Esophageal squamous cell carcinoma tissue and paired adjacent non-cancerous tissues (n=56) were collected from patients who were diagnosed with primary ESCC and received radical esophageal surgery without preoperative radiotherapy, chemotherapy or biotherapy from August 2012 to September 2013 at Changhai hospital, Second Military Medical University (Shanghai, China). The median age of patients at the time of admission was 58 years, range 43–72 years. In total, 15 patients were women, and 41 were men. Total RNA from 56 pairs of tissue samples were extracted for RT-qPCR detection. Total protein from 6 pairs of tissue samples were extracted for western blot detection. All tissues (tumor tissue and paired para-cancerous tissue) were embedded in paraffin immediately after surgery. Each sample was confirmed by pathological diagnosis and was followed up until March 2018. The use of esophageal cancer tissues and clinical data was approved by the Biomedical Ethics Committee of the Second Military Medical University (Shanghai, China). Informed consent from each patient before surgery was obtained.

IHC was used to evaluate the expression level of SOX12 in paraffin-embedded tissues. All specimens had been fixed in 10% buffered formalin and embedded in paraffin. The embedded tissues were cut into 4-µm-thick serial sections and stored at 4°C. After deparaffinization, rehydration and antigen retrieval, each slide was blocked with 10% (w/v) normal goat serum (product no. SP KIT-B1; Fuzhou Maixin Biotech Co., Ltd.) at room temperature for 1 h and then incubated with anti-SOX12 primary antibody (dilution, 1:100; cat. no. SAB4502835; Sigma-Aldrich; Merck KGaA) at 4°C overnight, followed by incubation with secondary antibodies (EliVision plus; product no. KIT-9901; Fuzhou Maixin Biotech Co., Ltd.) at room temperature for 1 h and diaminobenzidine dye (product no. DAB-0031 (20X); Fuzhou Maixin Biotech Co., Ltd.) at room temperature for 1–9 min. Counterstaining of the nucleus was performed with hematoxylin (product no. CTS-1099; Fuzhou Maixin Biotech Co., Ltd.) at room temperature for 10 min. Negative controls were prepared by replacing the primary antibody with PBS. Images were obtained with light microscope at a magnification of ×200. All tissues were reviewed by experienced pathologists. For each patient sample, the number and proportion of immunopositive cells was determined in 10 randomly selected high-power fields and immunoreactive scoring was performed. A score of >4 was considered as high expression.

The human ESCC cell lines Eca109, TE1 and the human normal esophageal epithelial cell line HEEC (used as control) were obtained from the Cell Bank of the Shanghai Institutes for Biological Sciences. The cells were maintained in DMEM medium supplemented with 10% FBS (HyClone; Cytiva) at 37°C in a 5% CO₂ incubator. The plasmids expressing short hairpin (sh)RNAs targeting SOX12 or scrambled shRNA (shRNA1: 5′-CATGGCGGATTACCCGGACTA; shRNA2: 5′-TCCGCAGTCTTACGAGGAGTC; scrambled shRNA: 5′-GAGCTCTCTCTGCACATTCTT;) were purchased from Guangzhou RiboBio Co., Ltd. (product nos. stB0007967A-1-5, stB0007967B-1–5 and stB0007967C-1-5). Cell culture was performed in six well plates transfected with a total of 4 µg shRNA using Lipofectamine^® 2000 transfection reagent (Invitrogen; Thermo Fisher Scientific, Inc.), used according to the manufacturer's protocol, that was added in each well when the cell density reached 50–60%, and culture was continued at 37°C for 48–72 h according to the experimental requirements.

Recombinant human SOX12 protein (cat. no. H00006666-Q01; Abnova) and WP1066 (cat. no. HY-15312; MCE), a potent inhibitor of the JAK2/STAT3 signaling pathway were purchased and used according to the manufacturer's instructions. Recombinant human SOX12 protein (10 µl) was co-cultured with shRNA1 in the colony formation, Transwell and cellular immunofluorescence assays. WP1066 (5 µl), an inhibitor of the JAK2/STAT3 signaling pathway, was added to the cell culture medium in the colony formation, Transwell and western blotting assays. DMSO was used as a solvent control.

After transfection with the indicated plasmids, Eca109 and TE1 cells were collected and re-suspended in 1.5 ml complete culture medium supplemented with 0.45% low-melting point agarose (Invitrogen; Thermo Fisher Scientific, Inc.). The cells (1×10³) were inoculated in 35-mm tissue culture dishes containing 1.5 ml culture medium and agarose (0.75%) on the bottom layer. The dishes were cultured for two weeks. Cell colonies were fixed with 4% formaldehyde solution for 15 min and stained with 0.005% crystal violet for 10–30 min at room temperature. The number of colonies with more than 10 cells were counted by light microscope at a magnification of ×10. Experimental groups are representative of at least three independent experiments.

