The miR-187 Induced Bone Reconstruction and Healing in a Mouse Model of Osteoporosis, and Accelerated Osteoblastic Differentiation of Human Multipotent Stromal Cells by Targeting BARX2

Background: Multiple microRNAs (miRNAs) have been proven to regulate osteogenic differentiation by affecting the Runx2 signaling pathway. The intervention of miRNA can delay the progress of osteoporosis (OP) and induce fracture repair by affecting bone regeneration. However, the function and mechanism of miR-187 in osteoporotic fractures are still unknown. Methods: We rst established the OP mouse model. Next, the BMD value was certied by iDXA. The miR-187 level in the OP mice and serum of OP patients was identied through qRT-PCR. Bone repair and bone healing were assessed through toluidine blue staining and X-ray, and BARX2 expression was also conrmed. Osteogenesis-related proteins, ALP activity, and the matrix mineralization state were evaluated by western blot, ALP staining, and Alizarin Red staining in hMSCs after transfection with miR-187 mimics, miR-187 inhibitor, or human BarH-like homeobox 2 (BARX2) siRNA. Moreover, the interplay between miR-187 and BARX2 was identied through the dual-luciferase reporter. Results: The BMD value was notably reduced in the OP mice, and miR-187 was markedly downregulated in the OP mice and serum of OP patients. Meanwhile, we proved that miR-187 induced bone reconstruction and healing, and downregulated BARX2 in the OP mouse model. We also proved that BARX2 was a direct target of miR-187, and could be signicantly downregulated by miR-187. Furthermore, miR-187 induced osteogenic differentiation of hMSCs by targeting BARX2. Conclusions: The miR-187 might have a signicant therapeutic effect in osteoporotic fractures. miR-187 accelerated osteogenic differentiation of hMSCs by directly regulating BARX2.


Background
Osteoporosis (OP) is a universal metabolic bone disease and has been recognized as a major public health problem worldwide [1]. Postmenopausal osteoporosis is one of the most widespread types of osteoporosis [2]. Osteoporosis is characterized by reduced bone mass, decreased bone mineral density, and destruction of bone microstructure, which can result in increased bone fragility, and even an increased risk of fracture [3,4]. Currently, osteoporotic fractures have become the most serious complication of osteoporosis [5]. Studies have testi ed that the mortality rate associated with diseases caused by osteoporotic fractures have exceeded the combined death rates of the three major gynecological tumors (breast cancer, cervical cancer, and uterine body cancer), and the related death rates have been increasing year by year [6,7]. According to statistics, the occurrence risk of osteoporotic fracture is very high, being about 40-50% for females and 13-22% for males [8,9]. Currently, OP therapy drugs are broadly divided into basic drugs and anti-absorption drugs. These drugs cannot induce bone formation and have considerable side effects [10,11]. Therefore, further studies on the molecular mechanisms of osteoporosis are crucial for therapies for osteoporosis.
The growth of bone is manifested mainly by the enhancement of osteogenic differentiation and the weakening of osteoclast differentiation [12]. Osteoporosis is caused mainly by a decrease in osteogenic differentiation and an increase in osteoclast differentiation [13]. Human marrow-derived mesenchymal stem cells (hMSCs) have the potential to differentiate into a variety of cell types. They can also differentiate into osteoblasts and adipocytes under natural conditions, both of which maintain a dynamic balance [14,15]. Research has also shown that a disruption in this balance can generate metabolic bone diseases, such as osteoporosis [10,15]. An increasing amount of research has also con rmed that boneforming drugs will have a broad application prospect in OP [16,17]. Therefore, exploring the mechanism of osteogenic differentiation may provide a new strategy for the treatment of OP.
MicroRNAs (miRNAs) are a class of non-coding single-stranded RNA with a length of 22 nucleotides encoded by endogenous genes [18,19]. The miRNAs can mediate speci c gene silencing, cause inhibition and degradation of target gene mRNAs, and, thus, regulate protein synthesis [20]. Recent research has testi ed that miRNAs can be involved in the development process of osteoporosis by regulating the balance between osteogenesis and osteoclasts [21,22]. Among them, miR-187 has been shown to signi cantly contribute to a wide range of diseases, such as glaucoma [23], cancer [24][25][26][27], and psoriasis [28], etc. In accordance with the literature, miR-187 was highly expressed in the non-fracture group versus fracture group [29]. However, the function and mechanism of miR-187 in OP has not been explored.
In our study, we further investigated the expression of miR-187 in the OP mouse model and the serum of osteoporosis patients. We determined the in uence of miR-187 on bone repair and bone healing in the OP mouse model, and also on the osteogenic differentiation of hMSCs. We veri ed the interplay between miR-187 and human BarH-like homeobox 2 (BARX2) in OP.

