HIF1α Promotes BMP9-Mediated Osteoblastic Differentiation and Vascularization by Interacting with CBFA1

Bone morphogenetic protein 9 (BMP9) as the most potent osteogenic molecule which initiates the differentiation of stem cells into the osteoblast lineage and regulates angiogenesis, remains unclear how BMP9-regulated angiogenic signaling is coupled to the osteogenic pathway. Hypoxia-inducible factor 1α (HIF1α) is critical for vascularization and osteogenic differentiation and the CBFA1, known as runt-related transcription factor 2 (Runx2) which plays a regulatory role in osteogenesis. This study investigated the combined effect of HIF1α and Runx2 on BMP9-induced osteogenic and angiogenic differentiation of the immortalized mouse embryonic fibroblasts (iMEFs). The effect of HIF1α and Runx2 on the osteogenic and angiogenic differentiation of iMEFs was evaluated. The relationship between HIF1α- and Runx2-mediated angiogenesis during BMP9-regulated osteogenic differentiation of iMEFs was evaluated by ChIP assays. We demonstrated that exogenous expression of HIF1α and Runx2 is coupled to potentiate BMP9-induced osteogenic and angiogenic differentiation both in vitro and animal model. Chromatin immunoprecipitation assays (ChIP) showed that Runx2 is a downstream target of HIF1α that regulates BMP9-mediated osteogenesis and angiogenic differentiation. Our findings reveal that HIF1α immediately regulates Runx2 and may originate an essential regulatory thread to harmonize osteogenic and angiogenic differentiation in iMEFs, and this coupling between HIF1α and Runx2 is essential for bone healing.


Introduction
Mesenchymal progenitor cells (MPCs) exhibit self-renewal capacity and rapidly proliferate and differentiate into multiple lineages including osteogenic, chondrogenic, myogenic, and adipogenic cell lines [1][2][3][4]. Mouse embryonic fibroblasts (MEFs) are mesenchymal stem cells that could be able to undergo multidirectional differentiation potential which is into osteogenic, chondrogenic, and angiogenic cell lines [5]. Immortalized mouse embryonic fibroblasts (iMEFs) could be capable of conditionally immortalized MEFs that possess increased proliferative ability and long-term maintenance of cell proliferation [6]. iMEFs can be induced to undergo phenotype reversal by Cre recombinase, in possession of most markers of MPCs and retain multiple direc-tional differentiation potential, as they can differentiate into osteogenic and angiogenic lineages under suitable conditions [7,8]. iMEFs can be used in regenerative medicine as effective seed cells to treat bone defects, nonunion, and osteogenic-related diseases due to their proliferation and differentiation abilities and stromal progenitor cell biology [9].
Osteogenic differentiation is a continuous cascade that recaps nearly entire of the molecular adjustments that occur throughout embryonic skeletal development [10]. Bone morphogenetic proteins (BMPs) consist of nearly 20 members that be part of the transforming growth factor beta (TGF-β) superfamily [11]. Recombinant human bone morphogenic proteins, including BMP4 and BMP7, have been extensively used in clinical applications for the efficient stimulation of bone formation and for treating bone defects [12].
It has been reported that BMP9 can be used for filling engineered bone and treating osseous defects. When combined with some biomaterials, the release of the growth factor can be controlled at the proper rate. BMP9 as one of the most potent BMPs, can induce osteogenic differentiation of multipotent progenitor cells [13]. The effect of BMP9induced osteoblast differentiation is enhanced by regulating several downstream targets, such as HIF1α [14], Runx2 [15], Hey1 [16], and COX-2 [17]. However, the connotative mechanism involved in BMP9 mediated osteogenesis was still unclear.
Hypoxia-inducible factor 1α (HIF1α) is a generally accepted angiogenesis cascade regulator that is involved in many biological and organic developmental processes, including tumor-related and skeletal development. Some studies also indicated that HIF1α helps promote BMP9induced angiogenic and osteogenic differentiation though BMP/Smad signaling in iMEFs. Runx2 is a specific transcription factor in mature bone cells that plays an important role in the formation and reconstruction of bone tissue. Runx2 regulates the differentiation of iMEFs into osteoblasts and promotes the maturation and vascularization of cartilage. Additionally, this protein is involved in the synthesis of extracellular matrix. However, it is currently unclear whether and how the angiogenic factor HIF1α is coupled with the osteogenic factor Runx2 during BMP9-induced osteogenesis in iMEFs. It is also ambiguous whether these factors mutually participate in this regulation.
