To investigate CNV-miRNAs in the Korean population, CNV regions from our previously reported CNV study were analyzed (22). Briefly, 4,694 samples that are part of the Korean Genome Epidemiology Study were genotyped with the NimbleGen HD2 3×720 K comparative genomic hybridization array. A total of 9,388 CNV regions were identified in human genome build hg18. Among them, 3,601 CNV regions tagged by highly correlated SNPs were provided as the content for the database. In the present study, miRBase, a biological database of miRNA sequences and annotations, was used to obtain human miRNA genome coordinates with the human genome build hg19 (23). Next, the genome coordinates of hg19 were converted to those of hg18 using LiftOver in the University of California, Santa Cruz genome browser (http://genome.ucsc.edu/cgi-bin/hgLiftOver). Of the 1,872 human miRNAs in miRBase, two miRNAs, hsa-mir-1273 and hsa-mir-6724, were excluded due to the absence of genome coordinate information. Moreover, as hsa-mir-511 has two different genome coordinates, both genomic positions were included in the miRNA list. Finally, 9,388 CNV regions with 1,871 miRNA regions were compared to identify CNV-miRNAs.

The human osteosarcoma MG-63 cell line was purchased from the American Type Culture Collection. The medium used for routine subculture was Dulbecco's modified Eagle's medium (Gibco Life Technologies, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum, penicillin (100 U/ml) and streptomycin (100 µg/ml). The cells were maintained at 37°C in a humidified 5% CO₂ incubator. Each 20 nM of miR-650 mimic (forward, 5′-AGGAGGCAGCGCUCUCAGGAC-3′ and reverse, 5′-GUCCUGAGAGCGCUGCCUCCU-3′; Bioneer, Daejeon, Korea) and ING4 small interfering RNA (order no. 1074590; Bioneer) were transfected into MG-63 cells with Lipofectamine RNAiMAX reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer's instructions. Subsequent to transfection, the cells were incubated for 48 h and then treated with 10 ng/ml IL1B (R&D Systems Inc., Minneapolis, MN, USA) prior to being harvested for further experiments.

Total RNA was isolated from cell lysates using the RNeasy Mini kit (Qiagen GmbH, Hilden, Germany). Total RNA (1 µg) was mixed with the AccuPower RocketScript Cycle RT Premix (Bioneer) for cDNA synthesis according to the manufacturer's instructions. The transcribed products were used to amplify target genes. The primer sequences for PCR were as follows: miR-650 forward, 5′-AGAGGAGGCAGCGCTCT-3′ and reverse, 5′-CAGTGCGTGTCGTGGAGT-3′; ING4 (order no. P279919; Bioneer); GAPDH (order no. P267613; Bioneer); and IL6 (order no. P211161; Bioneer). Amplification was performed using the Exicycler™ 96 Real-Time Quantitative PCR System (Bioneer) in a 20 µl reaction mixture containing 2 µl cDNA template (80 ng), 2.5 µl of each primer and 13 µl distilled water with AccuPower Greenstar qPCR Premix (Bioneer), including dNTP mixture. qPCR was performed under the following conditions: Initial denaturation at 95°C for 10 min; 40 cycles of 95°C for 10 sec, 60°C for 30 sec. Exicycler 3 analysis software (version 3.55.0; Bioneer) was used to calculate cycle threshold (Ct) values for all genes. Relative expression was calculated using the 2^−ΔΔCt method (24)

Immunoblotting analysis was performed as previously described (25). Briefly, the cultured cells were rinsed with phosphate-buffered saline, scraped into 100 µl RIPA cell lysis buffer (Cell Signaling Technology Inc., Danvers, MA, USA) and placed on ice for 1 h. Next, the cells were centrifuged and the supernatant was harvested. Aliquots (20 µg) of soluble proteins were separated with SDS-PAGE and transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). The membranes were incubated overnight at 4°C with specific antibodies: Rabbit polyclonal antibody for IκBα and GAPDH, and goat polyclonal antibody for ING4 (1:1,000; Santa Cruz Biotechnology Inc., Dallas, TX, USA). The blot was then incubated with the corresponding horseradish peroxidase-conjugated anti-rabbit IgG or anti-goat IgG (1:5000; Cell Signaling Technology Inc.). The immune complex was visualized with enhanced chemiluminescence reagent (GE Healthcare Life Sciences, Chalfont, UK), and image processing was performed using an image acquisition system (Fusion FX, Vilber Lourmat, Marne-la-Vallée, France).