Animal
Forty female speci c pathogens free (SPF) C57BL/6J mice were supplied by the Animal Experiment Center of the Institute of Radiation Medicine of the Chinese Academy of Medical Sciences. The animals were kept in a room with a room temperature of 15-28℃, humidity of 45-55%, and a 12-hour light/dark cycle. Mice had free access to drinking water and standard food. This study obtained the approval of the Animal Experiment Center of the Institute of Radiation Medicine of the Chinese Academy of Medical Sciences animal experiment ethics quali cation.
Establishment of the OP model After a week of routine feeding, all C57BL/6J mice were anesthetized through intraperitoneal injection of 3% pentobarbital sodium (30 mg/kg). After skin disinfection, the abdominal midline of the mice was cut about 1.5 cm, and bilateral fallopian tubes and ovaries were exposed. After ligation of the fallopian tubes, the ovaries and the ligated fallopian tubes were removed, and the wound was closed. Penicillin was injected into the abdominal cavity for 3 days. In the sham group, we removed only the same amount of adipose tissue around bilateral ovaries. After surgery, the resected ovarian tissues were collected, and part of the tissues was embedded and sectioned.
Establishment of the femoral fracture model After 4 weeks of regular feeding, the mice in the sham-operated group and OP model group were anesthetized using the same procedure. The femurs of the mice were exposed, a steel needle with a diameter of 0.45 mm was inserted retrograde from the knee joint for intramedullary xation. The fracture was formed by cutting off 1/3 of the femur with a scalpel. After disinfection, healing of the wound was regularly observed. Negative control (NC) and miR-187 lentivirus were obtained from the Wuhan Hualian Biotechnology Co. Ltd (Wuhan, China), and the OP mice were injected with NC or miR-187 lentivirus through the tail vein. At the end of the experiment, rats were sacri ced under anesthesia using 50 mg/kg sodium pentobarbital by intraperitoneal injection.

Clinical samples
The blood samples of 33 healthy controls and 33 OP patients and the bone marrow samples of OP patients were collected from the First A liated Hospital of Jinan University. The serum was isolated from the blood samples through centrifugation (3,000 × g for 5 min at 4 °C). Before collection, informed consent was provided from all patients. This study was allowed by the ethics committee of the The First A liated Hospital of Jinan University.

X-ray
Specimens of the fractured tibiae were collected and immediately photographed using the RADspeed digital X-ray photography system (40 kV, 1,2 mA) to assess bone healing.
Detection of bone mineral density (BMD) The iDXA (GE Healthcare Lunar, USA) was used to examine the BMD of the fractured tibiae.

ALP staining
In line with the instructions of the ALP staining kit (Beyotime, Shanghai, China; cat. no. P0321), the transfected hMSCs were washed using PBS and then xed for 30 min in 4% paraformaldehyde. After washing the hMSCs, the proper amount of dye solution was added. After washing, we observed the results under a microscope.

Alizarin Red staining
After washing the xed hMSCs in each group, alizarin red S solution (Waldeck GmbH) was added at 37 °C for 30 min. The dye results were observed using a microscope.
Quantitative real-time PCR (qRT-PCR) assay Total RNA was isolated by applying TRIzol reagent (Invitrogen, Carlsbad, CA, USA), and total RNAs were reverted to cDNAs following the instructions of the reverse transcription kit (Invitrogen). Based on the instructions of the SYBR Green PCR kit (Invitrogen), gene expression was examined on the ABI PRISM® 7500 System (Applied Biosystems, Foster City, CA, USA). The relative expression of the target genes was analyzed by 2 −△△CT .

Western blot assay
Total protein was separated through the application of RIPA buffer (Beyotime) containing 1% proteinase inhibitor. The BCA method was utilized to determine the protein concentration in each group. Then, 40 µg total protein from each group was isolated using 10% SDS-PAGE under 120 V constant voltage. The protein was transferred to a PVDF membrane (Roche, Santa Clara, CA, USA; cat. no. 3010040001) under the conditions of 200 mA for 90 min. After sealing, the membranes were incubated overnight at 4 °C with the corresponding primary antibodies, which were from Abcam (Burlingame, CA, USA). Then, the membranes were exposed to the secondary antibody for 1 h. The speci c proteins were developed using ECL chemiluminescence (KALANG; cat. no. KL-D3490).

Statistical analysis
All experiments were repeated at least three times, and the data are displayed as the mean ± SD. The data were counted via GraphPad Prism 5.0 software (GraphPad, San Diego, CA, USA) using Student's t-test or one-way ANOVA. P < 0.05 denotes statistical signi cance.

