Silencing miR-199a-3p, Which Targets Integrin Β8 Prevents Steroid-induced Avascular Necrosis of the Femoral Head by Regulating Osteogenesis and Adipogenesis

The main pathogenesis of steroid-induced avascular necrosis of the femoral head (SANFH) is closely connected with osteogenesis and adipogenesis. MicroRNAs have been proved to play prominent roles in the initiation and progression of SANFH. The present study investigated the effects of silencing miR-199a-3p on the prevention and early treatment of SANFH. RT-PCR was used to detect the expression of miR-199a-3p. Western blotting was used to detect protein expression. Alkaline phosphatase (ALP) activity, alizarin red S staining, and oil red O staining were used to study the effect of miR-199a-3p on the osteogenic and adipogenic differentiation of rat bone marrow-derived mesenchymal stem cells (rBMSCs) and MC3T3-E1 cells. A dual-luciferase reporter assay was used to conrm the target relation of miR-199a-3p and integrin β8 (ITGB8). Animal study was used to explore the effect of silencing miR-199a-3p in vivo. MC3T3-E1 to prevent the development of SANFH. conrmed the role in SANFH and identied new target molecules and the mechanisms by which miR-199a-3p suppressed osteogenesis and promoted adipogenesis.


Introduction
Steroid-induced avascular necrosis of the femoral head (SANFH), a prevalent, devasting, and refractory orthopedic disease induced by the systemic application of glucocorticoids1-3. The disease is characterized by femoral head collapse followed by substantial degeneration of the hip joint, severely restraining the life quality of patients4. Lack of e cient early diagnosis and treatment, the disease inescapably leads to a collapse of the femoral head followed by osteoarthritis of the hip joint5. Therefore, the patients are necessitated to undergo a hip arthroplasty6, 7. These conditions cause a heavy burden on society and families. However, to date, the detailed pathological mechanism of SANFH is still unclear.
So far, the pathogenesis of SANFH was closely connected with the imbalance between osteogenesis and adipogenesis of bone marrow-derived mesenchymal stem cells (BMSCs)8. BMSCs are multipotential cells and considered as the progenitor of osteoblasts in the femoral head, playing an important role in the growth, regeneration, and repairment of bone tissues9. It was reported that BMSCs pools in the femoral head are impaired, the osteogenic ability of BMSCs is reduced, and the osteoblasts are abnormal in SANFH patients10. Therefore, the enhancement of osteogenesis may contribute to the prevention and early therapy of SANFH.
MicroRNAs (miRNAs) are a type of small non-coding RNAs that could bind to the 3′ untranslated region (3′UTR) of target genes and ultimately cleaving the mRNAs or repressing translation of the target mRNAs11, 12. Recently, plenty of researches had revealed that several microRNAs were involved in the pathogenesis mechanisms of SANFH, most of which were associated with osteogenesis or adipogenesis13-16. Interestingly, recent studies have revealed that miR-199a-3p was closely associated with some orthopedic diseases including osteoporosis17, osteoarthritis18, and rheumatoid arthritis19. However, no researcher performed speci c experiments to explore if miR-199a-3p is involved in the pathophysiological process of SANFH and its underlying mechanism.
In the present study, the expression of miR-199a-3p was measured. Then, the effects of miR-199a-3p on osteogenesis of rBMSCs and MC3T3-E1 cells were investigated. Furthermore, the interactions between miR-199a-3p and integrin β8 (ITGB8) were veri ed. Finally, the preventative effects of miR-199a-3psilencing were examined in vitro and in vivo. These ndings provide clues for the prevention and treatment of SANFH.

Patients and bone tissues
All experiments were conducted according to the Declaration of Helsinki. Every experiment was approved by the Research Ethics Committee of the A liated Hospital of Chongqing Medical University. Each donor signed an informed consensus with the approval of the Research Ethics Committee of the A liated Hospital of Chongqing Medical University. Finally, 11 SANFH (stage III and IV) patients and 11 femoral neck fracture (FNF) patients who underwent a hip arthroplasty in the First A liated Hospital of Chongqing Medical University were recruited. The SANFH patients and FNF patients were diagnosed by Xray and computerized tomography (CT) and further con rmed by histological analysis. All of these femoral head samples were collected after resected from the femur and were immediately divided into two halves using a bone knife. A part of the femoral head was rapidly stored in the liquid nitrogen for the next experiments, while the other part of each sample was xed in 4% paraformaldehyde for the histological study.

