A total of 80 OSCC tissues and adjacent non-tumour tissues from patients at the School and Hospital of Stomatology, China Medical University (Shenyang, China) between January 2013 and January 2015 were selected. Patients with malignancies in other organs and those who had received any anti-cancer therapy were excluded. Clinical data including age, sex, primary tumour site, differentiation, tumour node metastasis (TNM) classification and recurrence were obtained from pathological and clinical records. The ethics committee of the School and Hospital of Stomatology, China Medical University approved the present study and written informed consent was obtained from patients providing tissue specimens.

Human OSCC cell lines (Cal-27, SCC-9 and SCC-25) and human keratinocytes cell line (HaCaT cells) were purchased from the American Type Culture Collection, and were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% foetal bovine serum (FBS) and a 1% penicillin-streptomycin solution in a humidified atmosphere containing 5% CO₂ at 37°C. The cell culture medium and FBS were purchased from Gibco (Thermo Fisher Scientific, Inc.). The HaCaT cell line was originated from normal skin tissue. The normal oral mucosal epithelial cell line is not widely available, and the structure of the skin is very similar to the structure of the oral mucosa, so HaCaT cells were used as the control, as previously described (16,17). The HaCaT cell line was authenticated by STR sequencing.

miR-198 mimics, small interfering RNA (siRNA/si-) targeting cyclin-dependent kinase 4 (CDK4) and negative controls (NCs) were designed and synthesised by Shanghai GenePharma Co., Ltd. Cells (Cal-27 and SCC-9) were seeded into 96-well plates in antibiotic-free growth medium at a density of 4×10³ cells/well. miR-198 mimics (100 nmol/µl; 5′-GGU CCA GAG GGG AGA UAG GUU C-3′) and the control (100 nmol/µl; 5′-UUC UCC GAA CGU GUC ACG UAA U-3′) were transfected when the cells reached 70-80% confluence. si-CDK4 (80 nmol/µl; sense, 5′-CAG UUC GUG AGG UGC UUU AC-3′ and antisense, 5′-GUA A AG CCA CCUC ACG AAC UG-3′) and si-Ctrl (80 nmol/µl; sense, 5′-UUC UCC GAG CGU GUC ACG UTT-3′ and antisense, 5′-ACG UGA CAC GUU CGG AGA ATT-3′) were transfected using Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.) when cells reached 30-40% confluence, in accordance with the manufacturer's protocol. By inserting CDK4 cDNA into the pcDNA3.1 vector (Shanghai GenePharma Co., Ltd.), the CDK4 overexpression plasmid (pcDNA-CDK4) was generated and its sequence was confirmed by Shanghai GenePharma Co., Ltd. An empty pcDNA3.1 vector was used as the NC. Using Lipofectamine 3000, pcDNA-CDK4 (500 ng/µl) was transfected into cells. After incubation at 37°C for 6 h, the culture medium was changed. After transfection for 48 h, the subsequent experimentation was carried out. Transfection efficiency was analysed via reverse transcription-quantitative PCR (RT-qPCR).

Total RNA was extracted from frozen tissues and Cal-27, SCC-9 and HaCat cells using TRIzol^® (Takara Biotechnology Co., Ltd.) according to the manufacturer's protocol. Using standard spectrophotometric methods, RNA concentration was quantified and its purity was determined.

