Controlled and Sequential Delivery of Stromal Derived Factor-1 α (SDF-1α) and Magnesium Ions from Bifunctional Hydrogel for Bone Regeneration

Bone healing is a complex process that requires the participation of cells and bioactive factors. Stromal derived factor-1 α (SDF-1α) and magnesium ions (Mg2+) both are significant bioactive factors for cell recruitment and osteogenesis during bone regeneration. Thus, a bifunctional hydrogel containing a sequential delivery system is fabricated to improve osteogenesis. During sequential delivery of the hydrogel, SDF-1α is predominantly released at the early stage of bone mesenchymal stem cells (BMSCs) recruitment, while Mg2+ are constantly delivered at a later stage to improve osteogenic differentiation of recruited cells. In addition, due to the early release of SDF-1α, the hydrogel showed strong BMSCs recruitment and proliferation activity. Mg2+ can not only induce up-regulation of osteogenic gene expression in vitro, but also promote bone tissue and angiogenesis in vivo. Taken together, the injection of xanthan gum-polydopamine crosslinked hydrogel co-loading SDF-1α and Mg2+ (XPMS hydrogel) provides a novel strategy to repair bone defects.


Introduction
Bone defects are usually caused by trauma, infection, and tumor resection. Currently, the gold-standard in clinical treatment of bone defects is autologous bone grafting. Although autologous bone grafting can be used to repair a wide range of bone defects, it has the disadvantages of limited bone tissue availability and donor site complications. Nanohydroxyapatite (nHA) is another bone graft substitute which is usually used to repair alveolar bone defects. However, poor angiogenic efficiency and inadequate osteogenesis are hindering the growth of new bone [1]. With a deeper understanding of bone healing, the number of studies recognizing the importance of cell recruitment in regeneration has increased [2][3][4]. Insufficient responding osteoblast occurs when bone healing process is interrupted, causing obstructed bone formation with competitive growth of fibrous tissues. Therefore, two strategies for establishing a new multi-functional bone tissue substitute are: (1) adapting the shape of the cavity of bone defects to prevent the poor bone healing caused by the fibrous tissues; and (2) rapidly recruiting sufficient quantity 1 g XG (Aladdin, Los Angeles, CA, USA) dissolved in 500 mL of DI water under magnetic stirring at room temperature overnight. Then 634 µL of 107 mg/mL Sodium periodate (Macklin Biochemical Co., Ltd., Shanghai, China) solution was added to reaction for 4 h. Then 24 µL of 1% ethylene glycol was added for 1 h to stop the reaction. Then the product was purified for 24 h with a dialysis bag for molecular weight 3500 D. Finally, the Xan-CHO product was freeze-dried in a freeze dryer (Alpha 1-4 LD plus, Christ, Osterode am Harz, Germany) and stored in a sealed plastic bag at 4 • C for further research.

Characterization
The morphology of Mg-NPs was observed by scanning electron microscope (SEM, Zeiss Merlin Compact, Oberkochen, Germany) and transmission electron microscopy (TEM, FEI Talos F200s, Hillsboro, OR, USA). Hydrodynamic size distribution was evaluated using a Zetasizer Nano ZS90 (Malvern Instruments Ltd., Malvern, Worcestershire, UK) at 25 • C. The surface morphology of XPMS hydrogel was characterized by SEM with energy dispersive spectrometer (SEM-EDS, X-MaxN, Oxford, UK). The mechanical properties of the hydrogels were measured on a rheometer (Physica MCR 301, Anton Poar, Graz, Austria) using a parallel plate with a diameter of 25 mm and a gap of 0.3 mm in an oscillatory mode at 37 • C. Shear stress sweep test of the XPMS hydrogel was conducted with a range of stress from 0 to 100 Pa at a fixed strain of 1%. From the stress sweep, the yield stress of the hydrogel can be obtained. Subsequently, the self-healing property of the XPMS hydrogel was confirmed by the continuous step-torque measurements. Briefly, the XPMS hydrogels were alternately subjected to a torque of 250-370 µN·m to break the hydrogel network and a torque of 4.5-4.6 µN·m to restore the hydrogel network. This process was repeated for two times.

The SDF-1α Releasing from XPMS Hydrogel
The release of SDF-1α from the XPMS hydrogel in vitro was quantified via ELISA. First, the gels were incubated in 15 mL PBS at 37 • C under different conditions. At each time point (1 d, 2 d, 3 d, 4 d, 5 d, 6 d, 7 d), 1 mL PBS was collected and 1 mL fresh PBS was replaced. Finally, the SDF-1α concentrations in the collected PBS were quantified with SDF-1α ELISA kit (Xitang Biotech Company, Shanghai, China) according to the manufacturer's instructions.

BMSCs Isolation and Culture
BMSCs were isolated and cultured from the femurs of 4 weeks old Sprague-Dawley rats (Animal Research Center, Sun Yat-sen University, Guangzhou, China). Then the isolated cells were then cultured in F12/DMEM (Gibco, CA, USA) supplemented with 10% FBS (Gibco, CA, USA), 100 units/mL streptomycins and penicillin (P/S, Gibco, CA, USA) at 37 • C in an atmosphere of 5% CO 2 . The culture medium was replaced every 3 days. In the following experiments, cells from passages 3 were employed. The BMSCs have been identified by the flow cytometry assay ( Figure S1), and the BMSCs are positive for CD90 (BD Biosciences, NY, USA) and CD44 (eBioscience, San Diego, CA, USA), but negative for CD34 (BD Biosciences, NY, USA) and CD45 (eBioscience, San Diego, CA, USA).