Results
The miR-187 was signi cantly downregulated in OP.
To determine the change in expression of miR-187 in OP, we rst performed a bilateral ovariectomy to establish the OP mouse model. First, we uncovered that the BMD value was observably reduced in the left femur of the mice in the OP model group compared to the sham group (P < 0.001, Fig. 1A). Second, our data proved that miR-187 expression was lower in the femoral bone of the OP mice than in the sham mice (P < 0.01, Fig. 1B). The level of miR-187 was markedly downregulated in the serum of OP patients (n = 33) relative to the healthy controls (n = 33) (P < 0.001, Fig. 1C). These results proved that the expression of miR-187 in OP was low.
The miR-187 markedly promoted the reconstruction and healing of bone and downregulated BARX2 in the OP mouse model.
To further study whether miR-187 has a signi cant effect during OP pathogenesis, we injected OP mice with miR-187 or NC lentivirus through the tail vein. The data from the qRT-PCR analysis exhibited a signi cant decline in the expression of miR-187 in the femoral bone of the OP mice versus the sham mice. This decline could be partially reversed by miR-187 overexpression in the femoral bone of the OP mice (P < 0.05, Fig. 2A). In addition, our results of toluidine blue staining disclosed that the relative bone repair rate was dramatically reduced in the OP mice compared with the sham mice. At the same time, overexpression of miR-187 signi cantly attenuated the reduction of the relative bone repair rate in the OP mice (P < 0.05, Fig. 2B). Also, the relative bone healing rate was notably decreased in the OP mice versus the sham mice. miR-187 overexpression could signi cantly reverse the decreased bone healing rate in the OP mice (P < 0.05, Fig. 2C). Moreover, we demonstrated that BARX2 expression was signi cantly elevated in the OP mice relative to the sham mice (P < 0.05, Fig. 2D). This elevation could be weakened by miR-187 overexpression in the OP mice (P < 0.001, Fig. 2E). So, these ndings manifested that miR-187 might have signi cantly promoted the reconstruction and healing of bone in the OP mice.
Subsequently, we investigated the impact of miR-187 on osteoblastic differentiation in hMSCs. We successfully isolated hMSCs from OP patients and then transfected the extracted hMSCs with miR-187 mimics or miR-187 inhibitor. qRT-PCR assay was conducted to evaluate the transfection e ciencies of miR-187 mimics and miR-187 inhibitor in hMSCs. As displayed in Fig. 3A, relative to their respective controls, miR-187 expression was markedly increased in the mimics group, and signi cantly decreased in the inhibitor group (P < 0.001). We con rmed that the osteogenic markers (OCN, OPN, RUNX2, BSP, and ALP) were upregulated in the miR-187 mimics group, and dramatically downregulated in the miR-187 inhibitor group versus the NC group (Fig. 3B). The results of ALP staining showed that miR-187 mimics enhanced ALP activity, while miR-187 inhibitors weakened ALP activity in hMSCs (P < 0.01, P < 0.001, Fig. 3C). Next, we applied Alizarin Red staining to examine the in uence of miR-187 on bone mineralization in hMSCs. As presented in Fig. 3D, the calcium nodules were signi cantly higher in the miR-187 mimics group than in the NC mimics group, and the calcium nodules were also signi cantly lower in the miR-187 inhibitor group than in the NC inhibitor group. These results implied that miR-187 facilitated osteoblast differentiation in hMSCs.
BARX2 was a target of miR-187.
Next, the mechanism by which miR-187 induces osteogenic differentiation was further explored. After experimental exploration, we discovered that BARX2 expression was signi cantly reduced at 14 and 28 days after induction of osteogenic differentiation in hMSCs (P < 0.01, P < 0.001, respectively, Fig. 4A and 4B). We applied bioinformatics analysis to predict the possible binding sites between BARX2 and miR-187, which are shown in Fig. 4C.
We further determined whether the pro-osteogenic effect of miR-187 on hMSCs was mediated by BARX2. hMSCs were individually transfected or co-transfected with miR-187 inhibitor or BARX2 siRNA. First, our qRT-PCR data displayed that miR-187 inhibitor signi cantly elevated the level of BARX2, while BARX2 expression was markedly reduced after BARX2 siRNA in miR-187 inhibitor-mediated hMSCs (P < 0.001, Fig. 5A and 5B). Second, the results of the Western blot also exhibited that miR-187 inhibitor downregulated the osteogenic markers (OCN, OPN, RUNX2, BSP, and ALP) in hMSCs. These downregulations, which were mediated by miR-187 inhibitor, could also be signi cantly reversed by BARX2 siRNA in hMSCs (Fig. 5B). Next, ALP staining uncovered that miR-187 inhibitor notably lowered the ALP activity, and the ALP activity, which was inhibited by miR-187 inhibitor, could be elevated by BARX2 knockdown in hMSCs (P < 0.01, P < 0.001, Fig. 5C). Alizarin Red staining results showed that the matrix mineralization state of hMSCs was reduced in the miR-187 inhibitor group versus the NC inhibitor group, and knockdown of BARX2 enhanced the matrix mineralization of hMSCs, which was suppressed by the miR-187 inhibitor (Fig. 5D). Therefore, we have shown that the effect of miR-187 on osteogenic differentiation was achieved by regulating BARX2 in hMSCs.

