Mg‐CS/HA Microscaffolds Display Excellent Biodegradability and Controlled Release of Si and Mg Bioactive Ions to Synergistically Promote Vascularized Bone Regeneration

For bone defect repair, it is critical to utilize biomaterials with pro‐angiogenic properties to enhance osteogenesis. Hydroxyapatite (HA)‐based materials widely used in clinical applications have shown much potential for bone repair. However, their predominant calcium phosphate (CaP) composition and poor biodegradability limit their angiogenic potential and hence osteogenic efficiency of HA‐based materials. Here, a magnesium ion‐doped calcium silicate/HA composite microscaffold (Mg‐CS/HA) is fabricated to enhance angiogenesis and osteogenic efficiency for bone repair. Incorporation of CS improved the biodegradability of the Mg‐CS/HA microscaffold, which could simultaneously release Si and Mg bioactive ions during the early stage of implantation, synergistically enhancing angiogenesis and osteogenic efficiency. In co‐culture systems, the synergistic effects of Si and Mg ions promote the “osteogenesis‐angiogenesis coupling effect.” In vivo, the Mg‐CS/HA microscaffold could significantly promote reconstruction of the vascular network and bone regeneration. This study thus provides a new strategy for coordinated release of bioactive ions to achieve synergistic effects on vascularized bone regeneration by HA‐based bone implant materials.


Introduction
More than two million bone grafts are implanted annually worldwide to cope with the challenges of clinical bone defect repair. [1] Due to limitations in obtaining autogenous bone grafts and related complications associated with donor site morbidity, various artificial bone graft materials have been developed in recent years. [2][3][4][5][6] Synthetic materials that have been recently used for bone regeneration include bio-ceramics, metal scaffolds, polymers, and composites. The composition and structure of Hydroxyapatite (HA) microscaffolds are similar to the inorganic phase of natural bone and teeth, and has a wide range of applications in the biomedical field. [7][8][9][10] However, the weak osteoinductivity of HA makes it difficult to activate biological processes required for high efficiency bone regeneration. [11] Bone regeneration involves simultaneous activation and complex interaction between the angiogenic and osteogenic pathways, which critically determines the osteoinductivity of implant materials. [12][13][14][15][16] Hence, it is imperative to target relevant biological processes to improve the angiogenic properties of HA microscaffolds for enhancing its bone defect repair efficacy.
To overcome the limited pro-angiogenic properties of bone repair materials (such as Bio-Oss and tricalcium phosphate) that are currently used in the clinic, much interests have been focused on improving the angiogenic properties of HA by bioactive ion doping, which has attracted much attention due to their stable and extensive biological effects. [17,18] Various ions, such as Si, Mg, Sr and Cu ions, etc., have been used for doping HA to promote vascularized bone regeneration. [19][20][21][22] It is worth noting that these bioactive ions often have extensive biological effects in vivo. [23][24][25][26][27][28] In particular, Mg ion is an essential element for physiological functions, and has been demonstrated to be able to promote angiogenesis, while Si ion can promote the early mineralization process of osteogenesis. [29][30][31][32][33] This thus suggests that the synergistic effects of cooperative ion doping within the HA microscaffold could be a more effective way of promoting vascularized bone regeneration by activating the "osteogenesisangiogenesis coupling effect." The excellent biodegradability of bone implant materials is indispensable for achieving efficient release of bioactive ions. [34,35] The disadvantage of the HA material is its poor degradability, even though much work has been done on HA properties and application areas, the degradation behavior of HA and release of doped ions under bone defect conditions are still not controllable. [36][37][38] Hence, a better strategy is still needed to control the degradability of HA and provide an effective synergistic ionic microenvironment for bone defect repair. The calcium silicate (CS) ceramics and their composites have been regarded as potential candidates for artificial bone graft materials due to their similar mechanical properties to natural bone tissues, excellent bioactivity and biodegradable properties. [39][40][41] Our previous studies have also revealed that Mg doped CS (Mg-CS) possessed excellent biodegradability and could efficiently release Si 4+  Mg 2+ at the same time. [42] Therefore, the combination of HA microscaffold and Mg-CS could be a promising method to control the degradation performance and provide a synergistic ionic microenvironment for vascularized bone regeneration.
In this study, a granular bioactive bone graft material -Mg ion doped CS/HA composite microscaffold (Mg-CS/HA) was obtained by compositing nano-sized HA (n-HA) in CS sol-gel doped with Mg ion. The Mg-CS/HA microscaffold could rapidly release a large amount of Si and Mg ions in simulated body fluid (SBF) to induce mineralization. The simultaneous release of Si and Mg ions from the Mg-CS/HA microscaffold could facilitate the early phase of vascularization of human umbilical vein endothelial cells (HUVECs) and enhance osteogenic differentiation of rat bone marrow mesenchymal stem cells (rBMSCs) via promotion of the "osteogenesis-angiogenesis coupling effect." The optimal degradation rate of Mg-CS/HA efficiently builds up an ionic microenvironment that significantly promotes early angiogenesis and osteogenesis within the bone defect area. This study thus provides a promising strategy to combine and synergize cooperative effects of bioactive ions with excellent degradability of the Mg-CS/HA microscaffold to achieve efficient vascularized bone regeneration.

