Mesenchymal stem cells (MSCs) with multipotential differentiation capacity can differentiate into bone cells under specific conditions and can be used to treat osteonecrosis (ON) of the femoral head (ONFH) through cell transplantation. The current study aims to explore the role of bone marrow (BM) MSCs (BMSCs)-derived exosomes carrying microRNA-122-5p (miR-122-5p) in ONFH rabbit models.
First, rabbit models with ONFH were established. ONFH-related miRNAs were screened using the Gene Expression Omnibus (GEO) database. A gain-of-function study was performed to investigate the effect of miR-122-5p on osteoblasts and BMSCs and effects of exosomes carrying miR-122-5p on ONFH. Co-culture experiments for osteoblasts and BMSCs were performed to examine the role of exosomal miR-122-5p in osteoblast proliferation and osteogenesis. The target relationship between miR-122-5p and Sprouty2 (SPRY2) was tested.
MiR-122, significantly decreased in ONFH in the GSE89587 expression profile, was screened. MiR-122-5p negatively regulated SPRY2 and elevated the activity of receptor tyrosine kinase (RTK), thereby promoting the proliferation and differentiation of osteoblasts. In vivo experiments indicated that bone mineral density (BMD), trabecular bone volume (TBV), and mean trabecular plate thickness (MTPT) of femoral head were increased after over-expressing miR-122-5p in exosomes. Significant healing of necrotic femoral head was also observed.
Exosomes carrying over-expressed miR-122-5p attenuated ONFH development by down-regulating SPRY2 via the RTK/Ras/mitogen-activated protein kinase (MAPK) signaling pathway. Findings in the present study may provide miR-122-5p as a novel biomarker for ONFH treatment.
Osteoporosis is a common skeleton disorder characterized by elevated risk of bone fracture as a result of fragility . Osteonecrosis (ON) of the femoral head (ONFH), a debilitating bone disease, causes the collapse of the joint cartilage and femoral head, along with joint function defects due to disordered blood supply and abnormal fibrinolytic system . More importantly, ONFH is a progressive disorder with bone marrow (BM) and osteocyte death leading to the eventual collapse of the femoral head . In the initial stage, ONFH may be induced by damage or interruption of blood circulation in the femoral head, followed by cell necrosis, ultimately leading to hip pain and dysfunction . Thus far, numerous surgical procedures have been applied for ONFH treatment, such as the total hip arthroplasty and autogenous cell transplantation. However, no treatment can completely cure this disease [5,6]. A previous study reported a method of autologous implantation of BM-derived and cultured mesenchymal stem cells (MSCs) to delay or avoid femoral head collapse . Furthermore, recent studies have also highlighted the importance and implications of microRNAs (miRNAs) in the pathogenesis, prevention, and treatment of ONFH [8,9].
Inherently, miRNAs play critical roles in regulating MSC differentiation and other cellular activities, such as proliferation, survival, and migration . Interestingly, miRNAs also participate in the pathogenesis, prevention, and treatment of ONFH by regulating bone development and regeneration [8,11,12]. Additionally, altered expression of various miRNAs has been revealed to participate in the development of ONFH . For instance, miR-210 was suggested to be over-expressed in ON and its expression was closely correlated to the pathogenesis of ON . Moreover, serum circulating miRNAs have also been indicated to serve as biomarkers of osteoporotic fracture . Exosomes are 40–100 nm nano-sized vesicles released from multiple cells into extracellular space, and exosomal RNAs can be carried by neighboring or distant cells when exosomes circulate, and subsequently mediate recipient cells . Furthermore, miR-122-5p has been previously demonstrated to influence the differentiation process of BM mesenchymal stem cells (BMSCs) into neuron-similar cells . The management of the exosomes derived from MSCs and exosomal-enclosed miRNAs have been reported to be responsible for rescuing tissue function and inducing protective in vitro effects in a diversity of diseases or injuries .
First, the targeting relationship between miR-122-5p and Sprouty2 (SPRY2) was determined by an initial bioinformatics prediction followed by a confirmatory dual-luciferase reporter assay. Notably, SPRY2 plays a role in the development of osteoporosis via interaction with fibroblast growth factor 23 [18,19]. Mitogen-activated protein kinase/extracellular signal-regulated kinase (MAPK/ERK) signaling pathway was also suggested to be involved in osteoporosis . SPRY2 functions as a tumor suppressor gene in multiple myeloma cells by inhibiting the activation of the MAPK/ERK signaling pathway . Furthermore, the correlation of miR-122-5p to the differentiation of MSCs into neuron-like cells has also been highlighted . In the current study, we hypothesized that BMSCs-derived exosomes carrying miR-122-5p exert protective effects on ONFH via interaction with SPRY2.
Materials and methods
First, the ONFH miRNA expression profile GSE89587, consisting of serum miRNA expression profiles of ten no-necrosis patients and ten patients with traumatic ONFH, was retrieved from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/). The annotation platform of the obtained profile was GPL21439 miRCURY LNA microRNA Array, the 7th generation-hsa, mmu and rno (miRBase 21; probeID version). Next, the affy package in R language  was applied for background correction and normalization processing of each profile data, and differential expression analysis was performed using the limma package . The screening threshold of differentially expressed miRNAs was set as P-value <0.05 and |Log Fold Change| > 2. Subsequently, the heatmap of the differentially expressed miRNAs was plotted. Additionally, TargetScan (http://www.targetscan.org/vert_71/), starBase (http://starbase.sysu.edu.cn/index.php), mirDIP (http://ophid.utoronto.ca/mirDIP/), miRDB (http://www.mirdb.org/), DIANA (http://diana.imis.athena-innovation.gr/DianaTools/index.php?r=microT_CDS/index), and miRWalk (http://mirwalk.umm.uni-heidelberg.de/) were employed for prediction of the target genes of the differentially expressed miRNAs. The predicted results from the aforementioned websites were analyzed using jvenn (http://jvenn.toulouse.inra.fr/app/example.html).
