Biomaterials 74 (2016) 155e166

Contents lists available at ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

Healing of osteoporotic bone defects by baculovirus-engineered bone marrow-derived MSCs expressing MicroRNA sponges Kuei-Chang Li a, 1, Yu-Han Chang b, c, 1, Chia-Lin Yeh a, Yu-Chen Hu a, * a

Department of Chemical Engineering, National Tsing Hua University, Hsinchu 300, Taiwan Center for Tissue Engineering, Chang Gung Memorial Hospital, Taoyuan 333, Taiwan c Department of Orthopaedic, Chang Gung Memorial Hospital, Taoyuan 333, Taiwan b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 4 July 2015 Received in revised form 25 September 2015 Accepted 29 September 2015

Fractures associated with osteoporosis are a worldwide health problem. To augment osteoporotic bone healing, we aimed to develop a cell/gene therapy approach in combination with miRNA manipulation. We unraveled aberrant overexpression of miR-140* and miR-214 in the bone marrow-derived MSCs isolated from ovariectomized (OVX) rats (OVX-BMSCs). To suppress the miRNA levels, we constructed hybrid baculovirus vectors expressing miRNA sponges to antagonize miR-140* or miR-214. Engineering OVX-BMSCs with the hybrid vectors persistently attenuated the cellular miR-140*/miR-214 levels, which promoted the OVX-BMSCs osteogenesis and augmented the ability of OVX-BMSCs to repress osteoclast maturation in vitro. Notably, suppressing miR-214 exerted more potent osteoinductive effects. In the osteoporotic rat models with a critical-size bone defect at the femoral metaphysis, implanting the OVXBMSCs ectopically expressing BMP2 failed to heal the defect, which underscored the difficulty to heal osteoporotic bone defects. Nonetheless, allotransplantation of the miR-214 sponges-expressing OVXBMSCs healed the defect and ameliorated the bone quality (density, trabecular number, trabecular thickness and trabecular space) at 4 weeks post-implantation. Co-expressing BMP2 and miR-214 sponges in OVX-BMSCs further synergistically substantiated the healing. The baculovirus-engineered OVX-BMSCs that expressed miR-214 sponge, with or without BMP2 expression, thus paved a new avenue to the treatment of osteoporotic bone defects. © 2015 Published by Elsevier Ltd.

Keywords: Osteoporosis Bone defect MicroRNA miR-214 BMP2 miRNA sponge

1. Introduction Bone is a dynamic tissue that requires coordinated homeostasis of bone-forming osteoblasts and bone-resorbing osteoclasts to maintain constant bone mass and volume [1]. Osteoporosis arises from the dysregulation of bone turnover such that bone resorption exceeds bone formation, leading to reduction of bone mass and micro-architectural deterioration [1]. Osteoporosis affects as many as 75 million people in the US, Japan and Europe and z30e50% of women and z15e30% of men with osteoporosis have a fracture during their lifetime [2]. Since osteoporotic fractures are common (>9 million per year globally [1]), the lost productivity and healthcare expenditures have rendered osteoporosis a worldwide health problem [3,4]. To date, antiresorptive

* Corresponding author. E-mail address: [email protected] (Y.-C. Hu). 1 These two authors contributed equally to this work. http://dx.doi.org/10.1016/j.biomaterials.2015.09.046 0142-9612/© 2015 Published by Elsevier Ltd.

agents that inhibit osteoclast activities (e.g. bisphosphonates, calcitonin and estrogen) and anabolic agents (e.g. recombinant human parathyroid hormone) are available for osteoporosis treatment [4]. However, the repair of bone defects following fracture draws less attention. MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression by translation inhibition or mRNA degradation, through binding to the target mRNAs. MiRNAs modulate diverse biological processes, including osteoclastogenesis, osteogenesis, bone formation and can alter the bone phenotype (for review see Refs. [5e7]). For instance, miR-23a, miR-133, miR-335 and miR3077-5p inhibit osteogenesis by directly interacting with the 30 untranslated region (30 UTR) of osteogenic transcription factor Runx2 [8]. Conversely, miR-21 targets the Fas ligand to suppress osteoclastic apoptosis, thereby supporting osteoclastogenesis [9]. However, little is known about the roles of miRNA in osteoporosis. Mesenchymal stem cells (MSCs) play critical roles in bone formation as bone marrow-derived MSCs (BMSCs) are capable of

156

K.-C. Li et al. / Biomaterials 74 (2016) 155e166

differentiating into osteoblasts and then osteocytes. Baculovirus (BV) is a non-pathogenic insect virus and can transduce BMSCs and adipose-derived stem cells (ASCs) at efficiencies exceeding 90% [10,11]. We have shown that transduction of BMSCs or ASCs with BV carrying a transgene encoding osteoinductive bone morphogenetic protein 2 (BMP2) or miRNA (e.g. miR-148b) can enhance the osteogenic differentiation and promote the healing of critical-size bone defects after cell implantation into animals [12e19]. Normally, BV confers transient transgene expression [15] as BV genome exists as an episome and degrades with time [20], which eases the safety concern [21] but meanwhile restricts its applications. To extend the transgene expression, we developed a hybrid BV system, which comprises two BV vectors: one BV expressing Cre recombinase and the other substrate BV harbors the transgene cassette flanked by two loxP sites [22]. After co-transduction of ASCs with the Cre/loxPbased hybrid BV system, the Cre recombinase excises the loxPflanking cassette off the substrate BV genome and catalyzes the formation of DNA minicircles, thereby prolonging the transgene (e.g. BMP2) expression [22]. Given the putative roles of miRNAs in osteoclastogenesis and osteogenesis and the need to heal osteoporotic fractures, we hypothesized that miRNA expression in osteoporotic BMSCs is dysregulated, hence impairing the ability of BMSCs to differentiate into osteoblasts for bone repair. Therefore, we created osteoporotic rat models by ovariectomy (OVX) and analyzed the miRNA expression profile in the OVX-BMSCs. We uncovered that miR-140* and miR-214 were overexpressed in OVX-BMSCs, and thus constructed Cre/loxP-based BV harboring miRNA sponges, in order to persistently knock down miR-140* or miR-214. Whether suppressing miR-140* or miR-214 was able to restore the osteogenesis of OVX-BMSCs and inhibit the osteoclast maturation was evaluated in vitro. We further assessed whether the BV-engineered, miRNA sponge-expressing OVX-BMSCs, with or without BMP2 co-expression, were able to heal critical-size bone defects in osteoporotic rats. 2. Materials and methods 2.1. Osteoporotic rat model, cells, BMSCs isolation and expansion All animal experiments were performed in compliance with the Guide for the Care and Use of Laboratory Animals (Ministry of Science and Technology, Taiwan). To generate osteoporotic rat models, the SpragueeDawley female rats (8 weeks old, BioLASCO, Taiwan) were subjected to bilateral ovariectomy (OVX) or sham operation (Sham) after anesthetization by intramuscular injection of Zoletil 50 (25 mg/kg body weight) and 2% Rompun® (0.15 ml/kg body weight). Under sterile conditions, a lumbar lateral incision was made between the caudal edge of the rib cage and the base of the tail. The ovary and part of the oviduct were removed with scissors, and the muscles and skin were sutured with resorbable suture. After ovariectomy, the rats were daily injected with methylpredinisolone hemisuccinate (1 mg/kg body weight/day, Sigma) for four consecutive weeks [23]. BMSCs were harvested from the hind limb of OVX rats or Sham rats and the cells were flushed by 22G needle attached to a 10-ml syringe containing DMEM medium (Invitrogen). The cells were filtered through a 70-mm filter mesh to remove bone spicules and cell clumps, resuspended in DMEM medium containing 10% fetal bovine serum (FBS, Hyclone), 100 IU/ml penicillin and 100 IU/ml streptomycin, incubated at 37  C with 5% CO2 in a humidified chamber and were passaged 3e5 times for experiments. RAW 264.7 cells were cultured in RPMI-1640 medium (Invitrogen) supplemented with 10% FBS, 100 IU/ml penicillin and 100 IU/ml streptomycin.

