Upregulation of microRNA-576-5p protects from steroid-induced avascular necrosis of the femoral head by suppressing ANXA2.

Steroid-induced avascular necrosis of the femoral head (SANFH) is a common orthopedic disease. Evidence has shown that microRNAs (miRNAs) played essential roles in the development of SANFH. Nevertheless, the role of miR-576-5p in SANFH remains unknown. The rabbit SANFH models were constructed by injection of horse serum and methylprednisolone. Bone mineral density (BMD) of the proximal femur (including the femoral head), pathological changes, bone cell apoptosis and expressions of OPG/RANK in femoral head bone tissue were assessed upon treatment of up-regulation of miR-576-5p or knockdown of ANXA2. Osteoblasts were extracted from SANFH rabbit femoral head and cultured. Proliferation, apoptosis and mineralization were tested upon treatment of up-regulation of miR-576-5p or knockdown of ANXA2. The targeting relationship between miR-576-5p and ANXA2 was verified. Up-regulated miR-576-5p or down-regulated ANXA2 inhibited the decrease of BMD, improved pathological changes, limited cell apoptosis and increased OPG/RANKL ratio in bone tissues of SANFH rabbits. Up-regulating miR-576-5p or down-regulating ANXA2 promoted proliferation and mineralization and inhibited apoptosis of osteoblasts from SANFH rabbits. In addition, ANXA2 was found to be a target gene of miR-576-5p. Furthermore, overexpression of ANXA2 abolished the protective role of elevated miR-576-5p against femoral head necrosis. Elevated miR-576-5p or reduced ANXA2 repressed the progression of SANFH. This study may provide novel biomarkers for SANFH diagnosis and treatment.

Introduction

Avascular necrosis of the femoral head (ANFH) is one of the most common orthopedic diseases in the world [1]. Steroid-induced ANFH (SANFH), a type of ANFH, is a prevalent complication resulted from long-term excessive use of steroids. If not received treatment in time, the patients would suffer from the collapsed structure of the femoral head (FH) [2]. SANFH is an event that is closely related with apoptosis of osteoblasts and osteocytes [3] and cell apoptosis and the apoptotic pathway are the primary focus when studying the pathogenesis of SANFH [4]. The occurrence of SANFH is related to the management method and dose of the steroids, and for most patients, the early diagnosis is hard to achieve [4]. With nonspecific symptoms in the early stage of this disease, patients usually lose the best chance of receiving nonsurgical therapy. Early diagnosis as well as treatment of SANFH remains vital [5]. Thus, it is important to find new therapies for SANFH.

In recent years, an amount of microRNAs (miRNAs) has been testified to human diseases. Many researches have demonstrated that miRNAs are associated with the procedures of SANFH, such as miR-206 [6] and miR-34a [7]. There are still a few researches that have reported that miR-576-5p is involved in glioblastoma [8] and pertussis [9], and its role in atrophic nonunion has also been testified [10]. Also, an altered expression level of hsa-miR-576-5p is observed in osteoarthritic human chondrocytes [11]. miR-576-3p is differentially expressed between osteoporotic patients with fractures and controls [12]. Meanwhile, miR-576-3p has also been demonstrated to play a role in coordinated guidance of osteoblastic differentiation [13]. Several miRNAs, such as miR‑576-5p, are found to be lowly expressed at the fracture sites in patients with atrophic nonunion [10]. Nevertheless, the functions of miR-576-5p in SANFH are still not completely explored. Annexin A2 (ANXA2), a calcium-dependent phospholipid binding protein, is prevalently expressed in human diseases, including esophageal squamous cell carcinoma [14], nasopharyngeal carcinoma [15] and breast cancer [16]. Meanwhile, Baldwin C et al. have also found that ANXA2 is implicated in the progression of sickle cell osteonecrosis [17] and altered ANXA2 is connected with the development of osteonecrosis [18]. Besides, ANXA2 is implicated to improve osteoblast growth and connect with bone mineral density and osteoporotic fracture [19]. Also, ANXA2 has promising potentials in treating osteoporosis, showing an association with bone mineral density [20] and peripheral blood monocyte-expressed ANXA2 involves in pathogenesis of osteoporosis [21]. Furthermore, osteoprotegerin (OPG) and receptor activator of nuclear factor-kappa B (RANK) have been noted in our study as well. It has been reported that the exploration of the mechanisms of these two bone metabolic biomarkers contributes to the recognition and treatment of SANFH [7]. The mechanism of miR-576-5p/ANXA2 axis in SANFH remains scarcely unexplored; therefore, this study was aimed to explore the impacts of miR-576-5p on the progression of SANFH via regulating ANXA2.