Establishment of ONFH rabbit models 
Raised with normal diet, 80 mature New Zealand rabbits (weight: 2.8 ± 0.4 kg) were arranged into sham group (n=40) and ONFH group (n=40). All rabbits were anesthetized with 3% pentobarbital (30 ml/kg) through the intraperitoneal injection . Next, with the greater trochanter of femur as reference, a lateral incision on unilateral hip joint was made under sterile conditions. The soft tissues were peeled off, followed by complete exposure of the capsula articularis. After abscission of the ligament, the femoral head was dislocated and fixed with a hooklet. The upper end of the femoral head was immediately frozen using crumpled gauze dipped in liquid nitrogen and then rewarmed in warm saline. The incision was sutured layer-by-layer after resetting the dislocated femoral head. Same surgical regimen was followed in the sham group with the exception dislocation of the femoral head. Post-operative antibiotics were administered to all rabbits to prevent infections . Four rabbits from each group were killed at the 2, 4, 6, 8, 10, and 12 week intervals after ONFH induction, respectively. The bilateral femurs were collected for the measurement of bone mineral density (BMD) in the central coronal plane of the femoral head using a BMD-400 single photon bone densitometer. In addition, histological examination was also performed.
The femoral head was completely sawed along the coronal plane and sectioned into 3-mm slices. Next, the tissues with size of 1.0 mm × 0.7 mm × 0.3 mm were collected from four areas including the cartilage surface at the weight-bearing area, subchondral necrotic center, the repair area (containing granulation tissue and newly formed bone), and head and neck areas which were relatively normal. The raw slices were paraffin-embedded and sliced into 6-μm pathological sections. The coronal plane was sewn, fixed with 10% formalin for 48 h, and decalcified with 10% formic acid solution for 48–72 h. After dehydration with gradient alcohol, the specimens were paraffin-embedded, cut into approximately 5-μm slices, stained with Hematoxylin–Eosin (HE), and observed under a light microscope.
Isolation and culture of BMSCs
A total of 10 ml BM was extracted from the fractured end of the femoral shaft using a 20-ml syringe containing 2000 IU heparin under sterile conditions, followed by quick mixing with heparin. Then the suspension was centrifuged at 257×g for 10 min and the supernatant containing adipose tissue was discarded. The pellets containing BM were rinsed three times with Dulbecco’s Modified Eagle’s Medium (DMEM) and resuspended in 15 ml medium. Then the suspension was mixed with equal volume of Ficoll-Paque™ Plus lymphocyte separation solution and centrifuged at 715×g for 20 min. After centrifugation, the nucleated cells were suspended in the interface or the upper layer of the liquid, whereas the majority of erythrocytes were precipitated to the bottom. In total, 30 ml nucleated cells at the interface were then collected with a pipette and rinsed with phosphate-buffered saline (PBS), followed by centrifugation at 178×g for 8 min. The pellets containing nucleated cells were uniformly added with 5 ml cell culture medium. Then, 10 μl cell suspension from the above step was mixed with 490 μl PBS. Cells in PBS (10 μl) were moved on to the slide and counted under the microscope. Afterward, the cells were seeded into bottles at a density of 1 × 105 cells/bottle and incubated with 5 ml low glucose medium at 37°C with 5% CO2 and saturated humidity. After 24 h, BMSCs were observed to be adhered to the wells. Half of the medium was replaced after 24 h, and the full medium was replaced every 2–3 days until the cells reached approximately 80–90% confluence. The cell adhesion and growth were observed using an inverted phase contrast microscope. Subsequently, the BMSCs were detached with StemPro™ Accutase™ Cell Dissociation Reagent (A1110501, Thermo Fisher Scientific, San Jose, CA, U.S.A.) and passaged at a ratio of 1:2 to 1:3. Then cells were subject to flow cytometry analysis to detect the BMSC markers, CD29 (557332), CD34 (555822), CD44 (555478), CD45 (555482), CD71 (555536), and human leukocyte antigen (HLA)-DR (556643; all from BD Biosciences, San Jose, CA, U.S.A.) . The adipogenic and osteogenic differentiation of BMSCs were determined via Oil Red O staining and Alizarin Red staining. The formation of lipid droplets and calcium deposition were observed under an inverted microscope.
Fetal bovine serum (FBS) was ultra-centrifuged at 100000×g for 18 h to remove the exosomes from the serum . When cell confluence reached approximately 80%, BMSCs were rinsed two times with PBS and incubated with 10% exosome-free FBS for 48 h at 37°C with 5% CO2. After incubation, the supernatant was subject to sequential centrifugation at 4°C: × 300×g for 10 min, × 2000×g for 15 min, and × 5000×g for 15 min. The pellets were discarded after each centrifugation. Then the supernatant was centrifuged at × 12000×g for 30 min at 4°C, and the pellets were collected, followed by PBS rinsing. After being resuspended in PBS, the suspension was centrifuged at × 12000×g for 70 min at 4°C and the pellets were collected. After differential centrifugations, the supernatant was ultra-centrifuged at × 100000×g for 70 min and the pellets containing exosomes were collected. After being resuspended in PBS, the suspension was subject to another round of ultra-centrifugation at × 100000×g for 70 min. Finally, the pellets containing pure exosomes were collected.
Nanoparticle tracking analysis
A total of 20 μg exosomes were re-suspended in 1 ml PBS and whirled for 1 min to maintain uniform distribution. A nanoparticle tracking analyzer (NanoSight; Malvern Panalytical, Worcestershire, U.K.) was employed to directly measure the diameter of the exosomes and size distribution. Particles were tracked and measured on the basis of Brownian motion and the diffusion coefficient. After being isolated, the exosomes were diluted in 1 ml filtered PBS. Control medium and filtered PBS were used as controls. The Nanoparticle tracking analysis (NTA) measurement conditions were set at 23.75 ± 0.5°C with 25 frames per second for 60 s. All samples were subject to similar detection threshold. Each reaction was run in triplicate to obtain the mean value .
Transmission electron microscope observation
The prepared exosomes were fixed in 4% paraformaldehyde at 4°C for 2 h, rinsed three times with 0.1 mol/l PBS, and fixed with 1% osmium tetroxide for 2 h. And then, exosomes were routinely dehydrated with gradient ethanol and acetonum, followed by treatment with ethoxyline resin for infusion, embedment, and polymerization. Next, exosomes were sectioned into 0.5-μm semi-thin slices and observed under a light microscope. After confirming the location of exosomes, 60-nm ultra-thin sections were prepared for uranyl acetate and lead citrate staining. Transmission electron microscope (TEM) observation was carried out using JEM 1230 TEM (JEOL USA Inc., Peabody, MA, U.S.A.) at 110 kV and visualized with an UltraScan 4000 CCD camera and First Light Digital Camera Controller (Gatan Inc., Pleasanton, CA, U.S.A.) .