2.2. Recombinant BV preparation and transduction To construct donor plasmids (pBacECre, pBacLEW and pBacLEBW), the EF-1a promoter was PCR-amplified from pVITRO1neo-mcs (Invivogen) and cloned into pBacCre, pBacLCW and pBacLCBW [24] using XhoI/BamHI to replace the CMV promoter because EF-1a promoter was stronger than the CMV promoter in rat BMSCs (data not shown). To generate pBac140S (encoding miR140 sponge) and pBac214S (encoding miR-214 sponge), oligonucleotides encoding 10 tandem rno-miR-140* or hsa-miR-214 binding sites together with flanking 50 -NotI and 30 -AflII sites were chemically synthesized (Table S1) and cloned into pUC57 (Invitrogen) to yield pUC-miR140S and pUC-miR214S. The NotI-miR140S-AflII (or NotI-miR-214S-AflII) fragment was NotI/AflII digested from pUC-miR140S (or pUC-miR214S) and subcloned into pd2EGFP-N1 (Clontech) to yield pd2EGFP-140S (or pd2EGFP-214S) (Fig. S1). Next, the BamHI-d2EGFP-miR-140S-EcoRI (or BamHId2EGFP-miR-214S-EcoRI) fragment was BamHI/EcoRI digested from pd2EGFP-140S (or pd2EGFP-214S) and subcloned into pBacLEW to yield pBac140S (or pBac214S) (Fig. S2). All donor plasmids (pBacECre, pBacLEBW, pBac140S and pBac214S) were used to generate BV vectors (BacECre, BacLEBW, Bac140S and Bac214S) by using the Bac-To-Bac® system (Invitrogen) following the manufacturer's instructions. Virus titers were determined by end-point dilution method and are expressed as plaque forming units per milliliter (pfu/ml) [11]. BMSCs were transduced with BV vectors as described [11]. Briefly, rat BMSCs were seeded to 6-well plates (2  105 cells/well) or T-75 flasks (5  106 cells/flask), cultured overnight and washed twice with phosphate-buffered saline (PBS, pH 7.4) prior to transduction. Depending on the multiplicity of infection (MOI), a certain volume of virus supernatant was mixed with NaHCO3-free DMEM at a volumetric ratio of 1:4 (total volume was 0.5 and 2 ml in 6-well plates and T-75 flasks, respectively) and added to the cells. For mock transduction, virus-free TNM-FH medium was mixed with NaHCO3-free DMEM at a volumetric ratio of 1:4 and added to the cells. The cells were gently shaken on a rocking plate at room temperature for 6 h. After the transduction period, the virus mixture was replaced with the osteoinduction medium (DMEM containing 10% FBS, 100 IU/ml penicillin, 100 IU/ml streptomycin, 0.1 mM dexamethasone, 10 mM b-glycerol phosphate and 50 mM ascorbic acid 2-phosphate) containing 3 mM sodium butyrate (all from sigma). After 15 h of incubation at 37  C, the medium was replaced by fresh osteoinduction medium. 2.3. Quantitative real-time reverse-transcription PCR (qRT-PCR) Total RNA was isolated using the NucleoSpin RNA II kit (Machereye-Nagel) and reverse transcribed to cDNA using the Omniscript RT Kit (Qiagen). The osteogenic and osteoclastic genes were analyzed by qPCR using StepOnePlus Real-Time PCR Systems (Applied Biosystems) and gene-specific primers (Table S2). The mature miRNA was isolated using Trizol (Invitrogen) and analyzed using the TaqMan MicroRNA Assays kit (Applied Biosystems). The gene expression levels were normalized against that of U6 (for miRNAs) or gapdh (for osteogenic and osteoclastic genes) and referenced to that of the mock-transduced Sham-BMSCs (for miRNAs), mock-transduced OVX-BMSCs (for osteogenic genes) or Raw 264.7 co-cultured with mock-transduced Sham-BMSCs (for osteoclastic genes). 2.4. Alizarin red staining, calcium deposition assay and ELISA The transduced OVX-BMSCs and mock-transduced BMSCs (OVX or Sham) were cultured with osteoinduction medium and stained

K.-C. Li et al. / Biomaterials 74 (2016) 155e166

by Alizarin red which can chelate with calcium to form an alizarin redecalcium complex (red color). After removing the medium, the cells were washed twice with PBS and fixed with 4% paraformaldehyde for 10 min. After fixation, the cells were washed 3 times with deionized water and then stained with 40 mM Alizarin red (pH 4.2) for 10 min, followed by 3 PBS washes and microscopic observation. To quantify the calcium deposition, the cells were washed 3 times with PBS at 15 dpt, followed by incubation with 0.6 N HCl overnight. The supernatant was collected the next day for calcium phosphate deposition analysis (CALCIUM liquicolor, Human Inc.). The calcium deposition is expressed as nmol. For BMP2 and OPG ELISA analysis, the medium was collected at different dpt and analyzed using the Human BMP-2 DuoSet ELISA kit (R&D Systems) and Quantikine® osteoprotegerin ELISA kit (R&D Systems), respectively. 2.5. Co-culture assay for osteoclasts differentiation and TRAP staining The transduced OVX-BMSCs and mock-transduced BMSCs (OVX or Sham) were osteoinduced by culturing in the insert of 24-well Transwell plates (2  103 cells/well, 0.4 mm diameter pore, Corning) using the osteoinduction medium. After 15 days of culture, Raw 264.7 cells were seeded to the 24-well Transwell plates (5  103 cells/well) and co-cultured with the osteoinduced cells using the osteoclastic medium (osteoinduction medium containing 60 ng/ml recombinant mouse RANKL (R&D Systems)) for 5 days. After 5 days of co-culture, Raw 264.7 cells were stained for tartrateresistant acid phosphatase (TRAP) activity and quantified for the number of osteoclasts using the leukocyte acid phosphatase kit (Sigma) through forming insoluble diazonium fast garnet GBC salt deposits (purplish red color) when it accumulates rapidly at acid pH. TRAPþ multinuclear osteoclasts (>3 nuclei/cell) were counted using the bright-field microscope. 2.6. Preparation of BMSCs/gelatin constructs The gelatin sponge scaffolds were prepared by cutting the Spongostan™ gelatin sponge (porosity z 97%, cat#MS0003, Ethicon) into disks (diameter z 3 mm) and subsequent immersion in saline solution (thickness z 2 mm after immersion). The transduced and mock-transduced OVX-BMSCs were trypsinized, resuspended in DMEM medium, seeded into the gelatin sponge scaffolds (2  106 cells/scaffold) and allowed to adhere in the 12-well plate. After 1 h, the BMSCs/gelatin constructs were cultured in complete medium containing 3 mM sodium butyrate for 12 h and then implanted into the bone defect of osteoporotic rat. 2.7. Femoral metaphysis defect model The osteoporotic rat models were generated by ovariectomy and 4 weeks of methylpredinisolone hemisuccinate injection as described in Sec. 2.1. The osteoporotic rats were anesthetized by intramuscular injection of Zoletil 50 (25 mg/kg body weight) and 2% Rompun® (0.15 ml/kg body weight), and the distal epiphysis of the left femur was osteotomized using a 3-mm diameter trephine bur. The bone defect was filled with 1 BMSCs/gelatin implant (2  106 cells/animal), rinsed with 0.9% saline and the deep muscle layer and skin were closed. 2.8. mCT The rats were sacrificed at 2 or 4 weeks post-implantation and the femora were removed and scanned using an animal mCT imaging system (SkyScan 1174, Bruker, Belgium) at the tube voltage of