The migratory and invasive capacities of Eca109 and TE1 were assessed by using Transwell assays. A normal Transwell chamber was used for the migration assay, and a Transwell chamber whose membrane was pre-coated with Matrigel^® (cat. no. 356231; Corning; BD Bisociences) was used for the invasion assay. For each group, 5×10⁴ cells in 200 µl serum-free medium were added to each upper chamber, while 500 µl DMEM containing 20% FBS was added to the lower chambers. After incubation for 48 h, cells were fixed with 4% formaldehyde solution for 15 min and stained with 0.005% crystal violet for 10 min at room temperature and images were obtained by light microscope (magnification, ×200). Experimental groups are representative of at least three independent experiments.

Total RNA extraction from tissues and cell lines was performed with TRIzol™ reagent (cat. no. 15596026; Invitrogen; Thermo Fisher Scientific, Inc.) and then reverse-transcribed into complementary DNA immediately with the PrimeScript™ RT Reagent kit (cat. no. RR037B; Takara Bio, Inc.) according to the manufacturer's instructions. Fast SYBR™ Green Master Mix (cat. no. 4385612; Thermo Fisher Scientific;) was added to the PCR reaction mixture according to the manufacturers' protocols. The thermocycling conditions used for qPCR were as follows: 95°C for 2 min; 45 cycles at 95°C for 5 sec, 55°C for 20 sec and 72°C, 30 sec. A melting curve was established under the following reaction conditions: 95°C for 15 sec, 60°C for 1 min and 95°C for 15 sec. The experiment was performed in triplicate. GAPDH was used as internal control. The 2^−ΔΔCq method (14) was conducted to calculate the relative expression of the genes. The primer sequences used were as follows: SOX12 (NM_006943.3) forward, 5′-AGCACCCGTGTGACTCTTTCC-3′ and reverse, 5′-AGCAGAACCAAGCCCTGTCTC-3′; GAPDH (NM_001256799.1) forward, 5′-CACCCACTCCTCCACCTTTG-3′ and reverse, 5′-CCACCACCCTGTTGCTGTAG-3′.

Total protein from tissues and cell lines was extracted by using radioimmunoprecipitation assay lysis buffer (product no. P0013B; Beyotime Institute of Biotechnology). The total protein concentration was determined using the BCA method. A total of 30 µg of protein was separated by SDS-PAGE (10%) and transferred to a PVDF membrane. Non-specific binding sites were blocked by incubating with TBST (0.1% Tween 20) containing 5% (w/v) non-fat dried milk for 1 h at room temperature. The antibodies used were as follows: Anti-SOX12 primary antibody (product no. SAB1412152; dilution, 1:1,000; Sigma-Aldrich; Merck KGaA), anti-JAK primary antibody (product code ab108596; dilution, 1:5,000), anti-phosphorylated (p)-JAK at Tyr1007+1008 primary antibody (product code ab32101; dilution, 1:2,000), anti-STAT3 primary antibody (product code ab68153; dilution, 1:1,000), anti-p-STAT3 (phosphorylated at Tyr705) antibody (product code ab76315; dilution, 1:5,000), anti-GAPDH antibody (product code ab9482; dilution, 1:5,000; all from Abcam) and goat anti-rabbit immunoglobulin G H&L (cat. no. ab205718; dilution, 1:5,000; Abcam). Signals were visualized by ECL chemiluminescence (cat. no. 34095, Thermo Fisher Scientific, Inc.). The results were evaluated with Quantity One software (v4.6.6; Bio-Rad Laboratories, Inc.).

ESCC cell lines Eca109 and TE1 were seeded onto coverslips in six-well cell culture plates at 2×10⁴ cells/well with complete medium and transfected with different shRNAs as aforementioned. After culture for 48 h, the cells were rinsed three times with PBS carefully and then fixed with 4% formaldehyde solution for 15 min at room temperature. Cells were then permeabilized with Triton X-100 solution (1% v/v) for 10 min and blocked with non-immune goat serum for 1 h at room temperature. Cells were incubated with various antibodies as follows: SOX12 (product no. HPA055052; 1:100 dilution; Sigma-Aldrich; Merck KGaA), and the aforementioned antibodies p-JAK (phosphorylated at Tyr1007+1008; 1:100 dilution) and p-STAT3 (phosphorylated at Tyr705; dilution, 1:100) at 4°C overnight. Cells were then rinsed with PBS and incubated with cyanine 3-labeled (product no. A0516, 1:500 dilution) and Alexa 488-labeled (product no. A0423; 1:500 dilution) secondary antibodies (Beyotime Beyotime Institute of Biotechnology) for 1 h at room temperature, and stained with DAPI (1 µg/ml; Sigma-Aldrich; Merck KGaA) for 5 min at room temperature. Images were obtained using a fluorescence microscope (magnification, ×200).

