A total of 60 patients with TSCC who underwent surgical treatment in The Second Affiliated Hospital of Harbin Medical University between January 2014 and December 2016 were included in the present study. The inclusion criteria were: i) Patients with a clinically and histopathologically confirmed diagnosis of TSCC; ii) received surgical resection as initial treatment; iii) archived tumor tissue and paired adjacent tissues that were stored at the Pathological Department of the hospital; and iv) clinicopathological and follow-up data were complete and accessible. Patients were excluded if they underwent radiotherapy or chemotherapy prior to surgery. The present study was approved by the Institutional Review Board of The Second Affiliated Hospital of Harbin Medical University, and written informed consent was obtained from all patients or their guardians.

The demographic data and clinicopathological characteristics of patients, including age, gender, pathological grade, T stage, N stage, and tumor-node-metastasis (TNM) stage, which was evaluated according to the criteria of the 7th edition American Joint Committee on Cancer, were obtained from medical records. The survival data, which were used to calculate OS, were obtained from follow-up records (last follow-up date was 30/06/2018). The OS was defined as the duration from surgical treatment to patient mortality.

Immunohistochemistry and reverse transcription-quantitative polymerase chain reaction (RT-qPCR) were performed to detect ITGA7 expression levels; immuno- histochemistry was also conducted to detect CD133 expression levels in TSCC and paired adjacent tissues.

TSCC and paired adjacent tissue specimens from enrolled patients were acquired from the Pathological Department of the aforementioned hospital; the samples were paraffin-embedded. Then, immunohistochemistry was performed to assess ITGA7 and CD133 expression in tumor and paired adjacent normal tissues. Briefly, paraffin-embedded tissue sections (4-μm thick) were deparaffinized with xylene and rehydrated with ethanol. Heat-induced antigen retrieval was subsequently conducted in 0.01 mol/l sodium citrate buffer (pH 6.0) using a microwave, and endogenous peroxidase activity was inhibited with freshly prepared 3% H₂O₂. Following blocking using 1.5% normal goat serum (Shanghai Yeasen Biotechnology Co., Ltd.) at 37°C for 20 min, the sections were incubated at 4°C overnight with rabbit polyclonal ITGA7 antibody (1:200; cat. no. ab203254; Abcam) and rabbit polyclonal CD133 antibody (1:100; cat. no. ab19898; Abcam). The next day, tissue sections were incubated with horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin G antibody (diluted 1:1,000 in 3% bovine serum albumin; cat. no. ab6721; Abcam) at 37°C for 90 min. Subsequently, 3,3′-diaminobenzidine and hematoxylin were applied for staining of the sections, which were then sealed with neutral tree gum. ITGA7 and CD133 expression was evaluated under a light microscope.

ITGA7 and CD133 expression in TSCC and paired adjacent tissues was assessed by scores based on the average intensity and percentage of positively stained cells, as described previously (14). The intensity scores were graded as follows: 0, no staining; 1, weak staining, light yellow; 2, moderate staining, yellow brown; and 3, strong staining, brown. The percentage of stained cells was scored as: 0, 0%; 1, 1-25%; 2, 26-50%; 3, 51-75%; and 4, 76-100% positive cells. The total score was calculated by multiplying the density score and the percentage score; high expression was defined as a total score of ≥3, while low expression was defined as a total score of <3.

Four human TSCC cell lines and one normal human oral keratinocyte cell line were purchased through a third agent company (Shanghai QeeJen Bio-Tech Co., Ltd). In detail, the CAL-27 cell line was purchased from Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, the SCC-9 cell line was purchased from American Type Culture Collection, the HSC-4 cell line was purchased from Japanese Collection of Research Bioresources Cell Bank, the SCC-25 cell line was purchased from Cell Bank of Type Culture Collection of Chinese Academy of Sciences, and the HOK cell line was purchased from ScienCell Research Laboratories, Inc.