157

100 kV, 16 mm resolution. The 3D images were reconstructed using the CTVox software (Bruker, Belgium). For evaluation of bone regeneration and microarchitecture within the defect, volume of interest (VOI, diameter ¼ 3 mm) was drawn within the femoral metaphysis defect and analyzed using the CTAn software (Bruker, Belgium) to calculate the new bone formation (BV/TV), bone mineral density (BMD), trabecular thickness (Tb.Th), trabecular number (Tb.N) and distance between trabeculae (Tb.Sp). 2.9. Histological and immunohistochemical staining After mCT scanning, the femora were immersed in 0.5 M EDTA (pH 8.0) for 15e20 days for complete decalcification, and then dehydrated in a series of graded concentration of ethanol from 70% to 100%. After embedding in paraffin and coronal sectioning (thickness ¼ 3 mm) from the mid-shaft of femora to distal epiphysis of femora containing the defect site, the sections were stained with hematoxylin and eosin (H&E). The sections were de-paraffined and rehydrated for immunohistochemical staining specific for osteopontin and bone sialoprotein. The antigen retrieval was performed by incubation with trypsin for 1 h at 37  C. The primary and secondary antibodies were rabbit anti-rat osteopontin polyclonal Ab (1:200 dilution, Abcam), rabbit anti-rat bone sialoprotein polyclonal Ab (1:200 dilution, Abcam) and goat anti-rabbit, HRP-conjugated polyclonal Ab (1:200 dilution, Abcam), respectively. Finally, the samples were stained with 3,30 -Diaminobenzidine (DAB, Sigma), and then counterstained with Gill's Hematoxylin (Dako). The histochemical staining of TRAP was performed using leukocyte acid phosphatase kit (Sigma). 2.10. Statistical analysis All quantitative data were analyzed using one-way analysis of variance (ANOVA) or student's t-test using a two-tailed distribution. The in vitro data represent the means ± SD of at least 3 independent experiments. p < 0.05 was considered significant. 3. Results 3.1. miR-140*/miR-214 suppression in OVX-BMSCs by BV-mediated miRNA sponges expression To examine whether aberrant miRNA expression occurred in osteoporotic BMSCs, we first created the osteoporotic rat model by ovariectomy (OVX) and injection of methylprednisolone hemysuccinate for 4 weeks [23]. After confirming the occurrence of osteoporosis in OVX rats (Fig. S3), we isolated BMSCs from nonoperated (Sham) or ovariectomized rats. Because miR-23b, miR30a, miR-130a, miR-140* and miR-214 were recently found to be significantly upregulated in the femoral bones of aged individuals with fractures, which suggested the roles of these miRNAs in aging and osteoporosis [25], we performed qRT-PCR for these miRNAs. Fig. 1A unveils that, except miR-23b, miR-30b, miR-138, miR-140* and miR-214 levels were significantly (p < 0.05) higher in OVXBMSCs than in Sham-BMSCs. Of note, miR-140* and miR-214 levels were particularly upregulated in OVX-BMSCs, which prompted us to postulate that excessive miR-140* and/or miR-214 overexpression might play roles in the poor osteogenesis of OVXBMSCs and/or excessive osteoclast activities. To assess our hypothesis, we first constructed Cre/loxP-based BV vectors for sustained expression of miRNA sponges to decoy miR140* or miR-214 in OVX-BMSCs (Fig. 1B). BacECre expressed the Cre recombinase under the EF-1a promoter. The hybrid BV vectors Bac140S and Bac214S carried 10 tandem copies of sponges specifically antagonizing miR-140* and miR-214, respectively (Fig. 1B).

158

K.-C. Li et al. / Biomaterials 74 (2016) 155e166

Fig. 1. MiR-140*/miR-214 suppression in OVX-BMSCs by BV-mediated miRNA sponges expression. (A) Endogenous miR-23b, miR-30b, miR-138, miR-140* and miR-214 levels in OVX- and Sham-BMSCs. (B) Schematic illustration of BV vectors. (C) miR-140* and miR-214 levels after BV transduction. The miRNA levels were determined by TaqMan qRT-PCR using U6 small nuclear RNA as an internal control. The data were normalized to those of mock-transduced Sham-BMSCs. All data represent the means ± SD of three independent culture experiments. d2EGFP, destabilized enhanced green fluorescent protein; WPRE, woodchuck hepatitis posttranscriptional regulatory element.

The sponge arrays in Bac140S and Bac214S were located downstream of d2egfp gene and the gene cassettes were flanked by loxP sites, so that after co-transduction with BacECre the cassettes could be excised from the hybrid BV genome, forming DNA minicircles for prolonged expression. Additionally, we constructed a hybrid BV carrying a loxP-flanking BMP2 expression cassette (BacLEBW, Fig. 1B). We fine-tuned the co-transduction conditions and found that co-transduction of rat OVX-BMSCs with BacECre (MOI 100) and BacLEBW (MOI 150) for 6 h conferred enhanced and prolonged BMP2 expression without appreciable cell death (Fig. S4). Therefore, we co-transduced OVX-BMSCs with BacECre at MOI 100 and Bac140S (or Bac214S) at MOI 150 for 6 h. Mock-transduced OVX-BMSCs and Sham-BMSCs served as controls. As shown in Fig. 1C, mock-transduced OVX-BMSCs expressed significantly higher levels of miR-140* and miR-214 than mock-transduced Sham-BMSCs, but co-transduction of OVX-BMSCs with BacECre/ Bac140S (140S group) and BacECre/Bac214S (214S group) significantly knocked down the miR-140* and miR-214 levels, respectively. The miR-140* (Fig. S5A) and miR-214 (Fig. S5B) suppression persisted for at least 14 days.

3.2. Enhanced OVX-BMSCs osteogenesis by miR-140*/miR-214 sponges expression The osteogenesis capacity of BMSCs derived from osteoporotic individuals was reportedly compromised [26]. To explore whether suppressing miR-140* or miR-214 levels with miRNA sponges was able to restore the osteogenesis capability of OVX-BMSCs, ShamBMSCs were mock-transduced as a control (Sham group) while OVX-BMSCs were mock-transduced (Mock group) or co-transduced with BacECre/Bac140S (140S group) or BacECre/Bac214S (214S group), followed by culture in osteogenic medium. The levels of

osteogenic markers at early (runx2), middle (alkaline phosphatase, alp) and late (osteocalcin, ocn) stages were measured at different days post-transduction (dpt) using qRT-PCR and normalized to those of mock-transduced OVX-BMSCs at the same day (Fig. 2A). In comparison with the Mock group, both 140S and 214S groups significantly elevated the runx2, alp and ocn levels. Notably, the 214S group exerted significantly (p < 0.05) stronger effects than the 140S group, and resulted in statistically similar (p > 0.05) alp and ocn levels when compared with the Sham group. The Alizarin red staining (Fig. 2B) and calcium deposition assay (Fig. 2C) further attested that the 214S group substantially enhanced the mineralization and calcium deposition at 15 dpt and imparted stronger osteoinductive effects than the 140S group. These data implicated the potentials of miR-214 sponge expression to restore the OVXBMSCs osteogenic capability.