Determination of acetylcholinesterase activity
Exosomes extracted by multi-step ultra-centrifugations were diluted into 110 μl with PBS and added to a 96-well plate at 37.5 μl per well. Then, equal volumes of 0.1 mmol/l 5,5′-dithiobis-2-nitrobenzoic (DTNB) acid solution and 1.25 mmol/l acetylthiocholine iodide solution were added to the plate with a final volume of 300 μl. After 30 min of reaction, the optical density (OD) value was measured at an excitation wavelength of 412 nm using a microplate reader.
Pre-osteoblast hFOB1.191 cells (CRL-11372; American Type Culture Collection, Manassas, VA, U.S.A.). were maintained in DMEM/F12 medium containing 2 mM Glutamine and 10% FBS, and cultured with 5% CO2 at 37°C. Next, the hFOB1.191 cells were detached with trypsin, centrifuged at × 1000×g for 5 min, and resuspended in 3 ml DMEM. Then 1 ml suspension was diluted 20 times and fully mixed, and 10 μl cell suspension was counted using a hemacytometer. Meanwhile, osteoblasts and BMSCs were seeded into the co-culture chamber (pore size: 0.4 μm) at a ratio of 3:1. Especially, osteoblasts (1.2 × 105) were seeded in the apical chamber with 10% serum DMEM, and MSCs (0.4 × 105 cells) were seeded in the basolateral chamber with 15% serum DMEM. The co-culture chamber was then placed in a six-well plate and cultured for 4–5 days with the medium changed every 1–2 days. After co-culture, the apical chamber was removed and BMSCs were collected for subsequent experimentation.
Confocal fluorescence microscopy
A total of 20 μg exosome suspension was mixed with diluted DIR staining solution (1:1000), and allowed to stand for 15 min at 37°C. After PBS rinsing, the suspension was centrifuged at 100000×g for 70 min. Then, DIR-labeled exosomes were co-cultured with osteoblasts labeled with adenoviruses carrying green fluorescent protein (Ad-GFP), and the uptake of exosomes by osteoblasts was observed under a confocal fluorescence microscopy at the 6, 12, 18, 24, 36, and 48 h time intervals after co-culture.
Cell grouping and transfection
The osteoblasts were classified into the following four groups: the miR-122-5p-NC group (transfected with miR-122-5p negative control (NC) sequence), the miR-122-5p mimic group (transfected with miR-122-5p mimic), the miR-122-5p inhibitor + si-NC group (transfected with miR-122-5p inhibitors and si-NC), and the miR-122-5p inhibitor + si-SPRY2 group (transfected with miR-122-5p inhibitors and si-SPRY2).
Prior to transfection, osteoblasts at the logarithmic phase of growth were seeded in a six-well plate. Then, transfection was performed using a Lipofectamine 2000 kit (Invitrogen, Carlsbad, CA, U.S.A.) according to the instructions when cell density reached 30–50%. MiR-122-5p mimics, miR-122-5p inhibitor + si-SPRY2, and NC sequence (each 100 pmol) were diluted with 250 μl serum-free Opti-MEM (Gibco, Grand Island, NY, U.S.A.) and incubated for 5 min at room temperature, respectively. Meanwhile, 5 μl Lipofectamine 2000 was also diluted with another 250 μl serum-free medium Opti-MEM and incubated for 5 min at room temperature. After incubation, the above two solutions were mixed together and incubated for 20 min at room temperature. Then, the combined mixture was added to the wells containing osteoblasts according to the treatment groups. Incubation was performed at 37°C with 5% CO2, and the medium was replaced with complete medium after 6–8 h, followed by another incubation for 24–48 h.
Reverse transcription quantitative polymerase chain reaction
Total RNA was extracted using a TRIzol kit (Invitrogen, Carlsbad, CA, U.S.A.) and the concentration was measured by Nanodrop 2000 (Thermo Fisher Scientific, San Jose, CA, U.S.A.). Then, 1 μg RNA was reversely transcribed into complementary DNA (cDNA) using a PrimeScript™ RT reagent kit with gDNA Eraser kit (TaKaRa, Tokyo, Japan). After the addition of 5×g DNA Eraser Buffer and gDNA Eraser, DNA extraction reaction was performed at 42°C for 2 min. Afterward, cDNA was obtained by reverse transcription at 37°C for 15 min and at 85°C for 5 s. PCR experiments were performed using an ABI 7500 instrument (Thermo Fisher Scientific, San Jose, CA, U.S.A.) with the SYBR® Premix ExTaq™ (TliRNaseH Plus) kit (TaKaRa, Tokyo, Japan). Reaction conditions were as follows: pre-denaturation at 95°C for 10 min, 40 cycles of denaturation at 95°C for 15s, and annealing at 60°C for 30 s. U6 was utilized as the internal reference of miR-122-5p, whereas glyceraldehyde-3-phosphate dehydrogenase (GAPDH) served as the internal reference of SPRY2, runt-related transcription factor 2 (RUNX2), and Collagen I. The 2−ΔΔCt method was applied to calculate the relative expression of each factor and the formula was as follows: ΔΔCt = ΔCtexperimental group − ΔCtcontrol group and ΔCt = Cttarget gene − Ctinternal reference . Ct was the amplification cycles when the real-time fluorescence intensity of the reaction reached the set threshold. Primer sequences (Table 1) were obtained from Shanghai GenePharma Co., Ltd. (Shanghai, China). Three independent experiments were conducted.
|Collagen I||Forward: 5′-CTGGCCCCATTGGTAATGT-3′|
|Collagen I||Forward: 5′-CTGGCCCCATTGGTAATGT-3′|
Abbreviation: miR-122-5p, microRNA-122-5p.