The Preparation of Different Hydrogel Extraction
The indirect extraction method was used to evaluate samples' biocompatibility according to the ISO 10993-12 standard. Briefly, incubate the hydrogels for at least 48 h at 37 • C in F12/DMEM culture medium (containing 10%FBS, 1%P/S). After incubation, each extract solution of blank or drugs-loaded hydrogels samples was prepared by filtering through a 0.22 µm syringe filter (Merck, Darmstadt, Germany) before use. Experimental hydrogel extractions were grouped according to their physical and chemical modification, which divided into five groups: (1) XA group: extraction from pure XG; (2) XP group: extraction from PDA modified XG; (3) XPM group: extraction from Mg 2+ included PDA modified XG; (4) XS group: extraction from SDF-1α included pure XG; (5) XPMS group: extraction from SDF-1α and Mg 2+ included PDA modified XG. The recruitment of BMSCs in response to SDF-1α stimulation was investigated using a transwell migration model experiment. Briefly, cells were trypsinized and 200 µL cells suspension was seeded at a cell density of 1.5 × 10 4 cells per insert into the 8 µm pore transwell chamber (Corning, NY, USA) in a 24-well plate. Then, the chamber was inserted into each well of culture medium containing different concentrations of SDF-1α (0, 50, 100, 200 nM). After 24 h of incubation, the non-migration cells were completely wiped off from the top surface of the membrane using a cotton swab. After being fixed by 4% paraformaldehyde, the cells migrated to the undersurface of the membrane were stained by 0.1% crystal violet. Under the inverted microscope (Zeiss Axio Observer, Oberkochen, Germany), the average numbers of migrated cells were counted in at least five field from each well.

Functional Evaluation of Mg 2+ and SDF-1α on Chemotaxis of BMSCs
To validate whether SDF-1α and Mg 2+ have a synergistic effect on the recruitment of BMSCs, the chemotactic responses of BMSCs to SDF-1α (100 ng/mL) in the presence or absence of Mg 2+ (5 mM) were evaluated using a transwell assay. In a 24-well plate, cells were seeded at a density of 1.5 × 10 4 cells in the upper chamber. The chamber was then put into each well of culture medium containing SDF-1α and Mg 2+ alone or both. After 24 h of incubation, the non-migration cells were totally cleaned off the top surface of the membrane using a cotton swab. After being fixed by 4% paraformaldehyde, the cells migrated to the undersurface of the membrane were stained by 0.1% crystal violet. Under the inverted microscope (Zeiss Axio Observer, Oberkochen, Germany), the average numbers of migrated cells were counted in at least five fields from each well.

Cell Recruitment Activity of the XPMS Hydrogels In Vitro
The transwell migration assay was performed to evaluate the chemotactic responses of the BMSCs to the extraction of hydrogels. The upper chamber (pore size of 8 µm, Corning, NY, USA) was seeded with 1.5 × 10 4 cells. Lower chambers were supplemented with culture medium (10%FBS, 1%P/S) with different hydrogel extractions. The upper chamber was removed after 24 h of incubation, and the cells on the undersurface were stained with 0.1% crystal violet and counted. The average number of migrated cells in each group was calculated using five fields pictured by the inverted microscope (Zeiss Axio Observer, Oberkochen, Germany). The cell counting kit-8 assay (CCK-8, Dojindo, Kumamoto, Japan) was used to investigate the effect of Mg 2+ on BMSCs. In details, BMSCs were seeded at a density of 1 × 10 3 per well in 96-well plates and allowed to adhere and spread for 12 h. The BMSCs were then treated for 1, 3, and 5 days with cell culture medium containing F12/DMEM, 10% FBS, 1% (v/v) penicillin/streptomycin, and various amounts of added Mg 2+ (0, 2.5, 5, 10, 15, 20 mM). Thereafter, the cells were incubated with the CCK-8 solution for 2 h. A UV-spectrophotometer was used to measure the absorbance at 450 nm.

Cell Viability of the Hydrogels In Vitro
The biocompatibility of the hydrogels in vitro was investigated using the CCK-8 assay.
BMSCs were seeded at a density of 1 × 10 3 cells per well in 96-well plates and allowed to adhere and spread for 12 h. BMSCs were then treated with different extractions of hydrogels for 1, 3, and 5 days. After that, the cells were incubated with the CCK-8 solution for 2 h. The absorbance was measured at 450 nm using a UV-spectrophotometer. Live/dead assay was used to visualize the morphology and cell spreading of BMSCs under the extraction of hydrogels culturing for 1, 3, 5 days. Briefly, BMSCs were stained by calcein-AM and then pictured by microscope (Zeiss Axio Observer, Oberkochen, Germany) on days 1, 3, 5 after seeding. Calcium mineralization on the samples was evaluated with the use of Alizarin red S staining (ARS) at day 21. Briefly, cells were seeded in 6-well plates treated with OM (containing F12/DMEM, 10%FBS, 1% P/S, 10 × 10 −3 M β-glycerophosphate, 50 µg/mL L-ascorbic acid, and 10 × 10 −9 M dexamethasone) with the following various amounts of added Mg 2+ (0, 2.5, 5, 10, 15, 20 mM). After culturing for 21 days, cells were immersed in 4% paraformaldehyde for 15 min before being rinsed with PBS. The fixed cells were incubated in 2% ARS solution (Cyagen Bioscience Inc, Guangzhou, China) for 20 min. Finally, after thoroughly rinsing with deionized water, the stained cells were observed using an inverted microscope (Zeiss Axio Observer, Oberkochen, Germany).

