miR-344d-3p regulates osteogenic and adipogenic differentiation of mouse mandibular bone marrow mesenchymal stem cells

Experimental animals

Sixty female C57BL/6 mice aged 6 weeks (16 to 18g in weight), purchased from Chongqing Tengxin Biotechnology Co., LTD., License Number: SCXK (Beijing) 2019-0010, were randomly divided into Ovariectomies (OVX) group and sham-operated (Sham) group, with 30 mice in each group. The Affiliated Stomatological Hospital of Zunyi Medical University provided full approval for this research (No.: Lun Shen (2020) 2-473 and Lun Shen(2016)2-188). After separate cages, the animals were raised in SPF animal room with temperature of 26 °C and humidity of 50–65%. The animals could drink and eat freely. Mice were anesthetized with 1% pentobarbital sodium solution at 50mg/kg. Surgical methods of mice in OVX group: the middle and lower 1/3 of the back was skinned after anesthesia, and longitudinal incisions were made along the midline of the middle and lower short back of mice after iodophor disinfection. The skin, muscle and peritoneum were dissected in layers, the ovaries were found and removed, the remaining tissues were reset, and the muscle layer and skin layer were sutured contraptively. The procedure of SHAM group was the same as OVX group, but the ovaries were not removed. Mice were fed for 12 weeks for subsequent experiments.

Cell culture

Mice in the OVX group and the SHAM group aged 18 weeks after operation were sacrificed by cervical dislocation method, soaked in 75% alcohol, and then removed and placed in ultra-clean table about 5 min later. The mandible of mice was removed as soon as possible, and the attached soft tissues were removed. The mice were placed in α-MEM complete medium containing 20% fetal bovine serum, and the bone marrow cavity was rinsed and blown with a syringe. The remaining bone tissue of the mandible was cut and crushed until it became slag. The rinsed bone marrow together with the cut bone residue were placed into a 25 cm² plastic culture flask and cultured at 37 °C in an incubator containing 5% CO2. After the cell growth and fusion reached 80%, it was digested with trypsin for passage. These MBMSCs were used for cell differentiation identification and miRNAs sequencing analysis. Blasts, Mouse embryo osteoblast precursor cells (MC3T3-E1), Mouse embryo fibroblasts, 3T3-L1 and 293T cell lines (Zhongqiao Xinzhou Co., Ltd, Shanghai, China) were cultured in a 25 cm² plastic flask at 37 °C in an incubator with 5% CO2 concentration. MC3T3-E1 was cultured in 10% fetal bovine serum α-MEM complete medium. 3T3-E1 was cultured with DMEM complete medium containing 10% neonate calf serum, and 293T cell line was cultured with α-MEM complete medium containing 5% fetal calf serum. After cell growth and fusion reached 80%, trypsin was used for digestion and passage.

Cell transfection

Lipofectamine™ RNAiMAX (HanBio Co., Shanghai, China) transfected miRNAs on 24-well plates at a concentration of 40 nM (the procedure for transfection of miRNAs containing fluorescence is the same except that it needs to be carried out under dark conditions). 24 h before transfection, 2 × 10⁵ cells were inoculated in medium without antibiotics containing 400 µL per well for transfection when the degree of cell confluency was 50%. Appropriate amounts of miRNAs mimics, inhibitors and blank controls were absorbed and added into EP tubes containing non-double resistant medium, respectively, and then gently blown and mixed. Lipofectamine™ RNAiMAX was absorbed into an EP tube, gently blown and mixed, and stood for 10 min at room temperature. The transfection complex was added to the 24-well cell plate to supplement the medium without double antibiotics, so that the final concentration was 40 nM, and the cell plate was evenly mixed by shaking the cell plate before and after. Pure medium was used for transfection without additional antibiotics. 4 to 6 h after transfection, the medium with serum containing double antibody was replaced. 48–72 h after transfection, it can be used for subsequent experiments.

Induced cell differentiation

When the cell growth and fusion reached 80%, the induced cells differentiated into osteogenic and adipogenic directions, respectively. The osteogenic induction solution was as follows: 1% penicillomycin +0.1 µM dexamethasone +10 mM β-sodium glycerophosphate +0.2 mM vitamin C. Lipid induction solution was configured as follows: 1% penicillomycin +1 µM dexamethasone +0.5 mM IBMX + 2 mM insulin +200 µM indomethacin.

Alizarin red staining and quantification

After osteogenic induction for 21 days, the cells were fixed with 4% paraformaldehyde for 0.5 h, washed with PBS for 3 times, and then added appropriate alizarin red dye (Solarbio, China), and incubated for 5 min. After the dye was washed with PBS, the cells were observed under a microscope and photographed. 500 µl 10% cetylpyridinium chloride was added to the stained cells, and the absorbance value was measured at 570 nm after standing at room temperature for 15 min.

