MFG-E8, A Novel Target of Promoting Osteogenic Differentiation of Human Bone Marrow Mesenchymal Stem Cells

Background Fracture nonunion and bone defects are challenging for orthopedic surgeons. Milk fat globule-epidermal growth factor 8 (MFG-E8), a glycoprotein possibly secreted by macrophages in a fracture hematoma, participates in bone development. However, the role of MFG-E8 in the osteogenic differentiation of bone marrow mesenchymal stem cells (BMSCs) is unclear. Methods We investigated the osteogenic effect of MFG-E8 in vitro and in vivo. The CCK-8 assay was used to assess the effect of recombinant human MFG-E8 (rhMFG-E8) on the viability of human BMSCs (hBMSCs). Osteogenesis was investigated using real-time quantitative PCR, Western blotting, and immunouorescence. Alkaline phosphatase (ALP) and Alizarin red staining were used to evaluate ALP activity and mineralization, respectively. An enzyme-linked immunosorbent assay was conducted to evaluate the secretory MFG-E8 concentration. Knockdown and overexpression of MFG-E8 in hBMSCs were established via siRNA and lentivirus vector transfection, respectively. Exogenous rhMFG-E8 was used to verify the in vivo therapeutic effect in a tibia bone-defect model based on radiographic analysis and histological evaluation. signaling inhibitor. We evaluated the canonical Wnt/β-catenin signaling pathway, which is vital in the differentiation of BMSCs into osteoblasts. Related proteins, including GSK3β and β-catenin, were measured using Western blotting on day 1 of osteogenesis. The expression of p-GSK3β and active β-catenin (a-β-catenin) was increased by MFG-E8, whereas the expression of total GSK3β and total β-catenin (t-β-catenin) was unaffected 5A&B). Furthermore, the ratio of a-β-catenin to t-β-catenin was increased signicantly by GA 5B). Immunouorescence showed that the expression of a-β-catenin in the nucleus and cytoplasm was signicantly increased by MFG-E8 5C). Therefore, MFG-E8 promotes the osteogenic differentiation of hBMSCs via the GSK3β/β-catenin signaling pathway.


Introduction
The healing of bone defects or fracture nonunion caused by high-energy injury, infection, tumor resection, and fracture are challenging for orthopedic surgeons [1][2][3]. The nonunion rate of all fractures ranges from 1.9-10% [4,5]. Fracture healing is a complex and continuous process that includes hematoma formation and in ammatory responses as well as involves intracellular and extracellular signaling pathways [6,7]. The local hematoma formed after a fracture contains a variety of cells and cytokines and constitutes the fracture healing microenvironment, in which stem cells with osteogenic differentiation potential and cytokines affect the prognosis of the fracture [8,9]. The fracture hematoma plays an important role in fracture healing [10]. However, the mechanism by which the fracture hematoma affects the osteogenic differentiation of bone marrow mesenchymal stem cells (BMSCs) is unclear.
Milk fat globule-epidermal growth factor 8 (MFG-E8), also known as lactadherin, is a secreted glycoprotein with a molecular weight of 46 kDa [11]. MFG-E8 is expressed in epithelial cells, macrophages, dendritic cells, and broblasts [12], It mediates the clearance of apoptotic cells [13] and the maintenance and repair of intestinal epithelial function [14], exerts an anti-in ammatory effect [15,16], and regulates angiogenesis [17,18] In the musculoskeletal system, osteoblasts and monocyte-derived osteoclasts express MFG-E8 [19,20]. MFG-E8-de cient mice exhibited increased RANKL-induced osteoclastogenesis ex vivo [21,22]. Sinningen et al. also reported that MFG-E8-de cient mice exhibited impaired bone formation and mineralization compared with wild-type animals [23]. However, the role of MFG-E8 in the osteogenic differentiation of BMSCs is unclear. We investigated the effect of MFG-E8 on the osteogenic differentiation of human BMSCs (hBMSCs) and found that MFG-E8 promoted their osteogenic differentiation by regulating the GSK3β/β-catenin signaling pathway in vitro and in vivo.