Materials and methods

Ethics statement

All patients had provided written informed consent before the study. The protocols of this study were approved by the Ethic Committee of the First Affiliated Hospital of Anhui Medical University and based on the ethical principles for medical research. Animal experiments were in line with the Guide to the Management and Use of Laboratory Animals issued by the National Institutes of Health. The protocol of animal experiments was approved by the Institutional Animal Care and Use Committee of the First Affiliated Hospital of Anhui Medical University.

Study subjects

Femoral head (FH) samples were obtained from patients with SANFH and femoral neck fracture (FNF) who received total hip replacement in the First Affiliated Hospital of Anhui Medical University. The patients were included if they were diagnosed as SANFH in the ACRO III/IV stage by three doctors and underwent total hip joint replacement surgery. The patients were excluded if the causes of femoral head necrosis were unknown or involved multiple factors; so did those with severe liver and kidney failure. There were 8 males and 18 females in 26 patients with SANFH (aged 56.6 ± 4.3 years old), and 14 males and 12 females in 26 patients with FNF (aged 58.1 ± 3.39 years old) as a control. The age and gender of patients between the two groups performed no marked difference (both P> 0.05). The FH samples were collected and then preserved at −80°C.

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

The total RNAs were extracted by Trizol (Invitrogen Inc., Carlsbad, CA, USA), and the RNA concentration and purity were evaluated. As to miR-576-5p and U6, the PCR primers and Stem-loop specific reversely transcribed primers were acquired from Guangzhou RiboBio Co., Ltd. (Guangdong, China). Primer sequences were not provided here due to business factors. Random primers with MMLV reverse transcriptase (Invitrogen) were selected for the reverse transcription of ANXA2, Bcl-2 and Bax by glyceraldehyde phosphate dehydrogenase (GAPDH). The PCR primers were synthesized by Invitrogen. PCR reaction was performed using a SYBR Green PCR Master Mix (BioTeke Co., Ltd., Beijing, China) with U6 and GAPDH as the internal reference. The data were analyzed using the 2−ΔΔCt method [22].

Western blot analysis

Concentration of total proteins in the tissues and cells was measured, and 10% sodium dodecyl sulfate separation gel and spacer gel were prepared. The samples were boiled with buffer solution at 100°C for 5 min, followed by ice bath and centrifugation. Proteins were then transferred onto the polyvinylidene fluoride membrane and fixed by 5% skim milk powder. Next, the membranes were appended with primary antibodies: ANXA2 (1:500), OPG (1:1000), and RANK (1:1000, both from Santa Cruz Biotechnology, CA, USA). The membranes were then appended with secondary antibody, immersed in the enhanced chemiluminescent reaction fluid (Pierce, Rockford, IL, USA) and developed. GAPDH was used as the loading control. Images were analyzed using ImageJ2x software.

Establishment of SANFH rabbit models

Healthy adult New Zealand rabbits (2.6–3.2 kg, male and female in half) were obtained from the Laboratory Animal Center of Anhui Medical University (Anhui, China). Feeding conditions: single cage, free access to food and water, temperature at 22 ± 2°C, relative humidity at 55 ± 5%, 12-h day/night cycles. After acclimated for 1 w, the rabbits were accurately weighed. Rabbits were randomly selected and injected with horse serum (Sigma, USA) via ear vein at 10 mL/kg. Three weeks later, the rabbits were injected with horse serum through the ear vein at 6 mL/kg again. Two weeks later, the rabbits were intraperitoneally injected with methylprednisolone (Sigma) at 45 mL/kg (once a day, for 3 d). During the injections, 100,000 U of penicillin was given into the abdominal cavity of rabbits to prevent infection [23].

