Rabbit bone marrow mesenchymal sem cells were purchased from Cyagen Co. Ltd. (GuangZhou, China). Sodium alginate, methacrylic anhydride, 2-hydroxy4′-(2-hydroxyethoxy)-2-methylpropiophenone (Irgacure 2959, 98%) and sodium hydroxide (NaOH) were purchased from Sigma-Aldrich Co. Ltd. (MO, United States). 4-arm poly (ethylene glycol) acrylate was provided by Sinopeg Biotech Co. Ltd. (China). Ultrapure water was obtained from a Millipore Auto-pure system. All regents were used without further purification.

The methacrylated alginate was prepared according to a reported study (Choi et al., 2021). Briefly, ALG was dissolved with 20 ml of deionized water to produce a 2% (w/v) solution at 37°C overnight. Then, 20 ml of methacrylic anhydride was added dropwise to the system, and the solution was continuously stirred magnetically at room temperature for 3 days, with the pH periodically adjusted to 7 with aqueous NaOH (5 M). After the 3-days reaction period, the resulting solution was poured into 100 ml of ethanol (pre-chilled in advance at -20°C) to precipitate the ALG-MA product for about 8 h at RT. Finally, the precipitate was vacuum filtered, washed three times with ethanol, oven dried at 37°C, and stored at −20°C until use. From the addition of methacrylic anhydride, all the synthesis steps were carried out under dark conditions.

The PRP was prepared according to a reported study (Faramarzi et al., 2018). Briefly, rabbit ear margin vein blood was centrifuged at 250 g for 15 min to separate red blood cells from plasma. The upper plasma phase, including the interface, was centrifuged at 1600 g for 60 min to pellet the platelets. The upper three-quarters of the plasma phase were discarded to retain the remaining one-quarter. The obtained PRP was then stored at -80°C until further use. Further platelet activation was performed by repeated freeze-thaw cycles in subsequent experiments.

To generate hydrogels, alginate-methacrylic anhydride was dissolved separately in PBS (PRP group: 10% PRP [v/v]-ALGMA) to achieve a concentration of 1.67% (w/v) in borosilicate vials, combined with four arm-PEG at twice the solution’s mass, and incubated in a 37°C water bath until complete dissolution. The resulting solutions were added with I2959 (1% (w/v)). Then, 180 µL of the resulting solution was solidified in PDMS molds (diameter, 1.3 mm; height, 1 mm) to form sheets and subjected to UV-light irradiation for 90 s at 365 nm. Finally, these hydrogel disks were soaked in dopamine (2 mg/ml) and collagen for 60 s after being submerged in PBS or SBF for 24 h to remove the excess unreacted monomer. For in vitro experiments, all hydrogel disks were sterilized by ethylene oxide.

Twelve Male Newland rabbits with an average weight of 4 kg were obtained from the Laboratory Animal Center of Drum Tower Hospital affiliated to the Medical School of Nanjing University (China). All experimental protocols were approved by the ethics committee of Drum Tower Hospital affiliated to the Medical School of Nanjing University, and performed according to the Institutional Animal Care and Use Committee (IACUC) guidelines.

All experiments were performed with three replicates unless otherwise stated. Data are mean ± SD. Statistical analysis was performed with the Origin software (8.5 version). Asterisks in statistical analysis indicate statistically significant differences between the control and experimental groups (∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.005; ∗∗∗∗p < 0.001).

Methacrylated alginate (M-ALG) was synthesized from the reaction between the hydroxyl group of alginate and the epoxy group of methacrylic anhydride under alkaline conditions, the mechanisms of which include epoxide ring-opening, carbodiimide chemistry and transesterification (Araiza-Verduzco et al., 2020). The reaction is schematically shown in Figure 1A, and the structures of neat ALG and M-ALG were confirmed by 1 HNMR spectroscopy in D₂O (Figure 1B), depicting characteristic peaks between 3.50 and 5.20 ppm from both neat ALG and M-ALG due to their saccharide units. As expected, the spectrum of M-ALG displayed distinctive peaks corresponding to the vinyl (5.35 and 5.66 ppm) and methyl (1.80 ppm) hydrogens of methacrylate grafted groups, consistent with previous reports (Wang et al., 2015). Of note, these peaks could slightly shift due to different chemical environments (Burdick et al., 2005). Furthermore, the chemical structures of neat ALG and M-ALG were characterized by FTIR. As shown in Figure 1C, M-ALG′ spectrum exhibited two additional bands compared with that of ALG (arrows): occurrence of–CH stretching bands (2980–2850 cm⁻¹) and appearance of a shoulder (1721 cm⁻¹), which were attributed to stretching vibrations of the aliphatic chains’ –CH groups and the esters’ C=O group, respectively. Both groups were due to the grafting of methacrylate units (Mugnaini et al., 2021) (chemical structures in Figure 1A). These results confirmed that methacrylate was successfully grafted to neat ALG.

