluorescence images were captured with a fluorescence microscope and analyzed using ImageJ software.

2.12. Statistical analysis

All data were presented as mean ± standard deviation (SD). Statistical differences were assessed through one-way analysis of variance (ANOVA). The statistical analyses were conducted using GraphPad Prism (v8.0). A significance level of P < 0.05 was considered statistically significant.

3. Results

3.1. Morphology of AM

In this study, we employed a microfluidic chip to design and fabricate AM for the sustained release of BMP-2 (Fig. 2A). Upon optical microscope examination, our prepared AM exhibited a spherical shape with a consistent particle size, approximately 50∼60 μm in diameter (Fig. 2B). The spherical morphology and uniform particle size were advantageous for achieving a more even controlled release of drugs or growth factors. When observed under the electron microscope, the AM displayed a slightly elliptical spherical shape, and the particle size remained homogeneous. However, the diameter of the AM reduced to about 20∼30 μm compared to the optical microscope observation. This size reduction was attributed to the alcohol gradient dehydration of AM before the supercritical CO₂ fluid drying process (Fig. 2C).

Fig. 2.

Optical and SEM images of AM (A–C). Photograph of Gel/AM 3D printed scaffold (D). Mechanical property of Gel/AM bioinks in Gel/AM-0, Gel/AM-5, Gel/AM-10, and Gel/AM-15 groups through compressive test (E–G). SEM images of Gel/AM 3D printed scaffold in Gel/AM-0, Gel/AM-5, Gel/AM-10, and Gel/AM-15 groups; red arrows indicate AM (H–K). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.2. Mechanical property of Gel/AM composite bioinks

To investigate the mechanical properties of Gel/AM composite bioinks, we evaluated the compressive characteristics of all Gel/AM composite bioinks groups, including Gel/AM-0, Gel/AM-5, Gel/AM-10, and Gel/AM-15. The stress-strain curves and quantitative analysis of compressive strength and compressive modulus (Fig. 2E-F) revealed that the maximum compressive strengths for Gel/AM-0, Gel/AM-5, Gel/AM-10, and Gel/AM-15 composite bioinks were 33.29 KPa, 33.45 KPa, 36.08 KPa and 31.87 KPa on average, respectively. These values exhibited minimal differences and were not statistically significant (Fig. 2F). Additionally, the compression modulus of the composite bioinks, Gel/AM-0, Gel/AM-5, Gel/AM-10, and Gel/AM-15, averaged 16.39 KPa, 18.39 KPa, 17.05 KPa and 16.76 KPa, respectively, and these values were not statistically significant (Fig. 2G). These findings suggest that there was no significant alteration in the compressive strength and compressive modulus of the composite bioinks with the increased amounts of AM within a certain range.

3.3. Characterization of 3D printed Gel/AM scaffold

In this study, we utilized a 5 % GelMA hydrogel mixed with AM to form Gel/AM composite bioinks for the creation of 3D printed hydrogel scaffold with favorable porosity. From the images (Fig. 2D), it is evident that Gel/AM bioinks exhibit satisfactory printability, enabling the fabrication of 3D printed scaffold structures with up to 10 layers by optimizing parameter conditions. Scanning electron microscopy of Gel/AM scaffold from Gel/AM-0, Gel/AM-5, Gel/AM-10, and Gel/AM-15 groups revealed good porosity in the internal structure, with pore sizes ranging from about 100 to 300 μm. These pore sizes remained suitable for cell growth, stretching, and proliferation within the hydrogel. Additionally, AM with a size of about 20∼30 μm were observed scattered in the Gel/AM scaffold, appearing mostly as irregular oblate spheres with angular edges due to the extrusion process during 3D printing. In summary, considering the mechanical test results of Gel/AM composite bioinks, it is evident that, within a certain range, the addition of AM does not significantly alter the physicochemical properties of GelMA hydrogels. For the sake of experiment consistency, Gel/AM-10 composite bioinks were chosen for subsequent experiments to validate the biological properties. Gel/AM-10 exhibited slightly higher compression strength compared to other groups, and the distribution density of AM was relatively moderate in the 3D printed Gel/AM composite scaffold.