In this study, we investigated the role of HIF1α and Runx2 on BMP9-induced angiogenic and osteogenic differentiation in MEFs. We found that exogenous overexpression of HIF1α and Runx2 intensifies BMP9-induced osteogenic and angiogenic differentiation of iMEFs in vitro and in animal model, whereas RNA-mediated silencing of HIF1α and Runx2 strongly dulls BMP9-induced angiogenic and osteogenic signaling in iMEFs. Mechanistically, HIF1α can directly regulate Runx2 expression at the transcript and protein levels. Our findings no mere broad our understanding of the molecular events underlying BMP9-induced osteoblastspecific differentiation but also suggest that targeting the coupling of BMP9-induced angiogenic and osteogenic signaling may become a new approach for vascular-targeted therapy in bone tissue engineering.

Materials and Methods
2.1. Cell Culture and Chemicals. The experiment was approved by the Ethics Committee of the First Affiliated Hospital of Chongqing Medical University, and the written informed consent was obtained before surgery. HEK-293 cells and iMEFs were obtained from ATCC (Manassas, VA, USA). Cells were maintained in low-glucose Dulbecco's modified Eagle's medium (LG-DMEM) containing 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 mg/mL streptomycin. The cells were incubated at 37°C with 5% CO 2 . The cultured medium was changed every three to four days.

Recombinant Adenovirus Arrangement.
The recombinant adenoviruses were constructed with AdEasy technology [18,19]. To be short, the core regions of genes encoding BMP9, HIF1α, green fluorescent protein (GFP), red fluorescent protein (RFP), and Runx2 were amplified by quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR), cloned into shuttle vectors which could be subjected to amplify recombinant adenoviruses in HEK-293 cells. The target sites of siRNA against mouse HIF1α and Runx2 coding regions were amplified into the pSES adenoviral shuttle vector which form recombinant adenoviruses. The adenoviruses nominated Ad-BMP9, Ad-HIF1α, Ad-Runx2, Ad-Sim-HIF1α, and Ad-Sim-Runx2. Ad-BMP9 expressed GFP, while Ad-HIF1α, Ad-Sim-HIF1α, Ad-Runx2, and Ad-Sim-Runx2 expressed RFP as a visual sign for observing efficiency of infection. Analogous adenoviruses expressing monomeric GFP (Ad-GFP) served as negative controls.

Gene Expression of Osteogenesis-and
Angiogenesis-Related mRNA. The cells were cultured according to the experimental design. Total RNA was gained with TRIzol reaction agentia, and cDNA was extracted from RNA with a reverse transcription reaction reagent (RR047A, TAKARA, Japan). The first-strand cDNA products were diluted 5-10fold and used as templates for detection by real-time qRT-PCR (RR820A, TAKARA). The amplification conditions were set: predenaturation, denaturation, and annealing. GAPDH served as the internal control set. Expression levels of target genes from various groups were measured using the 2 ΔΔCT method. Gene expression of collected samples was conducted triplicate. The information of primers for PCR are shown in Table 1. 2.4. Chromatin Immunoprecipitation (ChIP) Assay. Subconfluent iMEFs were treated with infection of Ad-GFP or Ad-BMP9 and Ad-HIF1α. The cell samples were cross-linked after 48 hours of infection, then subjected to ChIP analysis according to the manufacturer's instructions (#9003, CST). The cells were incubated with a rabbit anti-mouse HIF1α (HIF1α67, Abcam) antibody or IgG to pull down the DNA-protein Mmixture. The Runx2 promoter sequence was determined using four pairs of primers consistent with the mouse Runx2 sites. The relevant primers and sequences of promoter information of the ChIP assay are listed in Supplementary Tables 1 and 2. 2.5. Alkaline Phosphatase (ALP) Expression Assay. ALP expression was measured by a modified automated chemiluminescence assay (BD Company, USA) and histochemical staining, according to the manufacturers' instruction [20,21]. Cells were treated with infection of Ad-GFP, Ad-HIF1α, Ad-BMP9, Ad-Runx2, and Ad-BMP9 combined with Ad-HIF1α, Ad-BMP9 combined with Ad-Runx2, Ad-BMP9 combined with Ad-SimHIF1α, or Ad-BMP9 combined with Ad-Sim-Runx2. In the chemiluminescence detection and quantification test, each sample was tested for three times, and the tests were repeated in at least three independent experiments. In addition, ALP expression was standardized 2 BioMed Research International to the total cellular protein level of cell samples. ALP expression was presented as the mean ± standard deviation.