The Effect of Hydrogel Extractions on Osteoinductivity of BMSCs In Vitro ALP Staining and Quantification in BMSCs Treated with Hydrogel Extraction
The activity of ALP in BMSCs treated with different hydrogel extractions was investigated. BMSCs were cultured in 6-well plates with a growth medium (F12/DMEM, 10%FBS, 1%P/S). The medium was replaced after 80% confluence with an osteogenic medium extracted from different hydrogels. After 7 days of culture, cells were rinsed with PBS (PH 7.4, Gibco, CA, USA), fixed in 4% paraformaldehyde, and incubated in p-nitrophenol buffer for 30 min according to the manufacturers' instructions (Yeason, Shanghai, China). The stained samples were examined with the inverted microscope (Zeiss Axio Observer, Oberkochen, Germany). For the colorimetric measurement of ALP activity, cells were lysed in a lysis buffer and the absorbance at 405 nm was determined by using p-nitrophenol phosphate assays (Beyotime, Shanghai, China) according to the protocol of manufacturers' procedure.

Alizarin Red S Staining of Mineralization in Hydrogel Extraction Treated BMSCs
Calcium mineralization on the samples was evaluated with the use of ARS at day 21. Briefly, cells were seeded in 6-well plates treated with OM (containing F12/DMEM, 10%FBS, 1%P/S, 10 × 10 −3 M β-glycerophosphate, 50 µg/mL L-ascorbic acid, 10 × 10 −9 M dexamethasone) extracted from different hydrogels. After culturing for 21 days, cells were immersed in 4% paraformaldehyde for 15 min and rinsed with PBS. The fixed cells were incubated in 2% ARS solution (Cyagen Bioscience Inc., Guangzhou, China) for 20 min. Finally, the stained cells were observed using an inverted microscope (Zeiss Axio Observer, Oberkochen, Germany) after thoroughly rinsed with deionized water.
Quantitative Real-Time Polymerase Chain Reaction (qRT-PCR) qRT-PCR analysis was used to evaluate the smaples' osteogenic capability. Briefly, cells treated with OM (containing F12/DMEM, 10%FBS, 1%P/S, 10 × 10 −3 M β-glycerophosphate, 50 µg/mL L-ascorbic acid, 10 × 10 −9 M dexamethasone) extracted from different hydrogels were lysed using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) after 7 or 14 days of culture. The total RNA concentrations were quantified with a Nanodrop spectrophotometer (Thermo-Fisher, Waltham, MA, USA). The first-strand cDNA was synthesized using RNA reverse transcription kit (Vazyme BioTech, Nanjing, China), and qRT-PCR was performed using a SYBR Green kit (Vazyme BioTech, Nanjing, China) according to the Q5 Quantstudio system protocol. Table 1 shows the primers for the targeted genes. Female Sprague-Dawley rats (SD rats) weighing 210-230 g were randomly assigned to one of five groups: (1) non-treated group; (2) Bio-oss group (Bio-oss ® hydroxyapatite granules, HA); (3) XPS group (HA + XPS); (4) XPM group (HA + XPM); (5) XPMS group (HA + XPMS). SD rats were provided by Guangdong Medical Laboratory Animal Center (Sun Yat-sen University, China) and maintained in the center. Animal were fed and handled in accordance with international guidelines on animal welfare and the standards of the Animal Ethical and Welfare Committee of Sun Yat-sen University. The animal experiments were approved by the Animal Ethical and Welfare Committee of Sun Yat-sen University. Briefly, SD rats were anesthetized by 1% sodium pentobarbital and 10% chloral hydrate. The rats' skin was incised bilaterally in the distal femoral epiphysis, and muscles were dissected bluntly to expose the femoral condyle. Then, using a trephine bur perpendicular to the femur's shaft axis at a slow speed and irrigated with saline solution to avoid heat necrosis, a 4 mm diameter bone defect was produced. The hydrogels were gently injected to fill the drilled defect with different components. Subsequently, the incision of the soft tissues was then closed, and the rats were given an intramuscular antibiotic injection 3 days after surgery. The rats in this study were sacrificed at 6 and 12 weeks respectively after implantation.

Micro-CT Analysis
The rats were sacrificed at 6 and 12 weeks by administering sodium pentobarbital (90-120 mg/kg), and the bilateral femoral condyles were fixed with formalin (10%) for 7 days. The distal femurs were then scanned using microcomputed tomography (micro-CT, Scanco Medical, Bassersdorf, Switzerland) at 70 kV and 200 µA. Once a region of interest was selected, and the center of the bone defect was reconstructed. Images following 3D construction were used to calculate the bone volume fraction (BV/TV), trabecular number (Tb.N), and trabecular separation (Tb.Sp) in the bone defect for analysis of the bone regeneration process within the defect.