To examine miR-198 expression, RNA (1 µg) was reverse transcribed using Hairpin-it™ miRNA and a U6 snRNA Normalisation Kit (Shanghai GenePharma Co., Ltd.). The RT reaction system contained the following: 4 µl 5X MMLV RT buffer, 0.75 µl dNTP, 1.2 µl miRNA and U6 snRNA RT primer mix, 0.2 µl MML reverse transcriptase (200 U/µl)², 1 µg RNA and RNase-free H₂O, which added up to 20 µl. The following conditions were employed for RT: 25°C for 30 min, 42°C for 30 min and 85°C for 5 min. The qPCR reaction system contained: 10 µl 2X Real-time PCR Master Mix, 0.4 µl miRNA and U6 snRNA specific primer set (10 µM)¹, 0.4 µl ROX reference dye (50X)³, 0.2 µl Taq DNA polymerase (5 U/µl), 2 µl miRNA RT product and sterilised H₂O, which added up to 20 µl. The thermocycling conditions were as follows: 95°C for 3 min, 95°C for 12 sec and 62°C for 40 sec. Gene expression was normalised to U6 expression, which was used as an internal control, and the relative expression level was calculated using the 2^−ΔΔCq method (18). The patients were divided into two groups based on the median level of miR-198 expression: High expression and low expression.

To determine the expression of E-cadherin, N-cadherin, vimentin, MMP-2, MMP-9 and CDK4, cDNA served as the template for PCR amplification using a cDNA synthesis kit (Takara Biotechnology Co., Ltd.), according to the manufacturer's protocol. The RT reaction system contained the following: 4 µl 5X PrimeScript RT Master Mix, 1 µg RNA and RNase-free H₂O, which added up to 20 µl. The following conditions were employed for RT: 37°C for 15 min and 85°C for 5 min. qPCR was conducted using SYBR Premix Ex Taq II kit (Takara Biotechnology Co., Ltd.) with the following reaction system: 10 µl SYBR Premix Ex Taq II, 1 µl cDNA, 0.5 µl forward primer, 0.5 µl reverse primer and 8 µl sterile water. The thermocycling conditions were as follows: 95°C for 1 min, 94°C for 30 sec, 58°C for 30 sec and 72°C for 10 sec. Primers were designed and synthesised by Takara Biotechnology Co., Ltd. Gene expression was normalised to GAPDH expression, which was used as an internal control, and the relative expression level was calculated using the 2^−ΔΔCq method.

Total protein from Cal-27 and SCC-9 cells were extracted using RIPA lysis buffer (CoWin Biosciences) containing PMSF (1 mM) and the concentration was determined using the bicinchoninic acid method. Subsequently, 50 µg total protein was separated via SDS-PAGE on a 10% gel, and subsequently separated proteins were transferred onto PVDF membranes (Sigma-Aldrich; Merck KGaA). Following blocking for 1 h with 5% non-fat milk at room temperature, the membranes were incubated with primary antibodies against E-cadherin (cat. no. 3195; 1:1,000; Cell Signaling Technology, Inc.), N-cadherin (cat. no. 13116; 1:500; Cell Signaling Technology, Inc.), vimentin (cat. no. 5741; 1:1,000; Cell Signaling Technology, Inc.), MMP-2 (cat. no. 40994; 1:1,000; Cell Signaling Technology, Inc.), MMP-9 (cat. no. 13667; 1:1,000; Cell Signaling Technology, Inc.), CDK4 (cat. no. 12790; 1:1,000; Cell Signaling Technology, Inc.) and GAPDH (cat. no. sc-47724; 1:5,000; Santa Cruz Biotechnology, Inc.) overnight at 4°C. The membranes were washed and then incubated with the secondary antibodies (cat. nos. sc-2357 and sc-2005; 1:5,000; Santa Cruz Biotechnology, Inc.) for 2 h at room temperature. The blots were visualised via ECL (Pierce; Thermo Fisher Scientific, Inc.). ImageJ software (version 1.48; National Institutes of Health) was used for densitometry.

The CDK4 expression data for OSCC was obtained from the Gene Expression Profiling Interactive Analysis (GEPIA) online database (http://gepia.cancer-pku.cn/). Tumour and normal samples in the GEPIA database were derived from The Cancer Genome Atlas (TCGA, https://tcga-data.nci.nih.gov/tcga/) and the Genotype-Tissue Expression (GTEx, https://www.gtexportal.org/home/) projects.