Western blot analysis
Total protein was extracted from cells of each group, and the protein concentration was determined using a bicinchoninic acid (BCA) kit (Thermo Fisher Scientific, San Jose, CA, U.S.A.). Then, 30 μg total protein was loaded into each lane, separated with sodium dodecyl sulfate/polyacrylamide gel electrophoresis (SDS/PAGE) at 80 V for 35 min and at 120 V for 45 min, and then transferred on to polyvinylidene fluoride (PVDF) membranes (Amersham, Little Chalfont, Buckinghamshire, U.K.). After being blocked with 5% skim milk powder at room temperature for 1 h, the membrane was incubated with following primary antibodies overnight, including CD63 (1:1000, ab134045), Hsp70 (1:1000, ab79852), Tsg101 (1:5000, ab30871), SPRY2 (1:1000, ab85670), vascular endothelial growth factor receptor (VEGFR)-1 (1:1000, ab2350), VEGFR-2 (1:5000, ab11939), platelet-derived growth factor receptor (PDGFR)-α (1:1000, ab65258), PDGFR-β (1:1000, ab16868), ERK (1:1000, ab17942), p-ERK (1:10000, ab50011), JNK (1:1000, ab199380), p-JNK (1:1000, ab47337), p38 (1:1000, ab31828) and p-p38 (1:1000, ab4822), and β-actin (1:5000, ab8227). All aforementioned antibodies were obtained from Abcam Inc. (Cambridge, MA, U.S.A.). After being rinsed three times with PBS containing 0.1% Tween-20 (PBST; 10 min each time), the membrane was incubated with the secondary horseradish peroxidase (HRP)-labeled goat anti-rabbit antibody (1:10000, Jackson, West Grove, PA, U.S.A.) for 1 h at room temperature. After being rinsed three times with PBST (10 min each time), the membrane was exposed to a chemiluminescence instrument (GE Healthcare, Pittsburgh, PA, U.S.A.). The relative expression of protein was measured using the ImagePro Plus 6.0 (Media Cybernetics, Silver Springs, MD, U.S.A.). Three independent experiments were conducted.
Alkaline phosphatase staining
Seven days after transfection, with cell medium removed, the BMSCs were rinsed two times with PBS and fixed with 10% neutral formaldehyde solution for 30 min at room temperature. After another two rinses with PBS, an Alkaline phosphatase (ALP) kit (CBA-300, Cell Biolabs, San Diego, CA, U.S.A.) was used for 45-min incubation at 37°C. After the staining solution was removed, cells were rinsed two times with PBS and air-dried at room temperature overnight.
Alizarin Red S staining
Cells at passage 4 were adjusted to a density of 1 × 105 cells/ml and seeded into a 24-well plate (1 ml per well). The medium was replaced every 3 days. On the 21st day, the cells were fixed, stained, and cleared according to the instructions of the Alizarin Red S staining kit. The staining of calcium nodules was observed under an inverted microscope.
Cells at the logarithmic phase of growth were cultured in a 96-well plate (4 × 103 to 1 × 105 cells per well) to the normal growth stage. The cell medium was used to dilute 5-ethynyl-2′-deoxyuridine (EdU) solution at a ratio of 1000:1, and then appropriate amount of 50 μM EdU medium was added to the cell culture plate (100 μl per well) for 2-h incubation. Once the medium was removed, cells in each well were incubated with 50 μl PBS containing 4% paraformaldehyde for 30 min at room temperature, followed by incubation with 50 μl of 2 mg/ml glycine for 5 min on a shaker. Next, cells in each well were incubated with 100 μl PBS containing 0.5% Triton X-100 on a decolorizing shaker for 10 min. After PBS rinsing for 5 min, cells in each well were incubated with 100 μl of 1× Apollo® staining reaction solution on a decolorizing shaker for 30 min at room temperature avoiding exposure to light. After the removal of the staining reaction solution, cells in each well were rinsed with 100 μl PBS containing 0.5% Triton X-100 on a decolorizing shaker two to three times (10 min each time). After the removal of penetrant, cells in each well were rinsed one to two times with 100 μl methanol (5 min each time) and PBS for 5 min. Meanwhile, Reagent F was diluted with the deionized water at a ratio of 1:100 to prepare 1× Hoechst33342 reaction solution, which was then stored without exposure to light. Finally, cells in each well were incubated with 100 μl of 1× Hoechst33342 reaction solution on a decolorizing shaker for 30 min at room temperature without exposure to light, and then rinsed one to three times with 100 μl PBS.
Cells at the logarithmic phase of growth were cultured for 72 h, rinsed with PBS, and detached with StemPro™ Accutase™ Cell Dissociation Reagent (A1110501, Thermo Fisher Scientific, San Jose, CA, U.S.A.). After being resuspended in a medium, the cells were subjected to centrifugation and fixed with 2% paraformaldehyde (in PBS) for 10 min at 37°C, followed by a 5-min centrifugation at × 1500×g at 37°C. After removal of the supernatant, cells were resuspended in PBS. Next, the cells were incubated with precooled 90% methanol on ice for 30 min, followed by another 5-min centrifugation at × 1500×g at 37°C. After being rinsed with the incubation buffer, cells were centrifuged at × 1500×g for 5 min at 37°C. Afterward, the cells were re-suspended in 90 μl incubation buffer and blocked for 10 min at room temperature. Then, cells were incubated with 10 μl antibody for 30 min at room temperature without exposure to light, re-suspended in 0.5 ml PBS, and analyzed using a flow cytometer. The experiment was repeated three times to obtain the mean value.
Dual luciferase reporter assay
The relationship between the SPRY2 gene and miR-122-5p was confirmed by the bioinformatics database microRNA.org (http://www.microrna.org/). Next, human embryonic kidney (HEK)-293T cells were cultured in DMEM containing 10% FBS at 37°C with 5% CO2. A cDNA fragment containing 3′-untranslated region (UTR) of SPRY2 (binding site to miR-122-5p) was inserted into the pmirGLO vector. A cDNA fragment of SPRY2 3′-UTR with a binding site mutation was constructed with DNA point mutation. It was then inserted into the pmirGLO vector and confirmed to be correct by sequencing. The pmirGLO-SPRY2 or pmirGLO-mutSPRY2 recombinant vector was co-transfected into HEK-293T cells with miR-122-5p mimic (over-expression sequence of miR-122-5p) or miR-NC. After incubation for 48 h, cells were collected and lysed. Then, 100 μl lysate was mixed with 100 μl Renilla luciferase assay working solution and subject to Renilla luciferase activity measurement. Meanwhile, 100 μl lysate was mixed with 100 μl firefly luciferase detection reagent for detecting the firefly luciferase activity. Subsequently, a multimode microplate reader SpectraMaxM5 (interval: 2 s; duration: 10 s) was employed to determine the activity of Renilla luciferase and firefly luciferase, respectively.
