A total of 51 HSCC biopsy samples with complete clinical and pathological data were obtained from patients who underwent operations from 2009 to 2015 in the Affiliated Hospital of Nantong University. Sixteen non-cancer HSCC tissue samples were recruited as controls. Prior to surgery, none of the patients received adjuvant chemotherapy or radiotherapy. The pathological analysis was carried out according to the TNM system of the American Joint Committee on Cancer (AJCC 2010). All the detailed information of these 51 HSCC patients and clinicopathological parameters in this series are shown in Table I. Furthermore, 7 HSCC and corresponding 7 para-carcinoma tissues were collected to confirm the expression of STMN1 protein. This study was approved by the Ethics Committee of the Affiliated Hospital of Nantong University and all the participants gave their informed consent.

Immunohistochemical staining was performed on 4-µm paraffin tissue sections mounted on slides and dried at 60°C for 8 h. The slides were deparaffinized in xylene and dehydrated conventionally, then pressure-cooked in sodium citrate buffer (Sangon Biotech, Shanghai, China) (pH 6.0) to facilitate antigen retrieval. After natural cooling, endogenous peroxidase was blocked with 3% hydrogen peroxide (ZSGB-BIO, Beijing, China). Non-specific bindings were blocked by treating slides with normal goat serum (ZSGB-BIO) for 40 min, the sections were subsequently incubated with rabbit anti-STMN1 polyclonal antibody (Abcam, Cambridge, MA, USA, dilution 1:250) or rabbit anti-Ki-67 polyclonal antibody (dilution 1:100, sc-15402; Santa Cruz Biotechnology, Inc. Santa Cruz, CA, USA), overnight at 4°C. After being washed with PBS, the sections were incubated for 30 min with the 2-step plus poly-HRP anti-mouse/rabbit IgG detection system (ZSGB-BIO). DAB detection kit (enhanced polymer) (ZSGB-BIO) was used for 3 min to show immunolabeling, resulting in a brown precipitate. Following the above, sections were counterstained with hematoxylin. Positive and negative controls were performed in parallel.

The immunostaining results were blindly evaluated by two experienced pathologists without knowledge of the clinicopathologic outcomes of these patients. Immunostaining for STMN1 was localized in the cytoplasm and, partly, in the nucleus, and immunostaining for Ki-67 was limited to the nucleus. The proportion of epithelial cells positively-stained was divided into two groups: category A scores from 1 to 4 [(A = 0 (0–4%); 1 (5–25%); 2 (26–50%); 3 (51–75%); 4 (76–100%)]. Intensity of staining was category B, scores 0 (negative); 1 (weak); 2 (moderate); 3 (strong). For the IHC-score assessment, all fields were observed at ×400 magnification. A final score was calculated by multiplying A by B (minimum 0, maximum 12). Moreover, the average score of each case was obtained. Total scores were devided into 4 grades: negative, 0–3; weak, 3–6 (contain 3, 6); moderate, 6–9; strong, 9–12 (contain 9).

Human HSCC cell line, FaDu cells, was purchased from ATCC, Shanghai, China. FaDu cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum at 37°C in a humid atmosphere containing 5% CO₂. To silence the expression of STMN1 (accession no. NM 203401 form NCBI GenBank), FaDu cells were transfected with four siRNAs which were designed and obtained from Biomics Biotechnologies Co., Ltd. (Nantong, China). The sequence alignment (BLAST) was carried out to ensure no homology with other genes in human. Furthermore, the negative control siRNA (si-NC) was also designed. The targeted sequences are shown in Table II.

FaDu cells were cultured in 24-well chamber slides, washed three times with cold 1X PBS, fixed with 4% paraformaldehyde for 1 h at 4°C, washed three times in PBS, permeabilized with 0.1% Triton X-100 for 10 min (except for staining the cell membrane-bound E-cadherin), and blocked with Immunol Staining Blocking Buffer (Beyotime, China) for 1 h at room temperature. Cells were then incubated overnight at 4°C with anti-STMN1 (Abcam, 1:100 dilution), anti-E-cadherin (Ruiying Biological, Suzhou, China, 1:100 dilution) or anti-vimentin (Ruiying Biological, 1:100 dilution) primary antibody and subsequently with Alexa Fluor-conjugated secondary antibodies (1:1,000; Invitrogen Life Technologies, Carlsbad, CA, USA) and Hoechst (Sigma-Aldrich Co., St. Louis, MO, USA) for 2 h at room temperature. The immunofluorescence images were acquired with a fluorescence microscope.