3.4. The encapsulation of Cy3-BSA and FITC-BSA imitated VEGF and BMP-2 in Gel/AM scaffold

To visually assess the encapsulation and release of growth factors, namely VEGF and BMP-2, in Gel/AM scaffold, fluorescein isothiocyanate-labeled bovine serum albumin (FITC-BSA) and Cy3-fluorescent bovine serum albumin (Cy3-BSA) were employed to mimic BMP-2 and VEGF, respectively, and loaded into the Gel/AM composite scaffold (Fig. 3A.B).

Fig. 3.

Schematic of FITC-BSA Release in Alginate Microsphere (A). Schematic of FITC-BSA and Cy3-BSA Release in the Gel/AM 3D Printed Scaffold (B). Top and 3D Views of Confocal Fluorescence Images of FITC-BSA Release in Alginate Microspheres (C–E). Top and 3D Views of Confocal Fluorescence Images of FITC-BSA and Cy3-BSA Release in the Gel/AM 3D Printed Scaffold (F–H). In vitro Cumulative Release Profiles of BMP-2 from Simple AM without the Gel/AM 3D Printed Scaffold (I). In vitro Cumulative Release Profiles of BMP-2 and VEGF, respectively, from the AM and GelMA Hydrogel of the Gel/AM 3D Printed Scaffold (J).

AM loaded with FITC-BSA, simulating BMP-2, were immersed in ddH₂O for 14 days to monitor the remaining FITC-BSA within the AM. As depicted in Fig. 3, FITC-BSA was evenly distributed in the AM at Day 0; the residual amount of FITC-BSA notably reduced by Day 3, and only a minimal amount remained by Day 7. After 14 days, very little FITC-BSA was observed in the AM compared to Day 7. This indicates a gradual release of FITC-BSA over time, with sustained release lasting approximately 14 days (Fig. 3C-E).

The Gel/AM composite scaffold, loaded with FITC-BSA mimicking BMP-2 and Cy3-BSA representing VEGF, were subjected to a 14-day immersion in ddH₂O to assess the remaining FITC-BSA and Cy3-BSA within the scaffold. As illustrated in Fig. 3, Cy3-BSA was uniformly distributed in the Gel/AM composite scaffold at Day 0, and the FITC-BSA-loaded AM were also relatively evenly dispersed within the Gel/AM composite scaffold. A small amount of FITC-BSA was observed to be released from the AM, dispersing in the Gel/AM composite scaffold or on their surface at Day 0 (Fig. 3F). Over the course of 14 days, the residual amount of Cy3-BSA in the Gel/AM composite scaffold gradually decreased, with the majority of it being released by Day 14(Fig. 3H). However, a larger portion of the remaining FITC-BSA was retained in Gel/AM composite scaffold after 14 days, compared to the Cy3-BSA in Gel/AM composite scaffold or FITC-BSA in simple AM without being embedded into Gel/AM composite scaffold (Fig. 3E.H). These results emphasize that, by assembly Gel/AM scaffold with AM, the gradient encapsulations of growth factors would be achieved as BSA, which may enable Gel/AM scaffold to provide a sequential release of growth factors in a temporally and spatially coordinated manner.

3.5. The release profile of VEGF and BMP-2 in Gel/AM scaffold

In order to quantitatively analyze the in vitro release profile of growth factors in the Gel/AM composite scaffold, BMP-2 and VEGF were loaded into the Gel/AM composite scaffold to detect the in vitro release profile of BMP-2 and VEGF by the ELESA method, and release curves of the growth factors in the Gel/AM composite scaffold were plotted.

As depicted in Fig. 3I-J, in the simple AM group, a burst release of BMP-2 was observed on day 3, with cumulative release reaching 89.80 % within 7 days and 96.77 % by day 35. In contrast, although a similar initial burst of BMP-2 occurred on day 3 in the Gel/AM composite scaffold, its release was significantly attenuated, reaching only 54.28 % at 7 days and 62.50 % at 35 days. Meanwhile, VEGF in the Gel/AM composite scaffold was rapidly released, with 76.18 % released within the first 7 days and a cumulative release of 86.68 % by day 35. These results indicate that the Gel/AM composite scaffold facilitates early rapid release of VEGF while markedly prolonging the release profile of BMP-2, thereby achieving a more sustained release behavior.