Cell Culture, Regents and Antibodies
The hBMSCs provided by Cyagen Biosciences (HUXMA-01001; Guangzhou, China) had the potential to differentiate into osteoblasts, chondrocytes, and adipocytes under appropriate conditions. Adherent hBMSCs were cultured in asks with hBMSC growth medium (Cyagen Biosciences, Guangzhou, China) in an incubator at 37°C under 5% CO2 and were passaged after reaching 80% con uence. The medium was replaced every 3 days; cells from passages 3-5 was used in subsequent experiments.
Small Interfering RNA (siRNA) transfection targeting MFG-E8 To knock down the expression of MFG-E8 in hBMSCs, small interfering RNA (siRNA) transfection was performed. siRNAs for the human MFG-E8 gene were purchased from GenePharma Inc. (Shanghai, China). The sequences were as follows: siRNA 1, GGUUUAUGCGAGGAGAUUUTT; siRNA 2, GCCUUAAUGGACACGAAUUTT; siRNA 3, CCCACAAGAAGAACUUGUUTT. hBMSCs were cultured in six-well plates for 18 h prior to siRNA transfection. The medium was replaced with Opti-MEMTM I Reduced Serum Medium (Thermo Fisher Technology Co., Ltd., China) with 20 nM targeting siRNA or negative control using Lipo6000™ transfection reagent (Beyotime Biotechnology, Shanghai, China). After culturing for 6 h at 37°C under 5% CO2, the medium was exchanged for fresh hBMSC growth medium. The MFG-E8 mRNA and protein levels were determined using real-time PCR (RT-PCR) and Western blotting.
Approximately 50-60% con uent hBMSCs were incubated with lentiviral particles and 5 µg/mL polybrene in hBMSC growth medium at a multiplicity of infection of 50 (the optimum according to GFP expression after lentiviral GFP particle infection). For infection, hBMSCs were incubated with lentiviral particles and polybrene (5 µg/mL) in hBMSC growth medium. After about 18 h, the infection medium was exchanged for fresh growth medium. After 3 days, the cells were screened using puromycin (4 µg/mL) and passaged for use in subsequent experiments. The expression of MFG-E8 was detected using RT-PCR, Western blot, and immuno uorescence.
ALP staining and ALP activity assay ALP staining was used to investigate early mineralization. For ALP staining, cells were xed with 4% paraformaldehyde (Sangon Biotech, Shanghai, China) for 30 min. The cells were then washed with double distilled water (ddH2O) three times and stained using an Alkaline Phosphatase Color Development kit (Beyotime, Shanghai, China). ALP activity was determined using an ALP Activity Assay kit (Beyotime) according to the manufacturer's instructions. Brie y, cells were lysed for 1 h with radioimmunoprecipitation assay (RIPA) buffer. The appropriate amount of supernatant and 50 µL of chromogenic substrate (para-nitrophenyl phosphate) were added to wells of a 96-well plate, to which testing buffer was also added to a volume of 100 µL. We prepared standard samples (para-nitrophenol 0.5 mM) to generate an ALP standard curve. Next, the 96-well plate was incubated at 37°C for 5-10 min. Finally, to each well was added 100µL of reaction termination solution to stop the reaction, and the A405 was measured using a microplate reader.
Alizarin red staining and quanti cation assay Alizarin red staining (ARS; Cyagen Biosciences) was performed to assess late mineralization. For ARS, cells were xed in 4% paraformaldehyde for 20 min at room temperature and subsequently washed three times with ddH2O. Finally, the cells were treated with Alizarin red stain (0.5%, pH 4.1-4.2) for 20 min and rinsed with distilled water. To quantify the staining intensity, stained mineralized nodules were incubated with 10% cetylpyridinium chloride (Sigma, Shanghai, China), the solution was collected, and the A570 was measured using a microplate reader.
Total RNA was isolated from cells cultured with OIM using RNAiso reagent (TaKaRa Bio Inc., Dalian, China) and quanti ed by measuring the A260 (NanoDrop 2000; Thermo Fisher Scienti c, Waltham, MA). First-strand cDNA was synthesized using PrimeScript RT Master Mix (TaKaRa Bio Inc.) according to the manufacturer's instructions. Total RNA (≤ 1,000 ng) was reverse-transcribed into cDNA in a reaction volume of 20 µL using a Double-Strand cDNA Synthesis kit (TaKaRa Bio Inc.). The levels of mRNAs encoding COL1A1, RUNX2, OCN, Osterix, OPN, ALP, and GAPDH were determined using the StepOnePlus Real-Time PCR System (Applied Biosystems Inc., Warrington, UK) and SYBR Premix Ex Taq (TaKaRa Bio Inc.) with the following program: 95°C for 30 s followed by 40 cycles of 95°C for 5 s and 60°C for 30 s. GAPDH was used as an internal control and for normalization. DNA concentrations were calculated using the 2 − ΔΔCt method24. [24] The primers were synthesized by Sangon Biotech and are listed in Table 1. ELISA MFG-E8 levels in the hBMSC culture supernatants were assessed via enzyme-linked immunosorbent assay (ELISA) using the Human MFGE ELISA Kit (Boster Biological Technology) following the manufacturer's protocol.

