The preparations of cells
We obtained hBMSCs (P4) and RAW264.7 cells from Chinese Academy of Sciences. These two kinds of cells were respectively cultured in complete α-MEM and high glucose DMEM, which consists of 10% FBS (Gibco) and 1% penicillin as well as streptomycin. We divided the experiment into following groups: PBS group, Ti group (titanium particle), Ti+miR-181bm NC group (titanium particle; NC: miR-181b mimic negative control), Ti+miR-181bm group (titanium particle, miR-181b mimic), Ti-miR-181bi NC (titanium particle, miR-181b inhibitor negative control), Ti-miR-181bi (titanium particle, miR-181b inhibitor), si-PRKCD (PRKCD that is inhibited by siRNA). Then, RAW264.7 was seeded into 6-well plates, treated with the above conditions and incubated in the environment of 37°C and 5% CO2.
The identification of hBMSCs
For the characterization of hBMSCs, we obtained images of hBMSC by optical microscope for explore whether hBMSC could adhere to the plastic disk. Also, in vitro tri-lineage differentiation including adipogenesis, osteogenesis, chondrogenesis capacity were performed. Meanwhile, flow cytometry for detecting the surface markers of hBMSC was conducted. The adipogenesis (Oil Red O staining), osteogenesis (Alizarin Red staining), chondrogenesis (Alcian Blue staining) were assessed after being induced by osteoblast inducing medium for 3 weeks.
The detection of inflammatory factors by ELISA
After seeded into 24-well plates RAW264.7 cells were treated with PBS, Ti, Ti+miR-181bm NC and Ti+miR-181bm for 24 h. Subsequently, the cell supernatants were acquired and used for measuring the concentrations of inflammatory and anti-inflammatory factors including TNF-α, IL-6 and IL-10 ELISA kit (Anogen, Canada).
qRT-PCR analysis
The relative gene expression including miR-181b, Arg-1, iNOS, CCR7, CD206, ALP, RUNX2, OCN, VEGF and BMP-2 was detected by qRT-PCR. Firstly, we applied TRIzol reagent (Invitrogen) for extracting the total RNA of RAW264.7 and hBMSC. Then, PrimeScript RT reagent Kit (Takara; Vazyme Biotech) was used for reverse transcription for acquiring complementary DNA of the extracted total RNA. Subsequently, qRT-PCR analysis experiments were conducted by SYBR Green detection reagent system (Takara). Finally, the levels of relative gene expression were calculated by applying the 2^-(△△CT) and 18S RNA levels was used for normalization.
Cell proliferation and migration
The proliferation of hBMSCs were determined by CCK-8 assay. In short, 2×103 cells/well hBMSCs were seeded on 96-well plate. After cultured with equivalent conditioned culture medium extracted from PBS, Ti, Ti+miR-181bm NC, Ti+miR-181bm groups after 1, 3 and 7 days, we removed and replaced culture medium by fresh culture medium and CCK-8 solution (Dojindo, Japan). The absorbance was evaluated by ELISA plate reader (Epoch, BIO-TEK, USA) at 450 nm after 2 h at 37℃.
Migration ability was assessed by transwell assay. Specifically, 1.5×104 hBMSCs were seeded on the upper chamber of a transwell plate with basal culture medium (Millipore). And the lower chamber was added with conditioned medium extracted from PBS, Ti, Ti+miR-181bm NC, Ti+miR-181bm groups. 24 h later, we removed the upper chamber cells, fixed the lower chamber cells with 4% paraformaldehyde and stained the migrated cells by 0.5% crystal violet for 5min. Optical microscope (Olympus IX 70, Tokyo, Japan) was utilized for the observing the migrated cells before counted by Image J.
Osteogenic differentiation
We evaluated the osteogenesis by Alkaline phosphatase (ALP) staining, Alizarin red staining and detecting the relative gene levels of ALP, RUNX2 (runt-related transcription factor 2), and OCN (osteocalcin), VEGF, BMP-2. Osteogenic differentiation medium supplemented with conditioned medium extracted from PBS, Ti, Ti+miR-181bm NC, Ti+miR-181bm groups for inducing hBMSCs for 14 and 21 days respectively. The cells induced by osteogenic differentiation medium were fixed with 4% paraformaldehyde. ALP (Beyotime) or Alizarin Red (Cyagen) staining solution were used for the staining. The samples were then observed and photographed by inverted optical microscope (Olympus IX 70, Tokyo, Japan). The relative gene levels including ALP, RUNX-2, OCN were assessed by qRT-PCR.
The extraction and identification of exosome
Exo and Exo-181b were acquired from the supernatant by ultracentrifugation. Concretely, miR-181bm NC and miR-181bm were used for the transfection of hBMSCs with serum-free culture medium for 72 h. Then, we acquired and centrifuged the medium at 400 g for 6 min and 2,100 g for 25 min for discarding dead cells when the cell confluence reached about 80%. Subsequently, a 0.22-μm filter was used for the filtration of the supernatants for getting rid of the nanoscale non-exosome impurities (Micropore). After that, the supernatant was filtrated by Ultra-Clear™ tubes (Beckman Coulter, USA) for filtrating at 100,000 g by ultracentrifuge for 1.5 h twice. Eventually, we resuspended the required pellets by PBS and kept them in -80℃ refrigerator.
Exosomes are nanoscale extracellular vesicles and transmission electron microscopy (TEM, JEM-1400) was applied for the observation of the ultrastructure and shape. In addition, nanoparticle tracking analysis (NTA, ZetaView PMX 110, Particle Metrix) was used for detecting the size distribution as well as nanoparticle concentration. Western blotting was utilized for the detection of the specific markers of exosomes CD9, Tsg101 and Alix.