3.3. Inhibition of osteoclast maturation by miR-140*/miR-214 sponge expression Osteoclast maturation requires the interaction of receptor activator of nuclear factor kappa-B (RANK) and RANK ligand (RANKL). Osteoprotegerin (OPG) blocks the interaction of RANK with RANKL, thus inhibiting osteoclast maturation [27]. qRT-PCR analysis at 15 dpt (Fig. 3A) depicted that, when compared with the Mock group, the 214S group significantly lowered the rankl level while enhanced the opg level, leading to a significantly reduced rankl/opg ratio, a key indicator of the homeostasis between bone formation and resorption [28]. The 140S group did not suppress the rankl transcription but enhanced the opg transcription, thus reducing the rankl/opg ratio as well (Fig. 3A). Moreover, ELISA analysis of the supernatant (Fig. 3B) revealed that the 214S group expressed significantly higher levels of OPG than the Mock group.

K.-C. Li et al. / Biomaterials 74 (2016) 155e166

159

Fig. 2. Enhanced OVX-BMSCs osteogenesis and inhibition of osteoclast maturation by miR-140*/miR-214 sponges expression (A) Expression profiles of runx2, alkaline phosphatase (alp) and osteocalcin (ocn) genes after BV transduction. (B) Alizarin red staining at 15 dpt. (C) Calcium deposition assay at 15 dpt. The osteogenic gene expression profiles were measured at different dpt using qRT-PCR and normalized to those of the mock-transduced OVX-BMSCs at the same day. All data represent the means ± SD of three independent culture experiments.

To examine whether the OPG secreted by the BV-engineered OVX-BMSCs did inhibit the osteoclast maturation, rat OVXBMSCs were mock-transduced or transduced as in Fig. 3A and the Sham group served as a control. The cells were cultured in the transwell insert on the culture plate for 15 days, and then cocultured with RAW 264.7 cells (which are able to differentiate

into osteoclasts) in the plate containing osteoclast induction medium for 5 days [29]. At 20 dpt, the cells were stained for tartrate resistant acid phosphatase (TRAP, an indicator of osteoclast maturation, Fig. 3C) and quantified for TRAPþ multinuclear osteoclasts with >3 nuclei/cell (Fig. 3D). Fig. 3DeE demonstrate that both the 140S and 214S groups mitigated osteoclast maturation,

160

K.-C. Li et al. / Biomaterials 74 (2016) 155e166

Fig. 3. Inhibition of osteoclast maturation by miR-140*/miR-214 sponges expression. (A) Expression profiles of rankl and opg genes and rankl/opg ratio after BV transduction. (B) OPG expression profile as measured by ELISA. (C) Osteoclast maturation was determined by TRAP staining. (D) Quantitative TRAPþ multinuclear osteoclasts (>3 nuclei/cell). (E) Expression profiles of cathepsin K and osteoclast-associated receptor (OSCAR). The rankl and opg levels were measured at 15 dpt using qRT-PCR and normalized to those of the mocktransduced OVX-BMSCs. For osteoclast maturation assay, the BMSCs were seeded to the insert and co-cultured with Raw 264.7 cells in the 24-well Transwell plates in the medium containing 60 ng/ml RANKL. After 5 days co-culture, Raw 264.7 cells were stained for TRAP activity and quantified for the number of multinucleated osteoclasts. For cathepsin K and OSCAR genes expression, the gene profiles were measured at 20 dpt using qRT-PCR and normalized to those of Raw 264.7 co-cultured with mock-transduced Sham-BMSCs. All data represent the means ± SD of three independent culture experiments.

when compared with the Mock group. qRT-PCR analysis attested that both the 140S and 214S groups downregulated the levels of osteoclast maturation marker genes, cathepsin K and osteoclastassociated receptor (OSCAR), when compared with the Mock group (Fig. 3E). Fig. 3 collectively demonstrates that both miR-140* and miR-214 sponge-expressing OVX-BMSCs mitigated osteoclast maturation in a paracrine fashion.

3.4. In vivo osteoporotic bone repair: evaluation by mCT imaging To assess the potential of the BV-engineered OVX-BMSCs for bone healing in osteoporotic rats, we chose Bac214S because miR214 sponge expression more potently induced osteogenesis (Fig. 2). OVX-BMSCs were co-transduced with BacECre/BacLEBW (LEBW group), BacECre/Bac214S (214S group), or BacECre/

K.-C. Li et al. / Biomaterials 74 (2016) 155e166

161

Fig. 4. Bone healing as evaluated by mCT. OVX-BMSCs were mock-transduced (Mock group, n ¼ 6 at each time point), co-transduced with hybrid BV (LEBW, 214S or LEBW/214S group, n ¼ 8 for each group at each time point), seeded to gelatin scaffolds and implanted into the critical-size defect (3 mm in diameter) at the left femoral metaphysis in osteoporotic rats. The rats were sacrificed at 2 (2W) and 4 (4W) weeks post-implantation, and the left femora were removed for mCT analysis. The open arrowheads indicate the defect regions.

BacLEBW/Bac214S (LEBW/214S Group), seeded into gelatin scaffolds at 1 dpt, and implanted into the critical-size defect (3 mm in diameter) at the left femoral metaphysis in osteoporotic rats created as in Fig. 1 (n ¼ 8 for each group at each time point). As controls, OVX-BMSCs were mock-transduced, seeded and implanted in the same manner (Mock group, n ¼ 6 at each time point). The bone repair in each group was evaluated by mCT analysis of the left femora removed at weeks 2 (2W) and 4 (4W). Fig. 4 illustrates poor bone healing in the Mock and LEBW groups at both time points, indicating that OVX-BMSCs/gelatin constructs, with or without BMP2 expression, failed to heal the femoral metaphysis defects in osteoporotic rats. In contrast, the 214S group triggered new bone formation at 2W, and nearly filled the defect with osseous tissue at 4W. The LEBW/214S group not only elicited abundant new bone formation at 2W, but also completely filled the defect at 4W, which resembled the counterpart region in the nonoperated OVX rats (Fig. 4). The front and rear views of 3D reconstructed images (Fig. 5) further confirmed that the 214S and LEBW/ 214S groups repaired both exterior and interior sides of the defects at 4W. Of note, the LEBW/214S group also developed trabecular bone-like structure within the femoral metaphysis. Quantitative analysis using the mCT data further attested that all indicators of new bone formation, including the ratio of bone volume to total volume (BV/TV, Fig. 6A), bone mineral density (BMD, Fig. 6B), trabecular thickness (Tb.Th, Fig. 6C) and trabecular number (Tb.N, Fig. 6D), were significantly better in the 214S and LEBW/214S groups than in the Mock and LEBW groups at 4W. Notably, at 4W the LEBW/214S group was significantly (p < 0.05) better than the