The CAL-27 and SCC-9 cell lines were cultured in 90% RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Inc.) with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.). The HSC-4 cell line was cultured in 90% Eagle's minimal essential medium (MEM; Gibco; Thermo Fisher Scientific, Inc.) with 10% FBS. The SCC-25 cell line was cultured in 90% Dulbecco's modified Eagle's medium/Ham's F12 medium (DMEM/F12; Gibco; Thermo Fisher Scientific, Inc.) with 10% FBS. The HOK cell line was cultured in 90% oral keratinocyte medium (ScienCell Research Laboratories, Inc.) with 10% FBS. All the cells were cultured in incubators at 37°C under 95% air and 5% CO₂ conditions.

qPCR, western blotting and immunofluorescence were used for the detection of the mRNA and protein expression levels of ITGA7, and western blotting and immunofluorescence were used for the detection of the expression levels of CD133, in the human TSCC cell lines CAL-27, SCC-9, HSC-4 and SCC-25 and the normal human oral keratinocyte cell line HOK (as a normal control).

Control short hairpin RNA (shRNA) and ITGA7-targeting shRNA shuttle plasmids were constructed using the pGLV-U6 vector (Shanghai GenePharma Co., Ltd.), and transfected into 293T cells (American Type Culture Collection) together with envelope plasmids (Shanghai GenePharma Co., Ltd.) and packaging plasmids (Shanghai GenePharma Co., Ltd.) using HilyMax (Dojindo Molecular Technologies, Inc.). The cell supernatant was obtained at 48 and 72 h post-transfection. Following purification and concentrating, the corresponding lentiviruses were collected. The sequences used for the knockdown experiments were as follows: ITGA7 shRNA, forward 5′-CACCGCTGCCCACTCTACAGCTTTTCGAAAAAAGCTGTAGAGTGGGCAGC-3′ and reverse, 5′-AAAAGCTGCCCACTCTACAGCTTTTTTCGAAAAGCTGTAGAGTGGGCAGC-3′; control shRNA, forward 5′-CACCGTTCTCCGAACGTGTCACGTCGAAACGTGACACGTTCGGAGAA-3′ and reverse 5′-AAAATTCTCCGAACGTGTCACGTTTCGACGTGACACGTTCGGAGAAC-3′. Subsequently, using a multiplicity of infection of 10, control shRNA and ITGA7 shRNA lentiviruses were added to the medium for the transduction of CAL-27 and HSC-4 cells with 6 μg/ml polybrene (Sigma-Aldrich; Merck KGaA) for 24 h, followed by the addition of fresh complete medium for another 48 h. Then, cells were cultured with 8 μg/ml puromycin (Thermo Fisher Scientific, Inc.) for 7 days to construct stably transduced CAL-27 and HSC-4 cells. Cells transduced with the control shRNA lentivirus were termed as the negative control (NC) group, while the cells transduced with the ITGA7shRNA lentivirus were termed as the ITGA7(-) group.

Following the construction of stably transduced CAL-27 and HSC-4 cells, ITGA7 expression was detected by RT-qPCR, western blotting and flow cytometry in the NC and ITGA7(-) groups, in order to determine successful transduction.

In the NC and ITGA7(-) groups of stably transduced CAL-27 and HSC-4 cells, the apoptotic rate was detected using a fluorescein isothiocyanate (FITC) Annexin-V Apoptosis Detection kit II (BD Biosciences). Expression levels of the cell apoptotic markers cleaved (C)-Caspase 3 and Bcl-2 were detected via western blotting. In addition, an equal quantity of cells of the NC and ITGA7(-) groups were cultured for 72 h, and the proliferative ability was detected at 0, 24, 48 and 72 h via a Cell Counting Kit-8 (CCK-8) assay (Sigma-Aldrich; Merck KGaA).

In the NC and ITGA7(-) groups of stably transduced CAL-27 and HSC-4 cells, the CSC markers cluster of differentiation (CD)24, CD44 and CD133 were detected by RT-qPCR and western blotting. In addition, ITGA-positive cells, as well as ITGA7-negative cells, were isolated by flow cytometry, and the protein expression levels of ITGA7, CD24, CD44 and CD133 were detected by western blotting.

Cisplatin (5 μM; Sigma-Aldrich; Merck KGaA) was applied to the NC and ITGA7(-) groups of stably transduced CAL-27 and HSC-4 cells for 24 h. Then, the drug resistance ability was assessed by detecting cell viability and apoptosis. Briefly, cell viability was detected via a CCK-8 assay and the relative cell viability of the ITGA7 (-) group was calculated according to the reference values of the NC group. The rate of cell apoptosis was determined using an FITC Annexin-V Apoptosis Detection Kit II.

