A total of 49 paired OS tissues and adjacent normal bone tissues were obtained from the First Affiliated Hospital of Zhengzhou University (Zhengzhou, China) between July, 2015 and August, 2017. All patients received surgical resection and had not been treated with chemotherapy, radiotherapy, or any other therapy prior to surgery. Freshly resected tissue samples were immediately frozen in liquid nitrogen and stored at −80°C until use. The Ethics Committee of the First Affiliated Hospital of Zhengzhou University approved this study, and all patients provided written informed consent.

We purchased a normal human osteoblast (hFOB1.19) and 4 human OS cell lines (MG-63, SAOS-2, HOS and U2OS) from the American Type Culture Collection (ATCC, Manassas, VA, USA). All the cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and a 1% mixture of penicillin/streptomycin (all from Gibco/Invitrogen/Thermo Fisher Scientific, Waltham, MA, USA). The cells were maintained at 37°C in a humidified atmosphere containing 5% CO₂.

We obtained miR-873 mimics and negative control miRNA mimics (miR-NC) from RiboBio (Guangzhou, China). Small interfering RNA for HOXA9 (HOXA9 siRNA) and negative control siRNA (NC siRNA) were chemically synthesized by Shanghai GenePharma Co., Ltd. (Shanghai, China). The HOXA9 overexpression plasmid, pcDNA3.1-HOXA9 (pc-HOXA9), and an empty pcDNA3.1 plasmid were generated by Integrated Biotech Solutions (Shanghai, China). The cells were plated in 6-well plates at a density of 6×10⁵ cells/well. When the cells reached 70-80% confluence, the above-mentioned mimics, siRNA, or plasmids were used for cell transfection with Lipofectamine 2000 (Invitrogen/Thermo Fisher Scientific), according to the manufacturer’s instructions. Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) and cell counting kit-8 (CCK-8) assay, colony formation assay and tumor xenograft assay were performed on transfected cells following 24 h of incubation at 37°C with 5% CO₂. At 48 h post-transfection, flow cytometric analysis and Transwell migration and invasion assays were carried out. Following 72 h of culture, western blot analysis was used for the determination of protein expression.

Total RNA was isolated from the tissue samples and cells using TRIzol reagent (Invitrogen/Thermo Fisher Scientific). Total RNA was used to determine miR-873 expression by reverse transcription using a TaqMan MicroRNA Reverse Transcription kit (Applied Biosystems, Foster City, CA, USA). The temperature protocol for reverse transcription was as follows: 16°C for 30 min, 42°C for 30 min and 85°C for 5 min. Thereafter, we carried out quantitative PCR (qPCR) using a TaqMan MicroRNA PCR kit (Applied Biosystems). The temperature protocol for qPCR was as follows: 50°C for 2 min, 95°C for 10 min; 40 cycles of denaturation at 95°C for 15 sec; and annealing/extension at 60°C for 60 sec. To determine HOXA9 mRNA expression, we prepared complementary DNA from total RNA using a PrimeScript RT reagent kit (Takara Bio, Dalian, China). The thermocycling conditions for reverse transcription were as follows: 37°C for 15 min and 85°C for 5 sec. Subsequently, qPCR was performed using a SYBR Premix Ex Taq™ kit (Takara Bio). The thermocycling conditions for qPCR were as follows: 5 min at 95°C, followed by 40 cycles of 95°C for 30 sec and 65°C for 45 sec. The relative expression levels of miR-873 and HOXA9 mRNA were normalized to the endogenous controls, U6 small nuclear RNA and GAPDH, respectively. The primers were designed as follows: miR-873, 5′-CTGCACTCCCCCACCTG-3′ (forward) and 5′-GTGCAGGGTCCGAGGT-3′ (reverse); U6, 5′-GCTTCGGCAGCACATATACTAAAAT-3′ (forward) and 5′-CGCTT CACGAATTTGCGTGTCAT-3′ (reverse); HOXA9, 5′-CAA CAAAGACCGAGCAAA-3′ (forward) and 5′-ATACCTCCT CCATCAAAGC-3′ (reverse); and GAPDH, 5′-TCCCATCAC CATCTTCCA-3′ (forward) and 5′-CATCACGCCACAGTT TCC-3′ (reverse). Gene expression was analyzed using the 2^−ΔΔCq method (29).