Histological Evaluation
The samples were fixed in formalin and subsequently decalcified with ethylenediaminetetraacetic acid (EDTA, Servicebio, Wuhan, China) for 28 days with gentle shaking, and EDTA changes every 3 days. After decalcification, samples were embedded into the paraffin and serial sections (4 µm thick) were cut from each sample for hematoxylin/eosin (H&E) and Masson's trichrome staining.

Statistical Analysis
Statistical analyses results were presented as mean ± standard deviation and performed with IBM SPSS V. 25.0 (IBM SPSS). All of the experiments were performed in triplicate. One-way analysis of variance and Tukey's post hoc test were used to determine the significant differences among the groups. All statistical assessments were performed at a level of 0.05.

Preparation and Characterization of the SDF-1α/Mg-NPs/Gel (XPMS Hydrogel)
Mg-NPs were first synthesized by the complexation of dopamine and Mg 2+ under alkaline conditions. The morphology of Mg-NPs was revealed by scanning electron microscope (SEM) and transmission electron microscopy (TEM) observation (Figure 1a,b). Mg-NPs showed a spherical morphology and had an averaged diameter of 70 nm (Figure 1c), and the particle size in the dispersion was found to be stable in PBS conditions after 7 days at 37 • C ( Figure S2). The synthesis schematic showed that XPMS hydrogel was synthesized by the crosslinking between the aldehyde groups of modified XG and the amino groups of Mg-NPs through a Schiff base reaction, thus SDF-1α was encapsulated and spread in the gelatinous matrix during the cross-linking process (Figure 1d). The resultant XPMS hydrogel showed black and non-flowing appearance ( Figure 1e). As shown in Figure 1f, the SEM image illustrated the irregular and coarse surface structure of the freeze-dried XPMS hydrogel, which favors water ingress and the free drug can diffuse freely from the hydrogel. The overlapped and single element distributions in XPMS hydrogel were shown in Figure 1g. The homogeneous distributions of C, O, N, and Mg indicated that Mg-NPs were uniformly dispersed in the hydrogels. The infrared spectra of the hydrogels showed that the peak of the C=O group at 1732 cm −1 became smaller after the introduction of Mg-NPs, indicating the crosslinking between aldehyde groups with the amino groups under the Schiff base reaction ( Figure S3). Rheological measurements in the amplitude sweep mode revealed that both the storage modulus G and loss modulus G showed a plateau at a low shear stress level. The substantially higher of the storage modulus G than the loss modulus G evidenced the gel state of XPMS hydrogel. As the shear stress increased, the loss modulus G" decreased while the storage modulus G dropped more sharply, and then an intersection of G and G appeared. This trend indicated the XPMS hydrogel changes from gel to sol state, evidencing its shear thinning property ( Figure 1h). As showed in Figure 1I, the hydrogel networks were ruptured by applying a large torque (250-370 µN·m) for 50 s, which led to a liquid-like status (G > G ). On the contrary, the hydrogel restored the solid-like mechanics (G > G ) immediately once switching the torque to a smaller one (4.5-4.6 µN·m). The G s could be restored within seconds even after two cycles of network rupture, indicating the quick recovery of the XPMS hydrogel after network failure. Furthermore, the XPMS hydrogel can be remolded into different shapes after injection ( Figure S4), and its injectability may be useful for minimally invasive treatments. Then, the long-lasting release of SDF-1α from XPMS hydrogel was evaluated under different conditions (Figure 1j). Compared with SDF-1α/gel (XS hydrogel), the release rate of SDF-1α was remarkably reduced in the presence of XPMS hydrogel attributed to the dense structure by cross-linking between Mg-NPs and XG. XPMS hydrogel showed a much faster SDF-1α release under acidic condition, due to the degradation of Mg-NPs and consequently disruption of the intact gel structure. SDF-1α and Mg 2+ have different release kinetics, we expected the release rate of Mg 2+ from XPMS hydrogels was much slower than that of SDF-1α due to the enhanced stability of chemically bonded. As shown in Figure S5, a significantly slower and sustained release of Mg 2+ was evaluated under different conditions. Many studies have reported the synergistic effect of SDF-1α with other bioactive factors in bone repair [27,28]. We first focused on examining the SDF-1α and Mg 2+ regulation on BMSCs recruitment. Zhang et al. found that Mg 2+ also has the potential effect on SMSCs recruitment [29]. Therefore, we speculated cotreatment would improve BMSCs recruitment more than treatment with SDF-1α and Mg 2+ alone. Figure 2c,d showed that com-

