The α-modified Eagle's medium (α-MEM), foetal bovine serum (FBS) and penicillin/streptomycin were obtained from Gibco (Thermo Fisher Scientific, Inc., Waltham, MA, USA). The Cell Counting kit-8 (CCK-8) was purchased from Dojindo Molecular Technology (Kumamoto, Japan). Mouse RANKL and recombinant soluble mouse M-CSF were obtained from R&D Systems (Minneapolis, MN, USA). The tartrate-resistant acid phosphatase (TRAP) staining kit, TRIzol, chloral hydrate and ketamine were purchased from Sigma-Aldrich (St Louis, MO, USA). All other reagents were from Sigma-Aldrich, unless stated otherwise. The present study was approved by the ethics committee of Sir Run Run Shaw Hospital (Hangzhou, China).

Bone marrow macrophages (BMMs) were separated from the femurs and tibiae of two six-week-old female C57BL/6 mice (Shanghai SLAC Laboratory Animal Co., Ltd., Shanghai, China) as described previously (15). The isolated BMMs were then cultured in α-MEM containing 10% FBS and 30 ng/ml M-CSF for proliferation. Cells were harvested when reaching 90% confluence and seeded into 96-well plates at a density of 8×10³ cells/well or into six-well plates at a density of 8×10⁵ cells/well in triplicate. After incubation for 24 h, cells were treated with RANKL (50 ng/ml) for various durations (0, 1, 3, 5 or 7 days). Cells in 96-well plates were then fixed and stained for TRAP activity following the kit manufacturer's instructions. Cells in the six-well plates were dissolved in TRIzol for extraction of total RNA (including miRNA).

An OVX-induced osteoporosis model was used to establish various degrees of osteoporosis in vivo. All experimental procedures were approved by the Animal Care and Experiment Committee of Zhejiang University School of Medicine (Huangzhou, China), and the study was performed according to the guidelines for Ethical Conduct from the Care and Use of Nonhuman Animals in Research by the American Psychological Association. Healthy eight-week-old C57BL/6J mice (n=24) were subjected to sham operation (n=8) or ovariectomy (OVX) (n=16) under anesthesia using 8% chloral hydrate. The mice were housed in three cages in an animal facility with conditions of 22–24°C, 12-h light/dark cycle, 50–60% humidity with access to food and water ad libitum. Eight sham-operated mice and eight OVX mice were sacrificed by euthanasia with 0.8 ml/100 g body weight ketamine delivered by intraperitoneal injection at 1 month following surgery (sham and OVX 1). The remaining eight OVX mice were sacrificed as above at 2 months following surgery (OVX 2). The left tibiae were obtained and fixed in 4% paraformaldehyde for micro-computer tomography (CT) analysis, while the right tibiae were immediately snap-frozen in liquid nitrogen for RNA (miRNA and mRNA) isolation.

After being fixed with 4% paraformaldehyde for 48 h, the left tibiae of each mouse were washed with phosphate-buffered saline (PBS) three times and stored in 70% ethanol prior to scanning. The fixed tibiae were then analyzed using a high-resolution micro-CT (Skyscan 1072; Skyscan, Aartselaar, Belgium). The scanning protocol was set at an isometric resolution of 9 µm, and X-ray energy settings of 70 kV and 80 µA. After reconstruction, the region of interest of the tibia was set at 0.5 mm from the proximal growth plate and selected for further qualitative and quantitative analysis. Several structural parameters, including trabecular bone volume per total volume (BV/TV), mean trabecular thickness (Tb.Th), mean trabecular number (Tb.N) and mean trabecular separation (Tb.Sp) were measured by micro-CT as reported previously (16).

Total RNA was isolated from BMMs that had been treated with RANKL for 0 or 5 days using the RNeasy^® Mini Kit (Invitrogen; Thermo Fisher Scientific, Inc.). The samples were then submitted to LC Sciences (Houston, TX, USA) for microarray analysis. The probe content was consistent with the miRBase database version 14.0 (http://www.mirbase.org/). The fluorescence value was determined by subtracting the background. The ratio of the two subgroups (log2 transformed) and P-values were calculated using Student's t-test. MiRNAs with P<0.01 and fluorescence values >500 were selected.