RNA extraction
The bone tissues were crunched using a rongeur and ground into powders with liquid nitrogen. Following that, the miRNAs of bone tissues were extracted using a Biospin miRNA Extraction kit (BioFlux, China). Finally, Biodrop Ulite (UK) was used to detect the miRNA concentration.
Alkaline phosphatase (ALP) activity assay The rBMSCs and MC3T3-E1 cells were seeded into 6-well plates and transfected. Based on the previously reported methods, the ALP activity was determined using an ALP activity detective kit (Wanlei Bio, China) according to the protocol of the manufacturer using a microplate reader 20, 21.

Osteogenic differentiation and evaluation
The rBMSCs and MC3T3-E1 cells were seeded into 24-well plates and transfected. When the con uence point of cells reached 80%, the medium was changed as the osteogenic differentiation medium (ODM, Cyagen, USA). Alizarin red S (ARS) staining was used to evaluate the level of osteogenic differentiation.
After cultured in osteogenic differentiation for 3 weeks, the rBMSCs were xed, stained, and washed according to the instructions of the Alizarin Red S staining kit (Solarbio, China). Based on the previously reported methods, the stained mineralization nodules were dissolved with 10% cetylpyridinium chloride at 37°C for 30 min22. The solution was added to a 96-well plate, and the OD value was measured at 562 nm for quantitative analysis.

Adipocyte differentiation and Oil red O (ORO) staining
The rBMSCs and MC3T3-E1 cells were seeded into 12-well plates and transfected. When the con uence point of cells reached 100%, the growth medium was changed as the adipocyte differentiation medium (Cyagen, USA). After cultured in adipocyte differentiation medium for 4 weeks, the rBMSCs and MC3T3-E1 cells were stained with an Oil red O staining kit (Solarbio, China) and pictured using an inverted microscope. Based on the previously reported methods, the stained lipid deposits were extracted with isopropyl alcohol at room temperature for 30 min23. Finally, the solution was added to a 96-well plate, and the OD value was measured at 562 nm for quantitative analysis. were purchased from CST (U.S.A.), ITGB8 was purchased from Zenbio (China). After the application, the protein bands were washed with 1 TBST three times and then incubated with secondary antibodies (Goat anti-rabbit, 1:8,000, biosharp, China) for 1 hour. Then the bands were washed with 1 TBST three times. Finally, the protein was detected with an ECL reagent (Zen-Bio, China).

Animal study
All experimental and animal care procedures were approved by the Research Ethics Committee of the A liated Hospital of Chongqing Medical University and performed following the guidelines of the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals. A total of 16 female SD rats (8-week old, 180-200g) were enrolled in this study and randomly divided into 4 groups: the control rats (n=4), MPS rats (n=4), and MPS+ NC rats (n=4), MPS+antagomir-199a-3p (n=4) respectively injected with PBS, MPS, MPS+ NC and MPS+antagomir-199a-3p. The MPS was intramuscularly injected into rats and the NC and antagomir-199a-3p were injected into rats via periosteum. Based on the previous studies, intramuscular injections of 20 mg/kg tetracycline (Solarbio), 10 mg/kg calcein (Solarbio) and 30 mg/kg Alizarin red S (Solarbio) were performed at weeks 0, 2, and 4 during the experiment for uorescence staining24-25. After injected for 6 weeks, all rats were sacri ced to harvest the femoral heads. Then the femoral heads were xed in 4% paraformaldehyde and scanned by micro-Computed Tomography (microCT) and analyzed by immunohistochemistry and histology.
A micro-CT (Skyscan1174 X-Ray Microtomography, Bruker, Belgium) was used to scan the rat femoral heads. After scanning, software N-Recon was used for 3-dimensional reconstruction of the femoral heads, and software CT-AN was used to analyze the osteogenic parameters including BV/TV (bone volume per tissue volume), Tb.Sp (trabecular separation), Tb.Th (trabecular thickness) and Tb. N (trabecular number).