PRP-incorporated interpenetrating polymeric network (IPN) hydrogels were synthesized through free radical polymerization reaction under UV-light irradiation in the presence of PRP (Figure 1D). Covalent crosslinks between M-ALG (methacrylate groups) and 4-Armed PEG-ACLT (acrylate groups) were formed after this reaction, where a biocompatible photoinitiator (I2959) was used to induce fluid-solid phase transformations. Tetra-functionalized PEG and M-ALG crosslinkers were used due to their high crosslinking efficiency, and the addition of PRP improved biological activities (Andersen et al., 2021; Xu et al., 2021). The as-prepared hydrogels were further crosslinked by exposure to SBF that contained Ca²⁺ to interact with ALG, forming mineralized PRP-incorporated IPN hydrogels. Photographs of as-prepared hydrogels, including PBS-PRP (-), PBS-PRP (+), SBF-PRP (-) and SBF-PRP (+) were shown in Supplementary Figure S1, depicting the hydrogel turning pale yellow after the incorporation of PRP. As expected, the volume of the hydrogel alone or combined with PRP biomineralized by SBF was smaller than that of the corresponding hydrogel treated with PBS, indicating that biological mineralization hindered the swelling of hydrogels. In addition, lyophilized PRP-incorporated IPN hydrogels were morphologically characterized by SEM. As illustrated in Figure 2, all IPN hydrogels had a porous structure with a pore size of about 20–50 μm. Different from the uniform pore size in the PBS-PRP (-) group, pore size in the PBS-PRP (+) group was slightly discrepant, and its surface appears clearly cross-linked fibers, indicating the formation of a platelet activation-dependent fibrin network which is consistent with the results of previous studies (Qian et al., 2022). In contrast to the PBS-PRP (-) group, hydrogel pores in the mineralized group are smaller while the pore wall thickness is significantly increased, which is related to the presence of more crosslinks within the scaffold of the mineralized group. Furthermore, EDS mapping based on SEM images was also present to verify the biological mineralization between Ca²⁺ in SBF and ALG. As shown in Supplementary Figure S2 and Supplementary Table S1, all IPN hydrogels contained O, Na, Cl and K. Meanwhile, Ca was only presence in SBF-PRP (-) and SBF-PRP (+), which was attributed to successful biological mineralization. In particular, we expected the porous microstructures are suitable for containing of cells, thereby accommodating a large number of cells and facilitating multiple cellular functions, including cell attachment, proliferation and differentiation (Li L. et al., 2020; Huang et al., 2021).

It is known that bioactive materials used as bone repair scaffolds should possess several features, including high surface area/volume ratio, proper mechanical properties, excellent biocompatibility and enhanced osteoconductivity (Dong et al., 2019; Wang et al., 2020). By achieving these, scaffolds can provide cells with an appropriate three-dimensional (3D) environment, enabling cells to exert normal functions such as proliferation and migration, and subsequently support the newly formed bone (Sun et al., 2016; Zhu et al., 2021). Hence, the physicochemical properties of the as-prepared IPN hydrogels were assessed. Swelling rate, one of the most important properties that can reflect the crosslinking degree, affinity toward H₂O molecules, ionization degree of functional groups and properties of swelling medium such as ionic strength and temperature (Kim et al., 2013) were firstly measured. As shown in Figure 3A, swelling rate decreased after the biological mineralization process, suggesting ionic crosslinking between Ca²⁺ and ALG occurred. Furthermore, we also studied the compressive mechanical properties of as-prepared IPN M-ALG/PEG hydrogels by standard mechanical tests. The obtained stress-strain curves based on the compression-crack test were shown in Supplementary Figure S3A, which indicated that SBF-PRP (+) demonstrated the enhanced Young’ modulus (11.36 ± 1.02 Kpa) compared with the other hydrogels (Figure 3B). Strikingly, the elongations at break (Figure 3C) and tensile strength (Figure 3D) of hydrogels increased after the PRP incorporation and the biological mineralization, which substantially improved the load bearing ability of the hydrogel. Furthermore, the energy dissipation levels of hydrogels were investigated through compression-relaxation cycles under 50% strain. As shown in Supplementary Figure S3B, all hydrogels demonstrated slight energy dissipation, with no significant difference among them, which could be attributed to the rupture of physical cross-linkers. As expect, the calculated dissipation of hydrogels (Figrue 3E) was significantly decreased after the incorporation of PRP and the biological mineralization process. The fast recovery ability of hydrogels was then tested by applying continuous compression-relaxation cycles for 10 cycles (Supplementary Figure S3C). The results revealed that the tensile strength of each cycle was slightly decreased but remained at more than 85% after 10 continuous cycles, suggesting the remarkably fast recovery ability of IPN hydrogels. In next, the rheological properties of the as-prepared hydrogels were investigated. Strain-dependent oscillatory shear rheology (Figure 3F) exhibited a strain at yield strength of 10%, which was the cross point of G′ and G″, representing the transition of the hydrogel from solid to liquid state. Besides, the frequency-sweep test (Figure 3G) demonstrated that the storage modulus (G′) was larger than the loss modulus (G″), suggesting that the fabricated hydrogels exhibited a solid-like behavior. Furthermore, step-strain measurements (Supplementary Figure S4) demonstrated that all hydrogels were destroyed partly and subsequently recovered completely and rapidly from a strain of 300%–0.1% for four cycles of breaking and reforming, revealing the dominant elastic nature of these hydrogels. Furthermore, the accumulative release of TGF-β from PRP-incorporated IPN hydrogels was investigated in phosphate buffer saline (PBS) over an incubation period of 14 days, which was quantified by enzyme-linked immuno sorbent assay (ELSA). The obtained release curves are shown in Figure 3H, sustained release of TGF-β over the investigated period could be observed for both PBS-PRP (+) and SBF-PRP (+). As expect, SBF-PRP (+) demonstrated relatively slower release due to the biological mineralization layer partly prevented the release of TGF-β. These results indicated that the as-prepared PRP-incorporated IPN hydrogels possess substantially improved mechanical strength and sustained release of growth factor, constituting potential repair scaffolds for bone tissue regeneration.