The mRNA expression of STMN1, E-cadherin, vimentin in FaDu cells after STMN1-siRNA transfection were quantified by real-time RT-PCR. According to the manufacturer's instructions, total RNA was extracted using a TRIzol^® reagent (Vazyme, USA) and 2 µg RNA was reverse-transcribed into cDNA samples by using a Transcriptor First Strand cDNA Synthesis kit (Roche, Germany). Then AceQ^® qPCR SYBR Green Master Mix kit (Vazyme) was used for RT-qPCR. The primers used were obtained from Biomics Biotechnologies Co. Ltd. and are shown in Table III. GAPDH was used as the reference gene. The PCR conditions consisted of 10 min at 95°C 1 cycle, 15 sec at 95°C, 30 sec at 60°C and 30 sec at 72°C 40 cycles. The experiment was performed in triplicate. The 2^−ΔΔCT method was implemented to analyze the results (10).

The cells or tissue samples were lysed in lysis buffer (PMSF:RIPA = 1:100, Beyotime) and then centrifuged at 10,000 rpm for 10 min at 4°C. Protein concentrations were measured with BCA protein assay kit (Pierce, USA). The supernatant was diluted in 5X SDS-polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer [50 mM Tris-HCl (pH 6.8), 10% (w/v) SDS, 0.5% (w/v) bromophenol blue, 50% (w/v) glycerol, 5% (w/v) β-mercaptoethanol] and boiled. Protein samples (20 µg) were separated by 10% (E-cadherin), 12% (vimentin, β-actin) and 15% (STMN1) polyacrylamide gels and electrophoresed, then electroblotted onto polyvinylidine difluoride filter (PVDF) membranes (Millipore, USA), followed by blocking with 5% skim milk in TBST (20 mM Tris, 150 mM NaCl, 0.05% Tween-20, pH 7.5) for 2 h at room temperature. The membranes were incubated with primary antibody overnight at 4°C (anti-STMN1: 1:5,000, Abcam), anti-E-cadherin: 1:300, anti-vimentin: 1:300 (Ruiying Biological), anti-β-actin: 1:2500 (Santa Cruz). After being washed three times with TBST, horseradish peroxidase (HRP)-conjugated secondary antibodies (goat anti-rabbit IgG-HRP with 1:1,000 dilution for STMN1, E-cadherin and vimentin; goat anti-mouse IgG-HRP with 1:2,000 dilution for β-actin) were used for 1.5 h at room temperature. After three additional washes with TBST for 45 min. The immunoreactive bands were detected with ECL plus kit (ZSbio, Beijing, China).

FaDu cells (5×10³ cells/well) were plated into a 96-well plate containing complete medium and incubated for 24 h before transfection. After treatment with siRNAs for 0, 12, 24, 48 and 72 h, 10 µl Cell Counting Kit-8 (CCK-8) kit solution (Beyotime) was added to each well containing original medium in a final volume of 100 µl and the plate was incubated for 1.5 h in an incubator at 37°C. Then the optical density (OD) was measured at 450 nm by using a Fluorescence Spectrophotometer (Hitachi, Japan). Three wells were used for each experimental condition and all the experiments were independently repeated three times.

For cell cycle analysis, FaDu cells were harvested at 48 h post-transfection, 6-cm plates. Cells were collected using cold PBS and fixed with 70% cold ethanol over 24 h in −20°C and treated with 1 ng/ml RNaseA for 20 min at 4°C. Cellular DNA was stained with 0.5 mg/ml propidium iodide (PI, 50 mg/ml; Becton-Dickinson, San Jose, CA, USA) for 20 min at 4°C in the dark. The cells were then sorted by FACSCalibur Flow Cytometer (Becton-Dickinson, San Jose, CA, USA) and CellQuest acquisition and analysis programs (Becton-Dickinson, San Diego, CA, USA). The experiments were replicated three times.