Oil red O staining and quantification

After lipogenic induction for 10 days, the cells were fixed with 4% paraformaldehyde for 0.5 h, rinsed with PBS for 3 times, followed by adding 60% isopropyl alcohol for 10 min, immersed with oil red O dye for 15 min (Solarbio, Beijing, China), rinsed with PBS for 3 times, adding hematoxylin dye for 30 s, rinsed with PBS for 3 times, Oil red O Buffer was added and stood for 1 min. The dye was washed with PBS and observed under a microscope and photographed. Isopropyl alcohol was added to the stained cells, and the absorbance value was measured at 520 nm wavelength after standing at room temperature for 10 min.

qRT-PCR

According to the TRIzol specification, RNA was extracted from the sample (Yang et al., 2019b), and the system was configured with a reverse transcription kit (Bioengineering, China) to complete the reverse transcription reaction and obtain cDNA. The extracted RNA was subjected to agarose gel electrophoresis experiment, and then the electrophoresis samples were observed on the ultraviolet transmission tester. The bands of 28S and 18S samples were bright, clear and sharp, without trailing phenomenon, and the ratio was close to or more than 2:1. RNA with OD 260/280 > 1.8 and OD 260/230 > 1.5 can be used in subsequent experiments. Total RNA in RT reaction solution was 2 µg, and cDNA in PCR reaction solution was 2  µl. qRT-PCR amplification was performed using U6 and β-actin as internal reference for miRNAs and mRNA respectively (Biobiotics, China). Using tailing reaction to detect the expression of micrornas.

Western blot

Cells were lysed (RIPA: phosphatase inhibitor: PMSF = 100:1): Protein quantification and sample preparation were performed by BCA method (Solarbio, Shanghai, China). The protein loading volume was 20 µg. Gel preparation was performed by SDS-PAGE gel rapid preparation kit (Thermo Fisher, Waltham, MA, USA). After sample loading, electrophoresis, membrane transfer, sealing and primary antibody incubation were performed by steps (Abcam, Waltham, MA, USA). On the second day, second antibody incubation and exposure were completed (Millipore Corporation, China). After the bands were exposed, frame selection and gray value detection of swimming lanes were carried out by Image Lab software, and the protein data of target genes (such as RUNX2 and PPAR-γ) were compared with their internal reference, and then multiples of these data were compared with the data of control respectively.

Data analysis

SPSS 18.0 (SPSS, Inc., Chicago, IL, USA) and GraphPad Prism 9.0 (GraphPad Software, San Diego, CA, USA) was used for statistical analysis of data, and independent sample T test was used for comparison between the two groups, and one-way ANOVA was used for comparison between the three groups (nonparametric test and two-tailed test). All experiments were repeated at least three times, and representative experiments were shown. Differences were considered significant at P < 0.05.

MBMSCs were imbalanced in osteogenic and adipogenic differentiation after POP

Alizarin red staining results of MBMSCs in the SHAM group and the OVX group after osteogenic induction showed that the number and volume of calcified nodules of MBMSCs in the SHAM group were larger than those in the OVX group (Fig. 1A). Quantitative results showed that the mineralization content of MBMSCs in the SHAM group was higher (P < 0.001) (Fig. 1B). qRT-PCR and WB results showed that the expression level of osteogenic related gene RUNX2 in MBMSCs in SHAM group was higher than that in OVX group (P < 0.001) (Figs. 1C and 1D). After adipogenic differentiation of MBMSCs in the SHAM group and OVX group, oil-red O staining results showed that the lipid droplets of MBMSCs in the OVX group were larger and more numerous than those in the SHAM group (Fig. 1E), and oil-red O quantitative results also showed that the lipid droplets in the OVX group were more abundant (P < 0.001) (Fig. 1F). qRT-PCR and WB results showed that the expression level of lippogenic differentiation related gene PPAR-γ in MBMSCs in OVX group was higher than that in SHAM group (P = 0.02) (Figs. 1G and 1H).

MBMSCs showed low expression of miR-344d-3p after POP

Microarray sequence analysis was performed on MBMSCs of SHAM group and OVX group to screen out more than 20 miRNAs with highly differentially expressed genes. Significance analysis of microarrays (SAM) was used for differentially expressed genes. MiRNAs whose false discovery rate (FDR) is controlled within 5% and the difference multiple is not less than 4 times are screened out, and these miRNAs are arranged from high to low according to the difference multiple. Through literature review and analysis, miR-344d-3p with higher expression (most down-regulated in OVX group) was not reported (Figs. 2A and 2B).