OSCC Cal-27 and SCC-9 cells were transfected and seeded on 96-well plates at a density of 2×10³ cells/well. After culture for 24, 48, 72 or 96 h, cell proliferation was tested via an MTS assay (Abcam), according to the manufacturer's protocol. The absorbance of each well was measured at 570 nm on a microplate reader.

OSCC Cal-27 and SCC-9 cells were transfected and seeded on 6-well plates at a density of 1×10³ cells/well and cultured for 7 days. The medium was removed and cells were washed with PBS twice, fixed in 100% methanol (1 ml) for 10 min and stained with 10% crystal violet solution for 20 min at room temperature, the colonies (defined as >10 cells) were counted and photographed.

OSCC Cal-27 and SCC-9 cells were transfected and seeded on 6-well plates at a density of 4×10⁵ cells/well and cultured for 48 h. According to the manufacturer's protocols, cells were incubated with Annexin V-Fluorescein Isothiocyanate/Propidium Iodide Apoptosis Detection kit (Dojindo Molecular Technologies, Inc.) in the dark. The results were analysed using a BD FACSVerse™ flow cytometer (BD Biosciences) equipped with the FlowJo version 10 software (FlowJo LLC). Early and late apoptosis was assessed.

OSCC Cal-27 and SCC-9 cells were transfected, resuspended in serum-free DMEM, and plated onto the upper chamber of the Transwell insert (Corning, Inc.) at a density of 5×10⁴ cells/well. DMEM containing 10% FBS (500 µl) was added into the lower chamber to act as a chemoattractant. For the invasion experiments, Transwell chambers precoated with Matrigel (Corning, Inc.) at 4°C for 3 h. Cells in the Transwell plates were cultured for 48 h at 37°C with 5% CO₂. The adherent cells on the upper surface of the insert membrane were removed using cotton tips. Cells that migrated into the lower chamber were stained with 10% crystal violet for 20 min at room temperature and quantitated by counting in five different areas under a light microscope.

According to a previous study (19), the OSCC Cal-27 and SCC-9 cell culture supernatant was collected after treatment. Using ELISA kits (cat. nos. MMP200 and DMP900; R&D Systems, Inc.) the concentrations of MMP-2 and MMP-9 were measured, according to the manufacturer's protocol.

The potential binding sites of miR-198 and 3′-untranslated region (UTR) of CDK4 were predicted by TargetScan 7.1 (www.targetscan.org/vert_71). miR-198 mimics, miR NC and CDK4-3′UTR mutant (CDK4-MUT) and CDK4-3′UTR wild-type (CDK4-WT) plasmids were constructed by Shanghai GenePharma Co., Ltd. OSCC Cal-27 and SCC-9 cells were seeded on 6-well plates at a density of 3×10⁴ cells/well. After 24 h, miR-198 mimics (5′-GGU CCA GAG GGG AGA UAG GUU C-3′) and miR NC (5′-UUC UCC GAA CGU GUC ACG UAA U-3′) were co-transfected with CDK4-MUT or -WT plasmids into OSCC cells using Lipofectamine 3000. After incubation for 48 h, according to the manufacturer's protocol, a Dual-Luciferase Reporter Assay System kit (Promega Corporation) was used. The relative luciferase activity was determined by normalisation with Renilla luciferase activity.

A total of 24 male BALB/c nude mice (age, 4-6 weeks; weight, 20-25 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd., (Charles River Laboratories). All animals were raised at 22-26°C, 40-70% relative humidity and a 12 h light/dark cycle. The animals were provided with water and food freely. All animals were randomly divided into two groups (experimental and control groups, 12 mice/group) and maintained in a specific pathogen-free environment. For the xenograft experiment, Cal-27 cells transfected with miR-198 mimics and miR-NC were detached and resuspended in serum-free medium. Subsequently, Cal-27 cells (1 ml; 1×10⁷ cells/ml) were implanted subcutaneously in the flanks of BALB/c nude mice (20). The tumour size was checked regularly at the indicated time points. The tumour-bearing mice were sacrificed 6 weeks after the operation by cervical dislocation and tumour weights were measured. The maximum tumour diameter was 13 mm, and volume was 1,100 mm³. All animal studies were conducted at the Animal Center of China Medical University (Shenyang, China) according to the protocols approved by the animal ethics committee (Experimental animal welfare and ethics committee of China Medical University, approval no. CMU2019182).