Data analyses were calculated by SPSS 18.0 statistical software (SPSS, Inc.) and values are expressed as the mean ± standard deviation. ANOVA and Dunnett's test were used for comparing the mean of multiple experimental groups and that of the control group. The chi-square test and Spearman's rank correlation were used to analyze the correlation between gene expression and clinical pathological data. Survival analysis was performed using the Kaplan-Meier method and significant differences in overall survival (OS) and disease-free survival (DFS) was determined with the log-rank test. Univariate and multivariate analyses were performed using Cox's proportional hazard model. Differences between multiple groups were analyzed using one-way analysis of variance. P<0.05 was considered to indicate a statistically significant difference. The experimental data were representative of three independent experiments.

In the present study, the protein level of SOX12 in 56 ESCC samples and adjacent non-cancerous tissues was examined by IHC. The results indicated that SOX12 was upregulated in ESCC tissues compared with adjacent non-cancerous tissues in 38 out of 56 cases (67.9%; Fig. 1A and B). Then 6 pairs of fresh tissue samples were collected and total proteins were extracted. The protein level of SOX12 in ESCC tissues and adjacent non-cancerous tissues was detected by western blot analysis. The results revealed that the protein levels of SOX12 were markedly upregulated in ESCC tissue samples compared with those in the corresponding non-cancerous tissues (Fig. 1C). In addition, the protein and mRNA expression of SOX12 in ESCC cell lines was evaluated by western blot and RT-qPCR analysis, respectively. Human normal esophageal epithelial cell line (HEEC) was used as a control. As presented in Fig. 1D and E, compared with that in the HEEC human normal esophageal epithelial cell line, SOX12 was significantly overexpressed in both Eca109 and TE1. Collectively, these results indicated that SOX12 was aberrantly overexpressed in ESCC cell lines and tissue samples.

To explore the function of SOX12 which regulates the biological characteristics of ESCC cells, three vectors carrying shRNA targeting SOX12, shRNA1, shRNA2 and shRNA3, were constructed and respectively transfected into Eca109 and TE1 cells. Two of the vectors, shRNA1 and shRNA2, were more effective at knocking down the expression of SOX12 (Fig. 2), as determined by RT-qPCR and western blot analysis. Colony formation and Transwell assays were then performed to examine the neoplastic capacity and motility of SOX12-silenced Eca109 and TE1 cells. The results revealed that downregulation of SOX12 could inhibit the viability and motility capacities of ESCC cells in vitro (Fig. 3). When recombinant SOX12 (10 µM) was co-cultured in transfected cells, the proliferation and motility of ESCC cells could be restored (Fig. S1A and B). It was therefore revealed that overexpression of SOX12 was closely related with the malignant biological behavior of ESCC cells.

Mutational activation of the JAK2/STAT3 signaling pathway is responsible for the malignant transformation and progression of various types of tumors (15). In this study, the protein expression levels of JAK2/STAT3 signaling proteins were measured by immunofluorescence and western blot assay. The results revealed that knockdown of SOX12 in ESCC cells decreased the levels of p-JAK2^Tyr1007+1008 and p-STAT3^Tyr705 (Fig. 4), which could be reversed by co-culture with recombinant SOX12 (10 µM) (Fig. S1C). In addition, WP1066 (5 µM) could inhibit colony formation and motility of ESCC cells in vitro (Fig. S2). Thus, it may be inferred that SOX12 promotes the malignant biological behavior of ESCC cells, including proliferation, invasion and migration, at least in part through activation of JAK2/STAT3 signaling pathway.

The correlation between SOX12 and clinicopathological factors of patients with ESCC was then evaluated. Statistical analysis indicated no significant difference in any of the clinicopathological characteristics of ESCC patients, including sex, age, degree of differentiation, tumor invasion and lymph node metastasis, between patients with high and low expression of SOX12 in their ESCC tissues (Table I). To explore the effect of SOX12 expression (high vs. low) on the clinical prognosis of ESCC patients, Kaplan-Meier survival analysis was used. As presented in Fig. 5, high expression of SOX12 was associated with lower OS (P=0.0341) and DFS (P=0.04) compared with low expression of SOX12. Accordingly, overexpression of SOX12 was closely associated to the poor prognosis and survival of patients.

Analysis of the clinical data by univariate logistic regression using Cox's proportional hazards model indicated that SOX12 is a significant prognostic factor for ESCC patients (P=0.016; Table II). Multivariate logistic regression analysis revealed that SOX12 is an independent prognostic factor for ESCC patients (P=0.044; Table II). The results of clinical analysis revealed that overexpression of SOX12 may affect the malignant characteristics of ESCC and poor prognosis.