2.6. Matrix Mineralization Detection. iMEFs were inoculated in 24-well plates for 6 h seeded at a density of 30% and then infected with Ad-BMP9, Ad-HIF1α, or Ad-Runx2. The osteoblastic differentiation-conditioned medium contained ascorbic acid and β-phosphoglycerol for 21 days. The matrix mineralization activity was evaluated by Alizarin Red S staining. The cell samples were soaked with 2.5% glutaraldehyde 15 min at room temperature, washed with PBS solution for three times. The mineralized tubercles were stained with 0.4% (v/v) Alizarin Red S dyestuff for 10 min, then washed with double distilled water for 15 min. Mineralization was observed under a microscope, and at least 3 independent experiments were conducted.
2.7. Nano-Glo Dual-Luciferase Reporter Assay. The iMEFs were seeded in Corning 24-well cell culture plates and transfected with 1 μg firefly luciferase vector, 1 μg Renilla luciferase vector, and 0.5 μg BMP9 binding element luciferase reporter (p12Xsbe-Luc) or Runx2 reporter plasmid (p6Xose2-Luc) per flask using Lipofectamine 2000 [22,23]. Twenty-four hours after transfection, the cells were reseeded in 24-well plates and treated with DMSO, HIF1α, and/or Runx2. After 36 h, the cell samples were lysed and subjected to a luciferase assay using a Promega Luciferase assay kit (E1500; Promega, Madison, WI, USA). Luciferase activities were normalized by detecting the concentration of total cellular protein.
2.8. Immunohistochemical Staining. Cells were infected with adenovirus and treated in accordance with the experimental design. The iMEFs were seeded onto sterile sections in 12well plates at a density of 10 4 cells/mL. The cells were washed 3 times with PBS for 10 min, then the cells were fixed with 4% paraformaldehyde at 37°C for 15 min, washed 3 times with PBS for 10 min, and permeabilized using 0.3% Triton X-100 for 30 minutes at 37°C. The cell samples were blocked with goat serum for 30 minutes after permeabilization. The samples were then incubated with primary antibodies overnight. The results of the expected protein were visualized by incubation with the appropriate fluorescently labeled secondary antibody. The sections were carefully removed and then mounted on slides with glycerol.
2.9. Stem Cell Implantation and Ectopic Ossification. iMEFs were transfected with appointed adenoviruses and prepared for injection at 5 × 10 6 cells into the side abdomen of athymic nude mice (4-7-week-old male, Sprague-Dawley) until fluorescence could be seen. Five weeks after injection, the mice were euthanized, and bony masses were obtained for micro-CT imaging and histologic staining and evaluation.
2.10. Microcomputed Tomographic Imaging Analysis. The bone masses were retrieved and scanned with SkyScan1174 X-ray microtomography (micro-CT) (Bruker Company, Belgium) after the animals were euthanized at 5 weeks. The 3D image reconstruction was analyzed with N-Recon software, and all image data analysis was performed using CT-SCAN software. Histomorphological parameters of bone formation including bone volume/total volume (BV/TV), trabecular number (Tb.N), trabecular separation (Tb.Sp), trabecular thickness (Tb.Th), and bone mineral density were detected.

Hematoxylin and Eosin (H&E) and Trichrome Staining.
The bone mass samples were collected and decalcified with 10% EDTA solution, then soaked in PBS for three times and fixed in 4% paraformaldehyde. The samples embedded in paraffin were subjected to H&E and Trichrome Masson's staining in accordance with manufacturers' instructions [24,25].