The culture medium of the cells that were transfected with the miR-650 mimic was collected at 6 h post-treatment with IL1B. The level of human IL6 was determined with an ELISA kit (R&D Systems Inc.).

The MG-63 cells were co-transfected with the pGL4.32 [luc2p/NF-κB-RE/Hygro] containing the nuclear factor kB (NFkB) response element, with pRL-TK expressing Renilla luciferase as an internal control (Promega, USA) and miR-650 mimic with Lipofectamine RNAiMAX reagent in OPTI-MEM (Life Technologies). Following transfection, the cells were incubated for 48 h and then treated with IL1B (10 ng/ml) prior to being harvested. Cell extracts were used for the dual-luciferase assay (Promega Corporation, Madison, WI, USA). Firefly luciferase activity as a reporter was normalized to Renilla activity to control for transfection efficiency.

Experimental results were analyzed using R software (version 3.0.2; http://www.r-project.org/). Statistical analysis was performed with a one-way analysis of variance, and data are expressed as the mean ± standard deviation. P<0.05 was considered to indicate a statistically significant difference.

From the comparative analysis of 9,388 CNV regions and 1,871 miRNA regions, 38 miRNAs on CNV regions were identified (Table I). Table I shows the genome coordinates of these 38 miRNAs. When several CNV regions with different break-points (i.e., start and end positions) encompassed the same miRNA, minimal overlapping regions between CNVs were used. Among the 38 miRNAs, the target genes of three (miR-650, miR-132 and miR-212) have previously been reported (13,26). From those three miRNAs, miR-650 was selected for this functional study, as its location overlapped with the exon region of another transcript, and as its target genes have been identified in several cancer cell types, but not in osteosarcoma cells (14,16,17,20,27.

Although ING4 is a target gene of miR-650 in other cancer cell types (13,17,20), it had not been evaluated in osteosarcoma cells. Thus, a miR-650 mimic was transfected to overexpress miR-650 in the MG-63 osteosarcoma cells (Fig. 1A). Overexpression of miR-650 decreased the expression of ING4 mRNA and protein (Fig. 1B and C). As MG-63 cells are osteoblast-like osteosarcoma cells, stimulation with IL1B may accelerate the production of inflammatory cytokines, such as tumor necrosis factor-α and IL6, through the NFκB signaling pathway (28). In the present study, the overexpression of miR-650 decreased the upregulation of ING4 mRNA expression that was induced by IL1B (Fig. 2A). A similar expression pattern was observed for protein levels (Fig. 2B). These results indicate that ING4 may be a target gene of miR-650 in osteosarcoma cells, similar to previous results in other cancer cells.

Overexpression of ING4 decreases the expression of IL6 in human umbilical vein endothelial cells (29). Therefore, we investigated the role of miR-650 in the production of IL6 induced by IL1B in MG-63 cells. As expected, IL6 mRNA and protein expression was increased by treatment with IL1B in the MG-63 cells and was more highly increased in the miR-650-overexpressing cells than in the control cells (Fig. 2C and D). ING4 may regulate the expression of IL6 by modulating NFkB activity, as reported previously in melanoma and brain tumors (29,30). Therefore, the present study investigated whether NFκB transcriptional activity is involved in the production of IL6 by miR-650 in MG-63 cells. NFκB transcriptional activity was increased by IL1B treatment and more highly increased in the miR-650-overexpressing cells than the control cells (Fig. 2E). Furthermore, IκBα protein levels were decreased by overexpression of miR-650 (Fig. 2F). These results indicate that miR-650 regulates the production of IL6 that is induced by IL1B by modulating ING4 expression and NFκB transcriptional activity in osteosarcoma cells. This is similar to the results found in other cells, such as brain tumor cells and melanoma cells (29,30).

The present study investigated whether the effect of miR-650 on the production of IL6 induced by IL1B was mediated by ING4. The production of IL6 mRNA and protein was more highly increased in the cells in which ING4 was knocked down compared with the negative control group (Fig. 3). These results indicate that miR-650 regulates the production of IL6 by modulating its target gene, ING4.