miR-424-5p shuttled by bone marrow stem cells-derived exosomes attenuates osteogenesis via regulating WIF1-mediated Wnt/β-catenin axis

Specimen collection and cell culture

The experiment of this study was authorized by the Research Ethics Committee of the Departments of Orthopedics and Pathology, 3201 Hospital. All patients who provided specimens for this study signed informed consent. Bone marrow specimens were obtained from 10 femoral head tissues with osteoporosis that were discarded during total hip arthroplasty (THA). Human BMSCs were isolated by the whole bone marrow adherent method. Then, the cells were cultured in RPMI1640 (Thermo Fisher Scientific, MA, USA) supplemented with 10% fetal bovine serum (FBS) (Thermo Fisher Scientific, MA, USA) and 1% penicillin/streptomycin (Invitrogen, CA, USA) at 37° C with 5% CO₂. During the logarithmic growth phase, 0.25% trypsin (Thermo Fisher HyClone, Utah, USA) was used for trypsinization and passage. Subsequently, BMSCs were inoculated into 100-mm tissue culture dishes and cultured up to 80% fusion rate to prepare exosomes. Afterward, the cells were cultured in a medium containing 2% exosomes-free serum (obtained through centrifugation at 4° C for 14 hours at 100000 rpm of serum) for 2 days.

Harvesting and identification of BMSCs-derived exosomes

As previously reported [13], the cell supernatant was collected and centrifuged at 4° C at 300 rpm (10 min), 1200 rpm (20 min), and 10,000 rpm (30 min), respectively. Then, the final supernatant was centrifuged at 4° C at 100,000 rpm for 1 hour. After the supernatant was removed, the exosome sediments were washed with plenty of cold PBS and centrifuged at 100,000 rpm at 4° C for 1 hour. Subsequently, the residual liquid was decanted and sucked out to obtain exosomes, which were reserved at -80° C until use. Finally, exosomes were examined with transmission electron microscopy (TEM, Tecnai, FEI, Hillsboro, OR, USA), and western blot was used to monitor exosome surface markers CD9 (1: 1000, Abcam, Cambridge, UK), CD63 (1: 1000, Abcam) and CD81 (1: 1000, Abcam).

Osteogenic differentiation in vitro

For osteoblast differentiation, BMSCs were inoculated in 12-well plates at 5×10⁴ cells/well for 24 hours before induction. Then, osteogenic differentiation was induced by culturing BMSCs in osteogenic differentiation medium (ODM, containing 20 mM β-glycerophosphate, 50 μg/mL ascorbic acid and 100 nM dexamethasone). The medium was altered every 3-4 days. Alizarin red staining (ARS) and qRT-PCR were conducted to determine the impact of exosomes on osteogenic differentiation. In addition, 2 μg/mL BMSCs-derived exosomes were added to BMSCs in the ODM to determine its effect on differentiation.

Cell transfection

Human BMSCs at the logarithmic growth stage were seeded into 6-well plates (5×10⁵/well). When the fusion rate of the cells reached 50%-60%, pcDNA empty vector (NC), pcDNA-WIF1 (WIF1), miR-424-5p inhibitor (miR-424-5p inhibitor group) or NC(NC group) synthetized by GenePharma (Shanghai, China) were transfected into BMSCs following the specification of the FuGENE®HD Transfection Reagent (Roche, Shanghai, China). After being incubated for 48 hours, the cells were collected, and the RNA or protein expression was tested by quantitative real-time polymerase chain reaction (qRT-PCR) and western blot.

Alizarin Red S (ARS) staining and quantitative analysis

ARS staining was applied to examine bone formation of BMSCs. In short, BMSCs were incubated with ARS staining buffer (Cyagen Bioscience, USA) for 20-30 min. Then, they were examined under an optical microscope (Nikon, Tokyo, Japan) to determine their salinity. Upon ARS dissolution, BMSCs were incubated with 100 mM cetylpyridinium chloride (Sigma, St Louis, MO, USA) for 1 hour. Finally, the optical density of the released ARS was measured at 570 nm with a microplate reader (Tecan, Mannedorf, Zurich, Switzerland).

