Osteosarcoma tissues and matched non-cancerous adjacent tissues (NATs) were obtained from 18 patients (9 males and 9 females; mean age, 22.7 years; age range, 15–35 years) histopathologically diagnosed with osteosarcoma at the First Affiliated Hospital of China Medical University (Shenyang, China). No patients in the study received any preoperative treatment, and all patients underwent resection of osteosarcoma at the time of diagnosis between June 2013 and June 2016. Patients with histological grade IIB/III osteosarcoma were included, and patients with any other primary disease were excluded. The paired and osteosarcoma tissues were immediately preserved in liquid nitrogen and maintained at −80°C. All associated clinical data, including age, sex and tumor node metastasis stage were obtained from the medical records of patients. The present study received approval by the Ethics Committee of the First Affiliated Hospital of China Medical University and written informed consent was obtained from all patients.

Human osteosarcoma U2OS (cat. no. TCHu88), Saos-2 (cat. no. TCHu114), MG63 (cat. no. TCHu124), 293T (cat. no. GNHu17) and human osteoblast hFOB1.19 (cat. no. GNHu14) cells were purchased from the Institute of Biochemistry and Cell Research Center at the Chinese Academy of Sciences (Shanghai, China). The U2OS, Saos-2 and MG63 cells were cultured in high-glucose Dulbecco's modified Eagle medium (DMEM) containing 10% fetal bovine serum (both from HyClone; GE Healthcare Life Sciences, Logan City, UT, USA) and hFOB1.19 cells were cultured in DMEM/F12 medium (HyClone; GE Healthcare Life Sciences) containing 10% fetal bovine serum. All cells were cultured at 37°C in a humidified atmosphere with 5% CO₂.

miR-106a gene-expression plasmid (miR-106a), miR-106a gene-inhibition plasmid packaged into letivirus core vector (hU6-MCS-Ubiquitin-EGFP-IRES-puromycin) (miR-106a-inhibitor), mock negative control (miR-NC), VNN2 wild type gene over-expression plasmid (GV272/VNN2), VNN2 mutant type gene over-expression plasmid (GV272/VNN2-mut), VNN2 negative control plasmid (GV272/VNN2-NC) were all chemically synthesized by Shanghai GeneChem Co., Ltd. (Shanghai, China). miR-106a-inhibitor was used as the experimental group and miR-NC as the control group. The experimental group and control group were treated with lentivirus, which was inactivated and served as a carrier. Therefore, the blank control group was removed (20). The sequences were as follows: miR-106a-inhibitor, 5′-GTAAGAAGTGCTTACATTGCAG-3′; miR-NC, 5′-TTCTCCGAACGTGTCACGT-3′; and miR-106a, 5′-ACGGGCCCTCTAGACTCGAGTGTTTTAACCAGGTGAGTC-3′. The sequence of miR-106a binding site in GV272/VNN2 3′UTR was 5′-GCCATTGCAAA-3′, and GV272/VNN2-mut 3′UTR was 5′-GCACGGTACAA-3′. All plasmid DNA was extracted using an EndoFree Mini Plasmid kit (Tiangen Biotech Co., Ltd., Beijing, China). U2OS cells were transfected with the plasmid using X-tremeGENE HP (Roche Diagnostics, Basel, Switzerland) and infected at a multiplicity of infection of 10 for 12 h in the incubator, then incubated for 72 h to perform the subsequent trials.