214S group in terms of bone formation (BV/TV) and bone microstructure (BMD, Tb.Th. and Tb.N). Furthermore, the LEBW/214S group resulted in the shortest distance between trabeculae (Tb.Sp) among all groups at 4W (Fig. 6E). In comparison with the nonoperated group (OVX rats without osteotomy), the LEBW/214S group also conferred significantly (p < 0.05) higher BV/TV (27.1 ± 1.7% vs. 22.3 ± 4.8%), BMD (0.39 ± 0.01 g/cm3 vs. 0.34 ± 0.02 g/cm3), Tb.Th (1.90 ± 0.21 mm vs. 1.47 ± 0.12 mm) and Tb.N (2.08 ± 0.24/mm vs. 1.68 ± 0.06/mm), and shorter Tb.Sp (0.59 ± 0.07 mm vs. 0.78 ± 0.08 mm). These data altogether indicated that the OVX-BMSCs co-expressing BMP2 and miR-214 sponges substantially improved the quality of regenerated bone. 3.5. Osteoporotic bone repair: evaluation by histology To evaluate the bone healing histologically, the coronal femoral sections from the mid-shaft to distal epiphysis encompassing the defect site were subjected to H&E and immunohistochemical staining. As shown in Fig. 7A, the Mock and LEBW groups were filled with fibrous tissues without evident new bone formation. The 214S group was filled with trabecular bone in the defect site, yet the bone structure was loose. In contrast, at 4W the LEBW/214S group already gave rise to cortical bone-like structure. Osteopontin (OPN) and bone sialoprotein (BSP) are markers of osteoblasts and osteocytes, respectively [30]. Immunohistochemical staining demonstrated abundant deposition of OPN (Fig. 7B) and BSP (Fig. 7C) in the 214S, LEBW/214S and non-operated groups but not in the Mock and LEBW groups. Of note, we observed no

162

K.-C. Li et al. / Biomaterials 74 (2016) 155e166

Fig. 5. Bone healing as evaluated by 3D reconstructed images. The front and rear views of the femora are shown.

distinguishable difference between the LEBW/214S and nonoperated groups. Since successful bone repair hinges on bone remodeling that follows initial bone formation, we performed immunohistochemical staining for TRAP, a marker of osteoclast activity and bone remodeling. Fig. 7D illustrates denser TRAPþ staining in the 214S group than the non-operated group, suggesting that the 214S group was still undergoing active remodeling. Conversely, the LEBW/214S group exhibited similar TRAPþ staining when compared with the non-operated group, indicating that bone repair in the LEBW/214S group already progressed through the active bone remodeling stage, which accounted for the superior formation of cortical bone-like structure. 4. Discussion Fractures associated with osteoporosis has become a worldwide health problem [3], yet the repair of bone defects of osteoporotic

patients captures less attention. To enhance bone healing in osteoporotic animals, this study aimed to develop a cell/gene therapy approach in combination with miRNA manipulation. Although miRNA functions in osteogenesis, osteoblastogenesis and osteoclastogenesis have been explored, little is known about the roles of miRNA in osteoporosis in the clinical setting [7]. One study revealed that a mutation in pre-miR-2861 causes primary osteoporosis in 2 adolescent patients, because miR-2861 promotes osteoblast differentiation by repressing histone deacetylase 5 which would otherwise enhance the degradation of osteogenic transcription factor Runx2 [31]. Another study unveiled that miR-133a in human circulating monocyte is upregulated in patients with a low BMD [32]. Recently, 5 miRNAs (miR-21, miR-23a, miR-24, miR-25, miR-100) were discovered to be upregulated in the serum and bone tissue in osteoporotic patients, suggesting their roles in osteoporosis [2]. Here we unraveled the aberrant overexpression of 4 miRNAs in the BMSCs isolated from osteoporotic rats (OVX-BMSCs), among

K.-C. Li et al. / Biomaterials 74 (2016) 155e166

163

Fig. 6. Bone microarchitecture parameters analyzed from mCT imaging data. (A) The ratio of bone volume to total volume (BV/TV). (B) Bone mineral density (BMD). (C) Average trabecular thickness (Tb.Th). (D) Trabecular numbers per square millimeter (Tb.N). (E) Average distance between trabecular (Tb.Sp). The mCT images were analyzed using the CTAn software in the cylindrical volume of interest (VOI) for the calculation of bone microarchitecture parameters.

which miR-140* and miR-214 were particularly upregulated (Fig. 1A). Engineering of the OVX-BMSCs with the hybrid Cre/loxPbased BV vector encoding the miR-140* or miR-214 sponges (Fig. 1B) successfully expressed the sponges for at least 14 days (Fig. S5) and significantly attenuated the miR-140* and miR-214 levels in the OVX-BMSCs (Fig. 1C). Ablating the miR-140* or miR214 levels in vitro indeed promoted the osteogenesis of OVXBMSCs (Fig. 2) and, strikingly, augmented the ability of OVXBMSCs to repress osteoclast maturation (Fig. 3).

MiR-140 is highly expressed in undifferentiated human MSCs [33], tissue-engineered chondrocytes [34] and in articular cartilage [35], thus miR-140 has been considered a cartilage-specific miRNA [36]. Yet miR-140* is also involved in early bone development, probably by targeting Jagged 1 (JAG1) and Tmem119 [37]. JAG1 is a Notch one receptor ligand, which induces MSCs to differentiate into osteoblast through Runx2 and ALP activation [38] while Tmem119 induces Smad1/5 and Runx2 transcriptional activity, which promotes osteogenic differentiation [39]. As such, miR-140*

164

K.-C. Li et al. / Biomaterials 74 (2016) 155e166

Fig. 7. Histological and immunohistochemical staining evaluated at 4 weeks post-implantation. (A) H&E staining. (B) Osteopontin (OPN) immunohistochemical staining. (C) Bone sialoprotein (BSP) immunohistochemical staining. (D) TRAP staining. Black arrows indicate OPN; black arrowheads indicate BSP and the open arrowheads indicate TRAP. F, fibrous tissue; B, bone tissue.

overexpression might suppress JAG1 and Tmem119, leading to OVX-BMSCs’ attenuated osteogenesis capability (Figs. 2 and 3). Our data supported the roles of miR-140* in modulating bone homeostasis and implicated the potential of suppressing miR-140* as a novel approach to treating osteoporosis. In the osteoporotic rat models, implanting the OVX-BMSCs failed to heal the bone defect (Figs. 4e7). Transplanting the BVengineered OVX-BMSCs that persistently expressed BMP2 improved the healing, but failed to completely heal the bone defects. Since BV-engineered, BMP2-expressing BMSCs or ASCs can heal large critical-size bone defects in rabbits [14e17] and minipigs [18], the failure to repair the defects using the BMP2-expressing OVX-BMSCs in this study underscored the difficulty to heal bone defects in osteoporotic animals using OVX-BMSCs, which agreed with the poor bone healing capability in osteoporotic rats [40]. The failure of BMP2 expression to heal bone defects in osteoporotic animals might be at least partly attributed to the excessive miR-140 expression (Fig. 1A), which was reported to inhibit the BMP2associated pathway critical to bone formation [33]. Strikingly, suppressing the miR-214 levels imposed potent effects in promoting osteogenesis and impeding osteoclast maturation in vitro (Figs. 2 and 3). Implanting the OVX-BMSCs that expressed miR-214 sponges remarkably augmented the healing (Figs. 4 and 5) and substantially ameliorated the microarchitecture