qRT-PCR was used to detect the endogenous expression level of miR-344d-3p in MBMSCs in the SHAM group and OVX group, and it was found that miR-344d-3p was stably down-regulated in the OVX group, consistent with the chip results (P < 0.001) (Fig. 2C). Then, qRT-PCR was used to detect the expression levels of miR-344d-3p in mouse heart tissue, liver tissue, lung tissue, kidney tissue, pancreas tissue, adipose tissue and skeletal muscle, the values of OVX group were compared with those of SHAM group. The results showed that the expression levels of miR-344d-3p were significantly different in adipose tissue and low in other tissues, suggesting that miR-344d-3p may have adipose-specificity (Fig. 2D). Therefore, miR-344d-3p was selected as the target gene of this study.

miR-344d-3p has osteogenic induction to cells

Control, mimics and inhibitor with fluorescence of miR-344d-3p were transfected into MC3T3-E1 cell lines, respectively. After 48 h observation under inverted fluorescence microscope, green stains could be found in the cytoplasm (Fig. 3A). qRT-PCR results showed that in MC3T3-E1 cell line, the expression of miR-344d-3p in mimics group increased by about 30 times compared with control group (P < 0.001), and decreased by about 4 times in inhibitor group (P < 0.001) (Fig. 3B).

After the miR-344d-3p chemical synthesis was transfected into MC3T3-E1 cell line, alizarine red staining results showed that the number of calcified nodules formed in the miR-344d-3p mimics group was the largest with large and deep volume, followed by the control group and the inhibitor group (Fig. 3C). Alizarine red quantitative analysis showed that the number of calcified nodules in miR-344d-3p mimics group was higher than that in control group (P < 0.01), and the number of calcified nodules in miR-344d-3p inhibitor group was lower than that in control group (P < 0.001) (Fig. 3D). WB results showed that the miR-344d-3p mimics group had the highest protein expression of RUNX2 (about 1.27 times higher than the control group), followed by the control group and the inhibitor group (0.18 times higher than the control group) (Fig. 3E). After transfection of the miR-344d-3p chemical complex into MBMSCs, WB results showed that the miR-344d-3p mimics group had the highest protein expression of RUNX2 (about 1.45 times higher than the control group), followed by the control group. The inhibitor group was 0.85 times the control group again (Fig. 3F). These results all suggest that miR-344d-3p positively regulates the osteogenic differentiation ability of MC3T3-E1 and MBMSCs.

Knockdown miR-344d-3p has adipogenic induction to cells

Control, mimics and inhibitor with fluorescence of miR-344d-3p were transfected into 3T3-L1 cell lines respectively, and observed under inverted fluorescence microscope 48 h later, green stains could be found in the cytoplasm (Fig. 4A). qRT-PCR results showed that in 3T3-L1 cell line, the expression of miR-344d-3p increased about 17 times in mimics group compared with control group (P < 0.001), and decreased about 2 times in inhibitor group (P = 0.01) (Fig. 4B).

After transfection of Figure chemical synthesis into 3T3-L1 cell line, oil red O staining results showed that the number of lipid droplets formed by miR-344d-3p inhibitor group was the most large and deep, followed by control group and mimics group (Fig. 4C). Quantitative analysis of oil-red O showed that the inhibitor group miR-344d-3p had higher lipid droplets than control group (P = 0.02), and the miR-344d-3p mimics group had lower lipid droplets than control group (P = 0.02) (Fig. 4D). WB results showed that PPAR-γ protein expression in miR-344d-3p inhibitor group was the highest (2.2 times higher than control group), followed by control group and mimics group (0.76 times higher than control group) (Fig. 4E). When the miR-344d-3p chemical complex was transfected into MBMSCs, the protein expression of PPAR-γ in miR-344d-3p inhibitor group was the highest (2.43 times higher than control group), followed by control group. The mimics group was 0.66 times higher than the control group (Fig. 4F). These results suggest that miR-344d-3p can negatively regulate the adipogenic differentiation ability of 3T3-L1 and MBMSCs.

Knockdown of miR-344d-3p increases the expression of Dnmt3a genes

Based on the analysis of biological information on target gene prediction websites (miRBase, miRand, miRWalk and TargetScan), it is predicted that miR-344d-3p may bind to the 3′-UTR region of more than 500 genes. Based on its function judgment, it is inferred that the target gene should have corresponding inverse function. Through literature screening and prediction of binding sites, it was found that there was a binding site between “TTATAT” sequence and miR-344d-3p in the 3′-UTR region of Dnmt3a (Fig. 5A). After transfection of miR-344d-3p chemical complex into 293T cell lines, protein expression of Dnmt3a target gene was detected by WB, and Dnmt3a protein expression was significantly increased in miR-344d-3p inhibitor group, while mimics expression was down-regulated (Fig. 5B). The above results indicated that miR-344d-3p overexpression prevented Dnmt3a protein expression, while endogenous Dnmt3a protein level increased significantly after down-regulation of miR-344d-3p. This suggests that miR-344d-3p blocks the translation of Dnmt3a target gene protein, and Dnmt3a may be the target gene of miR-344d-3p.