In vivo experiments
Rabbits with ONFH were assigned into the EXO-miR-122-5p-NC or EXO-miR-122-5p agomir groups (ten rabbits in each group). Exosomes-derived from BMSCs transfected with miR-NC or miR-122-5p were extracted and injected into the rabbits via the caudal vein. X-ray examination and measurement for BMD by single photon were performed 8 weeks later. Then, rabbits in the EXO-miR-122-5p-NC and EXO-miR-122-5p agomir groups were killed, and the bilateral femurs were collected for histological examination. Subsequently, the HE-stained slices were observed under a light microscope. At the same time, a computer image analysis system was applied for measurement of the relative trabecular bone volume (TBV, %) and mean trabecular plate thickness (MTPT, μm).
The sections were rehydrated and probed with the primary antibody to CD31 (1:200; ab24590, Abcam Inc., Cambridge, MA, U.S.A.) at 4°C overnight, then incubated with biotinylated secondary antibody to IgG. Sections were treated with avidin–biotin peroxidase complex, developed with 3,3′-diaminobenzidine, and then stained with Hematoxylin. All sections were consistently maintained in medium, and sections incubated without primary antibodies were used as the control.
Statistical analyses were performed using the SPSS 21.0 software (IBM Corp., Armonk, NY, U.S.A.). For all data, normality and homogeneity of variance were tested. Measurement data were expressed as mean ± standard deviation. Comparisons between two groups were analyzed by the t test. Comparisons among multiple groups were analyzed by one-way analysis of variance (ANOVA). Mean values between two groups were analyzed by post hoc test, while comparisons between the sham and ONFH group at multiple time points were analyzed by repeated measures analysis of variance. A value of P<0.05 was considered to be statistically significant.
miR-122 participates in the development of ONFH by regulating SPRY2
R language was employed to screen the differentially expressed miRNAs from the GSE89587 ONFH profile, and a heatmap of the top ten differentially expressed miRNAs in the profile is shown in Figure 1A. The results revealed that the expression of has-miR-122-5p was lower in ONFH compared with the control group. Similarly, down-regulated expression of miR-122 has been documented in various cancers and suggested to serve as a suppressor in bladder cancer , gallbladder carcinoma , gastric cancer , and hepatocellular carcinoma (HCC) . In addition, miR-122 has also been highlighted as a potential therapeutic target for the treatment of bone-associated osteosarcoma , so the focus of the current study shifted to elucidate the effects of differential expression of miR-122-5p on ONFH. Subsequently, the target genes of miR-122-5p were predicted and analyzed using TargetScan, starBase, mirDIP, miRDB, DIANA, and miRWalk, respectively. The Venn map analysis confirmed the existence of two intersecting genes, SLC7A1 and SPRY2, indicating they might be regulated by miR-122-5p (Figure 1B). Moreover, previous evidences have highlighted that SPRY2 is the target gene of miR-122-5p  and inhibition of SPRY2 could promote osteoblast differentiation . Therefore, it could be concluded that miR-122 might affect ONFH by regulating the SPRY2 gene.
miR-122 may be involved in the progression of ONFH via the regulation of SPRY2
Up-regulated miR-122-5p promotes osteoblast proliferation and osteogenesis
Additionally, we examined the effect of miR-122-5p on the biological behavior of osteoblasts using reverse transcription quantitative polymerase chain reaction (RT-qPCR), ALP staining, EdU staining, and Alizarin Red S staining. RT-qPCR results indicated that the expression of RUNX2 and collagen I was significantly increased after treatment with over-expressed miR-122-5p (P<0.05; Figure 2A). Whereas, the results of ALP staining demonstrated denser deposition of large blue cytoplasmic particles in the miR-122-5p mimic group compared with the miR-122-5p mimic-NC group, accompanied with increased ALP (Figure 2B). Results of EdU labeling revealed that the positive expression of osteoblasts in the miR-122-5p mimic group was notably higher than that of the miR-122-5p mimic-NC group (Figure 2C; P<0.05). Results of Alizarin Red S staining confirmed that the formation of calcium nodule was present in both groups. However, more calcium salts were observed in the miR-122-5p mimic group in comparison with the miR-122-5p mimic-NC group (P<0.05; Figure 2D). The above results demonstrated over-expression of miR-122-5p could promote osteoblast proliferation and osteogenesis.
Up-regulated miR-122-5p facilitates osteoblast proliferation and osteogenesis
BMSCs possess the ability of adipogenic differentiation
The subcultured BMSCs in good condition were selected, wherein the surface antigens were identified using fluorescein isothiocyanate-labeled mouse anti-rabbit antibodies to CD29, CD34, CD44, CD45, CD71, HLA-DR via flow cytometry. Results of flow cytometry demonstrated that CD29 (97.14%), CD44 (97.46%), and CD71 (98.17%) were positive in BMSCs, whereas HLA-DR (0.39%), CD34 (1.72%), and CD45 (0.32%) were all negative (Figure 3A). It was reported that CD29, CD44, CD71 were BMSC markers, CD34 and CD45 were hematopoietic stem cell markers, and HLA-DR was primarily expressed in several antigen presenting cells, such as B lymphocytes, macrophages, and activated T lymphocytes. The above results confirmed the purity of BMSCs. In addition, Oil Red O staining revealed the presence of a large amount of lipid droplets in BMSCs, confirming the deposition of lipid components (Figure 3B). This finding indicated that BMSCs were capable of adipogenic differentiation. Meanwhile, heteromorphic BMSCs were also observed, i.e. fusiform-like or polygonal in shape. Furthermore, Alizarin Red S staining showed that BMSCs presented as colonies after 14-day culture and were covered with calcareous particles (Figure 3C). Additionally, BMSCs in the center were noted to be merged with the loss of typical cellular structure, while calcium nodules were observed with red staining. These results suggested that BMSCs were capable of adipogenic differentiation.