In vivo evaluation in animals
All animal experiments and procedures were conducted in accordance with the principles of the Institutional Animal Care Use Committee of the Second A liated Hospital of Zhejiang University and approved by the same committee. A rat tibial-defect model was used to assess the bone-forming ability of MFG-E8. [25,26]. Fifteen rats were divided randomly into three groups: the blank group, normal saline (NS) (negative control group treated with NS) group, and rhMFG-E8 group. First, rats were anesthetized via inhalation of 2-5% iso urane, with the anesthesia maintained via 2% iso urane inhalation during surgery. After anesthesia, an incision was made lateral to the tibia, away from the bone. An intramedullary xation pin (1.2-mm-diameter stainless-steel syringe needle) was inserted inside the medullary canal of the tibia for xation. Osteotomy to create a transverse 1.5-mm-wide defect approximately 7 mm from the proximal tibial growth plate was performed using an electronic saw. The same leg was used in each group. Next, rhMFG-E8 (20 µg) was injected locally at the fracture site on days 0, 3, 5, 8, and 11 (i.e., immediately and at 72-h intervals thereafter); NS was used as a vehicle control. Rats were sacri ced at 1 month after surgery, and samples were collected and xed in 4% paraformaldehyde for 72 h at room temperature.

Radiographic analysis
A range of 3 mm above and below the bone-defect area of tibia samples was scanned using the µCT-100 Imaging System (Scanco Medical, Brüttisellen, Switzerland) with the following parameters: 70 kVp; reconstruction matrix, 1,024; slice thickness, 14.8 µm; and exposure time, 300 ms. The trabecular bone volume fraction (BV/TV), mean trabecular thickness (Tb.Th), mean trabecular number (Tb.N), and mean trabecular separation (Tb.Sp) were evaluated via standard three-dimensional microstructural analysis [25,27].

Histological evaluation
After micro-computed tomography (CT), samples were decalci ed using 10% ethylene diamine tetra acetic acid (Sigma) in 0.1 M phosphate-buffered saline, with the solution changed once a week for 6 weeks, before embedding in para n. Serial sections of 3 µm thickness were cut and mounted on polylysine-coated slides and depara nized. The consecutive tissue sections were stained with hematoxylin and eosin (HE) or Masson's trichrome stain. [25,28] Images were obtained using a microscope (Leica DM4000B; Leica, Wetzlar, Germany).

Data and statistical analysis
Statistical analysis was performed using Prism (version 8.0; GraphPad Software, San Diego, CA). All experiments were conducted at least three times and the data are presented as the means ± SDs. Differences between two groups were analyzed using the two-tailed Student's t-test. For comparisons of more than two groups, one-way analysis of variance followed by Bonferroni post hoc tests was used. In all analyses, P < 0.05 was taken to indicate statistical signi cance.

Role of the funding source
The funding agencies had no further role in study design, in the collection, analysis and interpretation of data, in the writing of the report and in the decision to submit the paper for publication.

Results
Endogenous MFG-E8 expression increased signi cantly during osteogenic differentiation of hBMSCs To investigate the association between endogenous expression of MFG-E8 and osteogenic differentiation, we assayed MFG-E8 expression in undifferentiated and differentiated hBMSCs using RT-PCR, Western blotting, and ELISA. Compared to levels in undifferentiated hBMSCs, MFG-E8 mRNA and protein levels increased signi cantly after osteogenic differentiation on days 1, 2, and 3. (Fig. 1A&B) The secretory MFG-E8 protein level also increased signi cantly after osteogenic differentiation on days 1, 2, and 3 compared to undifferentiated hBMSCs (Fig. 1C). Therefore, endogenous MFG-E8 expression increased signi cantly during osteogenic differentiation of hBMSCs.

