The mechanism of miR-889 regulates osteogenesis in human bone marrow mesenchymal stem cells

This study was approved by the Institutional Ethical Committee of TongRen Hospital, Shanghai Jiaotong University School of Medicine. Written informed consent was signed by all included patients. Inclusion criteria were (1) bone mineral density (BMD) of at least 2.5 standard deviation (SD) below the peak mean BMD of healthy young women (− 2.5 T-score), (2) initial diagnosis of osteoporosis (OP) and receive no drug or hormone treatment, and (3) no thyroid disease. Finally, we included 6 women with OP and 5 healthy controls for further analyses of fresh femoral neck trabecular bone from osteoporotic undergoing hip replacement due to either osteoporotic fracture (OP group, n = 6) or osteoarthritis in the absence of OP (control group, n = 5).

Human BMMSCs were obtained from Cyagen company (Cyagen Biosciences; Guangzhou, China) and cultured in the α-MEM containing 10% fetal bovine serum (ThermoFisher, USA). For osteogenic differentiation, the BMSCs were cultured in an induction medium containing 50 mM ascorbic acid, 10 mM sodium b-glycerophosphate, and 10 nM dexamethasone according to previous report [17]. Cells were harvested at the indicated times for mRNA and miRNA extraction. Negative control (NC), miR-889 mimic, miR-889 inhibitor, and WNT7A siRNA were obtained from GenePharm, Tech, (Shanghai, China). The sequences of miR-889 mimics were the primary chain, 5′-ACACTCCAGCTGGGAATGGCTGTCCGTAGT-3′, and passenger chain, 5′-TGGTGTCGTGGAGTCG-3′. The concentrations of miR-889 mimics and miR-889 inhibitor used were 100 nmol/L.

Samples were harvested from OP patients and healthy controls. HBMSCs that used for osteogenic induction were collected after 21 days of intervention. After washing by PBS for three times and smashed. Then, total RNA is extracted by TRIzol reagent (Invitrogen, USA) according to the manufacturer’s instructions. Real-time PCR was performed using the miScript SYBR Green PCR kit (Qiagen) for miRNA expression analysis and a standard SYBR Green PCR kit (Takara, Tokyo, Japan). The relative changes of transcripts of interest were analyzed according to the 2^-△△ct method with GAPDH and U6 being housekeeper genes. Primer sequences are available in Table 1.

First, a Venn diagram was produced to show the potential common target genes of miR-889 in Targetscan (http://​www.​targetscan.​org/​vert_​72/​), miRDB (http://​mirdb.​org/​), and miRanda (http://​www.​microrna.​org/​microrna/​home.​do) databases. Then, DAVID Bioinformatics Resources 6.8 (https://​david.​ncifcrf.​gov/​) was used to verify the Gene Ontology (GO) function of these target genes. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis was also performed to identify the potential pathway of miR-889 that involved. Binding sites of miR-889 and WNT7A were performed by Targetscan databases.

Total proteins are extracted by 1 ml RIPA mixed with PMSF. The isolated protein concentration was determined using BCA Protein Assay Kit (LEAGENE, Beijing, China). Equally amount of protein (20 μg) was separated on SDS-PAGE, electrotransferred to polyvinylidene difluoride (PVDF) membranes. Following blocking within Tris-buffered saline and Tween 20 (TBST) containing 5% skim milk, membranes were incubated with primary antibodies against ALP, BMP-2, RUNX2, OPN, and OCN (1:1000) all from Santa Cruz (USA) overnight at 4 °C. Then, membranes were washed by TBST for 5 min for three times. Samples were incubated with HRP-conjugated second antibodies (1:5000; Santa Cruz, CA, USA) for 2 h to detect immunoreactive bands. Blotted bands were visualized with ECL solution (Boster, Wuhan, China) and exposed to films.

After osteogenic induction for at least 7 days, each plate was washed by PBS for three times and then fixed by 4% paraformaldehyde for 15 min. Then, each plate was coated by nitro-blue tetrazolium chloride (BCIP)/5-bromo-4-chloro-3′-indolyphosphate p-toluidine salt (NBT) solution for 5 min. Then, plates were washed by water to stop the reaction.

ARS was performed to detect calcium deposition in each group. After osteogenic induction for 14 days, each well was fixed by 4% paraformaldehyde for 15 min. Then, we added Alizarin Red solution (0.1%, Ph = 6.8) for 45 min. Stained cells were then photographed after used water to stop the reaction.

To identify the β-catenin activity in response to miRNA treatments, immunofluorescence staining was performed for NC, miR-889 inhibitor, and miR-889 inhibitor+WNT7Ai groups. In briefly, hBMSCs were transfected with NC, miR-889 inhibitor, and miR-889 inhibitor+WNT7Ai, respectively. After osteogenic induction for 7 days, differentiated hBMSCs were harvested and used for immunofluorescence staining. HBMSCs were fixed by 4% paraformaldehyde for 30 min and then washed by PBS foe three times. After blocking with 1% BSA for 30 min, hBMSCs were incubated with primary antibody for β-catenin at 37 °C for 2 h; then, second antibody was added for fluorescence microscope (Olympus, Japan).