3.6. Effects of Gel/AM bioprinted scaffold loaded with VEGF on HUVECs migration

To assess the impact of VEGF-loaded Gel/AM bioprinted scaffold on HUVEC cell migration, we conducted a scratch test and a transwell test. In Fig. 4(E–H), after HUVECs were co-cultured with Gel/AM scaffold from the VEGF + group and the VEGF- group for 24 h, the migration of HUVEC cells toward the scratch was more pronounced in the VEGF + group compared to the VEGF- group. The statistical analysis of the scratch test results in Fig. 4K further confirmed a significantly higher number of cells migrating toward the scratches in the experimental group compared to the control group, with the data being statistically significant (*p < 0.05). Additionally, as depicted in Fig. 4I-J, HUVEC cells in Gel/AM scaffold of the VEGF + group exhibited a significantly greater migration from the upper to the lower part of the chambers compared to the control group. Furthermore, quantitative analysis of the statistical data (Fig. 4L, ****p < 0.0001) revealed a significantly higher number of cells migrating in the experimental group compared to the control group. These findings from the scratch assay and transwell assay strongly suggested that VEGF-loaded Gel/AM bioprinted scaffold enhance the migration capability of HUVEC cells by facilitating the controlled release of VEGF.

Fig. 4.

Schematic of HUVEC migration in the Transwell assay (A–B). Schematic of HUVEC tube formation on the Gel/AM 3D bioprinted scaffold (C–D). Influence of VEGF released from the Gel/AM 3D printed scaffold loaded VEGF on HUVEC migration determined by scratch assay and Transwell assay (E–J). Quantification of HUVEC migration by scratch assay (*p < 0.05) (K). Quantification of HUVEC migration by Transwell assay (****p < 0.0001) (L). Images of HUVEC tube formation on the Gel/AM 3D bioprinted scaffold (M–N). Immunofluorescence staining results of CD31(O).

3.7. Effects of Gel/AM scaffold loaded with VEGF on HUVECs tube formation assay

To investigate the potential of Gel/AM bioprinted scaffold loaded with VEGF in promoting in vitro tube formation of HUVECs, HUVEC cells were incorporated into the Gel/AM bioprinted scaffold and cultured in ECM medium for 14 and 21 days. After 14 days of culture, In the VEGF- group, Few HUVECs aggregated around the pores of the Gel/AM bioprinted scaffold to form tubes. In contrast, the VEGF + group displayed more pronounced proliferation of HUVECs, with cells concentrated in the hydrogel scaffold. Additionally, more HUVECs aggregated into tubular structures around the pores of the Gel/AM bioprinted scaffold (Fig. 4M). After 21 days of culture, the Gel/AM bioprinted scaffold of the VEGF + group demonstrated enhanced proliferation of HUVECs, along with more conspicuous aggregation into tubular structures at the edges of the pores, compared to that of the VEGF- group (Fig. 4N). To further evaluate the angiogenic capacity of the scaffold material, we assessed the expression level of CD31, The experimental results demonstrated that the Gel/AM bioprinted scaffold of the VEGF + group exhibited significantly enhanced CD31 expression, indicating superior vascularization performance(Fig. 4O).These findings indicated that Gel/AM bioprinted scaffold loaded with VEGF significantly promote the tube-forming properties of HUVECs by releasing VEGF.