Physicochemical Structural Characterization of Mg-CS/HA Microscaffolds
We developed a new type of granular bioactive bone graft material with good biodegradability as well as capacity for sustained release of bioactive ions (i.e., Mg 2+ , Ca 2+ , Si 4+ ), by utilizing the template method with high temperature sintering process. As shown in Figure 1, a n-HA suspension was first synthesized by chemical precipitation method ( Figure S1, Supporting Information), mixed uniformly with CS and Mg-CS/HA sol-gel that was aged at 37°C for 3 days, and then impregnated within a porous polymer template with a pore size of 300-500 μm. After the impregnation process, the cancellous bone-like scaffold was prepared by sintering the coated template at 1100°C, which was crushed into particles with a size of 0.25-1.0 mm, prior to characterization of a series of properties. The micromorphology of three groups of materials was shown in Figure S2, Supporting Information. The surface of the CS/HA and Mg-CS/HA particles was uniform without significant HA aggregations. In the presence of both CS and Mg-CS sol-gel, there was good transition and bonding between the CS and the HA phase to form an integrated structure, while the sintered n-HA displayed loose morphology. CS/HA and Mg-CS/HA not only resembled cancellous bone structurally, but also significantly increased its mechanical strength as compared to the HA ( Figure S3, Supporting Information). The energy dispersive X-ray spectroscopy (EDX) showed that the doping content of Mg 2+ was 10 mol% of Ca 2+ , which was close to the feeding ratio, indicating that there was no loss of dopant ions during the preparation process ( Figure S2, Supporting Information). The results of Xray photoelectron spectroscopy (XPS) in Figure S4, Supporting Information, further confirmed the results of EDX. Based on the standard X-ray diffraction (XRD) pattern ( Figure S5, Supporting Information), we confirmed that the crystal structures of CS and HA coexisted (CS: PDF#27-0088; HA: PDF#09-0432) in CS/HA and Mg-CS/HA, and that Mg 2+ doping will form a small amount of new phase magnesium silicate (MgSiO 3 : PDF#35-0610) without destroying the crystalline phases of CS and HA.

Biological Activity and Ion Release Profile
To evaluate the ion release profiles of the HA, CS/HA, Mg-CS/HA microscaffolds, the materials were soaked both in SBF solution (pH = 7.4) at 37°C for 28 days and in deionized water at 37°C for 7 days. The contents of Ca 2+ and Si 4+ within the solutions of the CS/HA and Mg-CS/HA groups increased significantly in the beginning, due to degradation of CS and Mg-CS, followed by accumulation of Ca 2+ to a certain extent and mineralization deposition (Figure 2a, Figure S6, Supporting Information). To be specific, the dissolution of CS released Ca 2+ and Si 4+ ions into the solution, generating free SiO 3 2− and Si-OH on the material surfaces. Then, the condensation between the Si-OH groups and Si-O-Si accumulated on the surface of the material, forming a negatively charged silicon-rich layer. The negative layer further adsorbed Ca 2+ in the solution to form a Si-Ca layer, which then further adsorbed HPO 4 2− in the solution. As a result, calcium and phosphorus compounds accumulated on the surface of the material (Figure 2b). The HA particles degraded slowly and had difficulty in dissolving ions and absorbing Ca 2+ and HPO 4 2− . Therefore, the amount of mineral deposition on the HA surface was significantly less than those on CS/HA and Mg-CS/HA ( Figure S7, Supporting Information). Because the materials in each group contained highly crystalline HA, it was not possible to determine whether the calcium and phosphorus compounds formed were crystalline, based on the XRD pattern ( Figure S8, Supporting Information). However, numerous studies had shown that during the mineralization process, the calcium and phosphorus compounds formed at the initial stage were attributed to amorphous calcium phosphate (ACP), which would gradually transform into crystalline HA upon progression of mineralization, which reflected the in vitro osteogenic activity of the material. The introduction of CS could significantly improve the osteogenic activity of the composite bioceramic.