Total RNA from the paired specimens, and U2OS, Saos-2, MG63, human hFOB1.19 cells and infected U2OS cells were isolated using the QIAzol reagent (Qiagen, Inc., Valencia, CA, USA) according to the manufacturer's protocol. The stem-loop reverse-transcription primers were designed as follows: miR-106a, 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACCTACCT-3′ and U6 small nuclear RNA (U6), 5′-CTCAACTGGTGTCGTGGAGTCGGCAATTCAGTTGAGGGGACAAA-3′. The oligo dT primers were used for VNN2 and GAPDH genes. Reverse transcription was performed using a PrimeScript RT reagent kit (Takara Bio, Inc., Otsu, Japan) according to the manufacturer's protocol. The sequences of the PCR primers were designed as follows: miR-106a forward, 5′-CGCAAAAGTGCTTACAGTGCA-3′ and reverse, 5′-GTGCAGGGTCCGAGGT-3′; U6 forward, 5′-CTCGCTTCGGCAGCACA-3′ and reverse, 5′-AACGCTTCACGAATTTGCGT-3′; VNN2 forward, 5′-CCATAAGGTGGGCAAGAGTCA-3′ and reverse, 5′-CTCCGGCTTTTCAGGGACAT-3′; GAPDH forward, 5′-GCACCGTCAAGGCTGAGAAC-3′ and reverse, 5′-TGGTGAAGACGCCAGTGGA-3′. All PCR reactions were performed using an Applied Biosystems 7900HT Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc., Waltham, MA, USA) using the SYBR PrimeScript RT-PCR kit (Takara Bio, Inc.) according to the manufacturer's protocol. The thermocycling conditions maintained were as follows: 95°C for 30 sec; 95°C for 5 sec; and 60°C for 32 sec, for 40 cycles. To ensure the fidelity of PCR reactions results, the expression levels of miR-106a were normalized to the expression of U6, while the expression levels of VNN2 were normalized to the expression of GAPDH. The relative expression levels were calculated and normalized using the 2^−ΔΔCq method (21).

The proliferation ability of infected miR-106a-inhibitor and miR-106a-NC U2OS cells was measured using an MTS assay kit (Promega Corporation, Madison, WI, USA) according to the manufacturer's protocol. Cells were plated (2×10³ cells/well) in 96-well plates, and cultured for 24, 48, 72 and 96 h. In a 96-well plate, 20 µl MTS reagent was added for every 100 µl medium/well and then incubated at 37°C in 5% CO₂ for 4 h. The optical density value of each sample was recorded at the wavelength of 490 nm via a microplate reader.

The migration and invasion ability of infected U2OS cells were detected using Transwell assays. The upper and lower chambers of a 24-well plate (Corning Incorporated, Corning, NY, USA) were washed with serum-free DMEM, and then 40 µl Matrigel (BD Biosciences, San Jose, CA, USA) diluted 1:8 with serum-free medium was added to the upper chamber to evenly cover the surface of the polycarbonate membrane for the invasion assay, whereas no Matrigel was added for migration assay. The 24-well plate was put into an incubator for 4 h at 37°C to allow the Matrigel to solidify. A total of 200 µl DMEM containing 1×10⁵ cells was seeded on the top of the upper chamber in the Transwell invasion chamber while 600 µl DMEM with 10% FBS was added into the lower chamber. The Transwell invasion assay was put in a cell incubator at 37°C with 5% CO₂. After 48 h, the cells on the upper surface were wiped away using a wet cotton swab. For the adherent cells, the migrated cells attach to the other side of the membrane in a Transwell assay (22). Cells that passed through the upper chamber and attach to the lower surface were fixed using absolute ethanol for 15 min at room temperature and stained with 0.1% crystal violet for 15 min at room temperature. Visible cells were counted using an inverted phase contrast microscope at a magnification of ×400.

After infection for 72 h, cells were detached using EDTA-free trypsin and washed three times with ice-cold PBS. The cell apoptosis rate was detected using the FITC Annexin V Apoptosis Detection kit (BD Biosciences) according to the manufacturer's protocol using a flow cytometer.

Cells in the logarithmic phase were harvested following infection for 72 h and washed twice with ice-cold PBS, then fixed with 5 ml 70% ethanol overnight at −20°C. Fixed cells were washed twice with ice-cold PBS and then centrifuged at 800 × g for 15 min at 4°C. Cells were resuspended in 0.4 ml ice-cold PBS and subjected to 1 ml propidium iodide (PI)/Triton X-100 containing RNase staining for 30 min at 4°C followed by flow cytometry. Data were analyzed using the CellQuest software (version 7.5.3; BD Biosciences).