HUVEC Tube Formation
Assay. HUVECs were seeded in the enhanced endothelial conditioned medium (ECM), which is containing 10% FBS, 2 mM L-glutamine, 100 U/ mL penicillin, 100 μg/mL streptomycin, and 1% ECGS (Sigma, USA) for 4 h. The HUVECs were transferred into lower storey of the Transwell which is the preparation of Matrigel at a density of 10 5 cells per well. Then, the experimental iMEFs transfected with adenovirus were transferred into the upper wells. The Transwell was incubated at 37°C for 4 h. The tube area was quantified with the relative area of tube structures formed. The relative area of tubes was counted for three times per well and averaged from three images per well using Image J software (National Institutes of Health, USA).  BMP9  TACAACAGATACACAACGGACA  GATGTTGAAGATCAGGATGTGC  135  Runx2  CAGACCAGCAGCACTCCATATC  CCGTCAGCGTCAACACCATC  182  HIF1α  TGCTCATCAGTTGCCACTTCC  TGCCTTCATCTCATCTTCACTGTC  139  OPN  GACCGTCACTGCTAGTACACAAG  CCTTAGACTCACCGCTCTTCATG  194  BSP  AAGCACAGACTTTTGAGTTAGC  ACTTCTGCTTCTTCGTTCTCAT  145  COL-A1  TGAACGTGGTGTACAAGGTC  CCATCTTTACCAGGAGAACCAT  234  VEGF  CACGACAGAAGGAGAGCAGAAG  CTCAATCGGACGGCAGTAGC  82  ANGPT1  TTCTTCGCTGCCATTCTGACTCAC  GTTGTACTGCTCTGTCGCACTCTC  165  vWF  ACAGTAACATGGAGATGGCAGTG  TTGTGGCGTGTATGTGAGGATG  143 3 BioMed Research International 2.13. Subcutaneous PLGA-iMEF Hybrid Implantation to Detect Angiogenesis. iMEFs were transfected with specific adenovirus in accordance with the study design for 24 h and transferred onto poly (lactic-co-glycolic acid) (PLGA) scaffolds. Eighteen mice (6-week-old males; BALB/cAnN, Chongqing, China) weighing 18-25 g were anesthetized with 1% pentobarbital sodium (30 mg/kg), and then, PLGA scaffolds carrying transfected iMEFs were implanted into the subcutaneous region of nude mice. At the end of the fifth week, the implants were harvested. The PLGA-iMEF composites were collected and analyzed after 5 weeks. The PLGA scaffold-iMEF composites were retrieved. The samples were, respectively, decalcified and paraffin-embedded, then the sections were performed with immunohistochemical staining for CD31 (Abcam, USA), then incubated with the fluorophore-conjugated antibodies.
2.14. Statistical Analysis. The results were assessed; the data represented correspond to the mean ± SD if not stated. Experiment was conducted three times n = 3. Statistical analyses were performed using GraphPad Prism app (CA, USA). Student-Newman-Keuls T tests and one-way ANOVA were used to analyze the significant differences from various groups. Level of significance was determined as p < 0:05 significance.

HIF1α and Runx2 Promote BMP9-Mediated Osteoblastic
Differentiation of iMEFs. To determine whether HIF1α and Runx2 are critical targets of BMP9-mediated osteogenic and angiogenic signaling, we sought to explore whether exogenous expression of HIF1α and Runx2 participates in BMP9-induced osteogenic differentiation of iMEFs. Recombinant adenoviruses expressing HIF1α (Ad-HIF1α) or Runx2 (Ad-Runx2) were constructed. The results demonstrated that HIF1α and Runx2 effectively mediate transgene expression in iMEFs (Supplementary Figure 1A). Alkaline phosphatase (ALP), an indicator of early-stage osteogenic differentiation, was used to detect ALP activity. The results qualitatively and quantitatively showed that HIF1α and Runx2 exerted an intense synergistic effect on BMP9-induced ALP activity. On day 7, HIF1α and Runx2 significantly enhanced BMP9-mediated ALP expression, while exogenous HIF1α and Runx2 alone did not produce a marked effect on the early stage of osteogenic differentiation (Figures 1(a) and 1(c)). The calcium deposition and mineralization levels of iMEFs in the BMP9+HIF1α and BMP9+ Runx2 groups were higher than those in the HIF1α, Runx2, and GFP groups on day 21 (Figures 1(b) and 1(c)). Bone sialoprotein (BSP), collagen type I (COL-A1), and osteopontin (OPN) are essential indicators of osteogenic differentiation of iMEFs. To further confirm whether HIF1α and Runx2 are vital mediators of BMP9-mediated osteogenesis, we investigated the role of HIF1α and Runx2 on the mRNA expression of genes encoding BSP, COL-A1, and OPN. The results demonstrated that the mRNA expression levels of genes encoding BSP, COL-A1, and OPN in the BMP9+HIF1α group were significantly upregulated compared with the levels in the BMP9 group and the HIF1α group at day 7. Consistent with the Runx2-mediated effect on osteogenic differentiation, the mRNA levels of these factors in the BMP9+Runx2 group were greatly increased compared to the levels in the GFP and Runx2 groups on day 7 (Figure 2(a)). Remarkably, the expression of genes encoding BSP, COL-A1, and OPN was also lower in the HIF1α and Runx2 groups than in the GFP group. Taken together, these results strongly indicated that HIF1α and Runx2 can critically potentiate the BMP9-induced early and late osteoblastic differentiation of iMEFs in vitro.