Effect of Hydrogels on Recruitment of BMSCs In Vitro
Studies have shown that the activation of the SDF-1α/CXCR4 axis is conducive to the recruiting of BMSCs to the injured site [24]. One of the chemokines, SDF-1α, also known as CXCL12, is secreted around the tissue defects site due to the early inflammation and plays an important role in the recruitment of BMSCs to the defect site. Moreover, SDF-1α has been proved to enhance the BMSCs migration in a dose-dependent manner [25,26]. Figure 2a,b showed that the chemotactic active factor SDF-1α has a significant dose-dependent recruitment effect on BMSCs.   Many studies have reported the synergistic effect of SDF-1α with other bioactive factors in bone repair [27,28]. We first focused on examining the SDF-1α and Mg 2+ regulation on BMSCs recruitment. Zhang et al. found that Mg 2+ also has the potential effect on SMSCs recruitment [29]. Therefore, we speculated cotreatment would improve BMSCs recruitment more than treatment with SDF-1α and Mg 2+ alone. Figure 2c,d showed that compared to the blank control group, the Mg 2+ alone group had a significantly higher number of migrated BMSCs, while SDF-1α alone group was higher than that in the Mg 2+ alone group. These results indicated that the presence of Mg 2+ improve BMSCs recruitment to some extent, but the significance of SDF-1α in the collaboration with Mg 2+ also further suggests that SDF-1α delivery was essential for BMSCs recruitment.
SDF-1α encapsulated in crosslinked network and Mg 2+ complexed by PDA nanoparticles were released in a time-dependent pattern due to the sequential drug release of hydrogels, which may exert a good cell recruitment effect on BMSCs. In order to investigate the effect of sequential drug release hydrogel on cell recruitment, we used hydrogel extract to simulate the effect of hydrogels in situ releasing SDF-1α into the surrounding bone defect microenvironment ( Figure 2e). The number of migrated cells in the XPMS group containing SDF-1α and Mg 2+ increased significantly, showing that the combination of SDF-1α and Mg 2+ significantly improved the motility and migration of BMSCs. Collectively, the results demonstrated that SDF-1α and Mg 2+ delivered locally via XPMS hydrogel would have more beneficial biological effects on BMSCs recruitment.