Knockdown of MFG-E8 inhibited the osteogenic differentiation of hBMSCs and decreased calcium deposition
To investigate the role of MFG-E8 in osteogenesis, three siRNAs were used to knock down MFG-E8 in hBMSCs. The e ciency of MFG-E8 knockdown was measured using RT-PCR. siRNA 1 and siRNA 2 resulted in signi cant gene knockdown, particularly siRNA2 ( Fig. 2A). siRNA 2 also signi cantly decreased the MFG-E8 protein level (Fig. 2B) and so was used in subsequent experiments.
To explore the effect of MFG-E8 knockdown on the osteogenesis of hBMSCs, osteogenesis-related genes and proteins were investigated using RT-PCR and Western blotting. The expression of osteogenesisrelated genes (RUNX2, Osterix, OCN, and ALP) was signi cantly decreased compared with the NC group after 1 and 3 days of osteogenesis (Fig. 2C&D). The protein levels of RUNX2 and COL1A1 were decreased in the MFG-E8 siRNA group on days 1 and 3 of osteogenesis (Fig. 2E&F). Moreover, signi cantly decreased ALP activity and mineral deposition were detected in the MFG-E8 siRNA group based on ALP staining and ARS (Fig. 2G).
MFG-E8 overexpression increased the levels of osteogenesis-related proteins and enhanced ALP activity and calcium deposition To explore the role of endogenous MFG-E8 in the osteogenic differentiation of hBMSCs, MFG-E8 was overexpressed using lentiviral vectors. The e ciency of MFG-E8 overexpression was measured using RT-PCR, Western blotting, and immuno uorescence. The MFG-E8 mRNA and protein levels increased signi cantly (Fig. 3A&B), indicating successful MFG-E8 overexpression in hBMSCs.
The effect of endogenous MFG-E8 overexpression on osteogenesis was investigated. Compared to the OE-NC group (MFG-E8 overexpression control), the expression of osteogenesis-related genes such as RUNX2, COL1A1, Osterix, OCN, and ALP was signi cantly elevated in the OE group (MFG-E8 overexpression) on day 1 of osteogenesis (Fig. 3C). Western blotting showed that the RUNX2 and COL1A1 protein levels were increased in the OE group compared with the OE-NC group (Fig. 3D). ALP staining and ARS showed signi cantly increased ALP activity and calcium deposition in the OE group (Fig. 3E).

Exogenous MFG-E8 increased the expression levels of osteogenesis-related genes and proteins and enhanced calcium deposition
To assess the effect of exogenous MFG-E8 on osteogenic differentiation, BMSCs were cultured with OIM with rhMFG-E8 protein (0, 10, 100, and 1000 ng/mL). Cell viability was unaffected by rhMFG-E8 (Fig. S1). The expression of osteogenesis-related genes (RUNX2, COL1A1, Osterix, OPN, and ALP) increased signi cantly as the rhMFG-E8 concentration increased on days 1 and 3 of osteogenesis compared with the control group (Fig. 4A). Additionally, the expression of osteogenesis-related proteins (RUNX2 and COL1A1) was elevated relative to levels in the control group on days 1 and 3 of osteogenesis in a concentration-dependent manner (Fig. 4B). These results were con rmed in the immuno uorescence assays (Fig. 4C).
We investigated the in uence of rhMFG-E8 on early-stage mineralization during osteogenic differentiation. MFG-E8 signi cantly enhanced ALP activity on day 2 in a dose-dependent manner (Fig. 4D&F). Based on ARS, MFG-E8 signi cantly increased calcium deposition and mineralization on day 11 (Fig. 4E&G).