BMSCs possess the ability of adipogenic differentiation
Co-culture of BMSCs with osteoblasts promotes osteoblast proliferation and osteogenesis
Subsequently, miR-122-5p mimic was introduced in BMSCs to investigate the effects of miR-122-5p over-expression on the biological function of osteoblasts (Figure 4A). Results of RT-qPCR discovered that the expression of miR-122-5p, RUNX2, and collagen I was significantly increased in the BMSCs-miR-122-5p mimic group after co-culture with differentiated hFOB1.191 compared with the BMSCs-miR-122-5p-NC group (Figure 4B; P<0.05). ALP staining confirmed denser deposition of large blue cytoplasmic particles in the BMSCs-miR-122-5p mimic group compared with that in the BMSCs-miR-122-5p-NC group, indicating stronger expression of ALP (Figure 4C). In addition, EdU labeling showed that the positive expression of osteoblasts in the BMSCs-miR-122-5p mimic group was significantly higher than that of the BMSCs-miR-122-5p-NC group (Figure 4D; P<0.05). Moreover, results of Alizarin Red S staining identified the formation of calcium nodules in all groups with pink staining. However, more calcium salts were observed in the BMSCs-miR-122-5p mimic group than that in the BMSCs-miR-122-5p-NC group (Figure 4E). The above results demonstrated that co-culture of BMSCs with osteoblasts could promote osteoblast proliferation and osteogenesis.
BMSCs treated with up-regulated miR-122-5p promotes osteoblast proliferation and osteogenesis
BMSCs secrete exosomes into osteoblasts
Subsequently, the cell supernatant during the culture of BMSCs was collected, and the exosomes were extracted using ultracentrifugation and identified. TEM images identified that exosomes presented with disc- or cup-shaped, two-layer membrane structures with a diameter of approximately 500 nm (Figure 5A). This finding was further confirmed by NTA (Figure 5B). Meanwhile, Western blot analysis confirmed the presence of Hsp70-positive, CD63-positive, and Tsg101-positive exosomes (Figure 5C), while Calnexin was absent. Next, to further investigate the characteristics and potential functions of exosomes derived from BMSCs, miR-122-5p mimic was introduced in BMSCs which were labeled with DIR. Afterward, DIR-labeled exosomes in BMSCs were co-cultured with GFP-labeled osteoblasts. We observed that the exosomes successfully entered into the cells. Furthermore, confocal fluorescence microscopy indicated that DIR-labeled exosomes were present in osteoblasts, being primarily aggregated in the cytoplasm (Figure 5D). As the co-culture time prolonged, more and more osteoblasts were observed in red fluorescence, indicating the number of DIR-exosomes in osteoblasts were gradually increasing and became apparent after co-culturing for 48 h. In addition, the uptake of BMSCs-derived exosomes confirmed the involvement of exosomes in the regulation of the transcription or biological functions of osteoblasts. RT-qPCR results showed that the expression of miR-122-5p in exosomes in the BMSCs-miR-122-5p mimic group was significantly increased compared with the BMSCs-miR-122-5p-NC group (Figure 5E). Simultaneously, we detected expression of SPRY2 in osteoblasts co-cultured with BMSCs-derived exosomes. The SPRY2 expression was observed to be lower in the BMSCs-miR-122-5p mimic group in comparison with the control group (Figure 5F,G). These findings suggested that miR-122-5p could be carried by exosomes.
Successful isolation and culture of exosomes
GW4869 inhibits the release of BMSCs-derived exosomes and miR-122-5p expression in osteoblasts
To investigate whether miR-122-5p was transmitted into osteoblasts through exosomes in vitro and altered their biological function, the exosome-specific inhibitor GW4869 was added to the Transwell chamber co-culture system. Subsequently, the release of exosomes and the expression patterns of miR-122-5p in osteoblasts were assessed. Acetylcholinesterase (AchE) results revealed that the activity of AchE in the co-culture + GW4869 group was significantly reduced compared with that of the co-culture + DMSO group (Figure 6A; P<0.05). A decline in the expression of exosomal miR-122-5p was also observed in the co-culture + GW4869 group (Figure 6B; P<0.05). Meanwhile, co-culture results showed that the proliferation and differentiation capacity of osteoblasts were significantly decreased after the addition of GW4869 in comparison with that of the co-culture + DMSO group (P<0.05; Figure 6C,D). The above results confirmed that GW4869 suppressed the expression of miR-122-5p in osteoblasts by inhibiting the release of BMSCs-derived exosomes.
GW4869 inhibits the release of BMSC-derived exosomes and miR-122-5p expression in exosomes
SPRY2 is the target gene of miR-122-5p
The specific binding sites between the 3′-UTR of SPRY2 gene and the miR-122-5p sequence were analyzed using the microRNA.org website, and the results confirmed that SPRY2 was a target gene of miR-122-5p (Figure 7A). This finding was further confirmed by a dual luciferase reporter assay (Figure 7B), the results of which showed that over-expressed miR-122-5p significantly suppressed the luciferase activity of 3′-UTR in the wild-type compared with the BMSCs-miR-122-5p-NC group (P<0.05), whereas no differences in luciferase activity of 3′-UTR was observed in the mutant-type (P>0.05). The above results indicated that miR-122-5p could directly target and bind to SPRY2. Furthermore, RT-qPCR results revealed that the expression of SPRY2 was significantly increased in femoral head necrotic cells (P<0.05; Figure 7C). Meanwhile, results from RT-qPCR and Western blot analysis proved that SPRY2 expression was significantly decreased as a result of up-regulation of miR-122-5p (Figure 7D,E; P<0.05). However, increased expression of SPRY2 was noted after suppressing the expression of miR-122-5p (P<0.05). The above results confirmed that SPRY2 was the target gene of miR-122-5p.
SPRY2 is the target gene of miR-122-5p
MiR-122-5p promotes the proliferation and osteogenesis of osteoblasts by down-regulating SPRY2 and activating receptor tyrosine kinase activity
Next, we silenced the expression of SPRY2 to evaluate changes in osteoblast proliferation, differentiation, and bone formation (Figure 8A–G). The expression of collagen I was found to be significantly higher in the miR-122-5p inhibitor + si-SPRY2 group compared with the miR-122-5p-inhibitor + si-NC group (P<0.05), with denser precipitation of large blue particles. Meanwhile, ALP was observed to be strongly expressed with more calcium deposits, stronger receptor tyrosine kinase (RTK) activity, activated MAPK pathway, and higher positive rate of EdU. This finding indicated that miR-122-5p down-regulated SPRY2 to enhance RTK activity and activate the RTK/Ras/MAPK signaling pathway, thus promoting osteoblast proliferation, differentiation, and osteogenesis. It also highlighted the use of miR-122-5p as a potential therapeutic target for ONFH.
