The protocol for the present study was independently reviewed and approved by the Institutional Review Board (IRB) of Kaohsiung Veterans General Hospital (KSVGH; IRB no. VGHKS14-CT6-18, Kaohsiung, Taiwan). A total of 69 paired tumor tissues and adjacent normal tissue samples were collected from patients with oral cancer at the Department of Otolaryngology of Kaohsiung Veterans General Hospital (Kaohsiung, Taiwan) between 2010 and 2012. The requirements for written informed consent from the subjects were waived by the IRB of KSVGH, since all the data and specimens had been previously collected and analyzed anonymously. Pathological tumor-node-metastasis classification was determined at the time of the initial resection of the tumor in accordance with the 2002 guidelines of the American Joint Committee on Cancer staging system (29). Fresh tissue samples included tumor tissues and corresponding adjacent normal tissues from 69 OSCC patients. Following surgery, samples were immediately placed in RNAlater^® solution (Qiagen GmbH, Hilden, Germany) and stored at room temperature overnight prior to be frozen at −80°C.

Total RNA and DNA were extracted from tissues using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA), according to the manufacturer's protocol. Briefly, tissue samples were homogenized in 1 ml TRIzol and mixed with 0.2 ml chloroform (Avantor Performance Materials, Center Valley, PA, USA) for extracting protein and DNA, while RNA was precipitated with 0.6 ml isopropanol (Avantor Performance Materials). The remaining liquid was mixed with 0.5 ml back extraction buffer (4 M guanidine thiocyanate, 50 mM sodium citrate and 1 M Tris; pH 8.0; Avantor Performance Materials), and DNA was precipitated with 0.4 ml isopropanol. The concentration, purity and amount of total RNA and DNA were determined with a NanoDrop 1000 spectrophotometer (Nanodrop Technologies; Thermo Fisher Scientific, Inc.).

Total RNA (1 µg) was reverse transcribed using miR-specific stem-loop miR-196b-RT primer (Gemomics BioSci & Tech, New Taipei, Taiwan) and SuperScript^® III Reverse Transcriptase (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. The RT reaction included 1 µg RNA, 1X buffer, 2.5 mM dNTP and 0.5 mM stem-loop miR-196b RT primer. The reaction was performed using the following incubation conditions with a PCR thermocycler (TPersonal 20 Thermal Cycler; Biometra GmbH, Göttingen, Germany): 16°C for 30 min, followed by 50 cycles of 20°C for 30 sec, 42°C for 30 sec and 50°C for 1 sec. The complementary DNA was used at a dilution of 1:20 in water in the subsequent qPCR, which was performed using a specific forward primer and a universal reverse primer, under the following conditions: Incubation at 94°C for 10 min, followed by 40 cycles of 94°C for 15 sec and 60°C for 30 sec. Gene expression was detected with SYBR Green I (Applied Biosystems; Thermo Fisher Scientific, Inc.). The expression levels of miRNAs were normalized to those of U6 (ΔCq=miR-196b Cq-U6 Cq) as described by Livak and Schmittgen (30). The individual primers were synthesized by Qiagen China Co., Ltd. (Shanghai, China) used in the present study were as follows: miR-196b RT, foward 5′-CTCAACTGGTGTCGTGGAGTCGGCAATTCAGTTGAGCCCAACAA-3′ and gene specific forward primer, 5′-CGGCGGTAGGTAGTTTCCTGTT-3′; universal/miR-196b reverse, 5′-CTGGTGTCGTGGAGTCGGCAATTC-3′; and U6 forward, 5′-CTCGCTTCGGCAGCACA-3′ and reverse, 5′-AACGCTTCACGAATTTGCGT-3′.

Genomic DNA was extracted from oral tissue samples using TRIzol, and an aliquot (2 µg) was subjected to bisulfite conversion using the EZ DNA Methylation-Gold™ kit (Zymo Research Corporation, Irvine, CA, USA). Bisulfite conversion was achieved by incubating the reaction at 98°C for 10 min, 64°C for 2.5 h and 4°C for ≤20 h in a PCR thermocycler (Biometra GmbH).