The sphere formation ability of the NC and ITGA7(-) groups of stably transduced CAL-27 and HSC-4 cells was analyzed by a sphere formation assay. Briefly, transduced CAL-27 and HSC-4 cells were cultured in DMEM/F12 medium supplemented with 2% B27 (Gibco; Thermo Fisher Scientific, Inc.), 20 ng/ml epidermal growth factor (EGF; Sigma-Aldrich; Merck KGaA), 20 ng/ml basic fibroblast growth factor (bFGF; Gibco; Thermo Fisher Scientific, Inc.) and 4 μg/ml heparin (Sigma-Aldrich; Merck KGaA) for 10 days; spheres of a diameter >50 μm were counted under a light microscope (Olympus Corporation). The sphere formation ability was calculated by dividing the number of these spheres by the total number of seeded cells (200) ×1,000. In addition, the sphere formation assay was conducted using the extreme limiting dilution method. CAL-27 and HSC-4 cells were cultured in DMEM/F12 medium supplemented with 2% B27, 20 ng/ml EGF (Sigma-Aldrich; Merck KGaA), 20 ng/ml bFGF and 4 μg/ml heparin (Sigma-Aldrich; Merck KGaA) for 10 days in a 24 well-plate at densities of 1,000, 100 or 10 cells per well. Subsequently, spheres with a diameter >50 μm were counted under a light microscope (Olympus Corporation) and the sphere formation ability was calculated using an extreme limiting dilution analysis software (http://bioinf.wehi.edu.au/software/elda/) (15).

CAL-27 and HSC-4 CSCs were generated by establishing drug-resistant cells, followed by detection with a sphere formation assay (16). Briefly, 5 μM cisplatin was applied to CAL-27 and HSC-4 cells for 72 h, and then cisplatin-free medium was applied to cells for another 72 h; these processes were repeated until no effect of cisplatin on cell viability/proliferation was observed by the CCK-8 assay. A sphere formation assay was performed as aforementioned, and spheres of CAL-27 and HSC-4 CSCs were isolated by centrifugation (300 × g at room temperature for 3 min). To determine the successful establishment of CSCs, the expression levels of CSC markers CD24, CD44 and CD133, as well as ITGA7, were detected by RT-qPCR and western blotting in CAL-27/HSC-4 CSCs and normal parental CAL-27/HSC-4 cells.

TRIzol® reagent (Thermo Fisher Scientific, Inc.) was used for the extraction of total RNA. The ReverTra Ace® qPCR RT kit (Toyobo Life Science) was used to transcribe 1 μg RNA into cDNA. SYBR® Green Realtime PCR Master Mix (Toyobo Life Science) was used for qPCR, which was performed as follows: 95°C for 5 min, followed by 40 cycles of 95°C for 5 sec and 61°C for 30 sec. The sequences of primers employed for qPCR are listed in Table I. Relative fold changes in mRNA expression were calculated using the 2^–ΔΔCq method (17,18). GAPDH was used as the internal reference.

Radioimmunoprecipitation assay buffer (Thermo Fisher Scientific, Inc.) was used for the extraction of total protein, and a Pierce™ BCA Protein Assay kit (Thermo Fisher Scientific, Inc.) was used to determine protein concentration; a standard curve was used for adjustments. The protein sample (20 μg) was fractionated via a NuPAGE™ 4-12% Bis-Tris Protein Gel (Invitrogen; Thermo Fisher Scientific, Inc.), and then transferred to nitrocellulose filter membranes (EMD Millipore). Subsequently, the membranes were blocked using 5% skim milk for 2 h at 37°C, and were incubated with primary antibodies overnight at 4°C. After the secondary antibody was applied for 1 h at room temperature, Pierce™ Enhanced Chemiluminescence Plus Western Blotting Substrate (Thermo Fisher Scientific, Inc.) was used for the visualization of bands, which were then exposed using X-ray film (Kodak). GAPDH was used as the internal reference. The detailed information regarding the antibodies employed for western blot analysis are listed in Table II.

Cells were fixed with 4% paraformaldehyde for 10 min at room temperature, permeabilized with 0.2% Triton X-100 for 5 min and blocked with PBS containing 2% BSA for 1 h at room temperature. Cells were then incubated with primary antibodies targeting ITGA7 (1:200; cast. no. ab203254; Abcam) and anti-CD133 (1:100; cat. no. ab19898; Abcam) overnight at 4°C. Following three washes with PBS, cells were incubated with an FITC-conjugated goat anti-rabbit immunoglobulin G antibody (1:1,000; cat. no. ab6717; Abcam) for 1 h at 37°C. The nuclei were counterstained with DAPI. Images were obtained using a BX41 fluorescence microscope (Olympus Corporation) at x400 magnification.