Alkaline phosphatase (ALP) staining

The BMSCs were cultured in 24-well plates, and then the medium was discarded. Afterward, the cells were washed with phosphate-buffered saline (PBS), fixed with 95% ethanol, stained with ALP solution and cultured for 4 hours in an incubator at 37° C. Then, 2% cobalt sulfide and ammonium sulfide (Tianjin Tianli Chemical Reagent Co., Ltd, Tianjin, China) were added. Subsequently, the cells were incubated with 10 mM p-nitrophenyl phosphate (Meilunbio, Dalian, China) for 30 min for quantitative analysis. Finally, the release of p-nitrophenol was determined by the ALP colorimetric kit (BioVision, USA) to evaluate the ALP activity. The absorbance was measured at 420 nm with a microplate reader (Tecan, Mannedorf, Zurich, Switzerland).

QRT-PCR

Total cellular RNA was extracted with the TRIzol reagent, and then it was reverse-transcribed into cDNA with PrimeScript™ RT Reagent kit (Invitrogen, Shanghai, China) as the manufacturer's specifications. Bio-Rad CFX96 quantitative PCR system and SYBR were used for qRT-PCR following the manufacturer's regulations. The conditions for PCR were as follows: Initial denaturation for 5 min at 95° C, denaturation for 15 s at 95° C, and annealing for 30 s at 60° C. GAPDH and U6 served as internal references of genes and miRNAs, respectively. The 2 ^(-ΔΔCt) method was adopted for statistics. Each experiment was done in triplicate. The specific primer sequences were in Table 1.

Western blot and co-immunoprecipitation

Total protein was obtained from human BMSCs with RIPA lysis buffer (Beyotime Biotechnology, Shanghai, China). Protein quantification was carried out with the Bradford method, and the protein samples were boiled for 5 min, cooled on ice and centrifuged for 30 s. Then, the supernatant was taken for polyacrylamide gel electrophoresis, and transferred to the polyvinylidene fluoride (PVDF) membranes at 100 V for 1 hour. Afterward, the membranes were blocked with 5% skimmed milk at room temperature (RT) for 1 hour, and then incubated with the primary anti-ALP (ab83259, 1:1000, Abcam, MA, USA), RUNX2 (ab76956, 1:1000), OPN (ab214050, 1:1000, Abcam), OCN (ab93876, 1:500, Abcam), WIF1 (1:1000, ab155101, Abcam), Wnt (1:1000, ab219412, Abcam), β-catenin (1:1000, ab32572, Abcam) or β-actin (ab8226, 1:1000) overnight at 4° C. Next, the membranes were rinsed with TBST twice, and incubated with the luciferin-labeled secondary antibody for 1 hour at RT. After washing the membrane for three times, ECL chromo-developing agent (Millipore, Billerica, MA, USA) was applied for exposure, and the membrane scanner was adopted for imaging.

For co-immunoprecipitation, the cells were rinsed with PBS and lysed for 20 min in a 4° C lysis buffer containing 50 mM Tris-HCl, 150 mM NaCl, 1% NP-40, 10% glycerol, 0.1% SDS, and 2 mM Na₂EDTA. The lysates were then removed through centrifugation, and the protein was immunoprecipitated with the affinity antibodies and A+G agarose beads (Beyotime, China) at 4° C. The immunoprecipitation was washed 3 times with 1 mL lysis buffer, and then analyzed by standard western blot.

Dual-luciferase reporter assay

The TargetScan software predicted that WNT3a was a potential target of miR-424-5p. Wild type and mutant WNT3a-3’UTR reporter gene plasmids were constructed, and miR-424-5p mimics and their negative controls were transfected into BMSCs. The dual-luciferase reporter assay was caried out 48 hours later following the experimental specifications (Promega, Madison, WI, USA), and the relative fluorescence intensity of different treatment groups was determined according to the ratio of firefly fluorescence intensity/renilla fluorescence intensity examined by the microplate reader.

RNA binding protein immunoprecipitation (RIP) assay

RIP was implemented with Magna RIP Kit (EMD MilliPore, Billerica, MA, United States) as per manufacturer's guidelines. After the cells were treated with RIP lysis buffer, human Anti-Ago -2 antibody (microporous) or the control antibody (normal mouse immunoglobulin, Micropores) was added and incubated overnight at 4° C. QRT-PCR was conducted to verify the WNT level.

Statistical analysis

SPSS17.0 statistical software (SPSS Inc., Chicago, IL, USA) was employed for analysis. Measurement data were presented as mean ± standard deviation (x±s). One-way ANOVA was adopted for multi-factor comparison, and the t test was employed for pair-wise comparison. P<0.05 indicated statistical significance.

Ethics statement

Our study was approved by the Ethics Review Board of 3201 Hospital.

Data availability statement

The data sets used and analyzed during the current study are available from the corresponding author on reasonable request.