Animal grouping

The successfully modeled 70 New Zealand rabbits were separated into 7 groups and 10 rabbits in each group. Rabbits in the SANFH group were treated with horse serum and methylprednisolone, while those in the other groups were injected with the corresponding vectors (5.5 × 1011 vp/mL) around the hip joint (25 μL per side) at 1 w post SANFH induction, including the agomir-NC group: injection of adeno-associated virus (AAV) vector expressing miR-576-5p agomir NC; the agomir-miR-576-5p group: injection of AAV vector expressing miR-576-5p agomir; the shRNA-NC group: injection of silenced ANXA2 NC AAV vector; the ANXA2-shRNA group: injection of silenced ANXA2 AAV vector; the agomir-miR-576-5p + overexpressed (Oe)-NC group: injection of the AAV vector expressing miR-576-5p agomir and the overexpressed ANXA2 NC AAV vector; the agomir-miR-576-5p + Oe-ANXA2 group: injection of AAV vector expressing miR-576-5p agomir and overexpressed ANXA2 AAV vector. The control group (10 normal rabbits without any injection) was set. AAV vector expressing miR-576-5p agomir NC (AAV-agomir-NC), AAV vector expressing miR-576-5p agomir (AAV-miR-576-5p-agomir), AAV vector expressing shRNA-NC (AAV-shRNA-NC), AAV vector expressing ANXA2-shRNA (AAV-ANXA2-shRNA), overexpressed ANXA2 NC AAV vector (AAV-Oe-NC) and overexpressed ANXA2 AAV vector (AAV-Oe-ANXA2) were purchased from RiboBio (Guangzhou, China). Four weeks after SANFH induction, bone mineral density (BMD) of rabbits in each group was measured. Rabbits were euthanized with their FH extracted for subsequent experiments.

Bone densitometry

Four weeks after treatment,(BMD was detected by bone density measuring instrument (LUNAR-Prodigy, LNR2898). Rabbits were placed in the supine position, the lower limbs were fixed with tape, and the BMD of the proximal femur (including the FH, femoral neck and metaphysis) was scanned (projection distance of 0.5 m, dose of 0.1 mREM). The experiments were run in triplicate to take the average value.

Hematoxylin-eosin (HE) staining

HE staining was carried out to observe the pathological changes of FH. FH tissues were fixed in 4% paraformaldehyde, rinsed with PBS (three times, each 20 min) and distilled water (three times, each 20 min), and then transferred to 30 times the volume of 10% ethylenediaminetetraacetic acid solution (the solution was changed once a week). After complete decalcification for 4 w, the specimens were paraffin-embedded, sectioned into 4 μm, baked and sequentially immersed into xylene I, xylene II, absolute ethanol I, absolute ethanol II, and gradient alcohol (95, 80, and 70%). Then, the specimens were stained by hematoxylin (Huaster, Wuhan, China), followed by differentiation in 1% hydrochloric acid, as well as immersion in gradient alcohol (50, 70, and 80%). Next, the specimens were soaked in eosin (Solarbio, Beijing, China) and treated with 95% alcohol, absolute ethanol I, absolute ethanol II, xylene I and xylene II. The sealed specimens in neutral gum were observed by a microscope [24].

Terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end-labeling (TUNEL) staining

TUNEL staining was performed to observe the apoptosis of osteocytes. Paraffin sections were dewaxed, dehydrated and incubated with Proteinase K (Boster Biological Technology, Ltd., Wuhan, China). The sections were incubated with TUNEL solution (Beyotime, Shanghai, China) and observed under a fluorescence microscope (Olympus, Tokyo, Japan). The negative control sections were incubated with the labeling solution without terminal transferase while the positive control sections were incubated with 0.01 mg/mL bovine pancreas DNase I (Boster). The number of apoptotic cells was calculated by Image-Pro Plus 6.0 software. The experiments were run in triplicate to take the average value [25].

Isolation and culture of osteoblasts

The extraneous soft connective tissues on the surface of femur and tibia were removed. Then, the femur and tibia were put into a Dulbecco modified Eagle medium (DMEM, sigma) with 100 IU/ml penicillin and 100 μg/mL streptomycin (Gibco), cut into small pieces (1–3 mm2) and rinsed by 10 mL DMEM. Next, the samples were cultured in 30 mL DMEM containing 0.25% collagenase (sigma), which was terminated by DMEM containing 10% fetal bovine serum (sigma). After that, the samples were combined with 30  mL DMEM, 30% FBS, 100 IU/mL penicillin, 100  μg/mL streptomycin, 50  μg/mL l-ascorbate-2-phosphate, and 100 μM sodium pyruvate (Gibco) for 7-d culture. Followed by that, 15% FBS-DMEM was added, with medium changed on the 3rd and 6th d before further experiment [26].

Alkaline phosphatase (ALP) staining and Alizarin red staining

ALP staining was conducted to assess ALP activity: The osteoblasts were lysed in 0.05% Triton in 10 mM Tris-HCl (pH = 7.4) buffer and stained with the ALP staining kit (Beyotime).

To observe the level of osteogenic induction of cells, the level of mineralized calcium nodules was observed by Alizarin red staining. Osteoblasts at passage 3 were cultured for 2 w to form opaque calcified nodules. Alizarin red staining was used to determine calcium deposits in the culture. The cells were fixed with 10% neutral formalin buffer for 15 min, incubated with 2% Alizarin Red S (Sigma) for 30 min and observed under an inverted optical microscope.