Cells were harvested and washed with PBS, and then incubated with rabbit polyclonal antibody against ITGA7 (1:50; cat. no. ab203254; Abcam) in incubation buffer (PBS) for 1 h at 4°C. Then, cells were washed with PBS and incubated with a FITC-conjugated goat anti-rabbit immunoglobulin G antibody (1:1,000; cat. no. ab6717; Abcam) for 1 h at 4°C in the dark. Subsequently, the cells were washed again, resuspended in FACS buffer (PBS) and analyzed by flow cytometry (Flowjo Vision 7.6; BD Biosciences) using a FACS Canto II flow cytometer (BD Biosciences).

Statistical analysis was performed using SPSS 22.0 software (IBM Corp.) and GraphPad Prism 7.00 (GraphPad Software Inc.). Normal distributed continuous variables were presented as the mean + standard deviation, and categorized variables were presented as a percentage. Comparisons of ITGA7 expression between tumor and paired adjacent tissues were conducted via a McNemar test. Associations between ITGA7 expression and patient characteristics were determined by a χ² or Wilcoxon rank sum test. Differences between the OS of patients with low and high ITGA7 expression were determined via Kaplan-Meier analysis, followed a log-rank test. Univariate and multivariate Cox proportional hazards regression analyses were performed to determine the factors affecting OS. Comparisons among groups in cell experiments were conducted with one-way ANOVA followed by Dunnett's multiple comparisons test. Comparisons between two groups in cell experiments were performed with unpaired parametric t-test. P<0.05 was considered to indicate a statistically significant difference.

A total of 60 patients with TSCC were enrolled with a mean age of 55.2+10.5 years, comprising 44 males and 16 females (Table III). Regarding pathological grade, the number of patients with TSCC in grade (G)1, G2 and G3 were 8 (13.3%), 42 (70.0%) and 10 (16.7%), respectively. There were 10 (16.7%), 24 (40.0%), 24 (40.0%) and 2 (3.3%) patients with TSCC of TNM stage I, II, III and IV, respectively. The detailed information of other clinicopathological characteristics of the patients is presented in Table III.

Immunohistochemistry was conducted for the detection of ITGA7 expression in tumor and paired adjacent tissues; representative images of high and low ITGA7 staining are presented in Fig. 1A. The results from the immunohistochemistry analysis demonstrated that the protein expression levels of ITGA7 were increased in the tumor tissues compared with paired adjacent tissues (P=0.019; Fig. 1B). Similar results were obtained by RT-qPCR analysis for the ITGA7 mRNA expression levels (P<0.001; Fig. 1C). In addition, the expression of the CSC marker CD133 was detected by immunohistochemistry in the patient tissues (Fig. S1A). The results revealed that CD133 expression was upregulated in tumor tissues compared with adjacent normal tissues (P=0.048; Fig. S1B).

Increased protein expression of ITGA7 was associated with advanced pathological grade (P=0.021), and higher N (P=0.011) and TNM (P=0.005) stages. However, no association between ITGA7 protein expression and age (P=0.465 and 0.876) or gender (P= 0.876) was observed in patients with TSCC (Table IV). In addition, high ITGA7 mRNA expression was associated with advanced pathological grade (P=0.003), increased T stage (P=0.002) and higher TNM stage (P=0.010) in patients with TSCC (Table IV).

Kaplan-Meier analysis revealed that the OS was shorter in patients with high ITGA7 protein expression levels [mean OS: 36.2 months, 95% confidence interval (CI): 31.24-41.1 months] compared with those possessing low ITGA7 protein expression (mean OS: 48.1 months, 95% CI: 44.9-51.3 months; P=0.013; Fig. 2A). In addition, a similar trend was reported for patients with high ITGA7 mRNA expression (mean OS: 37.2 months, 95% CI: 31.6-42.9 months) compared with those with low ITGA7 mRNA expression (mean OS: 48.2 months, 95% CI: 45.3-51.3 months; P=0.007; Fig. 2B).