Biocompatibility of the Mg-CS/HA Microscaffold
To study the influence of the ionic microenvironment of different materials on cell activity, the CCK-8 assay was used to compare proliferation of rBMSCs cultured in the various dilutions of the extracts of the microscaffolds (Figure 3a). The results showed that with an increase of dilution ratio, the proliferation activity of cells exhibited a gradually increasing trend. In particular, 1/8 concentration extracts of HA and Mg-CS/HA showed the highest proliferation activity. Therefore, consistent with other research on biomaterial extracts, the 1/8 concentration extracts of different microscaffolds were analyzed in subsequent tests. [43,44] As shown in Figure 3b and Figure S9, Supporting Information, no obvious dead cells with red staining were observed among the different extract groups after being co-cultured at 1, 2, 3 days. The same trend was also found in apoptosis detection by flow cytometry on day 1, indicating that the extracts of the HA, CS/HA and Mg-CS/HA microscaffolds demonstrated excellent biocompatibility ( Figure 3c). Furthermore, the scanning electron microscope (SEM) images ( Figure S10, Supporting Information) obtained after culturing rBMSCs on different microscaffolds showed that rBMSCs presented a wider cell spreading area, which was considered to be a phenotype that was conducive to osteogenic differen- tiation of BMSCs. [45] Interestingly, compared with other groups, more completed mineralized nodules with spherical shapes were enriched on the Mg-CS/HA microscaffolds, which demonstrated that it possessed better bone regeneration efficiency in vitro. [46,47]

The Mg-CS/HA Microscaffold Extract Promotes the Osteogenic Differentiation of rBMSCs
To evaluate the osteogenic properties of the different extracts in vitro, the alkaline phosphatase (ALP) assay and alizarin red staining were carried out (Figure 4a-d). The results showed that compared with the HA and CS/HA group, more ALP staining area ( Figure 4a) and mineralization nodules ( Figure 4c) were observed in the Mg-CS/HA group. Quantitative analysis also revealed more significant ALP activities ( Figure 4b) and higher mineralization values (Figure 4d) in the Mg-CS/HA group compared with the other groups. Moreover, as shown in Figure 4f, the expression levels of the osteogenic-related genes Runx2 (runt related transcription factor 2), Opn (osteopontin), Col1 (collagen type I) and Bmp2 (bone morphogenetic protein 2) were upregulated in the Mg-CS/HA group compared with other groups. Consistently, the protein expression levels of RUNX2, OPN, COL1 and BMP2 were confirmed in Figure 4h. As RUNX2 plays a critical role in the osteogenesis of BMSCs, the immunofluorescence staining of RUNX2 was performed. It was found that more extensive green fluorescence emitted from the nuclei of rBMSCs in the Mg-CS/HA group (Figure 4g). Consistent with the above results, the SEM images ( Figure 4e) and the immunofluorescence im-ages of Vinculin ( Figure 4e) further confirmed that the rBMSCs incubated with the Mg-CS/HA extract displayed a more conductive cell phenotype for osteogenesis. It was reported that Mg 2+ could enhance the adhesion, spreading, proliferation, migration and differentiation of BMSCs. [29,48] Additionally, the Si 4+ could take effect on collagen synthesis, bridging collagen tissue, and biomineralization, which makes up for the functional deficiency of Mg 2+ . [32,33] Significantly, our results suggested that the synergistic effects of Mg 2+ and Si 4+ could significantly improve the osteogenesis of rBMSCs.

The Mg-CS/HA Microscaffold Extract Promotes the Angiogenesis of HUVECs
To further evaluate the angiogenic potential of HUVECs cultured in the Mg-CS/HA extract, a tube formation assay was performed. The results showed there were much more mesh-like circles in the Mg-CS/HA group after culturing the HUVECs for 4 and 8 h (Figure 5a, Figure S11, Supporting Information). Consistently, immunofluorescence staining of vascular endothelial growth factor (VEGF) showed the strongest fluorescence intensity in the Mg-CS/HA group, as compared to the other groups ( Figure S12, Supporting Information). It is considered that the recruitment and migration of endothelial cells (ECs) played a key role in vascularization. [30] Therefore, the effects of the Mg-CS/HA extracts on HUVECs recruitment were investigated using the transwell assay (Figure 5b). It was shown that much more HUVECs could be observed within the lower chamber after being incu- bated with the Mg-CS/HA extract after 3 and 6 h. Additionally, compared with the other groups, cells incubated with the Mg-CS/HA extracts exhibited the highest percentage of wound healing, up to 62.3 ± 0.53% (Figure 5c), which indicated that it could obviously promote the migration of HUVECs. These results suggested that the Si 4+ and Mg 2+ ions within the Mg-CS/HA extract could provide an optimal multi-ionic environment to promote the vascularization of HUVECs. It had previously been reported that Si 4+ can enhance vascularization of ECs by stimulating the expression of VEGF from ECs or neighboring cells. [49] Additionally, Mg 2+ could activate the same mechanism as VEGF inducing angiogenesis by stimulating nitric oxide production in ECs. [21] Our above results indicated that the enhancing effect on angiogenesis could be attributed to the osteogenesis and angiogenesis coupling effects of BMSCs and ECs co-mediated by Si and Mg ions.

Mechanisms of Osteogenesis-Angiogenesis Coupling Mediated by the Mg-CS/HA Microscaffold
Osteogenesis-angiogenesis coupling refers to the indirect or direct crosstalk effect between BMSCs and ECs, which plays a crit-ical role during the process of bone repair. [50] To validate the synergistic effects of Si 4+ and Mg 2+ ions released from the Mg-CS/HA microscaffold on the coupling of angiogenesis and osteogenesis, the indirect co-culture system of rBMSCs and HUVECs was utilized as an experimental model (Figure 6a,b). As shown in Figure 6c,d and Figure S13, Supporting Information, as compared with the negative control conditional medium (CM), HU-VECs cultured in CM of different extracts exhibited better tube formation ability. In particular, the highest expression levels of markers for tube formation (total tubule length, number of tubes and junctions) were observed in the conditional medium of the Mg-CS/HA (Mg-CS/HA-CM) group at 8 h. Moreover, the highest ALP expression levels were observed in the Mg-CS/HA-CM group on day 3 (Figure 6e,f). Studies indicated that at the initial stage, the released Mg 2+ and Si 4+ could facilitate angiogenesis by promoting the recruitment of ECs and stimulating BMSCs to secrete VEGF. [49,51] Hence, we presumed that this promoting effect on angiogenesis could be attributed to the osteogenesis and angiogenesis coupling effects of BMSCs and ECs, which are comediated by Si and Mg ions.