Cell grouping

The osteoblasts were classified into eight groups: the control group (osteoblasts from control rabbits); the SANFH group (osteoblasts from SANFH rabbits); the mimic NC group (osteoblasts from SANFH rabbits were transfected with miRNA mimic NC); the miR-576-5p mimic group (osteoblasts from SANFH rabbits were transfected with miR-576-5p mimic); the shRNA-NC group (osteoblasts from SANFH rabbits were transfected with shRNA-NC); the ANXA2-shRNA group (osteoblasts from SANFH rabbits were transfected with ANXA2-shRNA); the miR-576-5p mimic + Oe-NC group (osteoblasts from SANFH rabbits were transfected with miR-576-5p mimic and pcDNA-NC); the miR-576-5p mimic + Oe-ANXA2 group (osteoblasts from SANFH rabbits were transfected with miR-576-5p mimic and pcDNA-ANXA2). The mimic NC, miR-576-5p mimic, shRNA-NC, ANXA2-shRNA, pcDNA-NC and pcDNA-ANXA2 vector were acquired from RiboBio. Cells were seeded onto 6-well plates 24 h before transfection and then were transiently transfected by lipofectamine 2000 reagent (Invitrogen) when the cell confluence reached 50%. (Figure 1a)

Figure 1.

miR-576-5p expression is reduced in SANFH patients. A. RT-qPCR detected miR-576-5p expression in bone tissues of patients with SANFH and FNF; B. Western blot bands of ANXA2; C. Western blot analysis detected ANXA2 expression in bone tissues of patients with SANFH and FNF; D. Western blot analysis detected OPG/RANKL expression in bone tissues of patients with SANFH and FNF; the data between two groups were analyzed by t test, * P < 0.05.

3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2-H-tetrazolium bromide (MTT) assay

The MTT assay was carried out to measure cell proliferation. In short, cells were seeded onto 96-well plates at 5000 cells/well. After cells reached 80% confluence, each well was added with 20 μL MTT solution (5 μg/mL, Sigma) for 4-h incubation. Then, the supernatant was discarded, and cells were appended with dimethyl sulfoxide (150 μL/well, Sigma). The optical density value (OD490) was analyzed using a microplate reader.

Flow cytometry

AnnexinV-APC/PI double staining was applied to detect cell apoptosis. Cells were suspended in 250 μL binding buffer (diluted 1:4 with deionized water) to reach 1 × 106 cells/mL. Then, 100 μL cell suspension was added with 5 μL Annexin V-APC and 5 μL propidium iodide (PI, both from BD Biosciences, New Jersey, USA). Cells were mixed with 400 μL PBS before analysis by a flow cytometer (FACSCalibur, BD Biosciences). The data were evaluated by BD FACSCanto II (BD Biosciences) [27].

Dual luciferase reporter gene assay

The sequences of wild-type (WT) and mutant (MUT) ANXA2 3ʹUTR comprising miR-576-5p binding sites were synthesized and inserted into pMIR-REPORTTM Luciferase vector plasmid (Ambion, Austin, TX, USA) to establish ANXA2-WT and ANXA2-MUT. ANXA2-WT or ANXA2-MUT reporter plasmid, along with an miR-576-5p mimic or mimic NC was co-transfected into 293 T cells through Lipofectamine 2000 (Invitrogen).Cells were lysed after 48 h of transfection and centrifuged for 3–5 min to take the supernatant. Renilla luciferase buffer and firefly luciferase buffer were dissolved using the luciferase detection kit (Beyotime). The buffer (100 μL) was added with a substrate at 1:100 to prepare Renilla luciferase detection working solution. The sample (20–100 μL) from each group was added with 100 μL of the firefly luciferase detection reagent and detected by using Luminometer TD-20/20 Fluorometer (Promega, USA). The cell lysate was used a blank control. A Luciferase detection working solution (100 μL) was added, and a relative light unit (RLU) was measured. Renilla luciferase (Renilla) expression vector PRL-TK was considered as the internal reference. The RLU value obtained after the measurement was divided by the RLU value obtained after the measurement with Renilla luciferase to obtain the luciferase activity.

Statistical analysis

SPSS21.0 statistical software (IBM, Armonk, NY, USA) was applied to evaluate the data and calculate the average value and standard deviation. The t-test was performed for comparisons between two groups, and one-way analysis of variance (ANOVA) was used for comparisons among multiple groups. The P value less than 0.05 was indicative of a statistically significant difference.