MiR-122-5p promotes the proliferation and osteogenesis of osteoblasts by down-regulating SPRY2 and activating RTK activity
MiR-122-5p enhances RTK activity and activates the RTK/Ras/MAPK signaling pathway
Results of flow cytometry for detection of RTK activity revealed that the extent of RTK phosphorylation was significantly increased in the BMSCs-miR-122-5p mimic group compared with that in the BMSCs-miR-122-5p-NC group (Figure 9A; P<0.05). In addition, the expression patterns of RTK/Ras/MAPK signaling pathway-related proteins were analyzed using Western blot analysis (Figure 9B,C). Compared with the BMSCs-miR-122-5p-NC group, the expression of VEGFR-1, VEGFR-2, PDGFR-α, PDGFR-β, ERK, JNK, and p38 were all noted to be increased in the BMSCs-miR-122-5p group (P<0.05). These results indicated that miR-122-5p enhanced RTK activity and activated the RTK/Ras/MAPK signaling pathway.
MiR-122-5p activates the RTK/Ras/MAPK signaling pathway
Successful establishment of ONFH rabbit models
In order to further investigate the effect of BMSCs-derived exosomes over-expressing miR-122-5p in vivo, we established a model of ONFH. Two months after induction, symptoms of claudication were observed in rabbits in the ONFH group. Meanwhile, it was observed that BMD of femoral head in the ONFH group was significantly lower than that of the sham group 8 weeks after operation, indicating a reduction in the osteogenic capacity in the ONFH group (Table 2). X-ray imaging demonstrated uneven distribution, lower density, and slight depression in the articular cartilage plane of the femoral head in the ONFH group in comparison with the normal group (Figure 10A). HE-staining results showed that femoral head in the treated side exhibited a large amount of osteolysis accompanied with the narrow, shortened, or disordered trabecular bone structure. Degeneration and necrosis were also observed with the rough cartilage plane, and dispersive tissues were present in the interstitium. Meanwhile, reparative fibrous connective tissues were also noted at the necrotized sites (Figure 10B). All the above pathological findings indicated toward the successful establishment of ONFH rabbit models.
Successful establishment of ONFH rabbit models
|Weeks after operation||Number||Normal group||ONFH group||P-value|
|2||4||0.82 ± 0.06||0.78 ± 0.09||0.487|
|4||4||0.69 ± 0.08||0.67 ± 0.09||0.883|
|6||4||0.68 ± 0.09||0.66 ± 0.08||0.778|
|8||4||0.61 ± 0.05||0.41 ± 0.05||0.001|
|10||4||0.70 ± 0.07||0.50 ± 0.04||0.003|
|12||4||0.78 ± 0.08||0.56 ± 0.05||0.003|
|Weeks after operation||Number||Normal group||ONFH group||P-value|
|2||4||0.82 ± 0.06||0.78 ± 0.09||0.487|
|4||4||0.69 ± 0.08||0.67 ± 0.09||0.883|
|6||4||0.68 ± 0.09||0.66 ± 0.08||0.778|
|8||4||0.61 ± 0.05||0.41 ± 0.05||0.001|
|10||4||0.70 ± 0.07||0.50 ± 0.04||0.003|
|12||4||0.78 ± 0.08||0.56 ± 0.05||0.003|
Data were statistically analyzed by repeated measurement analysis of variance, n=4.
BMSCs-derived exosomes modified by miR-122-5p promotes the angiogenesis, repair, and healing in ONFH rabbits
After the treatment of ONFH models with BMSCs-derived exosomes, a computer-assisted X-ray density image analyzing system was employed to detect the relative BMD and single photon bone densitometer to BMD. As shown in Figure 11A–C, femoral head BMD in the EXO-miR-122-5p-agomir group was observed to be significantly increased than that of the EXO-agomir-NC group (P<0.05). The morphometric results showed that the relative TBV and MTPT of femoral head were markedly increased in the EXO-miR-122-5p-agomir group after 8 weeks of operation compared with that of the EXO-agomir-NC group (Figure 11D; P<0.05), indicating over-expression of miR-122-5p could promote osteogenesis. Meanwhile, osteolysis of the femoral head was noted in the treated side in the EXO-miR-122-5p-agomir group (Figure 11E) with the normal bone trabecula and smooth cartilage plane. Reparative fibrous connective tissues were also observed at the necrotic sites with limited degeneration and necrosis and significantly reduced number of cavities (Figure 11F). In addition, immunohistochemistry was applied to detect the angiogenesis in the necrotic area. CD31 staining was positive and the blood vessels presented with typical round or oval structures. The number of blood vessels in the EXO-miR-122-5p-agomir group was significantly higher than that in the EXO-agomir-NC group (Figure 11G). To sum up, BMSCs-derived exosomes modified by miR-122-5p slowed down the progression of ONFH by promoting local angiogenesis and preventing bone loss in ONFH rabbit.
miR-122-5p modified BMSC-derived exosomes exert a promotive influence on repair and healing in ONFH rabbits
ONFH, a disease of mesenchymal cells or bone cells, occurs commonly in young patients and could result in femoral head collapse and even the need of total hip replacement [41,42]. Interestingly, miRNA-carrying exosomes have been previously highlighted to influence the process of osteoporosis by selectively regulating osteoblast activity [43,44]. In the current study, we aimed to elucidate the underlying mechanism of miR-122-5p in ONFH by altering its expression in ONFH rabbit models. Of importance, BMSCs-derived exosomes may prevent ONFH by means of miRNA-122-5p-dependent mechanisms.