Histological analyses and immunohistochemistry (IHC)
The collected femoral heads were xed with 4% paraformaldehyde for a week and decalci ed with EDTA decalcifying solution. The H&E staining was nished and pictured using an optical microscope to detect the effect of miR-199a-3p on rat femoral heads. The expression of RUNX2 and CD31 in bone tissue was measured using IHC staining.
Immuno uorescence staining The femoral heads were xed in 4% formaldehyde for 72 h, dehydrated with a gradient of ethanol, and ether for 24 h at each step. The specimens were in ltrated with polymethylmethacrylate and then embedded. Then, 10 μm thick sections were made with a Leica hard tissue slicer. The images of the uorescent-labeled specimens were captured using a confocal laser scanning microscope (Leica, Heidelberg, Germany). The excitation/emission wavelengths were set as follows: 405/560-590 nm (tetracycline, blue), 543/580-670 nm (alizarin red, red), and 488/500-550 nm (calcium green, green).

Statistical analysis
Statistical analyses were performed using GraphPad PRISM8.0. All of the measurement data were expressed as means ± standard deviation (SD). For the comparison between two groups, the Student's Ttest method was used, while for the comparison among multiple groups, the one-way ANOVA and Tukey test methods were used. P-value ≤0.05 was considered statistically signi cant.