Bisulfite-converted genomic DNA was used for methylation analysis of CpG islands with specific methylation primers. The PCR conditions were as follows: 94°C for 10 min, followed by 40 cycles of 94°C for 30 sec, 60°C for 3 sec and 72°C for 40 sec, with a final extension at 72°C for 10 min. A PCR thermocycler (Biometra GmbH) and Hot Start Taq DNA polymerase (Qiagen GmbH) were used for the reaction. The methylation status of the genomic DNA of individual samples was examined by TaqI digestion (New England BioLabs, Inc., Ipswich, MA, USA). The digested PCR fragments were then separated on a 2% agarose gel (Invitrogen; Thermo Fisher Scientific, Inc.) immersed in 1X Tris-acetate-ethylenediaminetetraacetic acid buffer (Omics Bio, Taipei, Taiwan) using Mupid-2plus DNA Gel Electrophoresis system (Advance Co., Ltd., Tokyo, Japan). Following ethidium bromide (MDBio, INc., Taipei, Taiwan) staining, the gels were visualized and anlyzed by a gel documentation system (Uvitec Cambridge, Cambridge, UK) In addition, the PCR products were cloned into the pGEM^®-T vector (Promega Corporation, Madison, WI, USA), and several clones were randomly selected for sequencing using a T7 primer (5′-TAATACGACTCACTATAGGG-3′) by Qiagen China Co., Ltd.. The individual primers used in the COBRA assay were as follows: miR-196b mF (forward primer), 5′-GAGAGAAAGGTGGATTTTAATAATAGGA-3′ and mR (reverse primer), 5′-AACCTATAACTTCCCCTTCCTTAAC-3′.

Human oral cancer SAS and Cal-27 cell lines were kindly provided by Dr Michael Hsiao (Genomics Research Center, Academia Sinica, Taipei, Taiwan). The oral cancer cell lines, SASA and CAL27, were cultured in Dulbecco's modified Eagle's medium (Biological Industries Israel Beit-Haemek Ltd., Kibbutz Beit-Haemek, Israel) supplemented with 10% fetal bovine serum (GE Healthcare Life Sciences, Logan, UT, USA) and 1% penicillin/streptomycin (Sigma-Aldrich, St. Louis, MO, USA). Stable or transient SAS and CAL-27 cell cultures expressing miR-196b were generated by transfecting SAS and CAL-27 oral cancer cells with the pLKO.1 vector (Addgene, Inc., Cambridge, MA, USA) or with pLKO.1-miRNA-196b plasmid using Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.). At 24 h post-transfection, stable cells expressing miR-196b were generated using puromycin (Sigma-Aldrich) selection for 14 days. The levels of miR-196b expression were measured by RT-qPCR as described previously in the ‘Stem-loop reverse-transcription-quantitative polymerase chain reaction’ section.

The sequence complementary to miR-196b, 5′-CCCAACAACAGGAAACUACCUA-3′, and the negative control (NC) sequence, 5′-GUGUAACACGUCUAUACGCCCA-3′, referred to as anti-miR-196b and anti-miR-NC, respectively, were synthesized with 2′-O-methyl oligo modified bases exhibiting phosphorothioate on the first two and final four bases and a 3-cholesterol modification, which was added through a hydroxyprolinol linkage (Dharmacon; GE Healthcare Life Sciences, Chalfont, UK). Transfection of these antagomirs was performed using Lipofectamine^® RNAiMAX (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. Cells were incubated with 100 nM anti-miR-196b for 2 days to suppress miR-196b expression.