To obtain deeper insights into the underlying mechanisms by which the ionic microenvironment provided by the Mg- . Evaluating osteogenic differentiation of rBMSCs cultured in HA, CS/HA, and Mg-CS/HA extracts at the optimal concentration (1/8 dilution). a) ALP staining and b) quantitative analysis after rBMSCs were cultured for 3 and 7 days. c) Alizarin red staining and d) quantitative analysis after rBMSCs were cultured for 21 days. e) SEM images of rBMSCs being cultured for 12 h (scale bars, 50 μm), and immunofluorescent images of rBMSCs stained for Vinculin (green) observed under CLSM (scale bars, 25 μm). f) Gene expression in relation to osteogenic differentiation of rBMSCs: Runx2, Bmp2, Opn, and Col1. g) Immunofluorescence images of RUNX2 (green) and nuclei (DAPI, blue) in rBMSCs cultured for 3 d. h) Protein expression levels of RUNX2, BMP2, COL1, and OPN (3 and 7 d) in rBMSCs.
CS/HA microscaffold promotes osteogenesis-angiogenesis coupling, RNA sequencing (RNA-seq) was performed on HUVECs and rBMSCs incubated with non-osteogenic medium containing or without the Mg-CS/HA extract. Differentially expressed genes (DEGs) analysis was performed between the Mg-CS/HA group versus control group for HUVECs and rBMSCs, respectively. The volcano plot showed that there were significant differences in the expression of many genes between the two groups for both cell types, thus indicating that the ionic microenvironment could markedly affect the physiological behavior of HU-VECs and rBMSCs ( Figure S14, Supporting Information). Moreover, GO enrichment analysis revealed that the molecular function of "metal ion binding" participated in the functional regulation process of ionic microenvironment on both HUVECs and rBMSCs (Figure 7a,d). According to some studies, Si 4+ and Mg 2+ ions could activate intracellular related signaling pathways by binding with cell-surface ion channel-linked receptors, ultimately promoting the vascularization of ECs and osteogenic differentiation of BMSCs, which is in accordance with our findings. [48,50] Kyoto encyclopedia of genes and genomes (KEGG) analysis was further performed to explore the potential signal pathways involved in osteogenic and angiogenic processes after endocytosis of Si 4+ and Mg 2+ . DEGs were enriched in angiogenesis or osteogenesis-related signaling pathways including "Autophagy-animal," "Mitophagy-animal," "Metabolic pathways," and "MAPK signaling pathway," which were listed in Figure 7b,e, indicating that Si 4+ and Mg 2+ might play a vital  role in the metabolic process and extracellular environment of angiogenesis-osteogenesis effect by controlling molecular degradation and organelle renewal. Notably, the heatmap showed significant up-regulation of genes related to osteogenesis (such as IL6 and ATF3) in HUVECs and genes related to angiogenesis (such as Hif1a and Vegfa) in rBMSCs after culture with the Mg-CS/HA extract (Figure 7c,f). It was reported that Mg 2+ could promote the expression of HIF-1 and VEGF in BMSCs, which could facilitate angiogenesis. [51,52] At the same time, above results showed that Si 4+ and Mg 2+ could stimulate HUVECs to secrete IL6 that could enhance bone regeneration. [53] Hence, these results suggested that IL6, Hif1a and Vegfa might be promising targets for ionic microenvironment regulated osteogenesisangiogenesis coupling.

Mg-CS/HA Microscaffolds Promote Vascularized Bone Regeneration In Vivo
To confirm vascularized bone regeneration, different microscaffolds were implanted into the rat skull critical bone defect model (Figure 8a). Angiogenesis during the early stage of implantation (4 week) was evaluated by gross observations (Figure 8d (Figure 8f). Therefore, the Mg-CS/HA microscaffold could promote bone defect repair by facilitating vascular reconstruction at early stages of healing after implantation.

Conclusion
In this study we fabricated Mg-CS/HA microscaffold. This Mg-CS/HA microscaffold synergized the bioactive effects of Mg/Si ions and enabled controlled degradation of the HA-based microscaffold to achieve osteogenesis-angiogenesis coupling for enhanced vascularized bone regeneration. The synergistic effects of cooperative ion doping and incorporation of CS within the HA microscaffold display a complementary effect on the ion release rate and microscaffold degradation. The optimal degradation rate of the Mg-CS/HA that led to simultaneous release of Si and Mg ions facilitated the early phase vascularization of HU-VECs and osteogenic differentiation of rBMSCs. The synergistic combination of Mg and Si ion release from the Mg-CS/HA microscaffold could promote bone defect repair by facilitating vascular reconstruction during the early stages of bone healing after scaffold implantation. Genes related to osteogenesis (such as IL6 and ATF3) in HUVECs and genes related to angiogenesis (such as Hif1a and Vegfa) in rBMSCs were up-regulated in the Si 4+ and Mg 2+ ionic microenvironment within the vicinity of the implanted Mg-CS/HA microscaffold, which validated enhancement of the osteogenesis-angiogenesis coupling effect. Hence, the regulated degradation of the Mg-CS/HA microscaffold with cooperative Mg/Si ion release, demonstrated much promising potential for the application of HA-based materials in bone repair treatment through promotion of early vascularized bone regeneration.