Identification and characterization of BMSCs-derived exosomes

Flow cytometry analysis with cell surface specific markers was used to identify BMSCs (Figure 1A). It was found that CD90, CD44 and CD105 were positive, while CD31, CD34 and CD45 were negative in BMSCs. BMSCs-derived exosomes were isolated through centrifugation (Figure 1B), and the purified exosomes were observed by TEM. Western blot confirmed that BMSCs-derived exosomes expressed and enriched with known exosome markers CD9, CD63 and CD81 (Figure 1C).

Figure 1. Identification and characterization of BMSCs-derived exosomes. (A) Flow cytometry was conducted to test the surface markers of BMSCs. (B) Exosomes were extracted from BMSCs by TEM identification. (C) Western blot was conducted to monitor the exosome surface markers.

BMSCs-derived exosomes inhibited osteogenesis

Osteogenic BMSC cells from patients were treated with normal BMSCs-derived exosomes for 7 days. The levels of miR-424-5p and WIF1 were testified by qRT-PCR, which confirmed that miR-424-5p was overexpressed, while WIF1 was knocked down in exosomes from osteoporosis patients (Figure 2A, 2B). Then, we adopted western blot to determine the WIF1/Wnt/β-catenin protein expression, and discovered that compared with that in exosomes extracted from normal cells, WIF1 was knocked down, while Wnt and β-catenin were overexpressed in exosomes from osteoporosis patients (Figure 2C). Subsequently, the mRNA expression of OPN, OCN, and Runx2 in BMSCs were verified by qRT-PCR, and the results indicated that their levels in exosomes extracted from osteoporosis patients were lower than those in the control group (Figure 2D–2F), which was further demonstrated by western blot (Figure 2G). Besides, ALP staining showed that compared with BMSCs treated with exosomes extracted from normal, those treated with exosomes from osteoporosis patients had less blue cytoplasmic granules and decreased ALP (Figure 2H), suggesting that exosomes derived from BMSCs in osteoporosis patients inhibited osteogenesis.

Figure 2. Inhibitory effect of BMSCs-derived exosomes on osteogenesis. Osteogenic BMSC cells were treated with exosomes from normal bone and osteoporosis patients for 7 days, respectively. (A, B) The miR-424-5p and WIF1 levels were verified by qRT-PCR. (C) Western blot was implemented to examine the protein expression of WIF1/Wnt/β-catenin. (D–F) The expression of OPN, OCN and Runx2 mRNAs were compared by qRT-PCR. (G) Western blot was implemented to examine the protein expression of OPN, OCN and Runx2. (H) ALP staining. ***P<0.01, ***P<0.001 (vs. Con(Exo)).

Dampening the miR-424-5p in exosomes enhanced osteogenic differentiation of BMSCs

BMSCs-derived exosomes with miR-424-5p inhibitors were obtained, and then co-cultured with osteogenic BMSCs for 7 days. Then, the miR-424-5p and WIF1 expression were examined by qRT-PCR. As a result, compared with the OP(Exo)+ miR-NC group, miR-424-5p was knocked down and WIF1 was overexpressed after transfection with miR-424-5p inhibitors (Figure 3A, 3B). Meanwhile, western blot was performed to test the WIF1/Wnt/β-catenin expression, and it was found that compared with the OP(Exo)+ miR-NC group, WIF1 was overexpressed, while Wnt and β-catenin were knocked down after transfection of miR-424-5p inhibitors (Figure 3C). Subsequently, the mRNA expression of OPN, OCN, and Runx2 in BMSCs were compared by qRT-PCR, and the results illustrated that their levels were notably higher than those of the OP(Exo)+miR-NC group after transfection of miR-424-5p inhibitors (Figure 3D–3F). Also, western blot demonstrated the protein expression changes of OPN, OCN, and Runx2 were similar with that in mRNA (Figure 3G). Furthermore, ALP staining proved that compared with OP(Exo)+miR-NC group, the BMSCs transfected with miR-424-5p inhibitors had more blue cytoplasmic granules, accompanied by ALP elevation (Figure 3H), suggesting that dampening miR-424-5p expression in exosomes enhanced osteogenic differentiation of BMSCs.