A Transwell chamber was used (8 µm, 24-well format; BD Biosciences) to perform cell migration assays. For the migration assay, 5×10⁴ cells in 500 µl serum-free media were added to the upper chamber after transfection for 48 h, while medium containing 10% FBS was added to the lower chambers to act as a chemoattractant. The upper surfaces of the chamber were coated with a growth factor-reduced Matrigel matrix. Cells were incubated at 37°C with 5% CO₂ for 24 h, and the medium was removed from the upper chamber. The non-invaded cells on the upper side of the chamber were gently scraped off with a cotton swab, whereas the cells that had migrated through the membrane were stained with 4% paraformaldehyde and 0.1% crystal violet, imaged, and counted using an inverted microscope and analyzed by visualizing five random fields at a magnification of ×200. Each experiment was replicated three times.

Morphological changes of the FaDu cells were observed using an inverted microscope. The image was taken using a Leica microscope image system (Leica, Mannheim, Germany) at a magnification of ×200.

Statistical analyses were performed using SPSS statistical software package, version 22.0. The STMN1 IHC scores of each subgroup of clinicopathological parameters in Table I are shown as mean ± standard deviation (SD). Statistical significance was determined using t-test. Kaplan-Meier and log-rank tests were used for calculating survival curves. In addition, statistical analysis was performed using GraphPad Prism 5.0 statistical software, sigmaplot 10.0 and Adobe Photoshop CC. Statistical significance was assigned for P<0.05.

We first sought to investigate the role of STMN1 in HSCC. Sixteen human normal cases and 51 HSCC cases were observed by us through immunohistochemistry staining. STMN1 protein was localized in the cytoplasm and, partly, in the nucleus in all tumor tissues, while negative in all of the normal tissues.

With respect to the staining intensity, normal tissue (NT), cancer in situ (CIS), and invasive cancer (IC) displayed negative, moderate and strong immunoreactivity for STMN1 protein, respectively (Fig. 1A). As showed in Fig. 1C and D, STMN1 IHC score was statistically increased in IC tissue (7.05±3.13) compared with that in the CIS (4.13±1.38, P=0.009) or normal tissue (0, P<0.001). Moreover, the correlations between the expression of STMN1 and clinicopathological characteristics of the patients with HSCC are summarized in Table I (Fig. 2). Interestingly, IHC staining showed that the intensity of STMN1 expression was much stronger in poorly differentiated than well-differentiated samples (P<0.001, Fig. 2A and B). In addition, overexpression of STMN1 was significantly correlated with advanced clinical stage (stage III and IV compared with stage I and II, P<0.001, Fig. 2C), large tumor size (size >2 cm compared with those ≤2 cm, P=0.001, Fig. 2D), lymph node metastasis (metastasis to the lymph nodes compared with non-metastasis, P=0.008, Fig. 2E) and treatment (P=0.01), but not correlated with age (P=0.430), gender (P=0.824), tobacco smoking (P=0.242), or alcohol consumption (P=0.242).

To further verify the results of immunohistochemistry staining, STMN1 expression at protein levels in 7 HSCC tumor tissues (T1–T7) and their adjacent tissues (N1–N7) were detected using western blot analysis. Results showed that, compared with their adjacent tissues, the expression of STMN1 in HSCC tumor was significantly higher (Fig. 2F and H). Besides, the basic expression and distribution of STMN1 in FaDu cells were examined at protein levels. As is showed in Fig. 2G, the results of western blot analysis suggested that STMN1 is highly expressed in FaDu cells. Simultaneously, strong immunoreactivity for STMN1 protein in cytoplasm in FaDu cells was found by immunocytochemistry (Fig. 3D).