3.8. Osteogenic bioactivity of Gel/AM bioprinted scaffold loaded with BMP-2 and VEGF in vitro

To evaluate the osteogenic differentiation potential of a Gel/AM scaffold loaded with BMP-2 and VEGF, an in vitro assay was performed to induce osteogenic differentiation of BMSC. As illustrated in Fig. 5A, ARS staining demonstrated that both the BMP-2/VEGF/BMSC and BMP-2/BMSC groups promoted mineralization, with a more pronounced effect observed in the BMP-2/VEGF/BMSC group, Concurrently, results from alkaline phosphatase (ALP) staining showed analogous outcomes. Quantitative real-time polymerase chain reaction (qRT-PCR) analyses revealed that the BMP-2/VEGF/BMSC group significantly upregulated the expression levels of osteogenic-related mRNA markers compared to the other groups (Fig. 5B). Fig. 5C presented the results after 21 days of induced osteogenic differentiation culture. In all groups, BMSC within Gel/AM bioprinted scaffold exhibited robust growth, stretching, and proliferation. The highest fluorescent intensity of Runx2 was observed in the BMP-2/VEGF/BMSC group and the BMP-2/BMSC group, while the VEGF/BMSC and BMSC groups exhibited the least intensity. Similar trends were noted at Day 7 and Day 14 (Fig. S1A-D). Quantitative analysis (Fig. 5E) revealed that the expression of Runx2 in the BMP-2/VEGF/BMSC and BMP-2/BMSC groups consistently surpassed that in the VEGF/BMSC and BMSC groups. These findings suggested that exogenous BMP-2 promotes osteogenic differentiation by activating the Runx2 signaling pathway. Similarly, after 21 days of induced osteogenic differentiation culture, Fig. 5D displays Osterix fluorescent intensity, mirroring the patterns observed for Runx2. The highest Osterix intensity was seen in the BMP-2/VEGF/BMSC group and the BMP-2/BMSC group, while the VEGF/BMSC and BMSC groups showed the least intensity. Quantitative analysis (Fig. 5F) consistently indicated significantly higher expression of Osterix in the BMP-2/VEGF/BMSC and BMP-2/BMSC groups compared to the VEGF/BMSC and BMSC groups, with statistically significant differences. These results indicated that exogenous BMP-2 can simultaneously enhance osteogenic differentiation by activating the Osterix signaling pathway. These results demonstrated that the Gel/AM composite hydrogel 3D-printed scaffold system, which can release BMP-2 slowly and continuously, has a better ability to contribute to bone differentiation in vitro.

Fig. 5.

(A) Images of alkaline phosphatase (ALP) and Alizarin Red S(ARS) staining. (B) Osteogenesis-related mRNA indicators (Runx2, OSX, and OCN) were evaluated by qRT-PCR. (C, D) Images of fluorescent staining of DAPI (blue), Runx2 (green), Osterix (green) and Actin (red) of BMSC in the Gel/AM 3D bioprinted scaffold of each group at Day 21. (E, F) Quantitative analysis of the expression levels of Runx2 and Osterix in each group. (*P < 0.05, **P < 0.01). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.9. Osteogenic bioactivity of Gel/AM bioprinted scaffold loaded with BMP-2 and VEGF in vivo

To evaluate the in vivo degradation behavior of the material, the Gel/AM bioprinted scaffolds were labeled with DIR fluorescent dye and implanted subcutaneously. Retention was monitored using an IVIS imaging system. As shown in Fig. S2, the fluorescence intensity gradually decreased over the three-week period, a trend consistent with that observed in other comparable materials, indicating favorable in vivo degradation performance.

To assess the bioactivity of Gel/AM bioprinted scaffold loaded with BMP-2 and VEGF for repairing large segmental bone defects, we established a rat critical cranial defect model with a 6 mm diameter. Gel/AM bioprinted composite scaffold loaded with BMP-2 and VEGF were implanted, and bone regeneration was evaluated using micro-CT at 4 weeks, 8 weeks, and 12 weeks. As depicted in Fig. 6A, significantly increased new bone formation was observed in the bone defects of both the BMP-2/VEGF/BMSC group and the BMP-2/BMSC group. In the BMP-2/VEGF/BMSC group, the cranial defect area was largely filled with new bone at 4 weeks, nearly replaced by new bone at 8 weeks, and completely filled with new bone at 12 weeks. The BMP-2/BMSC group exhibited mostly new bone filling in the cranial defect area at 12 weeks. And the VEGF/BMSC group also displayed some new bone filling in the defect area compared to the BMSC and Empty groups, where only minimal new bone was generated after 12 weeks. Quantitative analysis of BV、BV/TV、Tb.N and Tb.Sp in the bone defect area (Fig. 6B-E) revealed significantly higher bone formation in both the BMP-2/VEGF/BMSC group and the BMP-2/BMSC group compared to the other groups, indicating that exogenous BMP-2 significantly promoted bone regeneration in the defect area. Additionally, the new bone in the BMP-2/VEGF/BMSC group was significantly higher than the BMP-2/BMSC group, especially at 8 weeks and 12 weeks, suggesting the synergistic and significant effect of sequential delivery of VEGF and BMP-2 in promoting bone regeneration.