3.2. Silencing HIF1α and Runx2 Inhibits BMP9-Mediated Osteogenesis. We further confirmed that HIF1α and Runx2 are essential mediators of BMP9-induced osteogenic signaling. We used recombinant adenovirus expressing a pool of three siRNAs poniting the coding regions of mouse HIF1α and Runx2 using pSOS system, generating Ad-Sim-HIF1α and Ad-Sim-Runx2. On day 7, the results showed that compared with that of the Sim-Runx2 group, ALP activity was greatly decreased in iMEFs after transfection with Ad-Sim-HIF1α. Qualitatively, ALP staining results presented similar results, and the expression of Sim-HIF1α drastically reduced ALP activity in iMEFs induced by BMP9 (Figures 3(a) and 3(c)). Moreover, Sim-HIF1α expression almost totally blunted BMP9-mediated bone mineralization in iMEFs, as shown by Alizarin Red S staining and the relative mineralization rate on day 21 (Figures 3(b) and 3(c)). Similarly, the qRT-PCR results illustrated that the expression of genes encoding osteogenesis-related factors such as BSP, COL-A1, and OPN was significantly downregulated in the Sim-HIF1α group compared to the BMP9 group at day 7 (Figure 2(a)). In short, these results strongly demonstrated that silencing HIF1α and Runx2 expression may exert a negative regulatory effect on the BMP9-induced osteogenic differentiation of iMEFs.

HIF1α and Runx2
Are Critical for Junctional Differentiation of BMP9-Induced Subcutaneous Bone Formation in Nude Mice. It still remains unknown whether HIF1α and Runx2 could affect bone formation in vivo. We hence utilized a subcutaneous bone formation model to detect the effects of HIF1α and/or Runx2 on BMP9induced ossification. The general morphology and micro-CT three-dimensional (3D) reconstruction results demonstrated that the general sizes of bony masses from the BMP9+HIF1α group were larger than those of bony masses from all other groups. The volume of osteogenic masses was smaller in the BMP9+Runx2 group than that in the BMP9 +HIF1α group, while the BMP9+Sim-HIF1α group showed a distinct decrease compared with the BMP9+Sim-Runx2 group (Figure 4(a)). The average mineral density displayed by the heatmap analysis revealed that HIF1α and Runx2 expression potentiated the average mineralization formed by BMP9-induced iMEFs. Notably, silencing exogenic HIF1α and Runx2 expression robustly decreased the average mineral density, but the effect of silencing HIF1α was greater than the effect of silencing Runx2 (Figure 4(c)). Quantitative results of bone parameters showed that the values including bone volume/total volume (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), and bone mineral density (BMD) were significantly higher in the BMP9 +HIF1α group than in the other groups. However, the values of these parameters in the BMP9+Sim-Runx2 group were significantly higher than those in the BMP9+ Sim-HIF1α group. There were no significant differences in trabecular spacing (Tb.Sp) among groups (Figure 4(b)).
The retrieved bone blocks were further subjected to histologic staining. The H&E and Masson's trichrome staining results demonstrated that compared with the BMP9 group, overexpression of HIF1α and Runx2 significantly promoted the number and quality of trabecular bone as well as mineralization, and HIF1α increased these values more than   Notably, less trabecular bone and bone matrix formation was observed in the BMP9+Sim-HIF1α group than in the BMP9+Sim-Runx2 group (Figure 4(d)).