For the cell proliferation assay, living cells (3×10³) were plated onto 96-well plates, and cell growth was determined 0–4 days later using WST-1 cell proliferation reagent (Roche Diagnostics, Basel, Switzerland). The absorbance value at 450 nm of each well was measured using a plate reader (Multiskan FC Microplate Photometer; Invitrogen™, Thermo Fisher Scientific, Inc.). Cells were also tested for their invasion ability in vitro using Transwell chambers (BD Biosciences, Franklin Lakes, NJ, USA). The lower side of the polycarbonate membranes (containing 8-µm pores) of the Transwell chamber was coated with Matrigel (1 or 0.2 µg/ml/well; BD Biosciences). Cells (5×10⁴) were then added onto the Matrigel layer in each well. Following incubation for 48 h at 37°C, the cells that had migrated through the Matrigel layer were visualized upon fixation with formaldehyde (BS Chemical, Kretinga, Lithuania) and staining with crystal violet (EMD Millipore, Billerica, MA, USA). The level of invasion was determined using a microscope (CKX41; Olympus Corporation, Tokyo, Japan). For the migration assay, cells (1.5×10⁶) were plated onto 6-well plates. A straight line was scratched into the cell monolayer in the middle of the well using a 200-ml pipette tip. Culture medium containing 10% fetal bovine serum was replaced with a serum-deprived culture medium, and the cells were then incubated at 37°C. Wound closure was monitored and photographed at 0, 6 and 12 h under a microscope.

Protien was was extracted from 3×10⁶ cells using 200 µl radioimmunoprecipitation assay buffer. A total of 30 µg of protein lysate was electrophoresed on 10% sodium dodecyl sulfate polyacrylamide gels (Sigma-Aldrich) and transferred onto polyvinylidene difluoride membranes (EMD Millipore). The membranes were blocked with blocking buffer [50 mM Tris-HCl (pH 7.6), 150 mM NaCl, 0.1% Tween 20, 5% non-fat dry milk, 0.05% sodium azide) for 1 h at room temperature. The membranes were incubated with the following primary antibodies overnight at 4°C: Polyclonal rabbit anti-fibronectin (dilution, 1:1,000; catalog no., GTX112794; GeneTex, Inc., Irvine, CA, USA); monoclonal rabbit anti-Ep-CAM (dilution, 1:2,000; catalog no., GTX61060; GeneTex, Inc.); polyclonal rabbit anti-vimentin (dilution, 1:2,000; catalog no., GTX100619; GeneTex, Inc.); polyclonal rabbit anti-Twist1/2 (dilution, 1:2,000; catalog no., GTX127310; GeneTex, Inc.); monoclonal rabbit anti-E-cadherin (dilution, 1:2,000; catalog no., GTX61329; GeneTex, Inc.); and monoclonal mouse anti-actin (dilution, 1:5,000, catalog no., MAB1501). Following washing three time for 10 min/time with Tris-buffered saline with Tween 20, the membranes were incubated with the following horseradish peroxidase-conjugated secondary antibodies for 1 hour at room temperature: Goat anti-mouse immunoglobulin (Ig)G (dilution, 1:20,000; catalog no., sc-2005; Santa Cruz Biotechnology, Inc., Dallas, TX, USA) and goat anti-rabbit IgG (dilution, 1:20,000; catalog no., sc-2004; Santa Cruz Biotechnology, Inc.). The target bands were visulized using WesternBright enhanced chemiluminescence reagent (Advansta, Inc., Menlo Park, CA, USA) and the results of the immunoreactions were analyzed with a BioSpectrum^® 500 Imaging System (Ultra-Violet Products Ltd., Cambridge, UK). The bands corresponding to individual genes were analyzed using ImageJ (imagej.nih.gov/ij/index.html) and the expression levels of these genes were normalized to those of actin.

The expression levels of miR-196b in all of the oral tissues were determined using stem-loop RT-qPCR. All reactions were performed in triplicate, and data were analyzed using the paired t test or Student's t test. The correlation between various clinicopathological characteristics and miR-196b expression levels (low vs. high) or methylation status (no vs. yes) was evaluated using the χ² test or Fisher's exact test. Cumulative survival curves were estimated using the Kaplan-Meier method, and comparisons between the survival curves were performed by using the log-rank test. P<0.05 was considered to indicate a statistically significant difference.