Effect of Hydrogels on Biocompatibility and Osteoinductivity In Vitro
Cell recruitment is the initial step in acquiring endogenous stem cells at a bone defect site, and maximizes the local regeneration ability. Furthermore, effective stem cell differentiation management is required to guarantee that recruited cells differentiate into the desired cell lineages [30,31]. Notably, Mg 2+ plays an important role in bone mineralization, promoting cell adhesion, proliferation, and osteogenic differentiation [32,33]. Many studies have reported that Mg 2+ have distinct respective effects at different concentrations. Lin et al. [34] proposed that Mg 2+ with a concentration gradient of 2.5 mM to 5 mM were optimal for inducing BMSCs osteogenic differentiation, and Wang et al. [35] also proposed that Mg 2+ with a concentration gradient of 6 mM and 10 mM were best for promoting cell adhesion, cell activity and osteogenic differentiation. Therefore, the concentration of Mg 2+ is a key factor in modulating the proliferation and osteogenic differentiation of BMSCs. Figure 3a showed that BMSCs were treated with various concentrations of Mg 2+ (0, 2.5, 5, 10, 15 and 20 mM) and cell proliferation of BMSCs was evaluated by CCK-8 assays after one, three, and five days. The results showed that 5 mM and 10 mM concentrations of Mg 2+ significantly promoted the proliferation of BMSCs on day three and day five, while higher concentrations greater than 10 mM inhibited the proliferation of BMSCs. The results were comparable to those previously reported in the literature, indicating that locally high concentrations of Mg 2+ would impede the mineralization of the bone matrix and damage the process of bone regeneration process [36,37], which was connected to the inhibition of BMSCs proliferation. Therefore, based on our finding, a concentration gradient of 5-10 mM of Mg 2+ may be a favorable condition for BMSCs proliferation and the subsequent osteogenic differentiation.
Furthermore, SDF-1α has been shown to contribute to the proliferation and the differentiation of BMSCs into osteoblastic cells mediated by Wnt/β-Catenin pathway [38]. Thus, Mg 2+ and SDF-1α were co-loaded into the hydrogels to investigate the synergistic effect on BMSCs proliferation. Figure 3b showed that, as compared to the XA group, the higher rate of proliferations of BMSCs were observed in XPM and XPMS group on day three, and there was no significant difference in XPM and XPMS group, supporting the notion that Mg 2+ delivered from hydrogels shows proliferation enhancement toward BMSCs. Meanwhile, morphological investigation was used to confirm the hydrogel extract's stimulation of BMSC proliferation (Figure 3c). All of the hydrogels showed significant cell density increases and no differences in cell appearance after incubation, which was consistent with the CCK-8 assays. Moreover, BMSCs in XPMS and XPM groups retained spindle shape and showed good proliferation. Collectively, these results suggest that the XPMS hydrogel is highly biocompatible and capable of preserving the survival and proliferation for recruited BMSCs in situ. The ability of Mg 2+ to enhance osteogenic differentiation was further investigated by analyzing the activity of ALP, which is an early marker of osteogenesis. The ALP staining of BMSCs was obviously intensified with the increase of Mg 2+ concentration, and particularly showed the best performance in 5-10 mM (Figure 4a). We also determined ALP activity to provide for quantitative analysis (Figure 4b). ALP expression was enhanced by up to 1.5-fold with 5 mM Mg 2+ , and show no difference with the higher concentration of Mg 2+ . ARS was used to detect calcium deposition as a late osteogenic marker (Figure 4c). The staining of calcium deposition was intensified as the concentration of Mg 2+ increased up to 5 mM. In contrast, the staining intensity decreased with further increase in Mg 2+ concentration, suggesting the inhibition of mineralization process. Over all, the osteogenic capacity of Mg 2+ upregulated with a concentration of 5 mM and significantly promoted osteogenic differentiation of BMSCs, which provide guidance for loading strategy in hydrogels. The ability of Mg 2+ to enhance osteogenic differentiation was further investigated by analyzing the activity of ALP, which is an early marker of osteogenesis. The ALP staining of BMSCs was obviously intensified with the increase of Mg 2+ concentration, and particularly showed the best performance in 5-10 mM (Figure 4a). We also determined ALP activity to provide for quantitative analysis (Figure 4b). ALP expression was enhanced by up to 1.5-fold with 5 mM Mg 2+ , and show no difference with the higher concentration of Mg 2+ . ARS was used to detect calcium deposition as a late osteogenic marker (Figure 4c). The staining of calcium deposition was intensified as the concentration of Mg 2+ increased up to 5 mM. In contrast, the staining intensity decreased with further increase in Mg 2+ concentration, suggesting the inhibition of mineralization process. Over all, the osteogenic capacity of Mg 2+ upregulated with a concentration of 5 mM and significantly promoted osteogenic differentiation of BMSCs, which provide guidance for loading strategy in hydrogels. Polymers 2022, 14, x FOR PEER REVIEW 1 Figure 4. In vitro effect of Mg 2+ on BMSCs osteogenic differentiation; (a) ALP staining and (b tification of BMSCs exposed to different concentrations of Mg 2+ for 7d; (c) ARS staining of exposed to different concentrations of Mg 2+ for 21d; *** p < 0.001.
Further, to quantify the oestoinductive ability of XPMS hydrogel, ALP stainin performed (Figure 5a). In comparison to the XA group, the ALP staining was inten in the XPM group and further amplified in the XPMS group. BMSCs in the XPM XPM groups exhibited significantly greater ALP activity. Moreover, ARS results ( 5b) also demonstrated that the calcium deposition staining in the XPM group was sified by incorporating Mg 2+ and exhibited the best performance in XPMS group, show a similar tendency with the ALP results. The results indicated that the combi of Mg 2+ and SDF-1α delivered from XPMS hydrogel induced a significantly greate motion on osteogenic differentiation of BMSCs. Meanwhile, previous studies have that SDF-1α stimulation increases the expression of the type I collagen gene (Col indicating that SDF-1α regulates BMSCs osteogenic differentiation. Subsequently PCR analysis was also used to evaluate the expression of osteogenic genes including I, Runx2, ALP, and OCN (Figure 5c). After culturing for seven days, the mRNA le COL-I were upregulated 2.3-fold in XPM group, with XPMS group showing the h expression (2.95-fold) as compared to XA group. The levels of Runx2 in XPM and groups were also increased 1.3-fold and 1.2-fold, respectively. The levels of ALP we regulated 1.4-fold in XPM and XPMS groups. The expression of OCN was signifi Further, to quantify the oestoinductive ability of XPMS hydrogel, ALP staining was performed (Figure 5a). In comparison to the XA group, the ALP staining was intensified in the XPM group and further amplified in the XPMS group. BMSCs in the XPMS and XPM groups exhibited significantly greater ALP activity. Moreover, ARS results (Figure 5b) also demonstrated that the calcium deposition staining in the XPM group was intensified by incorporating Mg 2+ and exhibited the best performance in XPMS group, which show a similar tendency with the ALP results. The results indicated that the combination of Mg 2+ and SDF-1α delivered from XPMS hydrogel induced a significantly greater promotion on osteogenic differentiation of BMSCs. Meanwhile, previous studies have shown that SDF-1α stimulation increases the expression of the type I collagen gene (Col I) [39], indicating that SDF-1α regulates BMSCs osteogenic differentiation. Subsequently, qRT-PCR analysis was also used to evaluate the expression of osteogenic genes including COL-I, Runx2, ALP, and OCN (Figure 5c). After culturing for seven days, the mRNA levels of COL-I were upregulated 2.3-fold in XPM group, with XPMS group showing the highest expression (2.95-fold) as compared to XA group. The levels of Runx2 in XPM and XPMS groups were also increased 1.3-fold and 1.2-fold, respectively. The levels of ALP were upregulated 1.4-fold in XPM and XPMS groups. The expression of OCN was significantly enhanced by XPM and XPMS hydrogels on day 14. These findings are consistent with previous report that SDF-1α can participate in osteogenic differentiation of BMSCs by upregulating the expression of COL-I. In addition, Mg 2+ delivered from the hydrogels plays a significant role in osteogenic differentiation of BMSCs, which suggest that the XPMS hydrogel could create an osteogenic microenvironment by delivering Mg 2+ and promoting the osteogenic effect of the recruited BMSCs. ers 2022, 14, x FOR PEER REVIEW 14 of enhanced by XPM and XPMS hydrogels on day 14. These findings are consistent w previous report that SDF-1α can participate in osteogenic differentiation of BMSCs by u regulating the expression of COL-I. In addition, Mg 2+ delivered from the hydrogels pla a significant role in osteogenic differentiation of BMSCs, which suggest that the XPM hydrogel could create an osteogenic microenvironment by delivering Mg 2+ and promoti the osteogenic effect of the recruited BMSCs.

