For reverse transcription-quantitiative polymerase chain reaction (RT-qPCR) and western blot analysis, 14 OTSCC tissues and matched adjacent non-cancerous tongue tissues (at a distance >2 cm from the tumour) were collected from patients who underwent half-tongue resection between May and July 2010. In addition, 133 paraffin-embedded samples of OTSCC specimens were collected between January 2000 and December 2002 for immunohistochemical assay. All tissue samples were collected prior to chemotherapy or radiotherapy and were histologically and clinically diagnosed at the Third Affiliated Hospital of Kunming Medical University (Kunming, China). The tumour stage of each patient was classified according to the 2010 AJCC staging system (22). The median follow-up time of patients whose tissues were subjected to an immunohistochemical assay was 67 months at the time of analysis and ranged from 12 to 124 months. This study was approved by the Ethics Committee of the Third Affiliated Hospital of Kunming Medical University. The samples were collected after receiving informed consent from the patients.

The mRNA of OTSCC (NTEC1, TCA8113, TSCCA, SCC-9, SCC-25 and CAL-27 cells) and matched adjacent noncancerous tissues was purified using TRIzol Reagent (Invitrogen, Thermo Fisher Scientific, Inc., Waltham, MA, USA), and 2 µg of each sample was reverse transcribed using a TIAN Script kit (Invitrogen, Thermo Fisher Scientific, Inc.). The following primers were used: Sense: 5′-GGTGTCTCTCGCAGATTCATC-3′ and antisense: 5′-TCAGTCTTCGGCTGAGGTTCT-3′ for Beclin-1; and sense: 5′-TGTTGCCATCAATGACCCCTT-3′ and antisense 5′-CTCCACGACGTACTCAGCG-3′ for glyceraldehyde 3-phosphate dehydrogenase (GAPDH), which was used as an endogenous control. All reactions were run in triplicate in three independent experiments. All primers were synthesized by Invitrogen, Thermo Fisher Scientific, Inc. PCR reactions were performed on a PTC-200 PCR system (Bio-Rad Laboratories, Hercules, CA, USA) using the following cyclical procedure: 10 min at 94°C, followed by 45 cycles of 15 sec at 94°C, 30 sec at 54°C, 30 sec at 72°C, and a final cycle at 72°C for 10 min. For OTSCC cells, RT-qPCR and data collection were performed using an ABI PRISM 7900 HT sequence detection system (Applied Biosystems, Thermo Fisher Scientific, Inc.).

Cells (NTEC1, TCA8113, TSCCA, SCC-9, SCC-25 and CAL-27 cells, and patients paired tissue cells) were lysed with 1X sodium dodecyl sulphate (SDS) sample buffer (62.5 mmol/l Tris-HCl, 2% SDS, 10% glycerol and 5% 2-mercaptoethanol). The protein concentration was determined using a bicinchoninic acid protein assay. A total of 25 µg protein was separated electrophoretically in 9% SDS-polyacrylamide gel electrophoresis, transferred to polyvinylidene fluoride membranes and incubated sequentially with primary rabbit monoclonal anti-Beclin-1 (cat. no. 2026-1; 1:4,000, Epitomics Inc., Burlingame, CA, USA), mouse monoclonal anti-LC3 (cat. no. AM1800a; 1:500, Abgent, San Diego, CA, USA), mouse monoclonal anti-p62/SQSTM1 (cat. no. sc-28359; 1:1,000, Santa Cruz Biotechnologies Inc., Santa Cruz, CA, USA), for 2 h at room temperature. After washing with TBS-T (Bio Basic Inc., Markham, ON, Canada), the membrane was incubated with a goat anti-mouse IgG (cat. no. 31430; 1:4,000) and goat-anti-rabbit IgG (cat. no. 31460; 1:3,000) secondary antibodies (Thermo Fisher Scientific, Inc. The membrane was washed and protein was detected using enhanced chemiluminescence reagent (Cell Signaling Technology, Beverly, MA, USA) and XAR film (Eastman Kodak Company, Rochester, NY, USA) according to the manufacturer's instructions. The membranes were then stripped and probed with an anti-GAPDH mouse monoclonal antibody (1:3,000, Santa Cruz Biotechnology, Inc., Dallas, TX, USA) to confirm equal loading of the samples. The western blot bands were scanned and were analyzed using the Quantity One 4.6.7 program (Bio-Rad Laboratories).