HIF1α and Runx2 Regulate Angiogenic Differentiation and Vasoformation during BMP9-Mediated Osteogenesis.
To further confirm the potential of angiogenic differentiation and vasoformation regulated by HIF1α and Runx2, we used qRT-PCR of the transfected iMEFs on day 7 and tube formation assays of HUVECs after 6 h of Transwell cultivation. The qRT-PCR assay revealed that the mRNA expression of genes encoding angiopoietin 1 (ANGPT1), vascular endothelial growth factor (VEGF), and von Willebrand factor (vWF) was significantly upregulated in the BMP9 +HIF1α group compared with the other groups (p < 0:05).
It is worth noting that the mRNA expression of ANGPT1, VEGF, and vWF was higher in the BMP9+Sim-Runx2 group than in the BMP9+ Sim-HIF1α group (Figure 2(a)). The phenotypes of isolated HUVECs at the third passage were identified, and they highly expressed CD31, VEGF, EMCN, and vWF (Supplementary Figure 2). Tube formation assays showed that HIF1α significantly increased the relative tube area. Runx2 also increased the tube area, albeit to a lower degree than HIF1α, both independently and in combination with BMP9. Moreover, the relative tube area in the BMP9+Sim-HIF1α group was greatly decreased compared with that in the BMP9+ Sim-Runx2 group (Figures 2(b)-2(d)). Taken together, these data indicate that HIF1α and Runx2 are able to jointly regulate BMP9-induced angiogenic signaling and vascularization, and Runx2 may be an essential downstream target of HIF1α that participates in angiogenesis.

Effects of HIF1α and Runx2 on BMP9-Induced Vessel
Invasion. To further investigate the effect of HIF1α and Runx2 on BMP9-induced angiogenesis in vivo, iMEFs were transfected with the appropriate adenovirus, seeded on PLGA scaffolds, and implanted in the dorsal subcutaneous tissue of male rats ( Figure 5(a)). The cells adhered well to the PLGA scaffold after seeding for 72 h ( Figure 5(b)). After gross observation of vascular invasion to the implants, PLGA scaffolds exhibited a vascularization effect, among which the vascularization ability in the BMP9 +HIF1α group was more apparent (Figure 5(c)). The implant histology demonstrated that PLGA scaffolds seeded with cells in the BMP9+HIF1α group showed greater levels of CD31 ( Figure 5(d)). Remarkably, CD31 expression in the BMP9+Sim-Runx2 group was higher than that in the BMP9+ Sim-HIF1α group ( Figure 5(e)). In summary, HIF1α significantly potentiated BMP9induced angiogenesis in vivo, and silencing HIF1α inhibited the angiogenesis of iMEFs.

Runx2 Is Directly Upregulated by HIF1α in BMP9-Stimulated iMEFs.
Our experiments confirmed that HIF1α and Runx2 are essential for osteoblastic differentiation of BMP9-stimulated iMEFs; therefore, we sought to determine whether HIF1α exerts any impact on BMP9induced Runx2 expression. We used a ChIP assay to determine the relationship between HIF1α and Runx2 in BMP9 signaling. Gel electrophoresis was used to analyze the pulled down composite using an anti-HIF1α antibody, and the observed length was between 100 and 200 bp (Figure 6(a)). The location of promoter-specific primers for Runx2 yielded the expected products upon transfection with BMP9+HIF1α. We found pairs of primers targeting Runx2 that amplified the expected fragments in the group transfected with BMP9+HIF1α ( Figure 6(b)). We further      9 BioMed Research International determined the coupling effect between HIF1α and Runx2 at the transcription level by using a luciferase reporter assay. The results indicated that the transcriptional activity of the Runx2 luciferase reporter was significantly upregulated by HIF1α upon BMP9 stimulation for 48 hours (Figure 6(c)). However, the transcriptional activity of the HIF1α luciferase reporter gene was not obviously different upon Runx2 expression (Figure 6(d)). These data indicated that the transcription factor HIF1α can activate the promoter of Runx2 to regulate the initiation and activation of Runx2 at the gene transcription level, and the osteogenic differentiation induced by HIF1α and Runx2 may also be mediated by the activation of the BMP9-induced signaling pathway in iMEFs.