All statistical analysis was performed using SPSS 21.0 software (IBM Corp.). Data are represented as the mean ± standard deviation based on at least three repeats. Comparisons between tumour and adjacent normal tissue expression were performed using a paired Student's t-test. For cell line experiments involving two groups, an unpaired t-test was conducted. One-way ANOVA followed by Tukey's post hoc test was performed to compare the differences between multiple groups. Using the Kaplan-Meier method and the log-rank test, survival curves were plotted and the overall survival (OS) and disease-free survival (DFS) outcomes were analysed. Pearson's correlation analysis was used to detect the correlation between miR-198 and CDK4 expression in OSCC tissues. Univariate and multivariate analyses were conducted according to the Cox regression model. P<0.05 was considered to indicate a statistically significant difference.

To investigate miR-198 expression in OSCC, miR-198 expression in OSCC tissues and the corresponding adjacent tissues was primarily analysed using RT-qPCR. It was found that miR-198 was downregulated in OSCC tissues compared with the normal tissues (Fig. 1A). The level of miR-198 expression in patients with metastasis was significantly decreased compared with patients with OSCC without metastasis (Fig. 1B). Moreover, RT-qPCR assays showed that miR-198 expression was reduced in all OSCC cell lines compared with HaCat cells (Fig. 1C). Expression was the most significantly reduced in the Cal-27 and SCC-9 cell lines, thus these cells were then chosen for further assays. Patients with OSCC with low miR-198 expression had poorer OS and DFS compared with patients with a high expression (Fig. 1D and E).

The cox proportional hazard model was used to determine the impact of miR-198 expression and the clinicopathological factors on the prognosis of patients with OSCC. Univariate Cox regression analysis showed that the differentiation (P=0.030 for OS; P=0.034 for DFS), TNM stage (P=0.039 for OS; P=0.031 for DFS) and miR-198 expression (P=0.025 for OS; P=0.022 for DFS) were significantly associated with poor OS and DFS. Multivariate analyses showed that the TNM stage (P=0.043 for OS; P=0.033 for DFS) and miR-198 expression (P=0.033 for OS; P=0.027 for DFS) were associated with poor OS and DFS (Table I).

RT-qPCR was used to detect transfection efficiency after cells were transfected with miR-198 mimics (Fig. 2A). MTS assays showed that compared with the miR-NC group, the proliferative activity of OSCC cells was decreased after transfection with miR-198 mimics at different observation time points (Fig. 2B). Consistent with this finding, the colony formation ability of OSCC cells was significantly reduced after miR-198 mimic transfection (Fig. 2C). Subsequently, the flow cytometry results showed that the percentage of apoptotic cells significantly increased after miR-198 mimics transfection (Fig. 2D). Xenograft experiments also showed a significant reduction in tumour size and volume when miR-198 was overexpressed (Fig. 2E). Taken together, these data indicated that miR-198 may function as a tumour suppressor in OSCC.

Transwell assays showed that compared with the miR-NC group, the number of invading OSCC cells was decreased significantly after transfection with miR-198 mimics (Fig. 3A). RT-qPCR and western blotting results showed that the overexpression of miR-198 significantly increased the level of E-cadherin, and decreased the levels of N-cadherin, vimentin, MMP-2 and MMP-9, in OSCC cells (Fig. 3B and C). ELISA showed that the concentration of MMP-2 and MMP-9 was decreased significantly after transfection with miR-198 mimics (Fig. 3D). Taken together, these data indicated that miR-198 may function as a tumour suppressor in OSCC.