of trabecular bone (Fig. 6) and bone remodeling (Fig. 7). MiR-214 has been shown to dictate the development of nervous system [41], teeth [42], pancreas [43] and hair follicle [44], to impede angiogenesis [45] and protect heart [46]. With regard to osteoporosis, deletion of miR-214 leads to skeletal abnormalities [47] and elevated miR-214 levels correlates with a lower degree of bone formation in bone specimens from aged patients with fractures [25]. Importantly, miR-214 inhibits osteoblast differentiation and bone formation in skeletal disorders and suppressing miR-214 by antagomirs promotes bone formation and increases bone mass in aged OVX-induced osteoporotic mice. The regulatory role of miR214 in osteogenesis is ascribed to the ability of miR-214 to target ATF-4, a transcription factor regulating osteoblast differentiation and function [25]. MiR-214 also suppresses osteogenic differentiation by targeting osterix [48], a transcription factor crucial for the differentiation of preosteoblasts into functional osteoblasts [3]. Conversely, our bioinformatic analysis using TargetScan 6.2 (http:// targetscan.org) indicated that miR-214 might target fibroblast growth factor receptor 1 (FGFR1), Wnt-induced secreted protein 1 (WISP-1) and E3 ubiquitin ligase Cbl. Downregulation of FGFR1 [49], WISP-1 [50] and Cbl [51] inhibit or reduce the levels of Runx2, which controls early osteoblasts commitment. Based on the aforementioned findings, we propose that aberrant overexpression of miR-214 in the OVX-BMSCs directly

K.-C. Li et al. / Biomaterials 74 (2016) 155e166

165

detected in human ovarian cancers and induces cell survival and chemoresistance through targeting the tumor suppressor PTEN, leading to the downregulation of PTEN and activation of the oncogenic Akt pathway [56]. In ovarian cancer stem cells (OCSC), miR-214 directly represses p53 and knockdown of miR-214 also mitigates OCSC properties [57]. Furthermore, miR-214 is frequently upregulated in osteosarcoma specimens and may function as an onco-miRNA in osteosarcoma [58]. As such, the approach we develop to persistently suppress miR-214 may be extended to the treatment of ovarian cancer and osteosarcoma. 5. Conclusions

Fig. 8. Putative mechanism by which miR-214 orchestrates osteogenesis and osteoclastogenesis. Aberrant overexpression of miR-214 in the OVX-BMSCs directly target ATF [25] and osterix [48], and might indirectly suppress Runx2 via the regulation of FGFR1, WISP-1 and Cbl, thereby blocking the differentiation of OVX-BMSCs towards preosteoblast and mature osteoblast. Antagonizing miR-214 by the BV-mediated miR214 sponge expression might enhance the activities of these osteogenic transcription factors, thereby promoting the differentiation of OVX-BMSCs towards osteoblast, as evidenced by the elevated ALP and OCN accumulation (Fig. 2). The enhanced BMSCs differentiation and ensuing osteoblast maturation led to elevated OPG secretion, which antagonizes RANKL and hence impede the osteoclast maturation (Fig. 3). FGFR1, fibroblast growth factor receptor 1; WISP-1, Wnt-induced secreted protein 1; Cbl, E3 ubiquitin ligase.

In summary, we demonstrated that the Cre/loxP-based BVmediated persistent expression of miR-140* and miR-214 sponges effectively augmented the osteogenic differentiation of OVX-BMSCs and mitigated osteoclast maturation, and suppressing miR-214 exerted more potent effects. Allotransplantation of the miR-214 sponges-expressing osteoporotic BMSCs into the critical-size defect at the femoral metaphysis of OVX oeteoporotic rat potentiated the bone healing, remodeling and ameliorated the bone quality (density, trabecular number, trabecular thickness and trabecular space) at 4W. Co-expressing BMP2 and miR-214 sponges in OVX-BMSCs synergistically substantiated the healing. To date, the majority of drugs for osteoporosis management either promote bone formation or inhibit bone resorption. The OVX-BMSCs-based cell therapy in conjunction with miR-214 sponge expression, with or without BMP2 expression, not only stimulates bone formation, but also mitigates excessive bone resorption, thus paving a new avenue to the treatment of osteoporotic bone defects. Acknowledgments

downregulated ATF and osterix as well as indirectly suppressed Runx2 via the regulation of FGFR1, WISP-1 and Cbl, thereby blocking the differentiation of OVX-BMSCs towards preosteoblast and mature osteoblast (Fig. 8). Antagonizing miR-214 by the BVmediated miR-214 sponge expression might enhance the activities of these osteogenic transcription factors, thereby promoting the differentiation of OVX-BMSCs towards osteoblast (Fig. 8). However, the exact molecular mechanisms necessitate further experimental validation. Besides enhancing osteogenesis, suppressing osteoclastogenesis is important in the treatment of osteoporosis and several miRNAs have been implicated in osteoclastogenesis. For instance, miR-21 is a pro-osteoclastogenic miRNA and may reduce bone resorption in pre-menopausal women with robust estrogen levels [9]. Recently, miR-214 has been also reported to promote osteoclastogenesis in bone marrow monocytes through PI3K/Akt pathway by targeting phosphatase and tensin homolog [52]. Here we also uncovered that knocking down miR-214 levels suppressed the osteoclast maturation in vitro and bone resorption in vivo. This might be attributed to the ameliorated differentiation of OVX-BMSCs, which secreted OPG (Fig. 3) to inhibit osteoclastogenesis in a paracrine fashion (Fig. 8). Intriguingly, co-expressing BMP2 and miR-214 sponge synergistically enhanced the bone formation in osteoporotic rats (Figs. 4e7). It is known that miR-214 suppresses b-catenin pathway [53] which directs osteogenic differentiation by enhancing mesenchymal cell responsiveness to osteogenic factors such as BMP2 [54]. As such, miR-214 sponge expression might repress miR214 which, in turn, activated the b-catenin pathway and enhanced the responsiveness of osteoporotic BMSCs to the BMP2-mediated osteoinduction, thereby further potentiating the bone repair. Aside from bone healing, miR-214 is involved in the induction of angiogenesis in endothelial cells [55]. MiR-214 overexpression is

This work was supported by the National Tsing Hua University (Toward World-Class University Project 104N2050E1 and NTHUCGMH Joint Research Program 103N2758E1), CGMH Intramural Project (CMRPG3E0441, CMRPG3E0061, CMRPG3E0062, CMRPG3B1542) and Ministry of Science and Technology (101-2628-E-007-009-MY3, 101-2923-E-007-002-MY3 and MOST 103-2221-E-007-093-MY3), Taiwan. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.biomaterials.2015.09.046. References [1] S. Muruganandan, H.J. Dranse, J.L. Rourke, N.M. McMullen, C.J. Sinal, Chemerin neutralization blocks hematopoietic stem cell osteoclastogenesis, Stem Cells 31 (2013) 2172e2182. [2] C. Seeliger, K. Karpinski, A.T. Haug, H. Vester, A. Schmitt, J.S. Bauer, et al., Five freely circulating miRNAs and bone tissue miRNAs are associated with osteoporotic fractures, J. Bone Min. Res. 29 (2014) 1718e1728. [3] K.M. Kim, S.K. Lim, Role of miRNAs in bone and their potential as therapeutic targets, Curr. Opin. Pharmacol. 16 (2014) 133e141. [4] C. Deal, Potential new drug targets for osteoporosis, Nat. Clin. Pract. Rheumatol. 5 (2009) 20e27. [5] K.-C. Li, Y.-C. Hu, Cartilage tissue engineering: recent advances and perspectives from gene regulation/therapy, Adv. Healthc. Mater. 4 (2015) 948e968. [6] J.B. Lian, G.S. Stein, A.J. van Wijnen, J.L. Stein, M.Q. Hassan, T. Gaur, et al., MicroRNA control of bone formation and homeostasis, Nat. Rev. Endocrinol. 8 (2012) 212e227. [7] A.J. van Wijnen, J. van de Peppel, J.P. van Leeuwen, J.B. Lian, G.S. Stein, J.J. Westendorf, et al., MicroRNA functions in osteogenesis and dysfunctions in osteoporosis, Curr. Osteoporos. Rep. 11 (2013) 72e82. [8] E.A. Clark, S. Kalomoiris, J.A. Nolta, F.A. Fierro, Concise review: MicroRNA function in multipotent mesenchymal stromal cells, Stem Cells 32 (2014) 1074e1082. [9] T. Sugatani, K.A. Hruska, Down-regulation of miR-21 biogenesis by estrogen

