OSCC cell lines, including OC3, KB, OEC-M1, HSC3 and SCC-4, were purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China) and were cultured according to the manufacturer's instructions. Normal human oral keratinocytes (HOK) were obtained from ScienCell (Carlsbad, CA, USA) and were grown in oral keratinocyte medium according to the manufacturer's instructions. The cells were maintained in a humidified atmosphere containing 5% CO₂ at 37°C. Five-week-old male BALB/c nude mice (weighing 25–30 g) were purchased from the Laboratory Animal Centre of the Medical School of Xi'an Jiaotong University (Xi'an, China) and were raised and handled according to the guideline of the Institutional Animal Care and Use Committee of Xi'an Jiaotong University.

Twenty paired adjacent non-tumorous and tumor tissues from patients undergoing surgical resection were frozen in liquid nitrogen. The present study was approved by and was in accordance with the guidelines of Institutional Human Experiment and the Ethics Committee of the First Affiliated Hospital, Medical School of Xi'an Jiaotong University (Xi'an, China).

Total RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA, USA) and was used to generate cDNA using M-MLV reverse transcriptase (Clontech, Palo Alto, CA, USA) and One Step PrimeScript miRNA cDNA Synthesis kit (Takara, Dalian, China) for mRNA and miRNAs analysis, respectively, according to the manufacturer's instructions. The gene expression was detected using SYBR-Green qPCR Master Mix (Thermo Fisher, Shanghai, China). The relative quantification of the gene expression level was compared with the internal referee GAPDH (for mRNA) or U6 SnRNA (for miRNAs) using the 2^−ΔΔCt method.

Lentivirus (LV)-pre-miR-138, LV-anti-mR-138 and their negative controls were purchased from GenePharma (Shanghai, China). For cell infection, the cells were cultured in a normal medium for 24 h and then cultured in a medium containing Polybrene (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA). The cells were then infected with pre-miR-13, LV-anti-mR-138s or their negative controls on a 0.5×10⁵ plaque-forming unit overnight. The old medium was discarded, and a new medium without Polybrene was added and incubated overnight. Stable clones were selected with puromycin dihydrochloride (Santa Cruz Biotechnology, Inc.).

Cell proliferation was detected by 3-(4,5)-dimethylthiahiazol(-z-y1)-3,5-di-phenytetrazo-liumromide (MTT) assay. Briefly, the cells infected with LV-miR-138 or LV-anti-miR-138 were seeded in 96-well plates at 1×10⁴ cells/well for 24, 48 and 72 h. The old medium was then replaced with fresh medium containing MTT [20 μl/well; 5 mg/ml diluted in phosphate-buffered saline (PBS)]. The cells were continually cultured for another 4 h, the medium was discarded and dimethylsulfoxide (150 μl/well; Sigma, St. Louis, MO, USA) was added to dissolve the formed formazan crystals. The optical density at 490 nm was then measured with a microtiter plate reader (Thermo Electron Corporation, Vantaa, Finland).

The cDNA fragment of 3′-UTR of YAP1 containing the putative binding site of miR-138 was subcloned into a pGL3 luciferase promoter vector (Promega, Madison, WI, USA). The cells firmly expressing pre-miR-138 or anti-miR-138 were transfected with 0.1 μg of pGL3-YAP1-3′-UTR and incubated for 48 h. The cells were collected and lysed. The luciferase activities were quantified using the Dual-Luciferase reporter assay kit (Promega) according to the manufacturer's instructions.

Proteins from cells or tumor tissues were extracted using a protein extraction kit (Applygen Technologies, Beijing, China). Protein concentrations in different samples were measured using the Bradford method. For protein separation, a total of 50 μg of protein was loaded on 12.5% sodium dodecyl sulfate polyacrylamide gel electrophoresis. The protein was then transferred to a nitrocellulose membrane (Bio-Rad, Hercules, CA, USA), which was blocked with 3.0% non-fat milk for 1 h at 37°C. Primary antibodies were added and incubated at 4°C overnight. Horseradish peroxidase-conjugated secondary antibodies (1:2,000; Bioss, Beijing, China) were added and incubated for 1 h at room temperature. After washing, the immunoreactive protein bands on the membrane were visualized using an enhanced chemiluminescence detection system (Amersham, Little Chalfont, UK). The primary antibodies used in these experiments, including anti-YAP1 and anti-GAPDH antibodies were purchased from Santa Cruz Biotechnology, Inc.