Fig. 6.

Representative images of new bone formation at 4 weeks, 8 weeks, 12 weeks post-surgery by micro-CT (A). Quantitative analysis of the new bone volume in each group (B–E). (*P < 0.05, **P < 0.01).

Simultaneously, we conducted an analysis of new bone regeneration through HE、Masson and immunofluorescence staining of cranial samples from each group. HE staining revealed the presence of fibrous connective tissue and new bone within the defect area in each group. The BMSC group and Empty group showed a predominance of fibrous connective tissue with minimal new bone. The VEGF/BMSC group exhibited abundant fibrous connective tissue and limited new bone. In contrast, both the BMP-2/VEGF/BMSC group and the BMP-2/BMSC group displayed substantial new bone with minimal fibrous connective tissue (Fig. 7A, Fig. S3.A). The observed amount of new bone regeneration within the defect area correlated with the reconstructed micro-CT images in each group. Maturity of the tissue in the cranial defect area was assessed through Masson trichrome staining, with mature bone marked by red staining and immature bone marked by blue staining. More mature or predominantly mature bone was evident in both the BMP-2/VEGF/BMSC group and the BMP-2/BMSC group, particularly at 8 weeks and 12 weeks. The VEGF/BMSC group at week 12 exhibited a limited amount of mature bone. In contrast, only a substantial amount of fibrous connective tissue was observed at any time in the BMSC and Empty groups (Fig. 7B, Fig. S3.B). Immunofluorescence staining for OCN revealed that both the BMP-2/VEGF/BMSC group and the BMP-2/BMSC group exhibited stronger OCN signals compared to the other groups, with the BMP-2/VEGF/BMSC group demonstrating the most potent osteogenic capacity(Fig. 7C,Fig. S4).

Fig. 7.

Histological analysis of bone regeneration with H&E and Masson's trichrome staining at 4 weeks and 12 weeks after surgery (A, B) (N:new bone tissue. F:fibrous tissue. B:old bone); OCN(green) immunofluorescence staining of bone regeneration at 4 weeks and 12 weeks after surgery (C). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

These results demonstrate that Gel/AM bioprinted scaffold loaded with BMP-2 and VEGF not only significantly enhance bone regeneration in the defect area but also facilitate the gradual transformation of new bone into mature bone by sequential delivery of VEGF and BMP-2. Notably, in the HE staining results, the GelMA hydrogel of Gel/AM bioprinted scaffold was largely degraded in all implantation groups at 4 weeks. However, remnants of AM of Gel/AM bioprinted scaffold were observed in some groups at 12 weeks. This suggested a slower degradation rate of AM compared to GelMA hydrogel in vivo, which also favored sequential release of VEGF and BMP-2 from Gel/AM bioprinted scaffol

3.10. Angiogenic bioactivity of Gel/AM bioprinted scaffold loaded BMP-2 and VEGF in vivo

Angiogenesis in the bone defect area was assessed by labeling vascular endothelial cell and vascular smooth muscle using CD31 and α-SMA immunofluorescence staining at 4 weeks, respectively. As evident in Fig. 8A, B, increased angiogenesis was observed in both the BMP-2/VEGF/BMSC group and the VEGF/BMSC group, while only a few blood vessels were observed in the BMP-2/BMSC group, the BMSC group, and the Empty group. Quantitative analysis of CD31 and α-SMA expression (Fig. 8C, D) revealed that the expression of α-SMA was significantly higher in the BMP-2/VEGF/BMSC group and the VEGF/BMSC group compared to the other groups, demonstrating a statistically significant difference.To better visualize the vascularization within the defect area, lower-magnification images have been included in the Supplementary Materials (Fig. S5) to facilitate a comprehensive assessment of angiogenic outcomes. This indicated that Gel/AM bioprinted composite scaffold loaded with VEGF significantly promote angiogenesis in the bone defect area in vivo by releasing VEGF.