Figure 3. Dampening miR-424-5p expression in exosomes enhanced osteogenic differentiation of BMSCs. MiR-424-5p inhibitors were transfected into BMSCs-derived exosomes, which were then co-cultured with BMSCs for 7 days. (A, B) The expression of miR-424-5p and WIF1 were detected by qRT-PCR. (C) Western blot was adopted to test the protein expression of WIF1/Wnt/β-catenin. (D–F) QRT-PCR was implemented to verify the mRNA expression of OPN, OCN and Runx2. (G) Western blot was employed to monitor the protein expression of OPN, OCN and Runx2. (H) ALP staining. **P<0.01, ***P<0.001 (vs. Con(Exo)). &&P<0.01, &&&P<0.001 (vs. Con(Exo)+miR-in).

miR-424-5p targeted WIF1

We adopted the online websites to search the target genes of miR-424-5p. Interestingly, StarBase (https://web.archive.org/web/20110222111721/http://starbase.sysu.edu.cn/) proved that WIF1 was a target of miR-424-5p (Figure 4A). In addition, dual-luciferase reporter assay proved that miR-424-5p bound to the wild-type WIF1 (Figure 4B). Hence, we conducted RIP to further clarify the relationship between WIF1 and miR-424-5p. As a result, after BMSCs were transfected with miR-424-5p, the amount of WIF1 precipitated in the Ago2 antibody group was far more than that in the IgG group, suggesting that WIF1 bound to Ago2 through miR-424-5p (Figure 4C). Subsequently, immunoprecipitation showed that WIF1 bound to WNT, and WIF1 was expressed in WNT, while not in IgG (Figure 4D). Finally, we transfected WIF1 overexpressed plasmids in BMSCs-derived exosomes. The result manifested that after transfection of WIF1 overexpressed plasmids, the miR-424-5p mRNA was down-regulated, while the WIF1 mRNA was up-regulated (Figure 4E, 4F). The above findings illustrated that miR-424-5p targeted WIF1 and was negatively correlated with it.

Figure 4. miR-424-5p targeted WIF1. (A) The downstream molecules of miR-424-5p were predicted by StarBase database (https://web.archive.org/web/20110222111721/http://starbase.sysu.edu.cn/), and miR-424-5p was found to be the target of WIF1. (B) The binding association between miR-424-5p and WIF1 was testified by dual-luciferase reporter assay. (C) RIP was conducted to determine the targeting correlation between miR-424-5p and WIF1. (D) IP was applied to examine the association between WIF1 and Wnt. (E, F) MiR-424-5p and WIF1 levels were tested by qRT-PCR. NSP>0.05, **P<0.01, ***P<0.001 (vs. miR-NC, vs. Con(Exo)). &&P<0.01, &&&P<0.001 (vs. Con(Exo)).

BMSCs-derived exosomes attenuated osteogenesis via regulating miR-424-5p/WIF1

BMSCs were divided into 4 groups based on different treatment methods, namely Con(Exo), OP(Exo), OP(Exo)+PCDNA-NC and OP(Exo)+PCDNA-WIF1. Interestingly, the mRNA levels of OPN, OCN and Runx2 in the OP(Exo) group were distinctly lower than that in the Con(Exo) group. Beside, the mRNA expression of OPN, OCN and Runx2 in the OP(Exo)+PCDNA-WIF1 group was increased compared with the OP(Exo)+PCDNA-NC group. However, mRNA levels of these genes did not change significantly compared with the Con(Exo) group (Figure 5A–5C). Similar results were obtained by ALP activity detection (Figure 5D). Also, western blot was implemented to compare the protein expression of OPN, OCN and Runx2, and the results were consistent with the gene expression trend (Figure 5E). Furthermore, western blot was used to monitor the expression of WIF1/Wnt/β-catenin. Interestingly, compared with the Con(Exo) group, WIF1 was down-regulated, while Wnt and β-catenin were up-regulated in the OP(Exo) group, and results were reversed after transfection of WIF1 overexpressed plasmid (Figure 5F). The above results suggested that WIF1 overexpression exosomes partially reversed the osteogenic effect of BMSCs-derived exosomes in osteoporosis patients.

Figure 5. BMSCs-derived exosomes inhibited osteogenesis via miR-424-5p/WIF1. The WIF1 overexpressed plasmid was transfected into BMSCs-derived exosomes. (A–C) The expression of OPN, OCN and Runx2 were compared by qRT-PCR. (D) ALP staining was used to examine ALP activity. (E) Western blot was applied to determine the protein expression of OPN, OCN and Runx2. (F) The protein expression of WIF1/Wnt/β-catenin was verified by western blot. **P<0.01, ***P<0.001 (vs. Con(Exo)). &&P<0.01, &&&P<0.001 (vs. OP(Exo)+PCDNA-NC).