Results
MiR-199a-3p may be involved in the pathological process of SANFH.
The hip joint X-ray and CT of the SANFH group and FNF group showed the necrosis, collapse, and deformation of the femoral head ( Fig.1 A, B). Figure 1C showed sectional images of femoral head specimens from the femoral neck fracture (FNF) group and SANFH group. The H&E staining showed normal bone trabecula in the FNF group and collapse, disorder, and necrosis of trabecula in the SANFH group (Fig.1D). IHC staining demonstrated that RUNX2 expression was notably declined in necrotic bone tissues, while PPARγ was upregulated. (Fig. 1E, F). RT-PCR showed that miR-199a-3p expression was upregulated in necrotic bone tissues and rBMSCs and MC3T3-E1 cells treated with 20UM dexamethasone (Fig. 1G, H, I). All of these data showed that miR-199a-3p may be involved in the pathological process of SANFH.
We found that the transfection of agomiR-199a-3p and antagomiR-199a-3p could separately increase and decrease the expression of miR-199a-3p in rBMSCs ( Fig. 2A) and MC3T3-E1 cells (Fig. 2C). Then, we examined the osteogenic and adipogenic markers and the results of western blotting showed that expression of OCN, ALP, and RUNX2 were declined in the the agomiR-199a-3p group, while the expression of PPARγ was increased (Fig. 2B, D). In addition, ALP activity demonstrated the suppressing effect of miR-199a-3p on the ALP activity of rBMSCs and MC3T3-E1 cells (Fig. 2E, F). Alizarin red S staining also exhibited a signi cant decrease of mineralization on the surface of rBMSCs and MC3T3-E1 cells in the agomiR-199a-3p group (Fig. 2 G, H, K, L). The Oil red O staining showed that the transfection of agomiR-199a-3p enhanced the formation of lipid droplets in rBMSCs, compared with those in the NC group and antagomir-199a-3p group (Fig. 2H, I, M, N). These data indicate that miR-199a-3p could inhibit the osteogenic differentiation of rBMSCs and MC3T3-E1 cells, enhancing adipogenic differentiation.
Integrin β8 (ITGB8) was the direct target of miR-199a-3p We found that there were two binding sites between miR-199a-3p and ITGB8 (Fig. 3A). Next, we constructed WT-ITGB8 and MT-ITGB8 plasmids (Fig. 3A) and conducted a dual-luciferase reporter assay. The results of the dual-luciferase reporter assay showed that the overexpression of miR-199a-3p signi cantly suppressed the luciferase activity of 3'-UTR in the WT-ITGB8 group, whereas no differences in luciferase activity of 3'-UTR were observed in the MT-ITGB8 group (Fig. 3B). These results indicated that miR-199a-3p could directly target and bind to ITGB8. Furthermore, the results of western blotting proved that ITGB8 expression was signi cantly reduced by force expression of miR-199a-3p in rBMSCs and MC3T3-E1 cells (Fig. 3C, D). The IHC staining also showed that ITGB8 was reduced in SANFH patients (Fig.3E). The above results con rmed that ITGB8 was the target gene of miR-199a-3p.M iR-199a-3p inhibited osteogenesis of rBMSCs and MC3T3-E1 cells by down-regulating ITGB8 and inactivating the ITGB8-FAK-ERK1/2-RUNX2 axis To further investigate the function of ITGB8, we silenced the expression of ITGB8 to study changes in osteogenic differentiation of rBMSCs and MC3T3-E1 cells. The silence of ITGB8 signi cantly inhibited osteogenic differentiation of rBMSCs and MC3T3-E1 cells, evidenced through downregulated osteogenic markers (Fig. 4A, B), reduced ALP activity (Fig. 4C, D), and decreased calcium deposits (Fig. 4E-H). We also found that the expression of PPARγ and the lipid droplets formation were negatively related to the expression of ITGB8 (Fig. 4A, B, I-L). In addition, the inhibition of miR-199a-3p (antagomiR-199a-3p) partly rescued the imbalance between osteogenic differentiation and adipogenic differentiation of rBMSCs and MC3T3-E1 cells induced by the interference of ITGB8.
From the KEGG database, we found that we found that integrin β8 was associated with the focal adhesion pathway. In this pathway, integrins phosphorylate the focal adhesion kinase (FAK) and activate it, p-FAK activates the Erk1/2, and then p-Erk1/2 interacts with RUNX2 and promote osteogenesis. According to the previous studies, the Erk1/2-RUNX2 pathway was closely associated with osteogenesis26-28. Therefore, we assumed that miR-199a-3p suppressed the osteogenic differentiation of rBMSCs and MC3T3-E1 cells by down-regulating the integrin β8-FAK-Erk1/2-Runx2 pathway. We next detected ITGB8, FAK, p-FAK, Erk1/2, p-Erk1/2, and RUNX2 (Fig. 4M, N). It was obvious that siITGB8 in uenced the expression of these proteins in rBMSCs and MC3T3-E1 cells compared with the NC group. However, The suppression of miR-199a-3p partly resolved the impairment of the ITGB8-FAK-Erk1/2-RUNX2 pathway caused by siITGB8. Taken together, these data suggest that the ITGB8-FAK-Erk1/2-RUNX2 pathway is involved in the osteogenesis of rBMSCs and MC3T3-E1 cells through interaction with miR-199a-3p.