166

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25] [26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

K.-C. Li et al. / Biomaterials 74 (2016) 155e166 action contributes to osteoclastic apoptosis, J. Cell Biochem. 114 (2013) 1217e1222. K.J. Airenne, Y.-C. Hu, T.A. Kost, R.H. Smith, R.M. Kotin, C. Ono, et al., Baculovirus: an insect-derived vector for diverse gene transfer applications, Mol. Ther. 21 (2013) 739e749. L.-Y. Sung, C.-L. Chen, S.-Y. Lin, K.-C. Li, C.-L. Yeh, G.-Y. Chen, et al., Efficient gene delivery into cell lines and stem cells using baculovirus, Nat. Protoc. 9 (2014) 1882e1899. C.-Y. Lin, Y.-H. Chang, C.-Y. Kao, C.-H. Lu, L.-Y. Sung, T.-C. Yen, et al., Augmented healing of critical-size calvarial defects by baculovirus-engineered MSCs that persistently express growth factors, Biomaterials 33 (2012) 3682e3692. C.-Y. Lin, Y.-H. Chang, K.-C. Li, C.-H. Lu, L.-Y. Sung, C.-L. Yeh, et al., The use of ASCs engineered to express BMP2 or TGF-b3 within scaffold constructs to promote calvarial bone repair, Biomaterials 34 (2013) 9401e9412. C.-Y. Lin, Y.-H. Chang, K.-J. Lin, T.-Z. Yen, C.-L. Tai, C.-Y. Chen, et al., The healing of critical-sized femoral segmental bone defects in rabbits using baculovirusengineered mesenchymal stem cells, Biomaterials 31 (2010) 3222e3230. C.-Y. Lin, Y.-H. Chang, L.-Y. Sung, C.-H. Lu, C.-L. Chen, S.-Y. Lin, et al., Long-term follow-up of massive segmental bone healing mediated by hybrid baculovirus-transduced ASCs: Focuses on bone remodeling and potential side effects, Tissue Eng. Part A 20 (2014) 1392e1402. C.-Y. Lin, K.-J. Lin, C.-Y. Kao, M.-C. Chen, T.-Z. Yen, W.-H. Lo, et al., The role of adipose-derived stem cells engineered with the persistently expressing hybrid baculovirus in the healing of massive bone defects, Biomaterials 32 (2011) 6505e6514. C.-Y. Lin, K.-J. Lin, K.-C. Li, L.-Y. Sung, S. Hsueh, C.-H. Lu, et al., Immune responses during healing of massive segmental femoral bone defects mediated by hybrid baculovirus-engineered ASCs, Biomaterials 33 (2012) 7422e7434. C.-Y. Lin, Y.-H. Wang, K.-C. Li, L.-Y. Sung, C.-L. Yeh, K.-J. Lin, et al., Healing of massive segmental femoral bone defects in minipigs by allogenic ASCs engineered with FLPo/Frt-based baculovirus vectors, Biomaterials 50 (2015) 98e106. Y.-H. Liao, Y.-H. Chang, L.-Y. Sung, K.-C. Li, C.-L. Yeh, T.-C. Yen, et al., Osteogenic differentiation of adipose-derived stem cells and calvarial defect repair using baculovirus-mediated co-expression of BMP-2 and miR-148b, Biomaterials 35 (2014) 4901e4910. Y.-C. Ho, Y.-C. Chung, S.-M. Hwang, K.-C. Wang, Y.-C. Hu, Transgene expression and differentiation of baculovirus-transduced human mesenchymal stem cells, J. Gene Med. 7 (2005) 860e868. C.-Y. Chen, H.-H. Wu, C.-P. Chen, S.-R. Chern, S.-M. Hwang, S.-F. Huang, et al., Biosafety assessment of human mesenchymal stem cells engineered by hybrid baculovirus vectors, Mol. Pharm. 8 (2011) 1505e1514. L.Y. Sung, C.L. Chen, S.Y. Lin, S.M. Hwang, C.H. Lu, K.C. Li, et al., Enhanced and prolonged baculovirus-mediated expression by incorporating recombinase system and in cis elements: a comparative study, Nucleic Acids Res. 41 (2013) e139. M. Bellido, L. Lugo, J.A. Roman-Blas, S. Castaneda, J.R. Caeiro, S. Dapia, et al., Subchondral bone microstructural damage by increased remodelling aggravates experimental osteoarthritis preceded by osteoporosis, Arthritis Res. Ther. 12 (2010) R152. C.Y. Lin, Y.H. Chang, K.C. Li, C.H. Lu, L.Y. Sung, C.L. Yeh, et al., The use of ASCs engineered to express BMP2 or TGF-beta3 within scaffold constructs to promote calvarial bone repair, Biomaterials 34 (2013) 9401e9412. X. Wang, B. Guo, Q. Li, J. Peng, Z. Yang, A. Wang, et al., miR-214 targets ATF4 to inhibit bone formation, Nat. Med. 19 (2013) 93e100. L. Cao, G. Liu, Y. Gan, Q. Fan, F. Yang, X. Zhang, et al., The use of autologous enriched bone marrow MSCs to enhance osteoporotic bone defect repair in long-term estrogen deficient goats, Biomaterials 33 (2012) 5076e5084. M.K. Osako, H. Nakagami, N. Koibuchi, H. Shimizu, F. Nakagami, H. Koriyama, et al., Estrogen inhibits vascular calcification via vascular RANKL system: common mechanism of osteoporosis and vascular calcification, Circ. Res. 107 (2010) 466e475. W. Yao, Z. Cheng, M. Shahnazari, W. Dai, M.L. Johnson, N.E. Lane, Overexpression of secreted frizzled-related protein 1 inhibits bone formation and attenuates parathyroid hormone bone anabolic effects, J. Bone Min. Res. 25 (2010) 190e199. N. Takahashi, N. Udagawa, Y. Kobayashi, T. Suda, Generation of osteoclasts in vitro, and assay of osteoclast activity, Methods Mol. Med. 135 (2007) 285e301. K.A. Staines, V.E. MacRae, C. Farquharson, The importance of the SIBLING family of proteins on skeletal mineralisation and bone remodelling, J. Endocrinol. 214 (2012) 241e255. H. Li, H. Xie, W. Liu, R. Hu, B. Huang, Y.F. Tan, et al., A novel microRNA targeting HDAC5 regulates osteoblast differentiation in mice and contributes to primary osteoporosis in humans, J. Clin. Invest. 119 (2009) 3666e3677. Y. Wang, L. Li, B.T. Moore, X.-H. Peng, X. Fang, J.M. Lappe, et al., MiR-133a in human circulating monocytes: a potential biomarker associated with postmenopausal osteoporosis, PLoS ONE 7 (2012) e34641. S. Hwang, S.K. Park, H.Y. Lee, S.W. Kim, J.S. Lee, E.K. Choi, et al., miR-140-5p