Experimental Section
Fabrication of HA, CS/HA, and Mg-CS/HA Microscaffolds: The nHA dispersion was prepared by co-precipitation method and then liquidsealed in ethanol. Briefly, (NH 4 ) 2 HPO 4 (4.25 g, 32 mmol) (chemical grade, Sigma, USA) and Ca(NO 3 ) 2 ·4H 2 O (13.8 g, 58 mmol) (chemical grade, Sigma, USA) were dissolved separately in deionized water (50 mL), then (NH 4 ) 2 HPO 4 solution was added slowly to Ca(NO 3 ) 2 ·4H 2 O solution at 90°C, with adjustment of the pH value to 10 with NH 3 ·H 2 O (50 mL). The reaction was allowed to continue for 180 min, and was then centrifugally washed to obtain nHA. The CS sol-gel was obtained by mixing Ca(NO 3 ) 2 ·4H 2 O (23.61 g, 100 mmol) with hydrolyzed TEOS (20.83 g, 100 mmol). The Mg ion-doped CS sol-gel was then obtained by further mixing Mg(NO 3 ) 2 ·6H 2 O (2.564 g, 10 mmol) with CS sol-gel, so as to replace part of the Ca(NO 3 ) 2 in the sol-gel. Briefly, the suspension nHA, CS/HA, as well as Mg-CS/HA were agitated at 37°C for 1 day under continuous stirring. Then, a commercially available polyurethane foam (pore size 300-500 μm) was used as the template, which was impregnated with the prepared sol-gel solution, squeezed to remove redundant liquid, dried in an 80°C oven for 10 min, and then retrieved. To thicken the coating, this procedure was repeated three times. After the sol-gel soaked templates were further dried in the 80°C oven for 24 h, these were heated at 5°C min −1 to 1100°C, and then sintered for 6 h to remove the polymeric template, to obtain the HA, CS/HA and Mg-CS/HA microscaffolds.
Characterization of the HA, CS/HA, and Mg-CS/HA Microscaffolds: The morphology and structure of the microscaffolds were characterized by scanning electron microscopy (SEM, S-250, UK), Energy dispersive spec-troscopy (EDS, Inca X-Max, UK), X-ray diffraction analysis (XRD, Rigaku D/max 2500 VB2+/PC, Japan) and X-ray photoelectron spectrometry (XPS, Thermo Kalpha, USA). The atomic force microscopy (AFM, Bruker Dimension ICON, Germany) was used to determine the Young's modulus of the microscaffolds. The SBF immersion test was carried out to investigate the bioactivity and the ion release profiles. The release of Si, Mg, Ca, and P ions from the HA, CS/HA, and Mg-CS/HA microscaffolds were determined by an inductively coupled plasma optical emission spectrometer (ICP-OES, PerkinElmer 8300, USA). 10 mg of each microscaffold was immersed in 15 mL of 1.5× SBF for 1, 3, 7, 14, 21, and 28 days at 37°C. Afterward, the solution was centrifuged and filtered, and the ionic dissolution products (IDPs) were measured by ICP-OES. After immersion, the samples were examined by SEM and XRD.
Preparation of Microscaffold Extracts: The extracts of HA, CS/HA and Mg-CS/HA microscaffolds were prepared according to the international organization for standardization protocol (ISO 10993-12). Briefly, a stock solution of 200 mg mL −1 was first prepared by adding the granules with sizes of 0.25-1.0 mm into the culture medium. After incubation at 37°C for 24 h, the mixture was then centrifuged and the supernatant was collected. Subsequently, the extract was sterilized by filtration through 0.2 mm filter membranes and stocked for further experiments. To investigate the effects of the proper ion concentration on cells, serially diluted extracts of 1/2, 1/4, 1/8, 1/16, 1/32, 1/64, and 1/128 concentrations were prepared by diluting the stock solution with serum-free culture medium.
Cell Culture: The rBMSCs and HUVECs used for the in vitro study, were purchased from ScienCell Research Laboratories. The rBMSCs were supplemented with 10% (v/v) fetal bovine serum (FBS, Thermo Scientific, USA) and 1% (v/v) penicillin/streptomycin (P/S, Life Technologies, USA). The HUVECs were cultured in endothelial cell medium (ECM, ScienCell, USA) with 5% (v/v) FBS and 1% (v/v) endothelial cell growth supplement/heparin kit (ECGS/H, Promocell, Germany). All the cells were incubated at 37°C within an atmosphere of 95% humidity and 5% CO 2 . Upon reaching 90% confluence, the cells were passaged with 0.25% (w/v) trypsin with EDTA. The third to eighth passage cells were used in this study.