Silencing of miR-199-3p rescued the suppression of osteogenesis of rBMSCs and MC3T3-E1 cells induced by dexamethasone
The experiments in gure 4 illustrated that antagomiR-199a-3p remedied the decreased osteogenesis of rBMSCs and MC3T3-E1 cells caused by inhibition of ITGB8. Then we investigated the effect of antagomiR-199a-3p (silence of miR-199a-3p) on glucocorticoid-induced impairment in rBMSCs and MC3T3-E1 cells. RT-PCR revealed that dexamethasone (DEX) could lead to the upregulation of miR-199a-3p in rBMSCs and MC3T3-E1 cells and the upregulation was rescued by the transfection of antagomiR-199a-3p (Fig. 5 A, C). After treated with dexamethasone for 3 days, the expression of OCN, ALP, ITGB8, p-FAK, p-Erk1/2, and RUNX2 were dramatically decreased with the augment of PPARγ (Fig. 5B, D). However, silencing antagomiR-199a-3p antagonized the above-mentioned trend. In addition, the ALP activity, Alizarin red S staining showed that the silence of antagomiR-199a-3p could restore the reduced ALP activity and mineralization of rBMSCs and MC3T3-E1 cells caused by dexamethasone antagomiR-199a-3p (Fig. 5 E-H, K-L). Oil red O staining demonstrated a similar protective effect of antagomiR-199a-3p ( Fig.   5I-J, M-N). These data suggested that antagomiR-199a-3p could remedy the damage of glucocorticoids on the differentiation of rBMSCs and MC3T3-E1 cells.
Silencing miR-199a-3p protect GC-induced SANFH damage in vivo Finally, antagomiR-199a-3p was injected into rats to observe its protective role against glucocorticoids in vivo (Fig. 6 A). Micro-CT scanning was conducted to evaluate the bone tissues of the rat femoral head. The results showed that the SANFH models were successfully conducted by the injection of methylprednisolone (MPS) and the collapse of the femoral was observed. By contrast, the treatment of antagomiR-199a-3p remarkably attenuated the pathological changes of SANFH (Fig. 6 B). In addition, MPS injection led to serious deterioration of trabecular parameters, such as BV/TV, Tb. N, Tb.Th and Tb.Sp. However, the silence of miR-199a-3p increased these parameters (Fig. 6C). The results of the H&E staining showed that there was more trabecular bone structure and less empty lacunae in the femoral head of the MPS+antagomiR-199a-3p group compared to the MPS+NC group (Fig. 6D, E). The IHC staining showed that ITGB8, RUNX2 were downregulated in the MPS group, while PPARγ was upregulated (Fig. 6F, G, H). The expression of ITGB8, RUNX2, and PPARγ was signi cantly restored by the silence of miR-199a-3p. Overall, these data suggested that antagomiR-199a-3p were able to treat GC-induced SANFH damage in vivo.
The dynamic bone formation and mineralization in the femoral head were displayed using uorochrome including tetracycline, Alizarin red S, and Calcein. The femoral head in the control group had a stronger uorochrome labeling with tetracycline (blue), alizarin red S (red), and calcein (green) in the femoral head.
However, the femoral head in the MPS group and MPS+NC group had signi cantly decreased new bone formation evidenced by much weaker uorochrome labeling in the femoral head (Figure 7). On the contrary, a much broader area of subchondral trabeculae was stained by uorochrome labeling in the MPS + antagomiR-199a-3p group, indicating the bene cial effect of antagomir-199a-3p on steroidinduced avascular necrosis of the femoral head (Figure 7).