Discussion
HIF1α is a well-established and vital regulator of angiogenesis, and it is necessary to elicit angiogenic signals to recruit new blood vessels and maintain many developmental processes, including skeletal development [26][27][28]. Runx2 is among the most vital transcription factors, is indispensable for osteogenic differentiation, and is responsible for the activation of osteogenesis marker genes [29,30]. However, it remains unclear how HIF1α is coupled with Runx2 during the process of BMP9-induced osteogenic differentiation and bone formation. In this study, we explored whether HIF1α directly regulated Runx2 and whether HIF1α-mediated angiogenesis affected BMP9-stimulated osteogenic differentiation of MPCs. We detected the effect of HIF1α and Runx2 on ALP activity as an early osteogenic factor, bone mineralization as a late osteogenic indicator, the expression of osteogenesis-and angiogenesis-related factors and consolidated subcutaneous ectopic bone formation of iMEFs in athymic nude mice, increased osteogenic volume, and the values of BV/TV, Tb.Th, Tb.N, and BMD; we also investigated HUVEC tube formation in prevascular-like structures 11 BioMed Research International after in vitro and in vivo PLGA subcutaneous implantation for angiopoiesis. Overall, the results demonstrated that the exogenous expression of HIF1α and Runx2 enhances BMP9-stimulated osteogenic differentiation, including ALP activity, calcium deposition, and the expression of osteogenesis-related factors such as BSP, COL-A1, OPN, and ectopic subcutaneous bone formation, yet silencing HIF1α and/or Runx2 strongly blocked BMP9-induced osteogenic signaling in iMEFs. Furthermore, while stable overexpression of HIF1α and Runx2 can potentiate the angiogenesis-related factors ANGPT1, VEGF, and vWF and support HUVEC orientation into prevascular-like structures in vitro and intensify robust vascularization in ectopic subcutaneous implantation of PLGA-transfected iMEFs, silencing HIF1α and Runx2 can profoundly inhibit these effects in vitro and in vivo. Mechanistically, HIF1α can directly regulate the activation of Runx2 at the transcriptional level, and HIF1α also exerts a synergistic effect by promoting angiogenic and osteogenic signaling pathways upon BMP9 induction in iMEFs. Therefore, our results strongly demonstrate that the coupling between HIF1α and Runx2 plays an essential role in BMP9-induced osteogenic and angiogenic differentiation.
HIF1α is a well-established factor that regulates adaptive responses, including capillary ingrowth and angiogenesis, to reduce oxygen utilization, both spatially and temporally. It was previously reported that oxygen-sensitive HIFα subunits are found in mammals: HIF1α, HIF2α, and HIF3α. HIF1α and HIF2α have been widely studied and are the most common [31]. There are two regulatory models of HIF1α. Under normoxia, HIF1α is rapidly hydroxylated and ubiquitinated by the E3 ubiquitin ligase complex, which contains the von Hippel-Lindau disease tumor suppressor (VHL) protein and undergoes rapid degradation. Another regulatory model of HIF1α is that HIF1α is hydroxylated at asparagine residues by inhibiting the HIF1 factor to prevent the transcriptional activity of HIF1α, which is accomplished by blocking the interaction of the transcriptional coactivator cAMP response element binding protein (CBP) and histone acetyltransferase p 300 (P 300 HAT) with HIF1α [32,33]. In contrast, under hypoxic conditions, HIF1α is stabilized by limited oxygen as a helper substrate for prolyl hydroxylase domain enzymes (PHDs), and the HIF1α protein hydroxylation rate is reduced by enhancing the transcriptional activation of PHD, HIF1, and CBP-p 300 coactivated complexes. Then, HIFα levels increase, and HIF1α target genes are expressed [34,35]. Recent studies have shown that stimulating angiogenesis plays a key role in the process of increased bone mass while the HIF1α-VEGF pathway is activated [33]. Some research found that VEGF enhances BMP-mediated stimulation of anabolic bone formation by affecting angiogenesis [36][37][38]. The possible nonautonomous mechanisms underlying the overexpression of HIF1α in mature mouse osteoblasts include the possibility that blocking the VHL protein significantly enhances angiogenesis and osteogenesis, including the activation of VEGF in vascular endothelial cells. Hence, in mice lacking HIF-1α, the trabecular bone volume and bone formation rate were significantly reduced, the cortical bone structure was changed, and vascular development in the long bones was decreased [39]. Our study showed that overexpression of exogenous HIF1α intensified BMP9-induced osteogenic differentiation and osteogenesis. Angiogenesis-related factors (e.g., VEGF, ANGPT1, CD31, and vWF) were upregulated, and the activity of prevascular structures and subcutaneous vasoformation were higher in the HIF1α-expressing group.