In brief, paraffin-embedded OTSCC specimens were cut into 4-µm sections and incubated at 65°C for 30 min. The sections were washed with xylene and rehydrated. Sections were submerged for 2 min into an EDTA buffer at 95°C and 90 kPa for antigen retrieval and treated with 3% hydrogen peroxide in methanol, followed by incubation with 1% rabbit serum albumin (Zhongshan Golden Bridge Biotechnology, Co., Ltd.). The specimens were incubated overnight at 4°C with anti-Beclin-1 (1:200, Epitomics Inc.). For the negative control, the primary antibody was replaced with the non-immune rabbit IgG (Zhongshan Golden Bridge Biotechnology, Co., Ltd.). of the same isotype. Immunohistochemical staining to determine the expression of Beclin-1 revealed diffuse, fine, granular cytoplasmic staining under an XSZ-H microscope (Chongqing Optical & Electrical Instrument Co., Ltd., Chongqing, China). The proportion of cells expressing Beclin-1 was scored as follows: 0, no expression; 1, 0–25%; 2, 26–50%; and 3, >50%. The intensity of cell staining was scored as either 0 (no expression), 1 (yellow) or 2 (brown). The intensity score and proportion score were added to yield the total score, which was divided into no/low expression (0 to 2) or high expression (3 to 5).

The NTEC1 normal tongue epithelial cell line was established by culturing normal tongue squamous epithelium from a healthy patient in keratinocyte/serum-free medium (Invitrogen, Thermo Fisher Scientific, Inc.). This cell line was a gift from Professor Musheng Zeng (Sun Yat-sen University Cancer Centre, Guangzhou, China). CAL-27, SCC-25 and SCC-9 OTSCC cell lines were purchased from the American Type Culture Collection (Manassas, VA, USA) and Tca8113 and TSCCA cells were obtained from the Committee of the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). With the exception of CAL-27, which was cultured in Dulbecco's modified Eagle's medium (Thermo Scientific, Inc.) supplemented with 10% foetal bovine serum (FBS, Thermo Fisher Scientific, Inc.), all cell lines were cultured in RPMI-1640 (Thermo Scientific, Inc.) supplemented with 5% FBS. All cells were cultured in penicillin (100 U/ml) and streptomycin (100 U/ml) at 37°C in a humidified 5% CO₂ incubator.

For the Beclin-1 expression plasmid, the full-length coding region of Beclin-1 cDNA was amplified by RT-qPCR, and the digested and purified PCR products were directly cloned into a lentivirus vehicle sinhcmv-mcs (r)-pre-cppt-IG (a gift from Professor Erik Zheng, Sun Yat-sen University Cancer Centre). The same method was used to construct a full-length LC3 expression plasmid using a pEGF vector (Invitrogen, Thermo Fisher Scientific, Inc.) to generate pEGF-C2-LC3. The primer sequences of the full-length coding region were as follows: Sense: 5′-CGCGGATCCATGGAAGGGTCTAAGACGT-3′ and antisense: 5′-CGGAATTCTCATTTGTTATAAAATTGTGAfor Beclin-1′; and sense: 5′-GAAGATCTTCATGCCCTCAGACCGGCC-3′ and antisense: 5′-CGGAATTCTCAGAAGCCGAAGGTTTCCT-3′ for LC3. All plasmids expressed enhanced green fluorescent protein (GFP) and were verified by sequencing. For the Beclin-1 overexpression plasmid, stably transfected cell lines were generated by lentivirus infection and selected using flow cytometry according to the manufacturer's instructions. The LC3 expression plasmid was transfected into OTSCC cell lines (TCA8113 and TSCCA) using Lipofectamine 2000 (Invitrogen, Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions.

The cell growth and viability of the stable OTSCC cell lines (TCA8113 and TSCCA cells) were assessed by using an MTT assay. Approximately 3,000 cells/well were seeded in a 96-well plate for 6 duplicate wells with different drug concentrations. The MTT assay was performed on days 1–5. After incubation, 20 µl of 5 mg/ml MTT (Sigma-Aldrich, St. Louis, MO, USA) was added to each well, followed by incubation for 4 h. After discarding the medium carefully, formazan crystals were dissolved in 150 µl dimethyl sulfoxide. The optical density was measured with a microplate spectrophotometer (Bio-Rad Laboratories) at 490 nm. Absorbance values were normalised to the percentage survival. Each experiment was performed in triplicate.