Our findings revealed that miR-122-5p was poorly expressed in rabbits with ONFH, and over-expression of miR-122-5p could contribute to osteoblast differentiation and osteogenesis, accompanied by an increase in RUNX2 and collagen I expressions in osteoblasts. Consistently, a previous study documented down-regulated expression of miR-122 in HCC patients infected with Hepatitis B virus, and indicated that miR-122 expression could regulate the proliferation of HCC . Similarly, breast cancer cells also exhibited down-regulated miR-122-5p, which serves as a tumor suppressor by regulating various signaling pathways, and we speculate likewise effects on ONFH . Moreover, another microRNA, miR-195-5p also exhibits decreased expression in ONFH . RUNX2, one of the earliest and most specific markers of osteogenesis, accelerates osteoblast differentiation, resulting in increased amounts of immature osteoblasts within the crucial period of bone development . Furthermore, Liu et al. proposed that the interactions between mediator MED23 and RUNX2 positively affect osteoblast differentiation and osteogenesis . In addition, a preliminary study suggested that RUNX2 functions as a transcription factor, which can elevate the expression of collagen I in osteoblasts . Meanwhile, RUNX2 was also found to be negatively modulated by miR-10 via the MAPK/ERK signaling pathway during osteogenic differentiation . In BMSCs, miR-210 could facilitate the expression of VEGF and stimulate osteoblast proliferation with a time-dependent manner [52,53]. Additionally, we verified that SPRY2 is a target gene of miR-122-5p by means of a dual luciferase reporter assay. In accordance with our findings, another study demonstrated that SPRY2, a negative regulator of the ERK/MAPK signaling pathway, is a direct target gene of miR-122-5p . This relationship is significant as miR-122-5p-induced down-regulation of SPRY2 has been previously demonstrated to promote the proliferation of keratinocytes in psoriasis .
Exosomes carrying miRNAs serve as crucial vehicles for intercellular communication, and poor circulating-exosome levels have been documented in ONFH . We founded that the release of exosomes could facilitate the repair of ONFH in a rabbit model. Co-culture of BMSCs with osteoblasts promoted osteoblast proliferation and osteogenesis. Stem cell-based therapy has been proven to be feasible for its efficient curative effect . Another study suggested that exosomes derived from human platelet-rich plasma can prevent glucocorticoid-induced apoptosis in ONFH via the Akt/Bad/B-cell lymphoma 2 signaling pathway under endoplasmic reticulum stress . Guo et al.  have suggested that exosomes released by human synovial-derived MSCs may serve as a protective factor for glucocorticoid-induced ONFH. Correspondingly, exosomes can be released from human pluripotent stem cell-derived BMSCs and can potentially reduce the risk of ONFH by improving angiogenesis . Moreover, we discovered that the exosome-specific inhibitor, GW4869 inhibited the release of BMSCs-derived exosomes and miR-122-5p expression in exosomes. The aforementioned findings and researches suggest that exosomes exert a protective role in ONFH and exosomes carrying miR-122-5p could promote the repair and healing of ONFH by accelerating angiogenesis as evidenced by increased number of blood vessels.
Our study also evidenced that over-expression of miR-122-5p could activate the RTK/Ras/MAPK pathway via up-regulation of RTK activity. RTK is associated with the conserved Ras/MAPK cassette and known to be a primary regulator of various vital cellular processes, including cell proliferation, differentiation and cell cycle [59,60]. Moreover, our findings demonstrated that symptoms of ONFH were relieved after down-regulation of SPRY2 and activation of RTK activity. Similarly, the targeting relationship between miRNA-23a and SPRY2 via the ERK signaling pathway in gastric cancer has been highlighted previously . Another study indicated that SPRY2 inactivates the Ras/MAPK signaling pathway and regulates different biological processes via the Casein kinase 1 [40,62,63]. According to the study of Wang et al. , the relationship between miR-122 and SPRY2 was determined using dual-luciferase reporter assays. MiR-122 is also known to modulate the ERK/MAPK signaling activity by repressing the expression of SPRY2, a known regulator of the RTK signaling pathway, to affect the duration and magnitude of ERK/MAPK activity .
In conclusion, our study demonstrated that BMSCs-derived exosomes carry miR-122-5p and promote osteoblast proliferation, differentiation and osteogenesis as well as angiogenesis in vivo. This function is exerted by suppression of the SPRY2 expression and activation of RTK activity via the RTK/Ras/MAPK signaling pathway. Our findings indicate exosome therapy as a potential therapeutic target for ONFH.
BMSCs have been used as a cellular therapeutic option for treatment of ONFH. Moreover, it has been documented that exosomal-encapsulated miRNAs play critical roles in a diversity of diseases or injuries.
We established ONFH rabbit models with expectation to study the function of BMSC-derived exosomal miR-122-5p in ONFH. Then, we observed that BMSC-derived containing up-regulated miR-122-5p attenuated ONFH development in rabbits by promoting osteoblast proliferation, differentiation, and osteogenesis via SPRY2.
The novel aforementioned findings asserted may open novel chapters for future ONFH treatments.
We would like to thank all participants enrolled in the present study.
All animal experiments were approved by the Ethics Committee of Wuhan Third Hospital, Tongren Hospital of Wuhan University and followed the Guide for the Care and Use of Laboratory Animals. All animal experiments were conducted in the Animal Management Center of Wuhan Third Hospital, Tongren Hospital of Wuhan University. All efforts were made to minimize the number and suffering of the included animals.
This work was supported by the National Natural Science Foundation of China [grant numbers 81472103, 81071463]; the Key Project of the Natural Science Foundation of Hubei Province [grant number 2015CFA079]; the Applied Basic Research Program of Wuhan Municipal Bureau of Science and Technology [grant number 2015061701011626]; the Wuhan Innovative Talent Development Funding Project; the Training Program for Young and Middle-aged Medical Talents in Wuhan Municipal Planning Commission; the Health Family Planning Research Fund of Wuhan City [grant number WX18M01]; and the Wuhan City ‘Huanghe Talent’ Program.
The authors declare that there are no competing interests associated with the manuscript.
Wen Liao, Yu Ning, and Hai-Jia Xu designed the study. Wen-Zhong Zou, Jing Hu, and Xiang-Zhong Liu collated the data and carried out the data analyses. Yi Yang and Zhang-Hua Li wrote the main manuscript text. All authors have read and approved the final submitted manuscript.
bone mineral density
BM mesenchymal stem cell
Dulbecco’s Modified Eagle’s Medium
extracellular signal-regulated kinase
fetal bovine serum
human embryonic kidney
human leukocyte antigen
mitogen-activated protein kinase
- miRNA, miR
mesenchymal stem cell
mean trabecular plate thickness
nanoparticle tracking analysis
ON of the femoral head
PBS containing 0.1% Tween-20
platelet-derived growth factor receptor
receptor tyrosine kinase
reverse transcription quantitative polymerase chain reaction
runt-related transcription factor 2
trabecular bone volume
transmission electron microscope
vascular endothelial growth factor receptor
These authors contributed equally to this work.