Univariate Cox proportional hazards regression model analysis was used to evaluate the effect of all clinicopathological factors on the OS of patients with TSCC. The results revealed that high ITGA7 protein (P=0.024) and mRNA expression (P=0.015), higher pathological grade (P= 0.049), advanced T stage (P=0.012), higher N stage (P=0.001), as well as advanced TNM stage (P<0.001) were associated with poorer OS in patients with TSCC (Table V). However, when multivariate Cox analysis was performed, only TNM stage could independently predict OS, while high ITGA7 expression (protein or mRNA) was not an independent factor for predicting shorter OS in patients with TSCC. This suggested that ITGA7 may be indirectly associated with poor prognosis by affecting tumor size, pathological grade or lymph node metastasis in patients with TSCC.

In order to investigate the function of ITGA7 in TSCC cells, in vitro experiments were performed. Firstly, the expression of ITGA7 was detected in several established TSCC cell lines and a normal human oral keratinocyte cell line. Compared to the normal HOK cells, both ITGA7 mRNA (Fig. 3A) and protein (Fig. 3B) expression levels were increased in the human TSCC cell lines CAL-27, SCC-9, HSC-4 and SCC-25.

In order to investigate the underlying mechanism of ITGA7 in CAL-27 and HSC-4 cells, control NC shRNA and ITGA7 shRNA lentiviruses were constructed and used to transduce these cell lines, hence generating the NC and ITGA7(-) cell groups, respectively. In CAL-27 cells, the mRNA (P<0.001; Fig. 4A) and protein (Fig. 4B) expression levels of ITGA7 were down- regulated in the ITGA7(-) group compared with the NC group. Additionally, a similar trend of ITGA7 expression at the mRNA (P<0.001; Fig. 4C) and protein (Fig. 4D) levels was observed between the ITGA7(-) and NC groups of HSC-4 cells. These findings suggested the successful construction of stably transduced ITGA7-silenced TSCC cell lines. In addition, the results of flow cytometry demonstrated that the percentage of ITGA7+ cells was decreased in the ITGA7(-) group compared with the NC group, for both the CAL-27 and HSC-4 cell lines (P<0.01; Fig. S2A-D).

The present study investigated the effects of ITGA7 knockdown on the proliferation and apoptosis of CAL-27 and HSC-4 cells. A CCK-8 assay revealed that cell proliferation was decreased in the ITGA7(-) group compared with the NC group at 48 (P<0.05) and 72 h (P<0.01) for CAL-27 cells (Fig. 5A), and at 48 (P<0.05) and 72 h (P<0.05) for HSC-4 cells (Fig. 5E). The rate of cell apoptosis was increased in the ITGA7(-) group compared with the NC group for CAL-27 cells (P<0.01; Fig. 5B and C) and HSC-4 cells (P<0.05; Fig. 5F and G). Western blot analysis revealed that the expression of the apoptotic protein marker C-Caspase 3 was increased, but the expression of the anti-apoptotic Bcl-2 was decreased, in the ITGA7(-) group compared with the NC group for CAL-27 cells (Fig. 5D) and HSC-4 cells (Fig. 5H). These findings indicated that ITGA7 knockdown inhibited cell proliferation, but promoted apoptosis in CAL-27 and HSC-4 cells.

To investigate the effects of ITGA7 knockdown on the sternness of TSCC cells, the expression of the common CSC markers CD24, CD44 and CD133 was examined in CAL-27 and HSC-4 cells. RT-qPCR and western blotting demonstrated that the mRNA and protein expression of CD24 (P<0.001; Fig. 6A and D) were increased, while that of CD44 (P<0.01; Fig. 6B and D) and CD133 (P<0.01; Fig. 6C and D) were decreased in the ITGA7(-) group compared with the NC group of CAL-27 cells. Additionally, similar trends were observed in the expression of CD24 (P<0.05; Fig. 6E and H), CD44 (P<0.05; Fig. 6F and H) and CD133 (P<0.01; Fig. 6G and H) between the ITGA7(-)and NC groups of HSC-4 cells. These results suggested that ITGA7 knockdown regulated the expression of common CSC markers in CAL-27 and HSC-4 cells.