For analysis of GFP LC3 puncta in cells overexpressing Beclin-1 and control cells, the cells were observed under a fluorescence microscope (×400) and representative cells were selected and photographed. The number of GFP-LC3 puncta per cell was quantified, and presented as the mean ± standard error of the mean from 50 randomly selected cells.

Cells (200/well; TCA8113 and TSCCA) were counted and plated in triplicate in six-well plates and were cultured with RPMI-1640 culture medium for ~10 days until the majority of the colonies had expanded to >50 cells. The plates were triple washed with cold phosphate-buffered saline, fixed in pre-chilled methanol for 10 min, and dyed with 0.5% crystal violet for 5 min at room temperature. After washing to remove excess stain, the plates were photographed. Images were analysed using Image J 1.42 software (National Institutes of Health, Bethesda, MD, USA), according to the manufacturer's instructions. At least three independent experiments were performed for each assay.

For the bottom layer, six-well plates were coated with a 0.6% agar in medium supplemented with 20% FBS. For the top layer, 5,000 cells were prepared in 0.3% agar with 10% FBS and seeded in triplicate. The plates were then incubated at 37°C in a humidified 5% CO₂ incubator for two weeks until colonies had formed. Each experiment was repeated at least three times. Colonies were photographed (final magnification ×200) under a phase-contrast microscope, and colonies >50 mm in diameter were counted under a light microscope.

All statistical analyses were performed using SPSS 16.0 statistical software package (SPSS Inc., Chicago, IL, USA). The χ² test for proportions was used to analyse the correlation between Beclin-1 expression and the clinicopathologic features of OTSCC. Survival curves were plotted using the Kaplan-Meier method and compared using the log-rank test. P<0.05 was considered to indicate a statistically significant difference.

The mRNA and protein expression of Beclin-1 were decreased in 78.6% of OTSCC tissues and cell lines relative to the matched adjacent noncancerous tissues and the NTEC1 normal tongue epithelial cell line, respectively (Fig. 1).

To further verify the expression of Beclin-1, an immunohistochemical assay was performed on 133 OTSCC specimens. As shown in Fig. 2, Beclin-1 protein was expressed in the cytoplasm. The decrease in Beclin-1 expression was closely correlated with poor differentiation, lymph node metastasis, and advanced clinical tumour-node-metastasis (TNM) stage of OTSCC (Table I). Notably, the five-year overall survival rate in the patients with no or low Beclin-1 expression was 47.5%, significantly shorter than that of the patients with high Beclin-1 expression (70.8%) (Fig. 3, P<0.01).

Beclin-1 is a pivotal gene in the process of autophagy. Therefore, the influence of Beclin-1 overexpression on autophagy upon starvation of OTSCC cells was investigated. As normal culture medium contains serum which may influence autophagy level, the cells were cultured in Earle's balanced salts (EBSS) to observe autophagy level. No significant difference in the expression levels of LC3-II and p62 was identified between the Beclin-1 overexpression cell line and the vehicle cell line subjected to starvation by culturing in EBSS for 1, 2, 4 or 8 h. The number of GFP-LC3 puncta was also counted in at least 50 cells after transient transfection of pEGF-C2-LC3, and no significant difference was observed between the cells overexpressing Beclin-1 and the vehicle group (12.66±0.32 vs. 11.98±0.34; P=0.7996) (Fig. 4). To determine whether Beclin-1 inhibited OTSCC cell proliferation, an MTT assay was performed, and the results suggested that Beclin-1 overexpression significantly inhibited the proliferation of OTSCC cell lines (Fig. 5A).

To further investigate whether Beclin-1 can inhibit the transformation of OTSCC cells, colony formation and soft-agar anchorage-independent growth assays were performed. In total, 200 cells were plated in triplicate wells of six-well plates. After 10 days of culture, the number of colonies formed by Beclin-1-expressing cells was significantly lower than that formed by negative-control cells or vehicle-control groups (Fig. 5B). Furthermore, Beclin-1-expressing cells formed significantly smaller and fewer colonies compared with the vector cells in the soft-agar assay (Fig. 5C).