A total of 106 esophageal carcinoma (EC) samples and 49 matched adjacent esophagus tissues were collected by endoscopic biopsies or surgical resection from 106 patients (Table I), and 17 normal esophagus mucosa tissues were obtained from surgical resections of trauma patients. The age of the EC patients ranged between 32 and 78, and the mean age was 59.8. The age of the trauma patients ranged from 32 to 72, and the mean age was 57.2. No patient had received any other treatment prior to surgery. These tissues were obtained from the Gastrointestinal Center at the Jiangyin People's Hospital, Medical School of University of Southeast of China (Jiangyin, China) from February 2010 to December 2012. This study was performed under the approval the Ethics Committees of Jiangyin People's Hospital and the patients gave written informed consent to participate. All diagnoses were based on pathological and/or cytological evidence. According to the classification criteria from the World Health Organization (23), the histological features of the specimens were evaluated by two senior pathologists. For immunohistochemistry, tissues were frozen in liquid nitrogen and maintained at −80°C immediately following endoscopic biopsies or surgical resection.

The expression patterns of CD63 in human esophageal cancer samples, the adjacent normal esophagus samples and the normal esophagus mucosa tissues were analyzed using immunohistochemistry. Sections of formalin-fixed and paraffin-embedded tissue with a thickness of 6 µm were deparaffinized and heated in citrate buffer (pH 6.0; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) at 95°C for 30 min as an antigen retrieval protocol. Endogenous peroxidase activity was blocked with 0.3% hydrogen peroxide, and non-specific-binding sites were reduced by incubating the sections at room temperature with 4% skim milk powder for 30 min. The sections were then incubated overnight at 4°C with a rabbit monoclonal antibody to CD63 protein (cat. no. ab134045; dilution, 1:1,000; Abcam, Cambridge, MA USA). Biotinylated secondary antibody mouse anti-rabbit immunoglobulin G-B (cat. no., sc-2491; dilution, 1:1,000, Santa Cruz Biotechnology, Santa Cruz, CA, USA) was then added to the sections followed by incubation for 30 min at room temperature. An avidin-biotin-peroxidase complex (Beyotime Institute of Biotechnology, Nantong, China) was added for an additional 30 min. Following treatment with a 3,3′-diaminobenzidine substrate-chromogen system (Dako; Agilent Technologies Inc., Santa Clara, CA, USA) for 8 min, counterstaining was performed with hematoxylin. All histological assessments were performed by the same pathologist. For statistical analysis, the sections with <10% of the whole tissue mass stained were classified as negative; the sections with 10–25% of the whole tissue mass stained were classified as weakly positive (+1); the sections with 25–75% of the whole tissue mass stained were defined as moderately positive (+2); the sections with >75% of the tissue stained positive were regarded as strongly positive (+3).

The human esophageal cancer cell line TE-1 was cultured in Dulbecco's modified Eagle's medium (Thermo Fisher Scientific Inc., Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.) and antibiotics (100 U/ml penicillin and 100 µg/streptomycin, Sigma-Aldrich; Merck KGaA). Cells were grown at 37°C in a 5% CO₂ atmosphere and were transfected according to the manufacturer's protocol using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) with CD63 siRNA (Santa Cruz Biotechnology Inc., Dallas, TX, USA).