To investigate the biological role of miR-196b in OSCC, the expression levels in OSCC tissues and adjacent normal tissues were examined by stem-loop RT-qPCR. As represented in Fig. 1A, the expression levels of miR-196b were significantly upregulated in OSCC tissues, compared with the corresponding adjacent normal tissues (64 of 69, 92.7%, P<0.001). In addition, the expression levels of miR-196b were significantly increased in tissues with DNA hypomethylation, compared with those with DNA hypermethylation in OSCC tissues (Fig. 1B; P=0.009). A previous study by the present authors demonstrated that DNA hypomethylation could lead to miR-196b overexpression in gastric cancer (19). Therefore, the present study examined whether the DNA methylation status of CpG islands contributed to miR-196b overexpression in OSCC. Using COBRA, the methylation status of the miR-196b promoter region was assayed in OSCC samples and corresponding adjacent normal samples from 69 OSCC patients. DNA hypomethylation was more frequently observed in oral cancer tissues than in adjacent normal tissues (32 of 69, 46.3%; Fig. 1C). The methylation status of individual CpG dinucleotides of the miR-196b promoter region was verified in selected patients using a bisulfate sequencing approach. Consistent with the COBRA results, the densities of methylated CpG dinucleotides were higher in normal tissues (methylation rate, 30%) than in tumor tissues (methylation rate, 0%; Fig. 1D). The data also indicated that CpG-rich regions of miR-196b were unmethylated in the oral cell lines SAS and CAL-27 (data not shown). These results indicated that abnormal DNA hypomethylation resulted in miR-196b overexpression in OSCC. The clinicopathological characteristics of the 69 independent OSCC patients are summarized in Table I. The data indicated that the expression levels and methylation status of miR-196b were not associated with any clinicopathological features (Tables II and III). Notably, the disease-specific survival (DSS) curves of patients with miR-196b promoters exhibiting unmethylation status were significantly lower than those displaying methylation status (P=0.035; Fig. 1E).

Since miR-196b has been demonstrated to be able to act as an oncogene or as a tumor suppressor in various types of human cancer (20,26), stable SAS and CAL-27 clones overexpressing miRNA-196b were established in the present study by transfecting SAS and CAL-27 cells with a precursor-miR-196b construct. As represented in Fig. 2A, miR-196b was significantly overexpressed in the two stable cell lines overexpressing miR-196b (P=0.37). The effect of miR-196b on the proliferation, migration and invasion abilities of SAS and CAL-27 cells was subsequently investigated. Ectopic overexpression of miR-196b did not induce a significant proliferation ability in these two cell lines (Fig. 2B). By contrast, the wound-healing assay indicated that the overexpression of miR-196b could significantly promote SAS and CAL-27 migration (P<0.01; Fig. 2C). Furthermore, the invasion ability significantly increased in miR-196b-expressing SAS and CAL-27 cells (SAS cells, P=0.003; CAL-27 cells, P=0.001; Fig. 2D). Endogenous expression of miR-196b was knocked down in the SAS and CAL-27 cells by transfecting the cells with anti-miR-NC or anti-miR-196b. As indicated in Fig. 3, knockdown of endogenous miR-196b significantly inhibited the cell migration and invasion abilities of SAS and CAL-27 cells (P<0.01; Fig. 3A and B).

Since the epithelial-mesenchymal transition (EMT) is a crucial process in promoting cancer cell metastasis, the present study examined whether miR-196b could induce EMT in SAS and CAL-27 cells. As presented in Fig. 3C, the protein expression of endogenous fibronectin and was clearly decreased in SAS cells in which miR-196b was downregulated. However, the expression levels of the other EMT markers were not clearly altered in SAS cells with miR-196b knockdown.

In summary, DNA hypomethylation resulted in overexpression of oncogenic miR-196b in OSCC tissues, compared with corresponding adjacent normal tissues. Furthermore, DNA hypomethylation may serve as an adverse prognostic marker for OSCC.