The transfected cells were incubated at 37°C in 5% CO₂ for 24 h, harvested, and seeded into 96-well plates at a density of 3×10³ cells/well. Cellular proliferation was determined at the following time points following inoculation: 0, 24, 48 and 72 h. Specifically, the cells were treated with 10 µl of CCK-8 solution (Dojindo Molecular Technologies, Inc., Kumamoto, Japan), then incubated them at 37°C in 5% CO₂ for a further 2 h. The absorbance was detected at 450 nm using an Enzyme Immunoassay Analyzer (Bio-Rad Laboratories, Hercules, CA, USA).

At 24 h following transfection, the cells were collected and digested using 0.25% trypsin (Gibco/Invitrogen/Thermo Fisher Scientific) to produce a suspension of single cells. A total of 1×10³ cells were seeded into 6-well plates, and incubated at 37°C in 5% CO₂ for 10 days. On day 11, the resulting colonies were washed with phosphate-buffered saline (PBS; Gibco/Invitrogen/Thermo Fisher Scientific), and fixed with 4% paraformaldehyde, and stained with methyl violet (Beyotime Institute of Biotechnology, Shanghai, China). Finally, the colonies were counted using an Olympus IX83 inverted microscope (Olympus Corp., Tokyo, Japan); a colony was defined as a group of >50 cells.

The proportion of apoptotic cells was evaluated using an Annexin V-fluorescein isothiocyanate (FITC) apoptosis detection kit (Biolegend, San Diego, CA, USA). Briefly, the transfected cells were cultured in 6-well plates for 48 h, collected, and washed 3 times with cold PBS. The cells were then resuspended in 100 µl of binding buffer, and double-labeled with 5 µl Annexin V-FITC and 5 µl propidium iodide. Following incubation at 37°C in the dark for 30 min, a flow cytometer (FACScan™; BD Biosciences, Franklin Lakes, NJ, USA) was used to determine the rate of cell apoptosis.

For these assays, 8-µm Transwell chambers (BD Biosciences) were used to determine the cell migratory ability. At 48 h following transfection, the cells were washed twice in PBS and suspended in FBS-free DMEM. A total of 5×104 cells were seeded into the upper compartments, and the lower compartments were covered with 500 µl of DMEM containing 10% FBS. Following cultivation for 48 h, the non-migrating or non-invading cells were wiped out using a cotton swab. The cells adhering to the lower surface of the Transwell insert were fixed with 4% paraformaldehyde, stained with 0.05% crystal violet (Beyotime Institute of Biotechnology), washed thrice with PBS, and air-dried. A Transwell invasion assay that was similar to the migration assay was carried out, except that the chambers were pre-coated with Matrigel (BD Biosciences) before the experiment began. Finally, the cells that had migrated or invaded were photographed, and we counted at least 5 representative fields/inserts under an inverted microscope (×200 magnification; IX83; Olympus Corp.).

We obtained 4-5-week-old female nude mice (weighing 18-19 g) from the Shanghai Laboratory Animal Center (Chinese Academy of Sciences, Shanghai, China). All nude mice were housed in sterile and pathogen-free conditions (25°C; 50% humidity; 10-h light/14-h dark cycle). The HOS cells were transfected with miR-873 mimics or the miR-NC. At 24 h following transfection, we collected the cells and subcutaneously injected them into the upper flanks of each nude mouse (n=4 for each group). We calculated the volume of each tumor xenograft using the following formula: 1/2 × tumor length x tumor width². We sacrificed all the nude mice 30 days after seeding the HOS cells (at the time of sacrifice, the mice weighed approximately 27 g), and excised and measured the tumors. All experimental procedures were approved by the Ethics Review Committee of the First Affiliated Hospital of Zhengzhou University.

We used the following three miRNA target prediction software packages to search for potential targets of miR-873: TargetScan 7.1 (http://www.targetscan.org/), miRanda (http://www.microrna.org) and miRDB (http://www.mirdb.org/).