Results

miR-576-5p level is reduced in SANFH patients

To explore the potential role of miR-576-5p in SANFH, we applied RT-qPCR and Western blot anaysis to detect miR-576-5p and ANXA2 levels in bone tissues of the SANFH group and control group. The results showed that the miR-576-5p level was reduced, while the ANXA2 level was increased in bone tissues in patients with SANFH. RANKL and its receptor OPG play a vital role in regulating bone formation [28]. Subsequently, RANKL and OPG levels in clinical samples were detected by Western blot analysis, and the findings suggested that the OPG/RANKL ratio was reduced in bone tissues of SANFH patients (Figure 1d).

miR-576-5p is down-regulated in SANFH rabbit

To further explore the potential mechanism of miR-576-5p affecting SANFH, a rabbit SANFH model was established by injecting horse serum plus methylprednisolone. BMD detection revealed that BMD was reduced in FH of SANFH rabbits (Figure 2a). Subsequently, the pathological condition of bone tissue was examined. HE staining revealed that there were no obvious abnormalities in the control group. The bone trabecula was neat and dense, arranged regularly, and the bone cells in the bone trabecula were clearly and evenly distributed. The bone trabecula of the SANFH model group was sparsely and disorderly arranged, thinned, and partially broken, and the number of normal bone cells in the trabecular bone decreased (Figure 2b). Next, we examined the effect of horse serum plus methylprednisolone on cell apoptosis. The findings manifested that the apoptotic rate and Bax expression of SANFH rabbits increased, while Bcl-2 expression decreased (Figure 2c, d). The findings proved the successful establishment of rabbit SANFH models. Finally, we detected miR-576-5p, ANXA2 and OPG/RANKL expression in bone tissues. It was observed that the miR-576-5p expression and OPG/RANKL ratio were reduced, whereas the ANXA2 expression was elevated in the FH of SANFH rabbits (figure 2e, f).

Figure 2.

miR-576-5p is down-regulated in SANFH rabbits. A. BMD of bone tissues in the control group and SANFH group; B. HE staining of bone tissues in the control group and SANFH group; C. TUNEL staining detected apoptosis rate in the control group and SANFH group; D. RT-qPCR detected Bax and Bcl-2 mRNA expression in bone tissues in the control group and SANFH group; E. RT-qPCR detected miR-576-5p expression in the control group and SANFH group; F. Western blot analysis detected ANXA2, OPG and RANKL protein expression in the control group and SANFH group;the data between two groups were analyzed by t test, * P < 0.05.

Elevated miR-576-5p attenuates SANFH in rabbits

Next, we overexpressed miR-576-5p in SANFH rabbits to find out the effect of miR-576-5p on SANFH. It was observed that agomir-miR-576-5p increased the miR-576-5p level and OPG/RANKL ratio and suppressed ANXA2 level in SANFH rabbits (Figure 3a, b). Afterward, it was noticed that the FH injury and necrosis were attenuated, BMD of FH was increased, Bcl-2 expression was heightened, and bone cell apoptosis rate and Bax expression were reduced in SANFH rabbits injected with AAV-miR-576-5p-agomir (Figure 3c). It was concluded that elevation of miR-576-5p could improve SANFH.

Figure 3.

Elevated miR-576-5p attenuates SANFH in rabbits. A. RT-qPCR detected miR-576-5p expression in the agomir-NC group and agomir-miR-576-5p group; B. Western blot analysis detected ANXA2, OPG and RANKL protein expression in the agomir-NC group and agomir-miR-576-5p group; C. BMD in the agomir-NC group and agomir-miR-576-5p group; D. HE staining of bone tissue in the agomir-NC group and agomir-miR-576-5p group; E. TUNEL staining detected apoptosis rate in the agomir-NC group and agomir-miR-576-5p group; F. RT-qPCR detected Bax and Bcl-2 mRNA expression in bone tissues in the agomir-NC group and agomir-miR-576-5p group; the data between two groups were analyzed by t test, * P < 0.05.

Silenced ANXA2 attenuates SANFH in rabbits

ANXA2 has been found to highly express in osteoporotic fractures and can inhibit the growth of osteoblasts [19]. Therefore, we speculated that ANXA2 may also play a role in SANFH. We found that ANXA2 was overexpressed in SANFH. We silenced ANXA2 in the SANFH rabbits (Figure 4a) and disclosed that after ANXA2 silencing, OPG/RANKL ratio in FH of SANFH rabbits increased. Silenced ANXA2 attenuated FH injury and necrosis, elevated BMD in FH, raised Bcl-2 expression, and suppressed bone cell apoptosis rate and Bax expression in SANFH rabbits (Figure 4c). Our findings indicated that silencing ANXA2 could relive SANFH.