In addition, immunofluorescence experiments were performed to detect ITGA7 and CD133 expression in TSCC cell lines and the normal human oral keratinocytes cell line, and the results revealed that ITGA7 and CD133 expression levels were markedly increased in human TSCC cell lines CAL-27, SCC-9, HSC-4 and SCC-25 compared with the normal HOK cells (Fig. S3). Furthermore, ITGA7-positive cells and ITGA7-negative cells were sorted, and the protein expressions of ITGA7, CD24, CD44 and CD133 were detected in the separate cell populations by western blot analysis. The results demonstrated that IT GA7, CD44 and CD133 protein expressions were increased, while CD24 protein expression was decreased in ITGA7-positive cells compared with ITGA7-negative cells in both CAL-27 and HSC-4 cell lines (Fig. S4). These findings indicated that high ITGA7 expression was associated with high expression of CSC markers in TSCC cells.

To further explore the role of ITGA7 in the sternness of TSCC cells, the effect of ITGA7 knockdown on drug resistance to cisplatin was investigated in CAL-27 and HSC-4 cells. A CCK-8 assay revealed that the relative cell viability decreased in the cisplatin + ITGA7(-) group compared with the cisplatin + NC group for CAL-27 cells (P<0.01; Fig. 7A) and HSC-4 cells (P<0.05; Fig. 7D). In addition, the rate of apoptosis increased in the cisplatin + ITGA7(-) group compared with the cisplatin + NC group of CAL-27 cells (P<0.01; Fig. 7B and C) and HSC-4 cells (P<0.05; Fig. 7E and F). These results suggested that ITGA7 knockdown decreased drug resistance to cisplatin in these cell lines.

To further validate the effects of ITGA7 knockdown on the sternness of TSCC cells, sphere formation assays were performed in CAL-27 and HSC-4 cells. The results demonstrated that the sphere formation ability was reduced in the ITGA7(-) group compared with the NC group for CAL-27 cells (P<0.01; Fig. 8A and B) and HSC-4 cells (P<0.05; Fig. 8C and D). In addition, an extreme limited dilution assay revealed that the sphere formation ability was reduced in the ITGA7(-) group compared with the NC group for CAL-27 and HSC-4 cells (both P<0.001; Table SI). These data indicated that ITGA7 knockdown decreased the sphere formation ability of CAL-27 and HSC-4 cells. Therefore, the aforementioned findings that ITGA7 regulated the expression of CSC markers, decreased drug resistance to cisplatin and suppressed sphere formation in CAL-27 and HSC-4 cells, indicated that ITGA7 knockdown may suppress TSCC cell stemness.

In order to explore whether ITGA7 is a potential marker for TSCC stem cells, the expression of ITGA7 in drug-resistant (R-) CAL-27 and R-HSC-4 CSCs was detected, and sphere formation assays were performed. A CCK-8 assay demonstrated no notable differences between the R-CAL-27 + cisplatin (Fig. 9A) and R-HSC-4 + cisplatin groups (Fig. 9I) at 0, 24, 48 and 72 h (P>0.05), which indicated the successful generation of drug-resistant cells. In addition, a sphere formation assay was performed for CAL-27 (Fig. 9B) and HSC-4 (Fig. 9J) cells. The number of spheres/1,000 cells was increased in the R-CAL-27 (P<0.01; Fig. 9C) and R-HSC-4 (P<0.01; Fig. 9K) cells compared with the corresponding parental control cell lines. Next, spheres from the R-CAL-27 and R-HSC-4 groups were isolated by centrifugation, serving as CAL-27 and HSC-4 CSCs. To further validate the establishment of CAL-27 and HSC-4 CSCs, the expression of common CSC markers was detected; the mRNA and protein expression levels of CD24 (Fig. 9D and H) were decreased, while those of CD44 (P<0.01; Fig. 9E and H) and CD133 (P<0.001; Fig. 9F and H) were increased in CAL-27 CSCs compared with parental CAL-27 cells. Additionally, similar trends in the expression of CD24 (P<0.01; Fig. 9L and P), CD44 (P<0.01; and Fig. 9M and P) and CD133 (P<0.001; Fig. 9N and P) were observed for HSC-4 CSCs compared with parental HSC-4 cells, which suggested that CAL-27 and HSC-4 CSCs were successfully obtained. Following the successful establishment of CAL-27 and HSC-4 CSCs, the expression of ITGA7 was detected in the TSCC stem cells. The results revealed that the mRNA and protein expression levels of ITGA7 were increased in CAL-27 CSCs compared with parental CAL-27 cells (P<0.001; Fig. 9G and H), and in HSC-4 CSCs compared with parental HSC-4 cells (P<0.01; Fig. 9O and P). These findings indicated that ITGA7 may be a potential marker of TSCC stem cells.