Cell Viability Assay: The proliferation of rBMSCs was evaluated using Cell Counting kit-8 (CCK-8, Dojindo, Japan) following the manufacturer's instructions. Specifically, the rBMSCs were seeded at a density of 2 × 10 3 cells per well in a 96-well plate. After 24 h, the culture medium was replaced by the diluted extracts. The medium without diluted extracts served as the control group. Then, each well was incubated with 10% (v/v) CCK-8 solution at 37°C for 1 h. Finally, the optical density (OD) was measured at a wavelength of 450 nm using a microplate reader (Thermo, USA).
A live/dead cell assay was conducted on day 1, 2, and 3. The rBM-SCs were washed in phosphate-buffered saline (PBS) and incubated for 30 min at 37°C with 2 mm calcein-AM (Sigma, USA) and 2 mm propidium iodide (PI, Sigma, USA) in PBS. After incubation, the samples were rinsed and viewed using a confocal laser scanning microscope (CLSM, TCS-SP8, Leica, Germany) under fluorescence excitation wavelengths of 488 and 562 nm, respectively.
An Annexin V-fluorescein isothiocyanate (FITC)/propidium iodide (PI) apoptosis detection kit (Solarbio, China) was utilized for the detection of rBMSCs apoptosis. Briefly, rBMSCs were seeded into 6-well plates at a density of 1 × 10 6 cells per well. After 24 h, the culture medium was replaced by 1/8 dilutions of HA, CS/HA, and Mg-CS/HA extracts. After another 24 h, the cells were harvested and washed with the binding buffer. Then, the cells were suspended in 200 μL of binding buffer at a cell density of 2 × 10 5 cells mL −1 , incubated with Annexin V-FITC (10 μg mL −1 , 5 μL) and incubated at 25°C for 15 min in dark. At the end of the incubation, rBM-SCs were washed with the binding buffer and centrifuged at 1000 rpm for 5 min, followed by suspension in 200 μL of binding buffer with 10 μL PI (20 μg mL −1 ). Then samples were put through a flow cytometer (Beckman, USA), with the collected data being analyzed by the FlowJo 7.6.1 software (FlowJo LLC, USA).
ALP Activity Assay: The rBMSCs were plated into 24-well plates at a density of 8 × 10 4 cells per well and cultured in osteogenic medium (Ori-Cell Osteogenesis Differentiation Kit RAXMX-90021, China) with 1/8 dilutions of HA, CS/HA and Mg-CS/HA extracts, as described above. Cells in the culture medium without extract served as the control group. On day 3 and 7, the cells were rinsed three times with PBS for ALP staining via a BCIP/NBT ALP color development kit (Beyotime, China). The ALP activity of rBMSCs was assessed using an Alkaline Phosphatase assay kit (Beyotime, China) according to the manufacturer's protocol. The total protein concentration was determined via a BCA Protein Assay Kit (Beyotime, China). The absorbance was measured at a wavelength of 405 nm, and values of ALP activity were read off a standard curve based on standard samples provided by the kit.
Alizarin Red Staining for Mineralization: The rBMSCs were plated into 6-well plates and cultured in osteogenic medium (OriCell Osteogenesis Differentiation Kit RAXMX-90021, China) with 1/8 dilutions of HA, CS/HA and Mg-CS/HA extracts. Cells in the culture medium without extract served as the control group. On day 21, the cells were fixed in 4% (w/v) paraformaldehyde for 15 min, and then stained in 1% (w/v) alizarin red s (pH 4.1-4.5, Sigma, USA) for 2 h at room temperature. After washing the cells in distilled water three times, images were captured. For quantitative analysis, the alizarin red stain on the specimen was dissolved in 10% (w/v) cetylpyridinium chloride (Sigma, USA) in 10 mm sodium phosphate to measure the absorbance values at 620 nm.
Attachment and Spreading of rBMSCs: The rBMSCs (5 × 10 3 cells per well) were plated into 6-well plates and cultured in HA, CS/HA, and Mg-CS/HA extracts. Cells in the culture medium without extract served as the control group. After 12 h, the attached cells were fixed in 2.5% (w/v) glutaraldehyde and serially dehydrated with an increasing ethanol gradient, air-dried within a hood, and sputtered with gold prior to imaging under SEM.
Immunofluorescent Staining: Briefly, the cells were seeded at a density of 4 × 10 3 cells per well in 24-well plates. After culturing for 24 h, the culture media were replaced by HA, CS/HA, and Mg-CS/HA extracts, respectively. Cells in the culture medium without extract served as the control group. The samples were harvested and then fixed in 4% (w/v) paraformaldehyde for 15 min and washed with PBS. Then the cells were permeabilized with 0.5% (w/v) Triton X-100 for 10 min followed by washing in PBS. After that, the cells were blocked in bovine serum albumin (BSA) for 1 h at room temperature, and incubated with the primary antibody overnight at 4°C. After being rinsed three times with PBS, the cells were incubated with the secondary antibody for 1 h. Finally, the cells were treated with DAPI (Beyotime, China). Images were captured by CLSM (TCS-SP8, Leica, Germany). The primary antibodies were Vinculin (ab129002, Abcam, UK), RUNX2 (ab236639, Abcam, UK), and VEGF (ab234110, Abcam, UK). The secondary antibody was IgG H&L (ab150077, Abcam, UK).