Fig. 8.

Fluorescence images of positive staining with CD31(red), α-SMA (green) and DAPI (blue) of the cranium at 4 weeks post-surgery (A, B). Quantitative analysis of the percentage of CD31 and α-SMA positive area in each group at 4 weeks post-surgery (C, D,*P < 0.05, **P < 0.01). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

4. Disscusion

The key of 3D bioprinting technology for bone regeneration lies in the construction of bionic composite scaffold that should mimic the orchestrated regulations about vasculogenic and osteogenic growth factors in the early and late stage of bone repair [13,21,23]. In this study, we developed Gel/AM bioink and bionic composite scaffold using GelMA hydrogel combined with embedded AM loaded with BMP-2 and VEGF, respectively. This design is intended to enable Gel/AM scaffold to provide a sequential release of growth factors in a temporally and spatially coordinated manner. On the one hand, by inserting AM microspheres embedded in GelMA hydrogel, the structure of gradient encapsulation of the growth factors was achieved to provide sequential and controllable release of VEGF and BMP-2. On the other hand, the difference in degradation rates of GelMA hydrogels and AM, as in HE staining results indicated, simultaneously resulted in sequential and controllable release of growth factors. Because GelMA hydrogels would be able to be rapidly degraded and release VEGF by collagenase in vivo, whereas AM, due to the lack of corresponding degradation enzymes in vivo, could only slowly hydrolyze and release BMP-2. And then, to verify the activity of the composite bionic scaffold in contributing to bone differentiation, the bionic composite scaffold of the BMSC group, the VEGF/BMSC group, the BMP-2/BMSC group, and the BMP-2/VEGF/BMSC group were subjected to in vitro osteogenic induction for 21 days and immunofluorescent staining for Runx2 and Osterix. Relative to the other groups, the BMP-2/VEGF/BMSC group could more effectively promote the expression of osteogenesis-related proteins. Moreover, to verify the ability of the bionic composite scaffold to promote angiogenesis and repair bone defects, the bionic composite scaffolds of the above groups were implanted into the cranial defect model for 4, 8, and 12 weeks. And it could be observed that, relative to other groups, the BMP-2/VEGF/BMSC composite bionic scaffold had more new bone formation. Meanwhile, fluorescence immunostaining for CD31 and α-SMA was performed on the cranial samples at 4 weeks postoperatively, the BMP-2/VEGF/BMSC group also expressed angiogenesis-related proteins more efficiently relative to the other groups. These results indicate that the Gel/AM bionic composite scaffold, which designed based on the sequential regulation of VEGF and BMP-2 in the early and late stage of bone repair, are capable of significantly contributing to bone differentiation, promoting angiogenesis, and repairing large bone defects.

In the result of the in vitro growth factor release kinetics, the sudden release of BMP-2 in the Gel/AM composite scaffold amounted to about 50.45 % on day 3, resulting in a certain loss of BMP-2. This is due to the presence of extrusion of the nozzle on the AM during the 3D printing of the scaffold, which caused a portion of the AM to be damaged or deformed. Therefore, in order to reduce the massive burst release of BMP-2, subsequent studies should optimize the printing conditions to reduce the extrusion or damage caused to AM. Moreover, the Gel/AM composite scaffold loaded with BMP-2 and VEGF demonstrated an early and rapid release of VEGF in the initial stage, followed by a relatively slow and sustained release of BMP-2 in the later stage to 35 days as in vitro study indicated. Notably, due to the presence of GelMA degradation and AM hydrolysis in vivo, the VEGF and BMP-2 release kinetics of Gel/AM composite scaffold in vivo would actually be somewhat different from the release kinetics detected in vitro. Following studies need to explore the in vivo VEGF and BMP-2 release kinetics of Gel/AM composite scaffold in depth.