Figure 4.

Silenced ANXA2 attenuates SANFH in rabbits. A. RT-qPCR detected miR-576-5p expression in the shRNA-NC group and ANXA2-shRNA group; B. Western blot analysis detected ANXA2, OPG and RANKL protein expression in the shRNA-NC group and ANXA2-shRNA group; C. BMD in the shRNA-NC group and ANXA2-shRNA group; D. HE staining of bone tissues in the shRNA-NC group and ANXA2-shRNA group; E. TUNEL staining detected apoptosis rate in the shRNA-NC group and ANXA2-shRNA group; F. RT-qPCR detected Bax and Bcl-2 mRNA expression in bone tissues in the shRNA-NC group and ANXA2-shRNA group; the data between two groups were analyzed by t test, * P < 0.05.

ANXA2 is the target gene of miR-576-5p

Next, we further explored whether ANXA2 was a potential target of miR-576-5p. The TargetScan website was employed to detect the binding sites of miR-576-5p on the ANXA2 3ʹUTR (Figure 5a), Dual luciferase reporter gene assay unraveled that relative to the NC, miR-576-5p mimic suppressed luciferase activity of 293 T cells transfected with WT ANXA2 3ʹUTR reporter plasmid but not 293 T cells transfected with MUT ANXA2 3ʹUTR reporter plasmid (Figure 5b).

Figure 5.

ANXA2 is the target gene of miR-576-5p. A. Bioinformatics website targetscan predicted the binding sites between miR-576-5p and ANXA2; B. Luciferase reporter gene experiment verified the targeting relation of miR-576-5p and ANXA2; C. Western blot analysis detected ANXA2, OPG and RANKL protein expression in the agomir-miR-576-5p + Oe-NC group and agomir-miR-576-5p + Oe-ANXA2 group; D. BMD of the agomir-miR-576-5p + Oe-NC group and agomir-miR-576-5p + Oe-ANXA2 group; E. HE staining of bone tissues in the agomir-miR-576-5p + Oe-NC group and agomir-miR-576-5p + Oe-ANXA2 group; F. TUNEL staining detected apoptosis rate in the agomir-miR-576-5p + Oe-NC group and agomir-miR-576-5p + Oe-ANXA2 group; G. RT-qPCR detected Bax and Bcl-2 mRNA expression in bone tissues in the agomir-miR-576-5p + Oe-NC group and agomir-miR-576-5p + Oe-ANXA2 group, The cell experiment was repeated three times; the data between two groups were analyzed by t test, * P < 0.05.

To further verify whether miR-576-5p binding to ANXA2 was involved in regulating SANFH, we injected AAV-miR-576-5p-agomir and AAV-Oe-ANXA2 into SANFH rabbits. Compared with the agomir-miR-576-5p + Oe-NC group, the agomir-miR-576-5p + Oe-ANXA2 group showed increased ANXA2 and reduced OPG/RANKL ratio, complained more severe FH injury and necrosis, decreased BMD in FH, reduced Bcl-2 expression, and elevated bone cell apoptosis rate and Bax expression (Figure 5c).

Elevated miR-576-5p promotes proliferation and mineralization, inhibits apoptosis and increases OPG/RANKL ratio of osteoblasts in SANFH

We further verified that miR-576-5p could competitively bind ANXA2 to regulate SANFH in vitro. Osteoblasts were isolated from rabbit bones and cultured. Displayed under the light microscope, the osteoblasts showed the shapes of triangle, fusiform and polygon (Figure 6a).

Figure 6.

Elevated miR-576-5p promotes proliferation and mineralization, inhibits apoptosis and increases OPG/RANKL ratio of osteoblasts in SANFH. A. Microscopic photograph of osteoblasts; B. ALP staining of osteoblasts; C. Alizarin red staining of osteoblasts; D. RT-qPCR detected miR-576-5p expression in the mimic NC group and miR-576-5p mimic group; E. Western blot analysis detected ANXA2, OPG and RANKL protein expression in osteoblasts in the control group, SANFH group, mimic NC group and miR-576-5p mimic group; F. MTT assay detected the proliferation ability of osteoblasts in the control group, SANFH group, mimic NC group and miR-576-5p mimic group; G. Flow cytometry detected the apoptosis of osteoblasts in the control group, SANFH group, mimic NC group and miR-576-5p mimic group; H. RT-qPCR detected Bax and Bcl-2 mRNA expression in osteoblasts in the control group, SANFH group, mimic NC group and miR-576-5p mimic group; I. Alizarin red staining detected bone cell mineralization ability in the control group, SANFH group, mimic NC group and miR-576-5p mimic group; the data between two groups were analyzed by One-way ANOVA, * P < 0.05.