The physical characterization of the scaffold was followed by cytocompatibility assessment in vitro. It is known that the integration of scaffolds with the natural bone by forming the apatite layer on its surface in vivo can be simulated in vitro by scaffold incubation in SBF (Lewandowska-Łańcucka et al., 2019). Considering the possible in vivo application of this scaffold in BTE, in vitro biomineralization of the PRP-hydrogel scaffold was performed using SBF. Subsequently, the cytocompatibility of scaffolds was assessed by co-incubating with cells and measuring cell viability, cell proliferation and cell adhesion assays. Cell morphology on different groups was depicted in Figure 4A. As demonstrated, cells can well spread, survive, and maintain their spindle-shaped morphology on the hydrogels, suggesting the excellent cytocompatibility of both non-mineralized and mineralized scaffolds (Figure 4B). It is worth to mention that higher cell attachment was found after 24 h of cell incubation on both non-mineralized and mineralized scaffolds in the presence of PRP compared with the control group. Also, cell morphology observation on all four groups after 48 h incubation was obtained by optical microscopy and confocal microscopy (Supplementary Figure S6). These results demonstrated that the incorporation of PRP facilities cell adhesion. In order to further evaluate the effect of PRP on cell adhesion, RT-PCR was performed to detect the expression of integrin β1, which is a classic transmembrane receptor that mediate the attachment between a cell and its surroundings. RT-PCR primer of integrin β1 was shown in revised Supplementary Table S2. BMSCs were cultivated on PBS-PRP (-), PBS-PRP (+), SBF-PRP (-) and SBF-PRP (+) hydrogels for 24 h and then collected respectively. Then the relative expression of integrin β1 was detected by RT-PCR. As shown in the revised Supplementary Figure S5, cells seeded on the PRP-containing hydrogels exhibit higher expression of integrin β1, with or without the presence of SBF. This is consistent with our previous results obtained by cytoskeleton staining in Figure 4A. Cell proliferation on the scaffolds was further quantified by the CCK-8 assay after 1, 3, 5, and 7 days of cell culture, respectively. As shown in Figure 4C, nearly no significant difference was observed at 1 and 3 days and the numerical results were shown in Supplementary Table S3 . Surprisingly, PRP-incorporated IPN hydrogel scaffolds had a higher cell proliferation rate than the neat hydrogel scaffolds after 5 and 7 days of culture (p < 0.05), which indicated that the incorporation of PRP into the alginate/ethylene-glycol hydrogel greatly promotes BMSC proliferation. In general, multiple studies have consistently confirmed that PRP increase cell proliferation in vitro (Gentile and Garcovich 2020), and this effect is mainly attributed to the release of growth factors from platelets (Martino et al., 2011). In this study, benefiting from the novel double-network structure of the hydrogel scaffold, biomolecules from PRP could more readily penetrate the scaffold and be continuously released without being affected by the mineralization of scaffolds.