[34]

[35]

[36]

[37]

[38] [39]

[40]

[41]

[42]

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50] [51]

[52]

[53]

[54]

[55]

[56]

[57]

[58]

suppresses BMP2-mediated osteogenesis in undifferentiated human mesenchymal stem cells, FEBS Lett. 588 (2014) 2957e2963. J. Yang, S.Y. Qin, C.Q. Yi, G. Ma, H. Zhu, W.R. Zhou, et al., MiR-140 is coexpressed with Wwp2-C transcript and activated by Sox9 to target Sp1 in maintaining the chondrocyte proliferation, FEBS Lett. 585 (2011) 2992e2997. M.E. Buechli, J. Lamarre, T.G. Koch, MicroRNA-140 expression during chondrogenic differentiation of equine cord blood-derived mesenchymal stromal cells, Stem Cells Dev. 22 (2013) 1288e1296. Y. Nakamura, J.B. Inloes, T. Katagiri, T. Kobayashi, Chondrocyte-specific microRNA-140 regulates endochondral bone development and targets Dnpep to modulate bone morphogenetic protein signaling, Mol. Cell Biol. 31 (2011) 3019e3028. T.A. Karlsen, R.B. Jakobsen, T.S. Mikkelsen, J.E. Brinchmann, microRNA-140 targets RALA and regulates chondrogenic differentiation of human mesenchymal stem cells by translational enhancement of SOX9 and ACAN, Stem Cells Dev. 23 (2014) 290e304. C.R. Hill, M. Yuasa, J. Schoenecker, S.L. Goudy, Jagged1 is essential for osteoblast development during maxillary ossification, Bone 62 (2014) 10e21. K. Tanaka, Y. Inoue, G.N. Hendy, L. Canaff, T. Katagiri, R. Kitazawa, et al., Interaction of Tmem119 and the bone morphogenetic protein pathway in the commitment of myoblastic into osteoblastic cells, Bone 51 (2012) 158e167. R.A. Oliver, Y. Yu, G. Yee, A.K. Low, A.D. Diwan, W.R. Walsh, Poor histological healing of a femoral fracture following 12 months of oestrogen deficiency in rats, Osteoporos. Int. 24 (2013) 2581e2589. H. Chen, R. Shalom-Feuerstein, J. Riley, S.D. Zhang, P. Tucci, M. Agostini, et al., miR-7 and miR-214 are specifically expressed during neuroblastoma differentiation, cortical development and embryonic stem cells differentiation, and control neurite outgrowth in vitro, Biochem. Biophys. Res. Commun. 394 (2010) 921e927. A. Sehic, S. Risnes, C. Khuu, Q.E. Khan, H. Osmundsen, Effects of in vivo transfection with anti-miR-214 on gene expression in murine molar tooth germ, Physiol. Genomics 43 (2011) 488e498. M.V. Joglekar, V.S. Parekh, A.A. Hardikar, New pancreas from old: microregulators of pancreas regeneration, Trends Endocrinol. Metab. 18 (2007) 393e400. M.I. Ahmed, M. Alam, V.U. Emelianov, K. Poterlowicz, A. Patel, A.A. Sharov, et al., MicroRNA-214 controls skin and hair follicle development by modulating the activity of the Wnt pathway, J. Cell Biol. 207 (2014) 549e567. A. van Mil, S. Grundmann, M.J. Goumans, Z. Lei, M.I. Oerlemans, S. Jaksani, et al., MicroRNA-214 inhibits angiogenesis by targeting quaking and reducing angiogenic growth factor release, Cardiovasc. Res. 93 (2012) 655e665. A.B. Aurora, A.I. Mahmoud, X. Luo, B.A. Johnson, E. van Rooij, S. Matsuzaki, et al., MicroRNA-214 protects the mouse heart from ischemic injury by controlling Ca2þ overload and cell death, J. Clin. Invest. 122 (2012) 1222e1232. T. Watanabe, T. Sato, T. Amano, Y. Kawamura, N. Kawamura, H. Kawaguchi, et al., Dnm3os, a non-coding RNA, is required for normal growth and skeletal development in mice, Dev. Dyn. 237 (2008) 3738e3748. K. Shi, J. Lu, Y. Zhao, L. Wang, J. Li, B. Qi, et al., MicroRNA-214 suppresses osteogenic differentiation of C2C12 myoblast cells by targeting Osterix, Bone 55 (2013) 487e494. Y. Chen, H.C. Whetstone, A. Youn, P. Nadesan, E.C. Chow, A.C. Lin, et al., Betacatenin signaling pathway is crucial for bone morphogenetic protein 2 to induce new bone formation, J. Biol. Chem. 282 (2007) 526e533. M. Ono, C.A. Inkson, T.M. Kilts, M.F. Young, WISP-1/CCN4 regulates osteogenesis by enhancing BMP-2 activity, J. Bone Min. Res. 26 (2011) 193e208. R.A. Salingcarnboriboon, P. Pavasant, M. Noda, Cbl-b enhances Runx2 protein stability and augments osteocalcin promoter activity in osteoblastic cell lines, J. Cell Physiol. 224 (2010) 743e747. C. Zhao, W. Sun, P. Zhang, S. Ling, Y. Li, D. Zhao, et al., miR-214 promotes osteoclastogenesis by targeting Pten/PI3k/Akt pathway, RNA Biol. 12 (2015) 343e353. H. Xia, L.L. Ooi, K.M. Hui, MiR-214 targets beta-catenin pathway to suppress invasion, stem-like traits and recurrence of human hepatocellular carcinoma, PLoS One 7 (2012) e44206. G. Mbalaviele, S. Sheikh, J.P. Stains, V.S. Salazar, S.L. Cheng, D. Chen, et al., Beta-catenin and BMP-2 synergize to promote osteoblast differentiation and new bone formation, J. Cell Biochem. 94 (2005) 403e418. B.W.M. van Balkom, O.G. de Jong, M. Smits, J. Brummelman, K. den Ouden, P.M. de Bree, et al., Endothelial cells require miR-214 to secrete exosomes that suppress senescence and induce angiogenesis in human and mouse endothelial cells, Blood 121 (2013) 3997e4006. H. Yang, W. Kong, L. He, J.J. Zhao, J.D. O'Donnell, J. Wang, et al., MicroRNA expression profiling in human ovarian cancer: miR-214 induces cell survival and cisplatin resistance by targeting PTEN, Cancer Res. 68 (2008) 425e433. C.X. Xu, M. Xu, L. Tan, H. Yang, J. Permuth-Wey, P.A. Kruk, et al., MicroRNA miR-214 regulates ovarian cancer cell stemness by targeting p53/Nanog, J. Biol. Chem. 287 (2012) 34970e34978. Z. Xu, T. Wang, miR-214 promotes the proliferation and invasion of osteosarcoma cells through direct suppression of LZTS1, Biochem. Biophys. Res. Commun. 449 (2014) 190e195.