Effect of Hydrogels on Osteoinductivity In Vivo
Previous studies have shown that gelatin hydrogels combined with SDF-1α and bo morphogenetic protein (BMP-2) was more effective to induce bone regeneration, whi may be attributed to the synergetic effect on intracellular Smad and Erk activation [2 Other researchers also found that SDF-1α may enhanced recruitment of endothelial pr genitor cells (EPCs) by SDF-1α/CXCR4 signaling, which helps to supply nutrient and su port bone repair [40]. SDF-1α is not sufficient for in-situ regeneration of critical bone d fects, additional osteogenic bioactive factors may also be required. Meanwhile, b material scaffolds containing Mg 2+ have been shown in numerous investigations to i prove osteogenesis [41][42][43]. Moreover, it is worth noting that Mg 2+ also have a recruiti effect on BMSCs via activating TRPM7/PI3K signaling pathway in recent literature [4 The therapeutic effect of the hydrogel system containing two bioactive factors was grea than that of the single component system. In order to mimic the process of natural bo healing, XPMS hydrogel possessing the double crosslinking structure was fabricated control the release sequence of each factor. The crosslinked structure of xanthan gum c prolong the duration of releasing the encapsulated SDF-1α, while the covalent bonding

Effect of Hydrogels on Osteoinductivity In Vivo
Previous studies have shown that gelatin hydrogels combined with SDF-1α and bone morphogenetic protein (BMP-2) was more effective to induce bone regeneration, which may be attributed to the synergetic effect on intracellular Smad and Erk activation [26]. Other researchers also found that SDF-1α may enhanced recruitment of endothelial progenitor cells (EPCs) by SDF-1α/CXCR4 signaling, which helps to supply nutrient and support bone repair [40]. SDF-1α is not sufficient for in-situ regeneration of critical bone defects, additional osteogenic bioactive factors may also be required. Meanwhile, biomaterial scaffolds containing Mg 2+ have been shown in numerous investigations to improve osteogenesis [41][42][43]. Moreover, it is worth noting that Mg 2+ also have a recruiting effect on BMSCs via activating TRPM7/PI3K signaling pathway in recent literature [44]. The therapeutic effect of the hydrogel system containing two bioactive factors was greater than that of the single component system. In order to mimic the process of natural bone healing, XPMS hydrogel possessing the double crosslinking structure was fabricated to control the release sequence of each factor. The crosslinked structure of xanthan gum can prolong the duration of releasing the encapsulated SDF-1α, while the covalent bonding of PDA and Mg 2+ creates another stable functional microstructure and release profile. Due to the distinct loading mechanisms of SDF-1α and Mg 2+ , SDF-1α is released early around the bone defect and recruits BMSCs and EPCs through SDF-1α/CXCR4 signaling activation, limiting the growth of fibrous tissues. Later, Mg 2+ are released in a sustained manner, which enhances the proliferation and osteogenic differentiation of recruited cells and accelerates mineralization of bone matrix throughout the bone healing life cycle.
Therefore, to validate the ability of XPMS hydrogel composite scaffolds to promote bone repair in vivo, composite biomaterials of injectable hydrogel containing different bioactive factors were injected into a rat femur defect, micro-CT and histological analysis of femur specimens were performed on 6 weeks and 12 weeks after in situ implantation. As shown in Figure 6a, the non-treated group's defects displayed a little new bone tissues at the margin, and we observed new bone formation in the Bio-oss group but substantial loss of bone powder in bone defects. The cause is likely to be blood clots forming that was insufficient to provide proper mechanical support on bone powder in the bone defect. In contrast, hydrogel-composite scaffolds showed significant new bone formation at 6 weeks and week 12. At week 6, the XPS and XPMS groups showed appreciable osteogenesis in the defect's center, forming a "bony bridge" between the bone powder and integrating into patchy. It was confirmed that SDF-1α greatly promote bone defect repair in the early stage by recruiting bone marrow mesenchymal stem cells. The composite hydrogel scaffolds containing the Mg 2+ (XPM and XPMS groups) showed apparent new bone formation at week 12, suggesting that Mg 2+ have a substantial role in promoting osteogenesis.
Further, BV/TV, Tb.N, and Tb.Sp were quantitatively evaluated by micro-CT images. As shown in Figure 6b, composite hydrogel scaffolds containing SDF-1α and Mg 2+ groups (XPMS group) displayed a noteworthy enhancement of BV/TV at week 12 compared to the non-treated group and the Bio-oss group. Meanwhile, the XPMS group had slightly greater bone parameters than the XPS or XPM groups, suggesting that SDF-1α may have a synergistic effect with Mg 2+ during bone healing in vivo, which was consistent with our findings in vitro. In contrast, when the XPMS group were compared to the Bio-oss group at week 6, there was no statistically significant difference in Tb.Sp. The trabecular separation of the composite hydrogel scaffold groups (XPS, XPM, and XPMS groups) were considerably lower than the Bio-oss group at 12 weeks after implantation. In conclusion, the composite hydrogel scaffolds may continuously promote the repair of bone defects and enhance osteogenic quality over time, indicating that the injectable bioactive composite hydrogel scaffolds have excellent osteogenesis performance.
The new formation of bone tissues was also evaluated histologically by hematoxylin and eosin (H&E) and Masson's trichrome staining as shown in Figure 7, and the results were largely consistent with micro-CT analysis. The defects in the non-treated group did not show obvious healing at six weeks, while the Bio-oss group and the composite hydrogel scaffolds group loaded with bioactive molecules (XPM, XPS, and XPMS groups) had more favorable newly woven bone tissue. Moreover, extensive capillaries were observed around the defects in the composite hydrogel scaffolds group. Angiogenesis and osteogenesis are highly correlated in the process of bone development and bone generation according to previous literature [45]. Our findings are in agreement with reports that SDF-1α or Mg 2+ encapsulated scaffolds have excellent angiogenesis function for bone reconstruction [46][47][48]. Furthermore, at six weeks postoperatively, a large number of osteoblasts were observed in the central area of the defect in the XPMS group as compared to XPS and XPM group, suggesting that the promotion of SDF-1α and Mg 2+ on the recruitment and proliferation of BMSCs around the defect site.
x FOR PEER REVIEW 16 of 20 The new formation of bone tissues was also evaluated histologically by hematoxylin and eosin (H&E) and Masson's trichrome staining as shown in Figure 7, and the results were largely consistent with micro-CT analysis. The defects in the non-treated group did not show obvious healing at six weeks, while the Bio-oss group and the composite hydrogel scaffolds group loaded with bioactive molecules (XPM, XPS, and XPMS groups) had more favorable newly woven bone tissue. Moreover, extensive capillaries were observed around the defects in the composite hydrogel scaffolds group. Angiogenesis and osteogenesis are highly correlated in the process of bone development and bone generation according to previous literature [45]. Our findings are in agreement with reports that SDF-1α or Mg 2+ encapsulated scaffolds have excellent angiogenesis function for bone reconstruction [46][47][48]. Furthermore, at six weeks postoperatively, a large number of osteoblasts were observed in the central area of the defect in the XPMS group as compared to XPS and XPM group, suggesting that the promotion of SDF-1α and Mg 2+ on the recruitment and proliferation of BMSCs around the defect site.
The histological results at 12 weeks showed that the defects in the non-treated group were filled with fibrous tissues, indicating that wound healing was delayed and showed The histological results at 12 weeks showed that the defects in the non-treated group were filled with fibrous tissues, indicating that wound healing was delayed and showed poor bone healing. In contrast, more woven bone tissues were observed within the defect in the composite hydrogel scaffolds group, and the fibers in the woven bone tissues were more compact and organized than those in the Bio-oss group. Meanwhile, the transition of woven bone tissues into mineralized bone tissues could be found in Masson's trichrome staining. The XPMS group showed a higher density of mineralized bone tissues growing connectively than the XPS and XPM groups, indicating that SDF-1α and Mg 2+ released from XPMS hydrogel had successfully accelerated the bone defect repair. Overall, XPMS composite hydrogel scaffolds not only enhance bone marrow mesenchymal stem cells and endothelial progenitor cells recruitment through SDF-1α chemotaxis in early release, but also cooperate with Mg 2+ released sequentially to induce osteogenesis and angiogenesis, carrying out the entire life cycle of bone restoration. connectively than the XPS and XPM groups, indicating that SDF-1α and Mg released from XPMS hydrogel had successfully accelerated the bone defect repair. Overall, XPMS composite hydrogel scaffolds not only enhance bone marrow mesenchymal stem cells and endothelial progenitor cells recruitment through SDF-1α chemotaxis in early release, but also cooperate with Mg 2+ released sequentially to induce osteogenesis and angiogenesis, carrying out the entire life cycle of bone restoration.