In contrast, silencing HIF1α impaired the osteoblast and angiogenic differentiation ability, as observed in the cells treated with Sim-HIF1α. Therefore, HIF1α signaling in anabolic bone formation may increase osteogenesis via a cellnonautonomous effect to integrate spatiotemporal expansion and revascularization and establishment of the blood supply in bone, although it does not preclude the existence of additional effectors. Moreover, we believe that HIF1α signaling is a critical accommodator of new vessels for bone regeneration.
Runx2 belongs to the small transcription factor family, and members of the Runx family contain the runt domain [40]. Runx2 is a vital regulator in bone development and is essential for osteoblast differentiation. Mutation of Runx2 is related to skeletal malformation syndromes, including cleidocranial dysplasia (CCD) [41,42]. Runx2 knockout mice lose all intramembranous and endochondral osteogenesis and die after birth due to lack of mineralization in the chest region, which results in breathing difficulties, even in the presence of BMP [43]. The DNA sequence 5 ′ -PuACCPuCA-3 ′ and its complementary sequence 5 ′ -TGPyGGTPy-3 ′ are recognized Another possible important branch of the regulatory loop between HIF1α and osteogenesis is that HIF1α is a vital factor that plays a positive role in vascularization by regulating the expression of vasoformation-related regulators, including VEGF, CD31, vWF, and ANGPT1. Our initial results support the notion that the transcript activity of Runx2 is regulated by and coupled to HIF1α, which is enhanced by BMP9. This loop not only strengthens the osteogenic differentiation of iMEFs and angiogenesis mediated by HIF1α but also enhances extracellular matrix deposition and mineralization, which boosts efficient bone regeneration.
by Runx2, which is capable of binding to the promoters of genes encoding type I collagen, osteopontin, osteocalcin, BSP, collagenase-3, VEGF, type X collagen, and others to enhance the expression of these proteins and boost the osteoblastic differentiation [44][45][46][47]. These results may indicate that Runx2 is important for regulating osteogenesis, chondrocyte hypertrophy, and vascular invasion of the developing skeleton [15,48]. Our results revealed that Runx2 overexpression is capable of promoting BMP9-induced early and late osteogenesis and ectopic subcutaneous bone formation and upregulating the expression of osteogenesis-and angiogenesis-related genes.
In addition, Runx2 organizes HUVECs into prevascular-like structures in vitro and promotes vascularization in vivo after PLGA implantation, even though Runx2-mediated osteogenesis and angiogenesis were less robust than those induced by HIF1α. Conversely, silencing Runx2 attenuated the effects of BMP9-stimulated osteogenesis and angiogenesis, further confirming that Runx2 is a key regulator of osteoblastic and angiopoietic differentiation and a molecular transducer in osteogenic biology. Even though bone formation is primarily mediated by osteoblasts, many other tissues in bone consist of vascular endothelium and/or sensory and motor nerves that help establish a conducive milieu for bone formation [49][50][51][52][53][54]. Our in vitro HUVEC tube formation assays and in vivo analysis of angiogenesis after PLGA implantation showed that the level of CD31 was increased by overexpression of HIF1α and Runx2. Furthermore, the expression of angiogenicrelated factors was higher after overexpressing HIF1α than that elicited by Runx2, and silencing HIF1α decreased the expression of these factors to a greater degree than did inhibiting Runx2.
We further evaluated the relationship between HIF1α and Runx2 at the gene level. Luciferase reporter assays showed that the transcription factor HIF1α was able to enhance luciferase expression downstream of the Runx2 promoter. The mRNA of Runx2 was shown to be pulled down with an anti-HIF1α antibody. Mechanistically, these data demonstrated that Runx2 is a downstream target of HIF1α and that HIF1α can directly regulate Runx2 (Figure 7).

Conclusion
In summary, this study recognized the ability of HIF1α and Runx2 to stimulate osteogenic and angiogenic differentiation of iMEFs and showed that silencing HIF1α and Runx2 inhibited BMP signaling. The coupling effect of HIF1α and