Quantitative Real-Time PCR Analysis: The rBMSCs with a density of 1 × 10 5 cells per well were plated in 6-well plates. After 24 h, the culture medium was replaced by osteogenic medium (OriCell Osteogenesis Differentiation Kit RAXMX-90021, China) with 1/8 dilutions of HA, CS/HA, and Mg-CS/HA extracts. Cells in the culture medium without extract served as the control group. After being cultured for 3 and 7 days, the gene expression levels of Runx2, Bmp2, Col1, and Opn were analyzed by real-time quantitative polymerase chain reaction (RT-qPCR). Typically, total RNA was extracted from each sample using Trizol reagent (Invitrogen, USA) following the manufacturer's instructions. The concentrations of the isolated RNA were determined at 260 nm using an ultramicrospectrophotometer (NanoDrop2000, USA). The isolated RNA was then reverse transcribed into cDNA by using the reverse transcription reagents kit (Takara, Japan). The RT-qPCR was carried out with a Maxima SYBR Green/ROX qPCR kit (Thermo, USA) using a Quantstudio 6 Flex thermocycler machine (Thermo, USA). The gene expression levels were calculated by the 2 −ΔΔCt method. The sequences of primers for Runx2, Bmp2, Col1, and Opn genes are listed in Table S1, Supporting Information. The relative expression lev-els of the genes of interest were normalized against the housekeeping gene GAPDH.
Western Blot Analysis: The rBMSCs were plated in 6-well plates. After 24 h, the culture medium was replaced by osteogenic medium (Ori-Cell Osteogenesis Differentiation Kit RAXMX-90021, China) with 1/8 dilutions of HA, CS/HA, and Mg-CS/HA extracts. Cells in the culture medium without extract served as the control group. Total protein was extracted by RIPA lysis buffer (Beyotime, China), containing protease and phosphatase inhibitors (Sigma-Aldrich, USA), for 30 min at 4°C. The cell lysates were then removed by centrifugation, and the protein concentration was determined using the bicinchoninic acid quantification kit (Beyotime, China). Then, 30 mg of sample was subjected to electrophoresis with 10% (w/v) SDS-PAGE gels (Beyotime, China), and subsequently transferred to a polyvinylidene difluoride membrane (Millipore, USA). The membranes were blocked with 5% (w/v) BSA for 1 h at room temperature and incubated overnight at 4°C with primary antibodies. After that, the membranes were then incubated with the secondary antibody. Finally, the membranes were visualized with enhanced chemiluminescence reagent (Beyotime, China). The band intensities were quantified using Image Lab software (Bio-Rad, USA). The primary antibodies were anti-GAPDH (ab8245, Abcam, UK), Anti-Osteopontin (ab63856, Abcam, UK), Anti-BMP2 (ab284387, Abcam, UK), Anti-Collagen I (ab270993, Abcam, UK), and Anti-RUNX2 (ab236639, Abcam, UK). The secondary antibody was HRP-labeled IgG (ab97051, Abcam, UK). GAPDH was used as the protein loading control. The protein expression levels were normalized to GAPDH.
Tube Formation Assays: Matrigel was pipetted onto pre-chilled 24-well plates (250 μL Matrigel per well) and left to solidify for 60 min at 37°C. Next, HUVECs were seeded onto the Matrigel coated plates at 2 × 10 4 cells per well suspended within culture medium containing 1/8 dilutions of HA, CS/HA and Mg-CS/HA extracts. After the 24-well plates were incubated at 37°C in a 5% CO 2 air incubator for 4 and 8 h, tubular structures were photographed, and the total tubule lengths, number of tubules and number of junctions were quantified and recorded. Three to four photographs from different areas were captured for the analysis of each sample. Triplicate samples (n = 3) were tested for each condition.
Cell Migration Assay: HUVECs (2 × 10 5 cells mL −1 ) were loaded into the upper chamber of a 24-well transwell plate (pore size, 8 μm, Corning, USA) after resuspension within the HA, CS/HA, and Mg-CS/HA extracts respectively, while 800 μL of ECM was added to the lower chamber. After 3 and 6 h of culture, cells on the upper surface of the chamber were removed with a cotton swab. Cells that had migrated to the lower surface of the chamber were fixed with 4% (w/v) paraformaldehyde, and stained for 15 min with 0.5% (w/v) crystal violet and observed by a phase contrast microscope (CKX 41, Olympus, Japan).