Homogenized tissue was obtained by grinding with liquid nitrogen. Total RNA (including miRNA) was isolated from cells and tissues by using the RNeasy^® Mini kit (Invitrogen; Thermo Fisher Scientific, Inc.) and reverse transcription of 1.0 µg total RNA was performed using the miRNA cDNA kit or the HiFiScript cDNA kit (CWBIO, Beijing, China), which were used to analyze the expression levels of miRNA and osteoclast-specific genes, respectively. Amplification reactions were set up in 20-µl reaction volumes containing amplification primers and UltraSYBR Mixture (with ROX) (CWBIO), and products were detected using an ABI 7500 sequencing detection system (Thermo Fisher Scientific, Inc.). A 1-µl volume of cDNA and 1-µl of primers (Sangon Biotech Co., Ltd., Shanghai, China) were used in each amplification reaction. The conditions of the reaction were as follows: Activation at 95°C for 10 min; 40 cycles of amplification, 95°C for 10 sec, 60°C for 24 sec and 72°C for 20 sec; and final extension at 72°C for 1 min. All reactions were run in triplicate and were normalized to the miRNA house-keeping gene U6, or mRNA house-keeping gene 18S and glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Expression was quantified using the ΔCq method, as described previously (17).

Primer sequences were as follows: 18S forward, 5′-CCTGCGGCTTAATTTGACTC-3′ and reverse, 5′-AACTAAGAACGGCCATGCAC-3′; GAPDH forward, 5′-ACCCAGAAGACTGTGGATGG-3′ and reverse, 5′-CACATTGGGGGTAGGAACAC-3′; TRAP forward, 5′-CCATTGTTAGCCACATACGG-3′ and reverse, 5′-CACTCAGCACATAGCCCACA-3′; Ctsk forward, 5′-TCCGCAATCCTTACCGAATA-3′ and reverse, 5′-AACTTGAACACCCACATCCTG-3′; NFATc1 forward, 5′-TCCACCCACTTCTGACTTCC-3′ and reverse, 5′-CTTCGCCCACTGATACGAG-3′; U6 forward, 5′-CTCGCTTCGGCAGCACA-3′ and reverse, 5′-AACGCTTCACGAATTTGCGT-3′; mmu-miR-27 forward, 5′-GTGGTTCACAGTGGCTAAG-3′; mmu-miR-21a-5p forward, 5′-GTTTGGTAGCTTATCAGACTGA-3′; mmu-let-7i-5p forward, 5′-GTTTGGTGAGGTAGTAGTTTGT-3′; mmu-miR-22-3p forward, 5′-GTGAAGCTGCCAGTTGAAG-3′; mmu-miR-378-3p forward, 5′-GTTGACTGGACTTGGAGTC-3′; mmu-miR-340-5p forward, 5′-GTGGGGTTATAAAGCAATGAGA-3′; mmu-miR-23a-5p forward, 5′-TTGGGGGTTCCTGGGGAT-3′; common reverse, GTGCAGGGTCCGAGGT.

GraphPad Prism^® software (GraphPad Inc., La Jolla, CA, USA) was used for all the statistical analyses. Values are expressed as the mean ± standard error of the mean. Analysis of variance followed by Dunnett's post hoc test was used for comparisons between groups. All the experiments were performed at least three times and P<0.05 was considered to indicate a statistically significant difference.