ALP staining unveiled that there were gray-black granules or block deposit in the cytoplasm of positive cells (Figure 6b). The cells were in overlapping growth after cultured for 21 d, calcium nodules were formed in some parts, red was shown in the alizarin red staining (Figure 6c), indicating that osteoblasts were successfully isolated.

Osteoblasts from SANFH rabbits transfected with miR-576-5p-mimic showed increased miR-576-5p level and OPG/RANKL ratio and reduced ANXA2 expression (Figure 6d, e). Subsequently, we tested the effect of overexpression of miR-576-5p on the proliferation ability, apoptosis and mineralization capacity of osteoblasts by MTT assay, flow cytometry, RT-qPCR, and alizarin red staining. It was demonstrated that the proliferation ability was obviously reduced, apoptosis was enhanced, Bcl-2 mRNA expression was reduced, Bax mRNA expression was elevated, and mineralization capacity was impaired in osteoblasts from SANFH rabbits. However, transfection of miR-576-5p mimic improved the above situation of osteoblasts from SANFH rabbits (figure 6f). The above results indicated that overexpressing miR-576-5p promoted osteoblast proliferation and mineralization, inhibited apoptosis and increased OPG/RANKL ratio.

ANXA2 knockdown promotes proliferation and mineralization, suppresses apoptosis and increases OPG/RANKL ratio of osteoblasts in SANFH

Next, we explored the effect of silencing ANXA2 on SANFH in vitro. ANXA2 expression was reduced, and the OPG/RANKL ratio increased after SANFH osteoblasts were transfected with ANXA2 shRNA (Figure 7a). Subsequently, it was noticed that the proliferation ability was enhanced, apoptosis was weakened, Bcl-2 mRNA expression was elevated, Bax mRNA expression was reduced, and mineralization capacity was promoted after SANFH osteoblasts transfected with ANXA2 shRNA (Figure 7b). The above results confirmed that that silencing ANXA2 promoted the proliferation and mineralization of osteoblasts, depressed apoptosis and increased the ratio of OPG/RANKL.

Figure 7.

ANXA2 knockdown promotes proliferation and mineralization, suppresses apoptosis and increases OPG/RANKL ratio of osteoblasts in SANFH. A. Western blot analysis detected ANXA2, OPG and RANKL protein expression in osteoblasts in the shRNA-NC group and ANXA2-shRNA group; B. MTT assay detected the proliferation ability of osteoblasts in the shRNA-NC group and ANXA2-shRNA group; C&D. Flow cytometry detected the apoptosis of osteoblasts in the shRNA-NC group and ANXA2-shRNA group; E. RT-qPCR detected Bax and Bcl-2 mRNA expression in osteoblasts in the shRNA-NC group and ANXA2-shRNA group; F. Alizarin red staining detected bone cell mineralization ability in the shRNA-NC group and ANXA2-shRNA group; the data between two groups were analyzed by t test, * P < 0.05.

Overexpression of ANXA2 weakens the in vitro effect of overexpression of miR-576-5p on SANFH

Next, we explored whether the miR-576-5p/ANXA2 axis affects SANFH in vitro. ANXA2 expression was enhanced, and the OPG/RANKL ratio was reduced in the miR-576-5p-mimic + Oe-ANXA2 group versus the miR-576-5p-mimic + Oe-NC group (Figure 8a). Subsequently, it was recognized that in contrast to the miR-576-5p mimic + Oe-NC group, proliferation ability and mineralization capacity impairment, apoptosis enhancement, Bcl-2 mRNA expression reduction and Bax mRNA expression elevation were seen in the miR-576-5p mimic + Oe-ANXA2 group (Figure 8b), hinting that miR-576-5p/ANXA2 axis affected the development of SANFH in vitro.

Figure 8.