Cells were seeded into a 24-well plate (1×10⁵ cells/well) and co-transfected with 0.4 µg of miR-106a or miR-NC vectors, and 0.1 µg of GV272/VNN2 or GV272/VNN2-mut vectors containing 3′-UTR as well as GV272/VNN2-NC vector. The firefly luciferase reporter gene was constructed into the vectors to quantitatively reflect the inhibitory effect. Cells were harvested 48 h after transfection for luciferase activity assays using the Reporter Assay system (Promega Corporation) according to the manufacturer's protocol.

Cells were harvested 72 h after infection, and total protein was extracted using radio immunoprecipitation assay lysis buffer and phenylmethanesulfonyl fluoride (both from Beyotime Institute of Biotechnology, Shanghai, China) according to the manufacturer's protocol. Total protein (30 µg/lane) were separated using 10% SDS-PAGE and electroblotted onto a 0.2-µm pore size polyvinylidene fluoride membrane (Beyotime Institute of Biotechnology). The membrane was blocked by 5% skim milk at room temperature for 2 h and then incubated overnight at 4°C with rabbit anti-human VNN2 antibody (1:1,000; cat. no., 25643-1-AP) or β-actin antibody (1:1,000; cat. no., 20536-1-AP) (both from ProteinTech Group, Inc., Chicago, IL, USA) as a control. Following washing of the membrane three times with Tris-buffered saline containing Tween-20 (1X TBST), the membrane was incubated for 2 h at room temperature with horseradish peroxidase-conjugated affinipure goat anti-rabbit IgG secondary antibody (1:5,000; cat. no., SA00001-2; ProteinTech Group, Inc.). Following washing of the membrane three times with TBST again, an ultrasensitive chemiluminescence solution was used from the BeyoECL Plus kit (Beyotime Biotechnology, Shanghai, China) to detect the protein bands according to the manufacturer's protocol (23). The bands were observed using MF-Chemisis 2.0 and GelCapture software (DNR Bio-Imaging Systems, Ltd., Neve Yamin, Israel).

Open access miRNA databases (TargetScan, http://www.targetscan.org/vert_72/; PicTarget, https://pictar.mdc-berlin.de/; and MicroCosm, http://www.ebi.ac.uk/enright-srv/microcosm/) were used for prediction of miR-106a target genes. SPSS software (version 22; IBM Corp., Armonk, NY, USA) was used to analyze all results. All statistical data are presented as mean ± standard deviation following three independent experiments. Student's t-tests were used when only two groups were present and one-way analysis of variance followed by the Scheffe post hoc test was used when more than two groups were present. The spearman's rank correlation test was used to measure the correlation between the expression of miR-106a and VNN2. P<0.05 was considered to indicate a statistically significant difference.

To quantify the expression of miR-106a in 18 osteosarcoma and non-cancerous adjacent tissues samples, RT-qPCR was performed. The results indicated that the miR-106a expression level was significantly increased in osteosarcoma tissues compared with NTAs (Fig. 1A). In the human U2OS, Saos-2 and MG63 cells, levels of miR-106a were significantly upregulated, compared with that in normal hFOB1.19 osteoblast cells (Fig. 1B). Since U2OS cells exhibited the highest miR-106a expression level among the three osteosarcoma cell lines, this cell line was chosen for further studies.

To explore the association of miR-106a expression with human U2OS cells, the cells were infected with lentivirus-mediated small interfering RNA (miR-106a-inhibitor). Infected cells labeled with green fluorescent protein were observed under a fluorescent microscope, and the infection efficiency was >90% when comparing the number of cells using the green fluorescent view in the same field (Fig. 2A and B). Infection efficiency was also determined by RT-qPCR at 72 h after infection (Fig. 2C).