Air pouch model
In our research, animal surgeries were approved by the Animal Care and Ethics Committee of Shanghai Sixth People's Hospital and were conducted in accordance with established guidelines. C57/BL6 mice were anesthetized by sodium pentobarbital and sterilized air was injected subcutaneously for establishing an air pouch model. Meanwhile, PBS, PT, PT+Exo and PT+Exo-181b were implanted into subcutaneous tissue for detecting the effect of Exo-181b on inflammation. After four days, we washed subcutaneous pouch by 3ml stain buffer for collecting cells. Then, the percentage of M1 and M2 macrophages was determined by flow cytometry.
The establishment of femoral bone defect model of rats.
Porous titanium alloy in our research has a pore size from 500 to 600 μm, which is too large for the sustained release of exosomes. Therefore, we applied a biocompatible, biodegradable commercial hyaluronic acid hydrogel for solving this problem. Thirty-two rats (12-week, male) were randomly divided into four groups containing porous titanium alloy group (PT), PT+HA hydrogel (PT-H), PT-H+exosomes group (PT-H-Exo), Ti-H-Exo+miR-181b group (PT-H-Exo-181b) (n=8). After anesthetizing the rats by 0.6% phenobarbital sodium at a dose of 1 ml/kg intraperitoneally, we made a knee medial incision and exposed the medial of femur condyles by moving the patella outward. Then, femoral condyles were drilled laterally with a 3.5-mm diameter trephine (Nouvag AG) to produce 5-mm-deep bone defect and implanted with PT, PT-H, PT-H-Exo and PT-H-Exo-181b. After that, the soft tissues were sutured layer by layer with 4-0 resorbable sutures postoperatively. When the anesthesia was finished, all the rats underwent operation were returned to the biosafety facility and guaranteed enough food and water. Each rat received antibiotics intramuscularly for 3 days. The rats were euthanized 3 months after the surgery and the femurs were harvested and fixed in 4% paraformaldehyde solution.
MicroCT
MicroCT (Skyscan 1176, Kontich, Belgium) was utilized for measuring new bone formation. The samples were scanned and the resolution was 18 μm. Afterwards, TV (total volume), BV (bone volume), BV/TV as well as BMD (bone mineral density) in the defects were assessed by CTAn image analysis software. Sagittal and three-dimensional reconstruction images were performed by Dataviewer and CTVox.
Sequential fluorescent labeling
Trichromatic fluorescent labeling was used for assessing the new bone formation. Specifically, 25 mg/kg tetracycline (TE, Sigma, yellow), 30 mg/kg alizarin reds (AL, Sigma, red) and 20 mg/kg calcein (CA, Sigma,green) were injected intraperitoneally respectively after 3, 6 and 9 weeks postoperatively.
Histological analysis
We dehydrated the harvested samples by gradient concentrations of alcohols and were immersed in xylene for transparency for 4 hours. Later, the samples were immersed in the infiltration liquid and embedding liquid for 2 days. Subsequently, all the samples were put into glass bottle full of embedding liquid for polymerization in 37℃ water bath. Then, we applied microtome for acquiring undecalcificated bones slicing (150μm thick, Leica). Two-photon confocal microscope (Leica) was used for observing the fluorescent labeling. The excitation/emission wavelengths for tetracycline, alizarin red and calcein were 405/560-590 nm, 543/580-670 nm, and 488/500-550 nm respectively. Later, the sections were polished to 50μm and were stained with van Gieson’s picrofuchsin for evaluating the new bone formation. Finally, the quantification of new bone formation area was calculated by Image J.
Bioinformation analysis
Four different databases including TargetScan, miRanda, miRDB and PicTar were used for predicting the target genes of miR-181b binding sites. Then we applied gene ontology database and chose the targets that were related to inflammatory responses. All the genes were intersected by Venn chart for obtaining the common genes.
Luciferase report
We purchased luciferase reporter plasmids containing the 3'-UTRs of PRKCD and the corresponding mutated 3'-UTR. After that, we inserted the 3'-UTRs and mutated 3'-UTR into pRL-CMV reporter vector (Promega, USA). Then, 96-well plates were used for the incubation of 293T cells for 24h. After that, a reporter plasmid and miR-181b mimics using Lipofectamine 2000 (11668-019, Invitrogen) was co-transfected into each well. Subsequently, we detected renilla and firefly luciferase activities through Dual-Luciferase Reporter Assay System (E1910, Promega).
Western blotting
Western blotting for the evaluation of PRKCD, AKT and p-AKT protein expression. Prechilled RIPA buffer was applied for lysing and harvesting the protein of RAW264.7 on ice. Subsequently, 5× loading buffer was applied for diluting the lysates. Then, we boiled the dilution at 95°C for about 5 min. Then, SDS-PAGE was used for separating different by molecular weight and the separated proteins were transferred onto a PVDF membrane (Merck-Millipore) and blocked with nonfat milk. Eventually, the primary antibodies and secondary antibodies were applied for incubating PVDF membrane overnight and for 1 h respectively. An ECL substrate kit was applied for visualizing the protein bands of the PVDF membrane.
Statistics
We used Student-Newman-Keuls post hoc tests, one-way ANOVA and two-way ANOVA for evaluating the statistical significance. Statistical analysis of the mean ± SEM was performed by Graphpad. P value < 0.05 was deemed significant.