Human wild-type (wt) HOXA9 3′-UTR containing the predicted miR-873 binding site and the mutant (mut) HOXA9 3′-UTR were amplified by Shanghai GenePharma Co., Ltd., and sub-cloned into the pMIR-REPORT vector (Promega, Madison, WI, USA). We named the generated luciferase reporter plasmids pMIR-wt-HOXA9-3′-UTR and pMIR-mut-HOXA9-3′-UTR. We seeded the cells into 24-well plates (1.5×10⁵ cells/well) one night prior to transfection. We introduced a mixture of pMIR-wt-HOXA9-3′-UTR, pMIR-mut-HOXA9-3′-UTR, and miR-873 mimics or miR-NC into the cells using Lipofectamine 2000 (Invitrogen/Thermo Fisher Scientific) in accordance with the manufacturer’s instructions. The transfected cells were lysed and assayed for the detection of luciferase activity at 48 h post-transfection using a dual-luciferase reporter assay system (Promega). The activity of Firefly luciferase was normalized to that of Renilla luciferase.

The homogenized tissues were solubilized and the cells were cultured with radioimmunoprecipitation assay lysis buffer (Beyotime Institute of Biotechnology) to isolate the total protein. The protein concentration was quantified using a BCA kit (Beyotime Institute of Biotechnology). Subsequently, equal amounts of protein were loaded onto 10% polyacrylamide gels for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), transferred them to PVDF membranes (EMD Millipore, Billerica, MA, USA), and blocked the proteins at room temperature for 2 h with 5% skimmed milk diluted in Tris-buffered saline (TBS) containing 0.1% Tween-20 (TBST). Following overnight incubation at 4°C with primary antibodies, the membranes were washed thrice with TBST and probed with horseradish peroxidase-conjugated immunoglobulin G secondary antibodies (ab205719 and ab6721; 1:5,000 dilution; Abcam, Cambridge, UK) for 2 h at room temperature. Western blot analysis was carried out using a Pierce ECL Western Blotting substrate (Invitrogen/Thermo Fisher Scientific), according to the manufacturer’s instructions. The following primary antibodies were used: Rabbit anti-human monoclonal HOXA9 antibody (ab140631; 1:1,000 dilution; Abcam), mouse anti-human monoclonal β-catenin antibody (sc-59737; 1:1,000 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA), mouse anti-human monoclonal p-β-catenin antibody (sc-57534; 1:1,000 dilution; Santa Cruz Biotechnology), rabbit anti-human monoclonal cyclin D1 antibody (ab134175; 1:1,000 dilution; Abcam) and rabbit anti-human GAPDH antibody (ab128915; 1:1,000 dilution; Abcam). The expression levels of the target proteins were normalized to those of GAPDH.

All data are expressed as the means ± standard error from at least 3 independent experiments. Differences between pairs of groups were analyzed using the Student’s t-test. One-way analysis of variance was used, followed by the Student-Newman-Keuls post hoc test to evaluate the differences between multiple groups. The correlation between the miR-873 and HOXA9 mRNA levels in the OS tissues was determined by Spearman’s correlation analysis. The Chi-square test was used to assess the association between miR-873 and the clinicopathological characteristics of the patients with OS. Statistical significance was set at P<0.05.

To investigate the clinical value of miR-873 in OS, we first detected miR-873 expression in 49 pairs of OS tissues and adjacent normal bone tissues using RT-qPCR analysis. The results revealed the decreased expression of miR-873 in the OS tissues compared to the adjacent normal bone tissues (P<0.05, Fig. 1A). Subsequently, we examined the association between miR-873 expression and the clinicopathological characteristics of the patients with OS. As shown in Table I, a low miR-873 expression was significantly associated with tumor size (P=0.015), clinical stage (P=0.032) and distant metastasis (P= 0.015). However, no obvious association between miR-873 expression and other clinicopathological characteristics of the OS patients was observed, i.e., age at diagnosis, sex and anatomical location (each P>0.05). RT-qPCR analysis was also carried out to measure the miR-873 expression levels in a panel of OS cell lines (MG-63, SAOS-2, HOS and U2OS). The results revealed that miR-873 expression was downregulated in all 4 OS cell lines compared to the normal human hFOB1.19 osteoblasts (P<0.05, Fig. 1B). These data suggest that miR-873 plays a crucial role in OS progression.