Conclusions
In conclusion, a novel injectable bone healing hydrogel was designed with excellent sequential controlled administration of SDF-1α and Mg 2+ . The XPMS hydrogel exhibited good injectable and mechanical properties after introducing Mg-NPs into XG via Schiff base reaction, thus enabling it appropriate for clinical usage. Furthermore, our results suggested that SDF-1α and Mg 2+ released in a controlled manner from XPMS hydrogels

Conclusions
In conclusion, a novel injectable bone healing hydrogel was designed with excellent sequential controlled administration of SDF-1α and Mg 2+ . The XPMS hydrogel exhibited good injectable and mechanical properties after introducing Mg-NPs into XG via Schiff base reaction, thus enabling it appropriate for clinical usage. Furthermore, our results suggested that SDF-1α and Mg 2+ released in a controlled manner from XPMS hydrogels synergistically stimulate osteogenesis, indicating that the underlying mechanism of cell recruitment enhancement at an early stage is involved. This point of view was further supported by a model of femoral bone defects in rats. In our work, the osteogenic capacity of XPMS hydrogels was investigated, and the role of cell recruitment during bone healing was highlighted. These findings may contribute to a new strategy for bone regeneration.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/polym14142872/s1, Figure S1: Identification of BMSCs in flow cytometry; Figure S2: Particle size variation of Mg-NPs for 7 days; Figure S3: IR spectra of XA and XPM; Figure S4: Injectable properties of XPMS hydrogel; Figure   Institutional Review Board Statement: The animal feeding and handling were carried out according to international guidelines on animal welfare and the standards of the Animal Ethical and Welfare Committee of Sun Yat-sen University. The animal study protocol was approved by the Animal Ethical and Welfare Committee of Sun Yat-sen University, based on decision number IACUC-DB-16-0411, on 2021-06/4.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.

Conflicts of Interest:
There are no conflict of interest to declare. The funders had no role in the design and implement of the research.