Discussion
OP is a frequent systemic degenerative disease of the skeletal system [30]. Osteoporosis not only increases the risk of fracture but also affects the stability of fracture xation and the healing process. Currently, osteoporosis has many pathogenic factors and a complicated pathogenesis [31]. There is no speci c drug to treat osteoporosis clinically. The application of animal models is a vital link in the study of the etiology, pathology, therapy, and prognosis of human osteoporosis [32]. Postmenopausal osteoporosis is the most familiar type of osteoporosis, accounting for 80% of the total number of patients [2]. Ovariectomized female mice have become the most commonly applied animal model of osteoporosis. Its advantages include a single modeling factor, high success rate, good repeatability, and high reliability [33]. In our study, we also established the osteoporosis mouse model through bilateral ovariectomy.
The hMSCs are pluripotent stem cells with self-renewal capacity, and they can differentiate into osteoblasts under appropriate culture conditions [34]. Multiple studies have manifested that hMSCs have crucial clinical application value in promoting bone healing and bone regeneration [35,36]. Therefore, studies on the osteogenic differentiation mechanism of hMSCs contribute to the clinical application of hMSCs in osteoporosis. In our study, we also successfully isolated hMSCs from OP patients.
Increasing evidence has suggested that miRNAs with the potential to regulate the 3 -UTR of speci c genes might be of great importance in the osteogenic differentiation of hMSCs [36][37][38][39]. For instance, miR-128 was con rmed to suppress the osteogenic differentiation of hMSCs by the VEGF pathway [38]; miR-7-5p could notably accelerate the osteogenic differentiation of hMSCs through CMKLR1 [39]; miR-224 was veri ed to induce osteoblastic differentiation of hMSCs by Rac1 [36]. Our current research revealed that miR-187 was signi cantly underexpressed in the femoral bones from OP mice and the serum of OP patients. These results are consistent with previous research in which miR-187 was signi cantly downregulated in osteoporotic fractures [40]. Meanwhile, we rst disclosed that miR-187 could dramatically accelerate bone healing and bone repair in the OP mouse model and induce osteogenic differentiation of hMSCs. Currently, miR-187 has also been shown to have a signi cant moderating effect on disease progression, especially cancer [26,[41][42][43][44]. These results suggested that miR-187 had a signi cant contribution to the disease process.
BARX2, as a homeodomain factor of the Bar family, has been con rmed to function in cell adhesion, cellular differentiation, and cytoskeletal remodeling [45,46]. Recent research has suggested that BARX2 is closely associated with multiple diseases, especially solid tumors [47][48][49]. Another study has also reported that BARX2 could be regulated by miR-187 in a targeted way in oral carcinoma [50]. In our study, we uncovered that with the BARX2 gene as a target, miR-187 could inhibit BARX2. Our data has also shown that BARX2 knockdown could reverse the effect on OP, which was brought about by the miR-187 inhibitor, suggesting that the miR-187 inhibitor prevented osteoblast differentiation in hMSCs by targeting BARX2.

Conclusions
In summary, we proved that miR-187 could markedly facilitate bone reconstruction and healing in a mouse model of OP in vivo and induce osteoblastic differentiation of hMSCs through BARX2 in vitro. Therefore, our results provide possible new targets for the treatment of osteoporotic fractures. They also provide novel evidence for further exploration of the interactions between miR-187 and BARX2 during osteogenic differentiation.

Consent for publication
Not applicable.

Availability of data and materials
The data used to support the ndings of this study are available from the corresponding author upon request.

Competing interests
The authors declare no competing interests.

Funding
This research did not receive any speci c grant from funding agencies in the public, commercial, or notfor-pro t sectors.
Authors' contributions JZ, FP, and ZGZ were responsible for the research design. BST and JL contributed to various aspects of data acquisition, analysis, and interpretation. JZ and ZGZ drafted the paper. All authors critically revised the paper. All authors read and approved the nal manuscript.     The 293T cells were cotransfected with miR-187 mimics or mimic NC and the luciferase constructs carrying BARX2, and the uorescence intensity was monitored using a luciferase assay. (E) The level of BARX2 was con rmed using qRT-PCR and Western blot assays in hMSCs after transfection with miR-187 mimics or miR-187 inhibitor. **P<0.01, ***P<0.001.