MFG-E8 promotes the osteogenic differentiation of hBMSCs via the GSK3β/β-catenin signaling pathway
We evaluated the canonical Wnt/β-catenin signaling pathway, which is vital in the differentiation of BMSCs into osteoblasts. Related proteins, including GSK3β and β-catenin, were measured using Western blotting on day 1 of osteogenesis. The expression of p-GSK3β and active β-catenin (a-β-catenin) was increased by MFG-E8, whereas the expression of total GSK3β and total β-catenin (t-β-catenin) was unaffected (Fig. 5A&B). Furthermore, the ratio of a-β-catenin to t-β-catenin was increased signi cantly by GA (Fig. 5B). Immuno uorescence showed that the expression of a-β-catenin in the nucleus and cytoplasm was signi cantly increased by MFG-E8 (Fig. 5C). Therefore, MFG-E8 promotes the osteogenic differentiation of hBMSCs via the GSK3β/β-catenin signaling pathway.
Enhanced osteogenic differentiation of hBMSCs caused by MFG-E8 was partially attenuated by a GSK3β/β-catenin signaling inhibitor AR-A014418 is a competitive and selective ATP inhibitor of GSK3β. The effect of AR-A014418 on GSK3β was explored using Western blotting. AR-A014418 at 10 and 20 µM signi cantly decreased p-GSK3β and a-β-catenin expression but did not alter that of total GSK3β and t-β-catenin (Fig. 6A). Therefore, 20 µM AR-A014418 was used in subsequent experiments.
AR-A014418 attenuated the MFG-E8-induced increases in RUNX2 and COL1A1 expression (1000 ng/mL) on day 1 of osteogenesis, as determined by Western blotting (Fig. 6B). Furthermore, the MFG-E8-induced increases in levels of Wnt signaling pathway-related proteins such as p-GSK3β and a-β-catenin were attenuated signi cantly (Fig. 6B). Immuno uorescence showed that the increased expression of RUNX2 and COL1A1 caused by MFG-E8 was decreased signi cantly by AR-A014418 (Fig. 6C). In addition, the enhanced mineralization and ALP activity induced by MFG-E8 were attenuated by AR-A014418, as determined by ARS and the ALP assay (Fig. 6D-G).

MFG-E8 accelerated bone healing in a rat tibial-defect model
To investigate the role of MFG-E8 in vivo, exogenous recombinant MFG-E8 protein was administered to rats with tibial defects. Micro-CT showed that MFG-E8 signi cantly accelerated bone fracture healing compared to the control and NS groups. MFG-E8 treatment signi cantly reduced the gap distance of cortical defects in comparison with the other two groups (Fig. 7A). Compared with the control and NS groups, the MFG-E8 group had signi cantly higher BV/TV, Tb.N, and Tb.Th values and a lower Tb.Sp value (Fig. 7B). Histological analysis using HE and Masson's trichrome staining demonstrated better cortical growth in the MFG-E8 group compared with the other two groups (Fig. 7C).