Outcomes are shown as the mean ± standard deviation (SD). Correlation between WNT7A and miR-889 was calculated by Pearson correlation coefficient. Statistical analysis was performed with t test between two groups and analysis of variance (ANOVA) if more than the two groups compared using SPSS statistics 21.0 (IBM Corp., Armonk, NY, USA). P values < 0.05 were considered as statistically significant.

As illustrated in Fig. 1, we found that the expression of miR-889 is increased in OP patients than the healthy control group. However, the WNT7A was decreased in OP patients than the healthy control group. Moreover, we found that the relative expression of WNT7A and miR-889 have a negative correlation (r = − 0.855, P = 0.001).

Figure 1b revealed the ALP and ARS staining; in accordance with general observation, ALP staining and red calcium nodules were decreased in miR-889 group, while increased in anti-miR-889 with statistically significant.

We further compared the miR-889 and WNT7A expression during osteogenic differentiation of hBMMSCs. The results are presented in Fig. 1c. Osteogenic-induced medium group was associated with an increased expression of miR-889, while with a reduction expression of WNT7A than the control group. Moreover, ALP staining and ARS results were presented in Fig. 1d, and we found that miR-889 could significantly decrease the ALP activity and calcium deposit than the control group while miR-889 inhibitor could significantly increase ALP activity and calcium deposit than the control group. As illustrated in As the time of osteogenic induction time prolong, the miR-889 expression was decreased (Fig. 1e), while WNT7A relative expression increased (Fig. 1f).

Firstly, we used three target gene prediction websites (miRanda, miRDB, and Targetscan) to verify the target genes of miR-889. Finally, we identify 666 target genes through Venn diagram (Fig. 2a). As listed in Fig. 2b, these 666 target genes gathered in the following Gene Ontology: SMAD binding, positive regulation of transcription from RNA polymerase II promoter, positive regulation of transcription, DNA templated, and postsynaptic membrane. Figure 2c shown the interaction between the target genes, and we further used MCODE model to identify the hub genes. Hub genes of the PPI were shown in Fig. 2e. The hub gene of the PPI was the FBXO22 gene. Figure 2d revealed that 8 potential pathways were enriched: TGF-beta signaling pathway, Wnt signaling pathway, cGMP-PKG signaling pathway, Rap1 signaling pathway, FoxO signaling pathway, phosphatidylinositol signaling pathway, and MAPK signaling pathway. Figure 2f revealed the binding sites of miR-889 and WNT7A.

To further determine the role of miR-889 in osteoblastic differentiation, the BMSCs were infected with negative control (Mic) or transfected with mimic (miR-889) or inhibitors (Anti miR-889) individually. Results were shown in Fig. 3. We found that compared with negative control, miR-889 mimic was associated with a decrease of the osteoblastic markers including ALP, BMP-2, RUNX-2, OPN, and OCN. And, when administrated with anti-miR-889, osteoblastic markers (ALP, BMP-2, RUNX-2, OPN, and OCN) were increased than the negative control and miR-889 mimic. Moreover, we measured Wnt-β catenin signaling pathway-related protein (Wnt7a and β-catenin). We found that miR-889 mimic could significantly reduce the Wnt7a and β-catenin expression. And anti-miR-889 significantly increased Wnt7a and β-catenin expression.

We further used negative control, miR-889 inhibitor, and miR-889 inhibitor combined with WNT7A inhibitor to further analyze the miR-889/Wnt-β catenin axis in inhibiting osteogenesis of hBMSCs. Compared with NC, miR-889 inhibitor decreased the protein level of osteogenic markers (ALP, BMP-2, RUNX-2, OPN, OCN, Fig. 4a) and Wnt-β catenin signaling pathway-related proteins (WNT7A, β-catenin, and LRP5, Fig. 4a). In contrast, when co-cultured miR-889 inhibitor and WNT7Ai, the protein level of osteogenic markers (ALP, BMP-2, RUNX-2, OPN, OCN) was decreased than miR-889 inhibitor alone. Furthermore, we used immunofluorescence to identify the expression of β-catenin in negative control, miR-889 inhibitor and miR-889 inhibitor+WNT7Ai groups (Fig. 4b). Results found that miR-889 inhibitor significantly increased the β-catenin expression level, while co-cultured with WNT7A inhibitor, the β-catenin expression level decreased to some extent. ARS staining was used to further identify miR-889/Wnt-β catenin signaling pathway involved in regulating the osteogenic differentiation of hBMMSCs (Fig. 4c).