Furthermore, specimens with positive STMN1 staining showed significantly higher frequencies of Ki-67 positivity (Fig. 1A). Spearman correlation analysis indicates a positive correlation between STMN1 expression and Ki-67 based on proliferative activity (R²=0.74, P=0.000; Fig. 1B). Thus, these findings indicate that overexpression of STMN1 is likely to be involved in the progression of HSCC.

In addtion, in light of our results that displayed a multifaceted expression of STMN1 in HSCC patients with different malignancy grade, we investigated the prognostic significance of STMN1 in HSCC using Kaplan-Meier analysis. Indeed, increased expression of STMN1 was significantly associated with worse prognoses. The choice of treatment is usually associated with clinical stage and the presence of lymph node metastasis, thus the expression of STMN1 was different in patients under different therapy. STMN1 samples highly or moderately stained indicated shorter overall survival and progression-free survival rate than those with STMN1 weakly or negatively stained in the 51 HSCC specimens (P=0.0019, P<0.0001; Fig. 2I and J).

To investigate the effect of STMN1 on HSCC, four siRNAs were designed to knock down the STMN1 expression in FaDu cells. The mRNA and protein levels were determined in treated FaDu cells after 48 h. As shown in Fig. 3A, compared with the control (CTR) and si-NC, after treatments with FaDu cells for 72 h, STMN1 expression was inhibited by siRNAs at the mRNA level achieving 87% (STMN1-si1), and the protein level achieving 90% (STMN1-si1) (Fig. 3B and C). Furthermore, the results of immunofluorescence indicated that the staining intensity of STMN1 in FaDu cells significantly diminished after treatment with STMN1-si1 (Fig. 3D and E).

To assess the effect of knocking down STMN1 in proliferation of FaDu cells, CCK-8 and flow cytometry assays were implemented. Results of CCK-8 assay showed that the cell proliferation of FaDu could be greatly reduced by treating with STMN1 siRNAs in comparison with si-NC or control (CTR) (Fig. 4C). We further examined the effect of STMN1 siRNAs on cell cycle distribution of FaDu cells. Flow cytometry analysis revealed that the knockdown of STMN1 expression in FaDu increased the percentage of cells in G1 phase (45.72±1.2 and 47.88±1.13 vs. 71.67±0.12 and 66.18±0.20%, respectively; P<0.05) whereas decreased the percentage of cells in S phase (35.46±1.35 and 31.96±2.14 vs. 9.77±0.31 and 13.15±0.09%, respectively; P<0.05; Fig. 4A and B). These results indicated that reduction of STMN1 expression in FaDu cells could inhibit cellular proliferation and promote cell cycle arrest.

The influence of STMN1 knockdown on the migration activity of FaDu cells was detected by Transwell assays. As is shown in Fig. 4E, the results indicated that there were fewer cells migrated to the bottom chambers in FaDu/STMN1-si1 (70.33±12.50). FaDu/STMN1-si2 (94.00±6.25) compared with FaDu/si-NC (114.67±8.74) and CTR (113.00±11.14). The representative micrographs are shown in Fig. 4D.

EMT is a crucial mechanism for migration of multiple tumor cells, with epithelial cells usually losing the capability of intercellular adhesion and acquiring a mesenchyme phenotype (11). To assess whether STMN1 promotes FaDu cell migration via stimulating EMT, EMT-associated proteins were detected by immunofluorescence in FaDu/STMN1-siRNA and FaDu/si-NC and CTR. Interestingly, results showed that the expression level of E-cadherin decreased while the expression level of vimentin increased by depleting STMN1 levels in FaDu cells (Fig. 5A), suggestive of a less epithelial phenotype and a more mesenchyme phenotype. These changes in mRNA and protein expression were confirmed by RT-qPCR and western blot analysis (Fig. 5B–D). During the EMT process, this is usually accompanied by a change in cell morphology (12). To explore whether knockdown of STMN1 has an impact on the morphology of FaDu cells, cell morphological changes were evaluated by an inverted microscope after treated with STMN1-si1 or si-NC. The results showed that cells treated with STMN1-si1 were changed from a more rounded, cell polarity to an elongated, with loss of cell-cell contacts, but there was no effect on cells treated with si-NC (Fig. 5E).