TargetScan showed that there was a miR-198 binding sequence in the 3′UTR of CDK4 (Fig. 4A). Dual-luciferase reporter assays showed that the overexpression of miR-198 could inhibit CDK4-WT reporter activity, but not the activity of the CDK4-MUT construct in OSCC cells, demonstrating that miR-198 could specifically target the CDK4-3′UTR by binding to the seed sequence (Fig. 4B). RT-qPCR showed that there were no obvious changes in CDK4 mRNA expression after transfection with miR-198 mimics (Fig. 4C). Western blotting showed that transfection with miR-198 mimics significantly decreased the expression of CDK4 protein (Fig. 4D). Taken together, these data indicated that miR-198 targets CDK4 directly in OSCC cells.

From the GEPIA2 database, it was found that CDK4 expression was higher in OSCC tissues than that in controls (Fig. 5A). RT-qPCR showed that CDK4 mRNA expression was upregulated in OSCC tissues compared with the normal tissues (Fig. 5B). Moreover, RT-qPCR showed that there were higher expression levels of CDK4 in OSCC cell lines compared with HaCat cells (Fig. 5C). Patients with OSCC with high CDK4 expression had poorer OS and DFS than patients with a low expression (Fig. 5D and E). RT-qPCR analysis also showed that miR-198 was moderately negatively correlated with CDK4 expression in patients with OSCC (Fig. 5F). Taken together, these data indicated that CDK4 expression may be involved in OSCC progression.

RT-qPCR was performed to detect the transfection efficiency after si-CDK4 was transfected (Fig. 6A). MTS assays showed that compared with the si-Ctrl group, the proliferative activity of OSCC cells was decreased after transfection of si-CDK4 at different observation time points (Fig. 6B). Consistent with this finding, the colony formation ability of OSCC cells was significantly reduced after transfection with si-CDK4 (Fig. 6C). Subsequently, the results of flow cytometry showed that the percentage of apoptotic cells was significantly increased after transfection with si-CDK4 (Fig. 6D). Taken together, these data indicated that CDK4 may function as a tumour promoter in OSCC.

Transwell assays showed that compared with the si-Ctrl group, the number of invading OSCC cells was significantly decreased after transfection with si-CDK4 (Fig. 7A). RT-qPCR and western blotting results showed that knockdown of CDK4 expression significantly increased the level of E-cadherin, and decreased the levels of N-cadherin, vimentin, MMP-2 and MMP-9, in OSCC cells (Fig. 7B and C). Collectively, these data indicated that CDK4 may function as a tumour promoter in OSCC.

A representation vector that encoded the entire CDK4 coding sequence and lacked the 3′UTR was constructed. RT-qPCR was used to detect the transfection efficiency after pcDNA-CDK4 was transfected (Fig. 8A). Then, pcDNA-CDK4 or its NC was co-transfected with the miR-198 mimic into Cal-27 cells. RT-qPCR showed that CDK4 expression was significantly increased after co-transfection of pcDNA-CDK4 + miR-198 mimics compared with the pcDNA3.1 + miR-198 mimic group (Fig. 8B). MTS assays showed that the concomitant overexpression of miR-198 and CDK4 abrogated the inhibitory effects of transfection with pcDNA3.1 + miR-198 mimic (Fig. 8C). Transwell assays also showed that CDK4 overexpression reversed the inhibitory effects of transfection with pcDNA3.1 + miR-198 mimic on the invasion of Cal-27 cells (Fig. 8D). Moreover, RT-qPCR and western blotting showed that the overexpression of CDK4 decreased the level of E-cadherin, and increased N-cadherin and vimentin, in Cal-27 cells compared with the pcDNA3.1 + miR-198 mimic group (Fig. 8E). Taken together, these findings revealed that CDK4 reversed the miR-198-mediated inhibitory effect in OSCC cells.