Discussion
To our knowledge, this is the rst report demonstrating the promotion of osteogenic differentiation of hBMSCs by MFG-E8 through the GSK3β/β-catenin signaling pathway. The levels of endogenous and secretory MFG-E8 increased signi cantly during osteogenic differentiation of hBMSCs. Therefore, MFG-E8 may be crucial for osteogenesis. Knockdown of MFG-E8 in hBMSCs inhibited their osteogenic differentiation and calcium deposition. Also, the overexpression of MFG-E8 or addition of exogenous rhMFG-E8 protein signi cantly increased the levels of osteogenesis-related genes and proteins, such as RUNX2, COL1A1, Osterix, ALP, OCN, and OPN, in hBMSCs. Furthermore, MFG-E8 increased ALP activity and calcium deposition, implying the enhancement of early and late osteogenesis. MFG-E8 promoted the osteogenic differentiation of hBMSCs by activating the Wnt/β-catenin pathway. MFG-E8 inhibited GSK3β via phosphorylation, thus increasing the a-β-catenin level and inducing its translocation into the nucleus to activate gene transcription. Therefore, MFG-E8 promotes the osteogenic differentiation of hBMSCs by regulating GSK3β/β-catenin signaling pathway.
Endogenous and secretory MFG-E8 protein levels increased signi cantly during the osteogenic differentiation of hBMSCs. This suggests that MFG-E8 plays an important role in the osteogenic differentiation of hBMSCs. MFG-E8 is expressed ubiquitously in various tissues and cells [12,29], and is universally distributed in mammals [12]. Therefore, the secretion and expression of MFG-E8 likely have complex regulatory effects. In the musculoskeletal system, MFG-E8 is expressed by osteoblasts, monocytes, and stromal cells [19,20]. Abe et al. [22] showed that MFG-E8 is expressed by and regulates osteoclasts, giant multinucleated cells that resorb bone during bone remodeling. The maintenance of bone homeostasis requires tight collaboration between osteoclasts and osteoblasts [30]. We found that when MFG-E8 expression is increased, it regulates the osteogenic differentiation of hBMSCs. The increase in MFG-E8 may be due to a positive feedback regulatory mechanism and may play other important regulatory roles in the osteogenic differentiation of hBMSCs.
MFG-E8 promoted the osteogenic differentiation of hBMSCs by activating the Wnt/β-catenin pathway. The canonical Wnt/β-catenin signaling pathway plays a key role in healing and promotes osteoblast function [31]. Activation of the Wnt pathway leads to the recruitment, phosphorylation, and inactivation of GSK3β, resulting in cytoplasmic β-catenin stabilization and its translocation into the nucleus. This results in the transcription of downstream osteogenesis-related genes and the promotion of osteogenic differentiation of BMSCs [32][33][34]. We found signi cantly increased expression of phosphorylated GSK3β and a-β-catenin. Immuno uorescence revealed a high level of β-catenin in the nucleus. Inactivation (phosphorylation) of GSK3β is an intermediate step after activation of the Wnt/β-catenin pathway [35,36]. GSK3β plays important roles in the growth factor, hedgehog, G-protein coupling ligand, cytokine, and Wnt pathways [37]. The activation or inactivation of GSK3β is involved in multiple signaling pathways, such as that centered on PI3K-AKT. The activation of PI3K/AKT signaling can inhibit GSK3β via its phosphorylation at Ser9. However, MFG-E8 did not signi cantly alter the expression of p-AKT during osteogenesis (Fig. S2.0). Therefore, other molecules or signaling pathways may be involved in the promotion of osteogenic differentiation by MFG-E8 via the GSK3β/β-catenin pathway.
Fracture hematomas were rst reported to play an important role in fracture healing in the rst half of the 20th century. Removal of the hematoma at the fracture site may delay fracture healing, and subcutaneous implantation of the hematoma can lead to heterotopic ossi cation [38,39]. However, their role in the initiation of regeneration is unclear. Little is known about the initial cellular and humoral composition of the fracture hematoma [39,40]. Wray et al. [41] found increased β-and γ-globulin in fracture hematoma supernatant. The increase in β-globulin may be due to hemoglobin, whereas the elevated γ-globulin fraction may indicate the involvement of immune cells. The fracture hematoma microenvironment contains various immune cells and cytokines [42]. The immune system in uences fracture healing while fractures in uence systemic immunity [43,44]. Macrophages are the most important immune cells in osteoimmunology [45]. Laplante et al.[46] found that MFG-E8 secreted by apoptotic endothelial and epithelial cells reprograms macrophages from an M1 (proin ammatory) to an M2 (anti-in ammatory) phenotype and promotes wound healing. Therefore, MFG-E8 may maintain bone homeostasis by regulating in ammation and immunity via various cell types, such as macrophages and lymphocytes.
This study has several limitations. First, we only explored the role of the canonical Wnt/β-catenin signaling pathway in regulating osteogenic differentiation by MFG-E8. Therefore, exploration of the noncanonical Wnt/β-catenin pathway and other potential links between Wnt/β-catenin and other signaling pathways by which MFG-E8 promotes osteogenic differentiation and regulates bone homeostasis is needed. Second, MFG-E8 promoted the osteogenic differentiation of hBMSCs but inhibited RANKLmediated osteoclast differentiation. The relationship between osteoblasts and osteoclasts needs to be investigated further. Third, bone healing involves endochondral ossi cation and intramembranous ossi cation [47]. We investigated intramembranous ossi cation; however, the effect of rhMFG-E8 on endochondral ossi cation is unclear. Future studies using bone-defect models should involve larger bone defects.

Conclusion
MFG-E8 promotes the osteogenic differentiation of hBMSCs by regulating the GSK3β/β-catenin signaling pathway and is thus a novel therapeutic target for bone homeostasis.   staining after 3 and 11 days of osteogenesis. Scar bar, 500μm. All data are expressed as the means ± SD. Reactions were performed in triplicate. *p < 0.05, **p < 0.01 compared to the control group.  measured by RT-PCR and western blot after treatment with various concentration of rhMFG-E8 on day 1 and 3 of osteogenesis. (C) Immuno uorescence staining for COL1A1 and RUNX2 protein after 1 day of osteogenesis. Scale bars, 100 μm. (D) ALP staining and quantitative assay on day 2 of osteogenic differentiation. Scale bar, 500 μm. (E) Mineralization was measured by ARS staining and quantitative assay after 11 days of osteogenesis. Scar bar, 500 μm. All data are expressed as the means ± SD.
Reactions were performed in triplicate. *p < 0.05, **p < 0.01 compared to the control group.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.
Supplemental le.pdf