ALP activity and ARS staining intensity were detected after seeding of BMSCs. Higher ALP activity (Figures 5A,C) and ARS intensity (Figures 5B,D) were quantitatively and statistically found on the BMSCs cultured on mineralized PRP-containing substrate compared with others, indicating an increased osteogenic differentiation capacity. We then cultured BMSCs on scaffolds and analyzed RUNX2 expression by immunofluorescence (IF) (Figure 6). Increased amount of RUNX2-positive cells were observed in the SBF-PRP (+) group compared with the other groups. Furthermore, we analyzed the expression levels of common osteogenesis markers (Tang et al., 2017), including BMP2, OPN, ALP and COL-I, and found that osteogenesis-related genes were significantly upregulated at both mRNA and protein levels in BMSCs cultured on the mineralized PRP-hydrogel scaffold (Figure 7). The raw data of western blot analysis was shown in Supplementary Figure S7 and semi quantitative analysis of western blot was shown in Supplementary Figure S8. These data demonstrated that PRP and mineralization of the scaffold synergically promote osteogenic differentiation in BMSCs. However, the osteogenic effects of PRP are complex. Some studies have suggested that PRP may facilitate bone formation in combination with MSCs from different species; others have shown that PRP decreases the osteogenic differentiation of MSCs (Gruber et al., 2004; Felka et al., 2010). This is probably because PRP highly varies from donor to donor, and each PRP preparation may differ in the concentrations of proteins and growth factors (Lim et al., 2013). Recent research reported that PRP, in a certain concentration range, promotes cell proliferation at the early stage of differentiation, but causes no impairment of osteogenic differentiation in BMSCs (Xu et al., 2015; Xie et al., 2020). This is partly consistent with our findings that PRP had a positive effect on the proliferation of BMSCs on PRP-hydrogel scaffolds whether mineralized or not, and stimulated the osteogenic differentiation of BMSCs, particularly those cultured on the mineralized PRP-hydrogel. It is known that differentiation and proliferation represent dichotomous aspects of cellular function, and proliferation is frequently associated with differentiation (Schröder et al., 2009; Zheng et al., 2021a; Zheng et al., 2021b). A possible explanation for our findings is the sustained release of PRP from the hydrogel, which has a beneficial effect on cell proliferation and the early stage of osteogenesis. Furthermore, the biomineralization of scaffolds has been shown to considerably promote osteogenic differentiation of stem cells (Kim et al., 2021). Because the PRP-hydrogel mineralizes in a manner similar to biological mineralization, this scaffold modification significantly enhanced osteogenic differentiation of BMSCs. Summary results of the abovementioned studies lead to the conclusion that such synergistic effects of physical structures, the double-network structure/biomineralization of scaffolds, and molecular supplements (PRP) could significantly promote osteogenesis and subsequent bone regeneration.

In vitro investigations have shown that SBF-PRP (+) hydrogel exhibited excellent biocompatibility and mechanical strength, and enhanced osteogenic activities. This prompted us to assess the in vivo bone regeneration performance of the fabricated hydrogels. Hence, we further evaluated the bone repair ability hydrogel with or without SBF-PRP in a rabbit model of femoral 5-mm defect (Zhang et al., 2021). SBF-PRP (-) and SBF-PRP (+) hydrogels were directly implanted into the femur defects of rabbits. For comparison, no operation was performed after creating the bone defects. At 2 months after the implantation of hydrogels, femurs were harvested and firstly observed by micro-CT. The reconstructed three-dimensional (3D) and sectional images of femoral condyles were shown in Figure 8A, indicating that there was still a large defect in the control group, while treatment with the SBF-PRP (+) hydrogel exhibited more newly formed cancellous bone compared with the SBF-PRP (-) group. For quantitative analysis, BMD, bone volume per total volume (BV/TV) and the trabecular parameters of the cancellous bone, including trabecular number (Tb.N), trabecular thickness (Tb.Th) and trabecular Spacing (TB.Sp) were assessed and the results were shown in Figure 8B. BMD in the control group was 0.77 ± 0.05 cm⁻¹, while the SBF-PRP (+) group had a value of 1.08 ± 0.05 cm⁻¹, which was higher than that of the SBF-PRP (-) group. A similar trend was observed for BV/TV. As for trabecular parameters, TB.N and TB. Th were significantly increased after treatment with the SBF-PRP (+) hydrogel compared with the SBF-PRP (-) and control groups, while TB. Sp was significantly decreased. These results confirmed that the SBF-PRP (+) hydrogel accelerates bone regeneration in vivo. Similar trends were observed in Masson staining. Taken together, these results suggested that the PRP-incorporated IPN hydrogels can rapidly and effectively promote bone regeneration.