The MTS assay used to investigate human U2OS cell proliferation following infection with miR-106a-inhibitor or NC was then performed. The MTS assay was performed at 24, 48, 72 and 96 h. No significant differences were identified in absorbance at 490 nm in the first 48 h. At 72 and 96 h, a significant decrease in the absorbance at 490 nm was identified in human U2OS cells infected with miR-106a-inhibitor compared with the NC-infected cells (P<0.05; Fig. 2D). This indicated that the knockdown of miR-106a expression may have suppressed cell proliferation in human U2OS cells.

Next, an invasion assay was performed to investigate the influence of knockdown of miR-106a of human U2OS cells on the migration and invasion ability. Cells that passed through the upper chamber and attached to the lower surface of the membrane were fixed (Fig. 2E) and quantified (Fig. 2F). The number of miR-106a-inhibitor-infected U2OS cells that passed through the upper chamber was significantly decreased compared with that of miR-NC-infected cells.

In order to explore the mechanisms underlying the inhibitory effect of miR-106a-inhibitor infection on cell proliferation and invasion, flow cytometric analysis of apoptosis and the cell cycle were applied (Fig. 3). As the Annexin V and PI staining results demonstrated, knockdown of miR-106a resulted in a significant increase in the apoptosis rate of human U2OS cells from 2.7 to 6.7% compared with miR-NC-infected cells (P<0.05; Fig. 3A). The cell cycle analysis demonstrated that knockdown of miR-106a significantly increased the percentage of cells in the G2/M phase from 12.80 to 22.54% and decreased the percentage in the S phase from 30.97 to 18.04% (both P<0.05; Fig. 3B).

To explore whether VNN2 is a candidate target gene of miR-106a, VNN2 protein expression levels were detected by western blot analysis in human U2OS cells following the knockdown of miR-106a (Fig. 2C). The results revealed a significant upregulation in VNN2 protein expression in knocked down U2OS cells compared with miR-NC-infected cells (Fig. 4A).

RT-qPCR was performed to analyze the effect of downregulating miR-106a expression on VNN2 expression at the transcriptional level. The expression level of VNN2 mRNA was significantly increased in miR-106a inhibitor-infected U2OS cells compared with that in NC-infected U2OS cells (P<0.05; Fig. 4B). An inverse association between miR-106a expression and VNN2 expression levels was verified via RT-qPCR and western blot analysis.

By using open access online databases (TargetScan, PicTarget and miRBase Targets), VNN2 was initially identified as a candidate target gene of miR-106a. A complementary site of miR-106a was identified in the VNN2 3′-UTR (Fig. 5A).

To confirm the possibility that miR-106a directly targets VNN2, the miR-106a binding region at the 3′-UTR of VNN2 or VNN2-mut mRNA were cloned downstream of the firefly luciferase reporter gene in GV272 vectors. miR-106a or miR-NC vectors and GV272 vectors were then co-transfected into 293T cells. The relative luciferase activity of the reporter containing VNN2 3′-UTR was significantly suppressed by 54±0.03% (P<0.05) when miR-106a vectors were co-transfected, whereas the relative luciferase activity of 3′-UTR-NC reporter was not affected (Fig. 5B). The results of luciferase activity assay indicated that miR-106a may target and suppress VNN2 gene expression by binding the 3′-UTR of VNN2 mRNA.

To quantify the expression of VNN2 in 18 osteosarcoma tissue and NTA samples, RT-qPCR was performed. The results indicated that the VNN2 expression level was significantly decreased in osteosarcoma tissues, compared with NTAs (Fig. 6A). Finally, the spearman's rank correlation test was used to analyze the expression of VNN2 and miR-106a in 18 osteosarcoma with non-cancerous adjacent tissues. The result indicated a negative correlation between them (R=−0.640, P<0.01; Fig. 6B).