Overexpression of ANXA2 weakens the in vitro effect of overexpression of miR-576-5p on SANFH. A. Western blot analysis detected ANXA2, OPG and RANKL protein expression in osteoblasts in the miR-576-5p mimic + Oe-NC group and miR-576-5p mimic + Oe-ANXA2 group; B. MTT assay detected the proliferation ability of osteoblasts in the miR-576-5p mimic + Oe-NC group and miR-576-5p mimic + Oe-ANXA2 group; C. Flow cytometry detected the apoptosis of osteoblasts in the miR-576-5p mimic + Oe-NC group and miR-576-5p mimic + Oe-ANXA2 group; D. RT-qPCR detected Bax and Bcl-2 mRNA expression in osteoblasts in the miR-576-5p mimic + Oe-NC group and miR-576-5p mimic + Oe-ANXA2 group; E. Alizarin red staining detected bone cell mineralization ability in the miR-576-5p mimic + Oe-NC group and miR-576-5p mimic + Oe-ANXA2 group; the data between two groups were analyzed by t test, * P < 0.05.

Discussion

Led by the use of glucocorticoid, SANFH is a serious complication that may cause osteoporotic fractures and aseptic necrosis of FH [29]. It has been demonstrated that the miRNAs play an essential part in leading molecules in RNA silencing [30]. Additionally, some studies have elucidated that the function of miR-576-5p may be correlated with human diseases, such as neuroendocrine tumors of the lung [31] and non-small cell lung cancer [32]. Nevertheless, there is little known about the role of miR-576-5p in SANFH involving ANXA2. Consequently, this research was determined to investigate the effects of miR-576-5p and ANXA2 on SANFH, and we have discovered that the overexpression of miR-576-5p and the reduction of ANXA2 could ameliorate osteocyte injury in SANFH.

miRNAs are a group of critical mediators in the process of SANFH, and targeted mediation of certain miRNAs may attenuate SANFH development [33]. In our research, we focused on miR-576-5p and found its inhibitory effects on SANFH development. miR-576-5p has been suggested to down-regulate in atrophic non‑union [10]. Also, it has been suggested that has-miR-576-5p has a dysregulation in patients with systemic lupus erythematosus [34] and lowly expressed hsa-miR-576-5p has been measured in osteoarthritis chondrocytes [11]. Just like previous reports, we discovered a lower level of miR-576-5p in patients and rabbit models with SANFH. A recent document has found that miR-576-5p is involved in the growth and proliferation signaling pathways in catch-up growth [35]. Also, miR-576-5p mimic treatment could enhance trophoblast cell proliferation in vitro, providing a potential therapy for preeclampsia [36]. Our research also found that elevation of miR-576-5p could suppress osteocyte injury in SANFH, as reflected by alleviated FH injury and necrosis, decreased BMD, reduced OPG/RANKL ratio and suppressed apoptosis in animal models, and induced proliferation and decelerated apoptosis of osteoblasts in SANFH.

Next, we further explored whether ANXA2 was a potential target gene of miR-576-5p. ANXA2 was found to be a target gene of miR-576-5p based on online prediction and dual luciferase reporter gene assay. Except for that, we have also found that ANXA2 was highly expressed in SANFH and silencing ANXA2 protected against SANFH. As reported previously, ANXA2 influenced bone formation by modulating osteoprogenitor proliferation and differentiation [37]. In line with this result, a literature that studied on pathogenesis of osteoporosis in humans has discovered that the elevation of ANXA2 has the ability to reduce the level of BMD [21]. Supported by another work, the ANXA2 level was measured to elevate in osteoporotic fracture and knocking down ANXA2 was promoting the osteoblastic growth [19]. Besides, an extant research has verified that the functioned as a sensor, expression of ANXA2 was in response to the tubular injury [38], and the protective impacts of ANXA2 has also been clarified in biliary atresia‑associated hepatic injury by a recent literature [39]. Besides, He et al. have unearthed that ANXA2 plays a vital part in colorectal cancer and hepatocarcinoma by promoting the cell proliferation [40]. Additionally, the effects of ANXA2 knockdown on enhanced apoptosis of human umbilical vein endothelial cells has also been reported by Jiang et al. [41]. All in all, the repressive functions of knocked down ANXA2 in disease were in line with our findings. The next step was to verify whether miR-576-5p binding to ANXA2 was involved in regulating SANFH, and the results suggested that overexpression of ANXA2 abolished the protective role of elevated miR-576-5p against femoral head necrosis.

To sum up, our study reflects that the elevation of miR-576-5p or the downregulation of ANXA2 plays a protective role in osteocyte injury resulted by SANFH and also represses the development of SANFH by promoting the proliferation and decelerating apoptosis of osteoblasts. More efforts remain to probe the mechanisms of miR-576-5p as well as ANXA2 in the progression of SANFH.