In order to obtain miRNAs during various stages of osteoclast differentiation, BMMs were treated with M-CSF and RANKL for 0, 1, 3, 5 or 7 days. As presented in Fig. 1A, BMMs were induced into mature osteoclasts following treatment with RANKL for 5 or 7 days. While no osteoclasts were present at days 0–3, the number of osteoclasts (TRAP-positive cells with >3 nuclei) was 103.93±3.55 per well and 150.67±6.13 per well following treatment of RANKL for 5 and 7 days, respectively (Fig. 1B). Furthermore, the area of osteoclasts was 56.30±2.43 µm² and 81.03±2.56 µm², respectively, following treatment with RANKL for 5 and 7 days (Fig. 1B). In addition, the expression levels of osteoclast-specific genes Ctsk, NFATc1 and TRAP were also assessed. As expected, the osteoclast-associated genes were upregulated in response to stimulation with RANKL. NFATc1 levels peaked at day three, while the expression of TRAP and Ctsk was low until day 3 and significantly increased on day 5 with a further increase on day 7 (Fig. 1C). These results suggested that osteoclast maturation follows stimulation with RANKL for 5–7 days.

Based on the successful induction of osteoclasts, the expression of miRNAs during osteoclastogenesis was investigated using total RNA extracted from BMMs (day 0) and mature osteoclasts (day 5) using microarray analysis. As listed in Table I, six miRNAs that were downregulated during osteoclastogenesis with a fold change of >1.5 were identified, including miR-340-5p, miR-23a-5p, miR-22-3p, miR-21-5p, miR-27a-3p and let-7i-5p. Largely consistent with the results of previous studies, the downregulation of further miRNAs, including miR-21, miR-155, miR-223 and miR-146a, was identified by microarray analysis (presented in Table II) (11,12,18–27).

Furthermore, changes in the expression of the six aberrantly expressed miRNAs during osteoclastogenesis was confirmed using RT-qPCR. As presented in Fig. 2, miR-21a-5p, miR-27a-3p, let-7i-5p, miR-22-3p, miR-340-5p and miR-23a-5p were observed to be downregulated during osteoclast differentiation. The expression of miR-27a-3p, miR-21a-5p and miR-23a-5p in matured osteoclasts was demonstrated to be ~70–80% lower than that in BMMs without RANKL stimulation, while the expression of let-7i-5p and miR-22-3p dropped to 50~70% in BMMs following treatment with RANKL for 5 or 7 days. Notably, the expression profiles of these miRNAs demonstrated a time-dependent decrease during osteoclast formation. These results confirmed the downregulation of a number of miRNAs during osteoclastogenesis.

An OVX-induced murine osteoporosis model was used to establish various degrees of osteoporosis in vivo. Microarchitectural changes of the tibiae were then detected by micro-CT (Fig. 3A). In accordance with other studies (21), OVX mice showed lower BV/TV and Tb.N, and higher Tb.Sp compared to sham-operated mice (Fig. 3B). In addition, the mice sacrificed two months after ovariectomy (OVX 2 group) displayed a reduced bone volume compared with that of the mice in the OVX 1 group, with BV/TVs of 7.89±0.27 and 4.60±0.44%, respectively (Fig. 3B). In addition, OVX 2 mice exhibited a lower Tb.N and a higher Tb.Sp than the OVX 1 mice (Fig. 3B). The Tb.Th value remained constant indicating that OVX-induced osteoporosis occurred predominantly by decreasing BV/TV and Tb.N. The changes in the factors of bone intactness indicated that the degree of osteoporosis in OVX 2 mice was higher than that in OVX 1 mice.

Based on the expression of miRNAs during osteoclast differentiation according to the microarray and RT-qPCR analyses of the present study, the expression of these miRNAs was assessed in the tibiae of a mouse model of osteoporosis. As shown in Fig. 4, miR-21a-5p, miR-27a-3p, let-7i-5p, miR-22-3p, miR-340-5p and miR-23a-5p were found to be downregulated in OVX 1 mice by at least 60% when compared to the sham-operated mice. Of note, the expression of all of these miRNAs was further declined in OVX 2 mice when compared to that in OVX 1 mice, indicating a time-dependent decrease in the expression of these miRNAs during the progression of osteoporosis.