Discussion
Resulted from the wide usage of glucocorticoids (GCs), SANFH has brought tremendous pressure to society1. The detailed mechanisms involved in the initiation and development of SANFH are complicated and remain unclear. Recently, considerable studies have revealed that a high concentration of GCs could inhibit osteogenesis and promote adipogenesis29-31. In the present study, we observed that the miR-199a-3p was dysregulated in vitro and in vivo. Moreover, we revealed the key role of the miR-199a-3p in mediating osteogenesis and adipogenesis of rBMSCs and MC3T3-E1 cells and provided fruitful and useful underlying targets to prevent the development of SANFH. In our study, we con rmed the harmful role in SANFH and identi ed new target molecules and the mechanisms by which miR-199a-3p suppressed osteogenesis and promoted adipogenesis.
Many differentially expressed miRNAs were identi ed in SANFH and the broad biological signi cance importance of miRNAs on osteogenesis and adipogenesis of SANFH has been investigated, including miR-155-5p16, miR-70832, miR-206-3p33, miR-26a34, and miR-144-3p35. The downstream target genes of these miRNAs include GSK3B, SAMD3, connexin43, EZH2, and FZD4. However, the role of miR-199a-3p in SANFH has never been studied before. In the present study, we rst found that in the necrotic bone tissues of patients with SANFH, the osteogenic marker was reduced while miR-199a-3p and PPARγ were notably up-regulated. To understand the role of miR-199a-3p in the pathogenesis of SANFH, NC, agomiR-199a-3p, and antagomir-199a-3p were transfected into rBMSCs and MC3T3-E1 cells to detect their effects on osteogenesis and adipogenesis. The data indicated that miR-199a-3p was negatively connected with osteogenesis and positively connected with adipogenesis. Moreover, it was observed that the silencing of miR-199a-3p could signi cantly reverse the impairment caused by glucocorticoids in vitro and in vivo. Thus, our ndings provide new insight into the regulation of osteogenesis, adipogenesis, and progression of SANFH.
To study the molecular mechanism by which miR-199a-3p regulates the differentiation of rBMSCs and MC3T3-E1 cells, we performed a search with TargetScan, which revealed that ITGB8 might be a possible target with 2 sites complementary to miR-199a-3p in its 3''UTR. To con rm this prediction, we conducted a dual-luciferase reporter assay and identi ed ITGB8 as a direct target of miR-199a-3p. Moreover, WB showed that upregulation of miR-199a-3p expression led to downregulation of ITGB8 at the protein level, whereas functional inhibition of miR-199a-3p led to derepression of ITGB8, strongly suggesting that ITGB8 is regulated by miR-199a-3p during osteogenic differentiation and adipogenic differentiation of rBMSCs and MC3T3-E1 cells. In addition, the silence of silencing of ITGB8 expression inhibited osteogenesis and angiogenesis, while these effects were rescued by antagomiR-199a-3p. ITGB8, also named integrin β8, is an important member of the integrin family. Previously, there was little research focused on the role of ITGB8 in osteogenesis and adipogenesis. It has been proved that the expression of ITGB8 was upregulated during the osteogenic differentiation of MSCs into osteoblasts and declined during adipogenesis36,37. However, no researcher conduct speci c to explore if ITGB8 could regulate osteoblastic differentiation and adipogenic differentiation as well as the role of ITGB8 in SANFH. In our study, our study rst proved that ITGB8 declined in necrotic bone tissues of patients with SANFH. In addition, we rst found that ITGB8 could regulate osteogenic differentiation of rBMSCs and MC3T3-E1 cells by interacting with the FAK-Erk1/2-RUNX2 pathway and proved the central role of the miR-199a-3p-ITGB8-FAK-Erk1/2-RUNX2 axis in SANFH. Moreover, it is believed that GCs could suppress the osteogenesis and promote adipogenesis by miR-199a-3p targeting ITGB8 and inactivating the ITGB8-FAK-Erk1/2-RUNX2 pathway.
Several studies have reported the direct use of miRNAs in the treatment of SANFH. Zuo et al. reported that miR-26a overexpressed in CD34+ stem cell-derived exosomes could protect the femoral head from damage caused by GCs by strengthening angiogenesis and osteogenesis38. Cao et al. reported that miR-224-5p silencing protected against GC-induced reductions of SMAD4 and promoted osteogenesis to alleviate SANFH in vitro and in vivo39. In this study, we directly used antagomir-199a-3p in rats and the utility of antagomiR-199a-3p could improve new bone formation in the rat femoral head, which could reduce the degree and incidence of SANFH. Moreover, Calcein and Alizarin Red S which could be absorbed and deposited in bone tissues were used to show the dynamic bone formation and the results showed that antagomir-199a-3p helped bone formation and mineralization and alleviate the harmful effect of glucocorticoids in vivo.

Conclusion
Taken together, these data highlighted that the upregulation of miR-199a-3p induced by GCs could reciprocally regulate the adipo-osteogenic differentiation of rBMSCs and MC3T3-E1 cells by targeting ITGB8. The miR-199a-3p-ITGB8-FAK-Erk1/2-RUNX2 axis is a potential signaling mechanism, and inhibition of miR-199a-3p expression could promote osteogenesis and inhibit adipogenesis to alleviate the occurrence of SANFH, which is a novel therapeutic target for the treatment of SAFH.   and MC3T3-E1 cells (N). *P <0.05, versus NC group; #P < 0.05, versus siITGB8 group. All data were expressed as mean ± SD.

Figure 5
Silencing of miR-199-3p rescued the suppression of osteogenesis of rBMSCs and MC3T3-E1 cells induced by dexamethasone. A, C RT-PCR was used to measure the expression of miR-199a-3p in rBMSCs   The protective function of antagomir-199a-3p against SANFH in the rat model. There was signi cantly decreased new bone formation in the MPS-treated group revealed by the uorochrome labeling, while the pharmacotherapy with antagomir-199a-3p critically restored bone formation capacity.