Wound Healing Assay: The in vitro wound healing ability of HUVECs was evaluated by using Ibidi μ-Dishs (35 mm) with 2-well silicon inserts (80 206, Ibidi, Germany). Briefly, HUVECs (5 × 10 5 cells mL −1 ) were seeded in each well of the inserts. After 24 h, the culture medium was replaced with HA, CS/HA, and Mg-CS/HA extracts, and then the silicon inserts were removed to create scratches with a width of 500 μm. Imaging of the scratch was carried out with a phase contrast microscope (CKX 41, Olympus, Japan). The cultures were further cultivated for 8 h, followed by a second round of observation under the phase contrast microscope. For each image, the size of the scratch area at 8 h was analyzed using ImageJ software. The percentage wound closure (%) was calculated as (A 0 −A n )/A 0 × 100%, where A 0 and A n represent the initial wound area and the residual wound area at the metering point, respectively.
Indirect Co-Cultures of rBMSCs and HUVECs: The rBMSCs (or HU-VECs) were cultured in 1/8 dilutions of HA, CS/HA and Mg-CS/HA extracts for 72 h. After centrifugation, the supernatant was collected and used as BMSC-CM (or HUVEC-CM). Schematic diagrams showing the coculture system of rBMSCs and HUVECs were presented in Figure 6a. The osteogenic differentiation of the rBMSCs in HUVEC-CM was assessed by determining the ALP activity, and the angiogenic property of the HUVECs in BMSC-CM was evaluated by the tube formation assay.
RNA Sequencing: The HUVECs and rBMSCs were incubated with Mg-CS/HA extract at the optimal concentration (1/8 dilution) for 3 and 7 days, respectively. Cells in culture medium without extract were used as the control group. The total RNA was extracted from cells using Trizol reagent. 1 μg of total RNA was used for library preparation. The poly(A) mRNA isolation was performed using Oligo(dT) beads. The mRNA fragmentation was performed using divalent cations and high temperature. Priming was performed using Random Primers. First strand cDNA and the second-strand cDNA were synthesized. Each sample was then amplified by PCR using P5 and P7 primers and the PCR products were validated. Then libraries with different indexes were multiplexed and loaded on an Illumina HiSeq/Illumina Novaseq/MGI2000 instrument for sequencing using a 2× 150 paired-end (PE) configuration according to the manufacturer's instructions.
Animals and Surgical Procedures: The animal surgical procedure was approved by the Institutional Animal Care and Use Committee of the Peking University (Approval number: LA2022611). The early in vivo bone forming ability of the microscaffolds were investigated in a rat cranial defect model (Sprague-Dawley, 6-week-old male). Pre-operatively, a total of sixty rats were checked for general health, weighed and randomized into four groups (control, HA, CS/HA and Mg-CS/HA). For establishing the cranial defect model, the rats were intraperitoneally anesthetized with phenobarbitol sodium (100 mg kg −1 ) and the cranial surface was exposed by a skin incision and elevation of the periosteum. Two full-thickness criticalsized bone defects in a circular form (5 mm diameter) were generated bilaterally using a trephine drill flushed with a saline solution. Bilateral bone defects were filled with the same type of microscaffold. The experimental groups were covered with collagen membranes (Heal-all, ZH-Bio, China) after implanting with HA, CS/HA, and Mg-CS/HA microscaffolds, and the defects were only covered with collagen membrane similar to the control group. The whole calvarias were then harvested for evaluation after 4 weeks implantation.
Micro-CT Analysis and Histological Evaluation: The Micro-CT (IN-VEON, SIEMENS, Germany) analysis was carried out at a voltage of 65 kV and an electric current of 67 mA. Briefly, the specimens were scanned with a thickness of 0.018 mm per slice in medium-resolution mode, with a 1024-reconstruction matrix, and 200 ms integration time. After reconstruction of the scanned images, the BV/TV and BMD values in the defect regions of the 4 weeks groups were used to evaluate new bone formation using the analysis software. Following the Micro-CT scan, the samples were decalcified in 10%(w/v) EDTA for 4 weeks and then embedded in paraffin parallel to the sectioned surface. Serial cross-sections of the decalcified samples were cut and processed for H&E and Masson's trichrome staining according to the manufacturer's instructions. Additionally, immunohistochemical staining was performed for detection of OCN (Servicebio, China) and CD31 (Proteintech, USA) expression within the samples.
Vascular Perfusion and Micro-CT Scanning: Laser Doppler imaging (LDI, Perimed, Sweden) was used to evaluate living vascular perfusion at 4 weeks after implantation. For evaluation of angiogenesis in the defect areas, rats in each group were perfused with Microfil (Flowtech, USA) at 4 weeks after implantation. In brief, after anesthesia, the rib cages of the rats were opened. Then, the rats were successively perfused with heparinized saline, 4% (w/v) paraformaldehyde, and 20 mL Microfil through the left ventricle. Finally, the rats were kept overnight at 4°C. After being photographed, the specimens were decalcified in decalcifying solution for 4 weeks, and then scanned with the Micro-CT. New blood vessel formation in the cranial defects was presented as 3D reconstructed images.
Statistical Analysis: All of the experiments were performed in triplicates and the quantitative data were presented as mean ± standard deviation (SD). Statistical analysis was conducted using SPSS 22.0 software (IBM Corp. Armonk, USA). One-way analysis of variance (ANOVA) and post-hoc LSD testing was used to analyze significant differences between groups. A value of p < 0.05 was considered statistically significant.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.