To clarify the role of miR-873 in the development of OS, we selected the HOS and U2OS cell lines, which expressed particularly low levels of miR-873 among the 4 OS cell lines examined (Fig. 1B), for functional experiments. We transfected the HOS and U2OS cells with miR-873 mimics or miR-NC. Following transfection, RT-qPCR analysis demonstrated that miR-873 was markedly overexpressed in the HOS and U2OS cells following transfection with the miR-873 mimics (P<0.05, Fig. 2A). We then carried out a CCK-8 assay to examine the proliferation of the HOS and U2OS cells transfected with miR-873 mimics or miR-NC. The results indicated that the proliferation of both cell lines was significantly suppressed when miR-873 was upregulated (P<0.05, Fig. 2B). Our subsequent evaluation of the colony formation ability revealed fewer colonies of the HOS and U2OS cells following transfection with miR-873 mimics (P<0.05, Fig. 2C). We also used flow cytometric analysis to verify whether the promotion of cell apoptosis suppressed cell proliferation. The results clearly revealed that the apoptotic rate increased in the miR-873 mimic-transfected HOS and U2OS cells compared to the cells transfected with miR-NC (P<0.05, Fig. 2D). These results indicate that miR-873 overexpression suppresses OS cell growth in vitro.

We carried out Transwell migration and invasion assays to determine whether miR-873 affects the metastatic ability of the OS cells. As shown in Fig. 3A, the ectopic expression of miR-873 suppressed the migration of the HOS and U2OS cells compared with the migration of the cells in the miR-NC groups (P<0.05). Furthermore, the invasive ability of the miR-873 mimic-transfected HOS and U2OS cells was markedly lower in comparison with that of the miR-NC-transfected cells (P<0.05, Fig. 3B). These results suggest that miR-873 suppresses OS cell metastasis in vitro.

To illustrate the molecular mechanisms responsible for the miR-873-induced regulation of cell growth and metastasis in OS, we used 3 miRNA target prediction software packages (TargetScan 7.1, miRanda, and miRDB) to predict the putative target(s) of miR-873. We identified an miR-873 binding site in the 3′-UTR of HOXA9 (Fig. 4A). Among hundreds of candidates, HOXA9 was selected for further identification owing to its importance in carcinogenesis and cancer progression (30-34). To confirm the prediction mentioned above, we carried out a luciferase reporter assay to verify whether miR-873 was able to recognize and target the 3′-UTR of HOXA9. The results revealed that the enhanced expression of miR-873 reduced the luciferase activity of the wild-type HOXA9 3′-UTR (P<0.05); however, this inhibition was not observed in the HOS and U2OS cells transfected with mutant HOXA9 3′-UTR (Fig. 4B).

We then measured HOXA9 expression in the OS tissues and assessed its association with the miR-873 levels. The results of RT-qPCR and western blot analyses revealed a substantially upregulated HOXA9 mRNA (P<0.05, Fig. 4C) and protein (P<0.05, Fig. 4D) expression in the OS tissues compared to the adjacent normal bone tissues. We also examined the correlation between the miR-873 and HOXA9 mRNA levels using Spearman’s correlation analysis, which established that HOXA9 mRNA expression inversely correlated with miR-873 expression in the OS tissues (r=-0.5412, P<0.0001; Fig. 4E). We found that the levels of HOXA9 mRNA (P<0.05, Fig. 4F) and HOXA9 protein (P<0.05, Fig. 4G) were significantly downregulated in the HOS and U2OS cells following the upregulation of miR-873 by transfection with miR-873 mimics. Collectively, these data identify HOXA9 as a direct target gene of miR-873 in OS cells.