In the animal experiments to validate the composite scaffold for repairing cranial defects, the new bone volume of the BMP-2/VEGF/BMSC group was significantly higher than that of the BMP-2/BMSC group, especially at weeks 8 and 12. Combined with the results of fluorescence immunization of pathological sections, angiogenesis-related proteins such as CD31 and α-SMA were significantly expressed in the BMP-2/VEGF/BMSC group relative to the other groups. These results revealed that exogenous VEGF could synergize with BMP-2 to promote bone regeneration in the defect area by promoting angiogenesis in the defect area [24,25]. However, for the skull samples at postoperative week 4, there was no statistically significant difference between the BMP-2/VEGF/BMSC group and the BMP-2/BMSC group, although the volume of new bone in the BMP-2/VEGF/BMSC group was higher than that in the BMP-2/BMSC group. The reason for this is speculated to be that the function of new blood vessels in the first 4 weeks may not be fully functional to some extent, and the availability of nutrients or other factors needed for bone regeneration is limited.

In addition, only a small amount of new bone generation was observed in the defect area in both the composite scaffold of the BMSC group and the Empty group, indicating that the Gel/AM composite scaffold loaded with BMSC did not have the effect of further promoting bone regeneration in the bone defect area. This result is considered to have two reasons. On the one hand, due to the destruction of blood vessels in the defect area, which makes the hypoxia and lack of nutrients, and ultimately leads to the possible death of a large number of implanted BMSC cells in a short period of time. On the other hand, since the density of our implanted BMSC cells was 1 × 10⁶ cells/mL, which is significantly less compared to the implanted cell density of 1 × 10⁷ cells/mL or 2 × 10⁷ cells/mL suggested by some studies [26,27]. These reasons ultimately lead to the possibility that very few BMSC could survive, grow, proliferate, and be induced to differentiate into osteoblasts to participate in bone regeneration. Therefore, subsequent studies are proposed to further address the hypoxic conditions in the bone defect region and to increase the number of cells implanted to improve cells survival in the implanted Gel/AM composite scaffold and to further enhance the bone defect repair activity of the composite scaffold.

Therefore, future studies should firstly further reveal the in vivo growth factor release kinetics of composite Gel/AM bioprinted scaffold. Secondly, optimize the printing conditions to improve the phenomenon of initial burst release of BMP-2 from AM in composite Gel/AM scaffold, and explore the in vivo growth factors release kinetics of Gel/AM composite scaffold in depth. In addition, the hypoxic conditions in the bone defect region need to be further improved and the number of cell implantation needs to be increased to improve the BMSC survival in implanted Gel/AM composite bioprinted scaffold in the bone defect region.

5. Conclusion

In our study, we have developed a bioink with staged growth factors releasing using GelMA combined with embedded AMs. The incorporation of the AMs within the GelMA maintains structural stability while enabling a controlled, gradient release of VEGF and BMP-2. Our in vitro and in vivo results highlight the efficacy of our bioink printed construct, which delivers VEGF and BMP-2 in a temporally and spatially coordinated manner, to stimulate bone regeneration. This approach significantly enhances the efficiency of treating bone defects, particularly those of a larger scale, by offering a promising avenue for therapeutic intervention.

CRediT authorship contribution statement

Enhui Zhou: Writing – original draft, Methodology, Investigation. Peipei He: Methodology, Investigation, Data curation. Zefeng Yang: Software, Data curation. Chunran Li: Visualization, Validation. Guofang Fang: Supervision. Jiachang Wu: Formal analysis. Weida Zhuang: Writing – review & editing, Supervision, Conceptualization. Hongxun Sang: Supervision, Resources, Project administration, Funding acquisition.

Declaration of competing interest

Each author certifies that there are no funding or commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest in connection with the submitted article related to the author or any immediate family members.

Appendix A. Supplementary data

The following are the Supplementary data to this article.

figs1.

figs2.

figs3.

figs4.

figs5.

Data availability

No data was used for the research described in the article.