OEC-M1 cells (1×10⁶ cells) infected with LV-pre-miR-138 or LV-anti-miR-138 diluted in 200 μl of PBS were inoculated subcutaneously into the left flank of nude mice. The size of the tumor (length and width) was measured every 3 days by a vernier caliper. The tumor volume was presented as length × width² × π/6. At the end of the experiment, the tumor was isolated, weighed and extracted for further analysis.

Data are presented as means ± standard deviation and processed using SPSS version 11.5 (SPSS, Inc., Chicago, IL, USA). Statistical differences were obtained by the Student's t-test or one-way analysis of variance followed by Bonferroni post hoc test. Differences were considered statistically significant at P<0.05.

To examine the role of miR-138 in OSCC, the expression profile of miR-138 in tumor and adjacent non-tumorous epithelia was quantified by RT-qPCR analysis. Generally, miR-138 was significantly downregulated in OSCC tumor tissues as compared with that in the adjacent non-tumorous tissues (Fig. 1A). In addition, the expression of miR-138 in several different OSCC cell lines was detected. Furthermore, the results showed that miR-138 expression was markedly reduced to different degrees in OC3, KB, OEC-M1, HSC3 and SCC-4 cell lines than that in the control HOK cells (Fig. 1B).

Considering the downregulated expression of miR-138 in OSCC, we investigated the function of miR-138 in OSCC cells. OSCC cells with miR-138 overexpression or downregulation were generated in OC3 (Fig. 2A) and OEC-M1 (Fig. 2B) cells by infecting with LV-pre-miR-138 or LV-anti-miR-138. MTT assay results showed that miR-138 overexpression significantly decreased the proliferation in OC3 cells, whereas miR-138 downregulation markedly increased OC3 cell proliferation (Fig. 2C). Additionally, similar results were obtained using OEC-M1 cells (Fig. 2D).

To investigate the potential mechanism of the miR-138 regulative role in OSCC, we predicted the target gene of miR-138 by bioinformatics analysis. Notably, we found that the 3′-UTR of YAP1, a candidate oncogene in various types of cancer (29), processed the putative binding sites for miR-138 (Fig. 3A). The wild-type and the mutants of 3′-UTR of YAP1 in the predicted binding sequences were constructed into luciferase reporters to confirm the direct binding between miR-138 and 3′-UTR of YAP1. These reporters were then transfected into HEK293 cells with miR-138 overexpression or downregulation by infection with LV-pre-miR-138 and LV-anti-138. The reporter assay exhibited that miR-138 overexpression significantly inhibited the luciferase activity in wild-type transfected cells, whereas miR-138 downregulation markedly increased the luciferase activity in wild-type transfected cells. Additionally, the aberrant expression of miR-138 had no obvious effect on the luciferase activity in the mutated transfected cells (Fig. 3B).

To verify the interaction between miR-138 and YAP1, we observed the effect of the ectopic expression of miR-138 on YAP1 expression in OSCC cells. RT-qPCR analysis revealed that miR-138 overexpression significantly decreased the mRNA expression of YAP1 in OC3 (Fig. 4A) and OEC-M1 (Fig. 4B). By contrast, miR-138 inhibition markedly increased the YAP1 mRNA expression in OC3 and OEC-M1 cells. Similar results were obtained in the western blot analysis, which demonstrated that the protein expression of YAP1 was repressed by miR-138 overexpression or promoted by miR-138 downregulation (Fig. 4C and D). To assess whether miR-138 regulated the cell proliferation of OSCC cells by action on YAP1, we transfected YAP1 overexpression vectors, which harbored no specific miR-138 binding specific sequences in 3′-UTR in miR-138-overexpressing cells. The results showed that the overexpression of YAP1 significantly abrogated the inhibitory effect of miR-138 on the cell proliferation of OSCC cells (Fig. 4E and F).

To confirm the above findings of miR-138, an in vivo xenograft assay was performed using overexpressed stable OC3 cells to test the inhibitory effect of miR-138 on tumor growth in vivo. Compared with the control group, the overexpression of miR-138 significantly reduced the average tumor weight (Fig. 5A) and tumor volume (Fig. 5B) in nude mice subcutaneously injected with OC3 cells. Moreover, the expression of miR-138 and YAP1 in harvested tumor tissues were detected. The results showed that miR-138 was significantly increased in the tumors derived from miR-138-overexpressing OC3 cells (Fig. 5C), whereas the protein expression level of YAP1 was significantly decreased (Fig. 5D).