To define the functional role of HOXA9 in OS cells, we introduced HOXA9 siRNA into the HOS and U2OS cells to knockdown HOXA9 expression. Again, NC siRNA was the negative control. The results of western blot analysis demonstrated that HOXA9 expression was markedly downregulated in the HOS and U2OS cells transfected with HOXA9 siRNA (P<0.05, Fig. 5A). The results of CCK-8 and colony formation assays indicated that HOXA9 downregulation significantly inhibited both the proliferation (P<0.05, Fig. 5B) and colony formation (P<0.05, Fig. 5C) of the HOS and U2OS cells. The apoptotic rate of the HOS and U2OS cells also increased following transfection with HOXA9 siRNA, as detected by flow cytometric analysis (P<0.05, Fig. 5D). Furthermore, the results of Transwell migration and invasion assays revealed that the silencing of HOXA9 suppressed the migratory (P<0.05, Fig. 5E) and invasive (P<0.05, Fig. 5F) capabilities of both cell lines. Therefore, HOXA9 knockdown affected the OS cells in a similar manner to miR-873 upregulation, further suggesting that HOXA9 is a direct downstream target of miR-873.

We carried out rescue experiments to further ascertain whether HOXA9 targeting by miR-873 is responsible for the tumor suppressive role of miR-873 in OS cells. We constructed a HOXA9 overexpression plasmid, pcDNA3.1-HOXA9 (pc-HOXA9). We then co-transfected the HOS and U2OS cells with pc-HOXA9 and the miR-873 mimics. pc-HOXA9 co-transfection reversed the suppressive effects on HOXA9 protein expression induced by transfection with miR-873 mimics (P<0.05, Fig. 6A). The results of functional assays indicated that miR-873 overexpression suppressed cell proliferation (P<0.05, Fig. 6B) and colony formation (P<0.05, Fig. 6C), while simultaneously inducing apoptosis (P<0.05, Fig. 6D) and suppressing cell migration (P<0.05, Fig. 7A) and invasion (P<0.05, Fig. 7B) in vitro. However, the restored expression of HOXA9 abolished all the effects described above. These results clearly demonstrate that miR-873 inhibits the aggressive phenotype of OS cells by directly targeting HOXA9, and that the downregulation of HOXA9 by miR-873 is essential for miR-873-induced tumor suppression in OS cells.

HOXA9 is involved in the modulation of the Wnt/β-catenin signaling pathway (34,35). Therefore, in this study, we attempted to determine whether the effects of miR-873 on HOXA9 have an impact on the activation of the Wnt/β-catenin signaling pathway in OS cells. We transfected the HOS and U2OS cells with miR-873 in combination with pc-HOXA9 or pcDNA3.1. Following transfection, we used western blot analysis to determine the expression levels of certain proteins that are involved in the Wnt/β-catenin pathway, such as p-β-catenin, β-catenin and cyclin D1. The results revealed that the expression levels of both p-β-catenin and cyclin D1 decreased in the HOS and U2OS cells when miR-873 was overexpressed. However, the ectopic overexpression of miR-873 had no effect on total β-catenin protein expression. Furthermore, HOXA9 overexpression was able to restore p-β-catenin and cyclin D1 expression in the miR-873-transfected HOS and U2OS cells (Fig. 8). These results suggest that miR-873 deactivates the Wnt/β-catenin pathway in OS cells by directly targeting HOXA9.

We used a tumor xenograft assay to determine the influence of miR-873 on OS cell tumor growth in vivo. We subcutaneously injected HOS cells transfected with miR-873 mimics or miR-NC into the upper flanks of nude mice. The volumes (P<0.05, Fig. 9A and B) and weights (P<0.05, Fig. 9C) of the tumor xenografts derived from the miR-873-overexpressing cells were significantly lower than those derived from the miR-NC-transfected cells. Moreover, there was a sharp increase in the selective expression of miR-873 in the tumor xenografts of the miR-873 mimics group, as determined by RT-qPCR analysis (P<0.05, Fig. 9D). Furthermore, western blot analysis of the tumor xenografts revealed the marked downregulation of HOXA9, p-β-catenin and cyclin D1 expression in the miR-873 mimics group (Fig. 9E). These data suggest that miR-873 suppresses the growth of OS cells in vivo, and identify the downregulation of HOXA9 and the subsequent inactivation of the Wnt/β-catenin pathway as possible underlying mechanisms.