Esophageal carcinoma TE-1 cells were divided into three groups: Control (with no transfection protocol), NC (cells transfected with empty plasmid), and siCD63 (cells transfected with small interfering RNA targeting CD63). Following treatment, cells were lysed with radioimmunoprecipitation assay lysis buffer (50 mM Tris-HCl, pH 7.4; 150 mM NaCl; 1% Triton X-100; 1% sodium deoxycholate; 0.1% SDS) and protease inhibitor cocktail (P8340; Sigma-Aldrich; Merck KGaA) for 30 min at 4°C. BCA protein assay (Bio-Rad Laboratories, Hercules, CA, USA) was used to determine the protein concentration. Equal quantities of protein lysates (40 µg total protein) were electrophoretically separated by 10% SDS-PAGE and transferred to polyvinylidene difluoride membranes (EMD Millipore, Billerica, MA, USA). Following blocking with 5% nonfat milk in TBST (20 mM Tris, pH 7.5; 500 mM NaCl; 0.05% Tween-20) for 1 h at room temperature, the membranes were then incubated overnight at 4°C with the primary antibody to CD63 (cat. no., ab134045; dilution, 1:1,000, Abcam), E-cadherin (cat. no., sc-59778; dilution, 1:200, Santa Cruz Biotechnology), vimentin (cat. no. sc-6260; dilution, 1:200, Santa Cruz Biotechnology), MMP-2 (cat. no. sc-13594; dilution, 1:200, Santa Cruz Biotechnology) and MMP-9 (cat. no. sc-21733; dilution, 1:200, Santa Cruz Biotechnology). β-Actin (cat. no. sc-8432; dilution, 1:200, Santa Cruz Biotechnology) was used as a loading control. Following washing four times with TBST, the membranes were incubated at room temperature with a horseradish peroxidase-conjugated secondary antibody (cat. nos. sc-2005, sc-2357; dilution, 1:1,000, Santa Cruz Biotechnology) for 2 h, and visualized by chemiluminescence using an enhanced chemiluminescence system (GE Healthcare Life Sciences, Chalfont, UK).

Esophageal carcinoma TE-1 cells were seeded in 6-well plates and allowed to reach a confluent state, then were divided into three groups: Control (with no transfection protocol), NC (cells transfected with empty plasmid), and siCD63 (cells transfected with small interfering RNA targeting CD63). Following treatment, the monolayer was scratched with the tip of a 200 µl pipette. The floating and detached cells were removed by washing twice with PBS. Subsequently, fresh serum-free medium was added. At 0 and 24 h, images were captured to assess the extent of cell migration using an Olympus BX 40 Light Microscope (Olympus Corporation, Tokyo, Japan).

The invasive potential of TE-1 cells was assessed using 24-well Matrigel invasion chambers (pore size 8 µm, Corning Costar Corporation, Corning, NY, USA). Inserts were pre-coated with 40 µl Matrigel (1:4 dilution; BD Biosciences, San Jose, CA, USA). Then, 5×10⁴ cells/ml pre-transfected with CD63 siRNA in serum-free medium were added to the upper chambers. The lower chambers contained 500 µl medium supplemented 10% FBS. Following incubation for 24 h at 37°C, the cells remaining in the upper chambers were removed with PBS-moistened cotton swabs, and the invading cells on the underside of the insert filter were fixed with 3.7% paraformaldehyde for 30 min at room temperature and stained with Giemsa for 1 h at room temperature. Images of 5 random microscopic fields gathered by an Olympus BX 40 Light Microscope (Olympus Corporation) were captured to calculate the average number of invaded cells.

Mean ± standard error of the mean of several independent experiments were calculated. One-way analysis of variance was applied to analyze the data, and then the significance of differences among different groups was evaluated by Student-Newman-Keuls post hoc test. Statistical tests were performed using Graph-Pad PRISM version 6.0 (Graph-Pad Software, San Diego, CA). P<0.05 was considered to indicate a statistically significant difference.

To investigate the variation in expression levels between the normal and cancerous esophageal tissues, CD63 protein expression levels were evaluated by immunohistochemistry staining in 106 paraffin-embedded human esophageal cancer tissues, 49 adjacent esophagus tissues and the 17 normal esophagus mucosa tissues. Overall positive staining for CD63 was frequently observed in the cytoplasm of EC tissues, whereas no staining of CD63 was observed in the adjacent esophagus tissues and the normal esophagus mucosa tissues (Fig. 1A). The sections were subsequently scored by a pathologist; there were significant differences in CD63 expression between esophageal cancer tissues, adjacent esophagus tissues and the 17 normal esophagus mucosa tissues (Fig. 1B, P<0.01). This demonstrated that CD63 expression was increased in esophageal carcinoma and may be a regulator in the development of EC.

The CD63 expression differences in the EC tissues in different clinical stages were subsequently analyzed. The average CD63 expression level in earlier stage (I and II) esophageal carcinoma was higher than the level in advanced stage (III and IV) (Fig. 2; P<0.01). CD63 expression levels were also found to be associated with lymph node metastasis in EC cases. Weak positive staining of EC tissues was detected in patients with lymph node metastasis. However, in the patients without metastatic lymph nodes, the CD63 staining was strongly positive (Fig. 3). This suggests that CD63 may have an important role in the lymph node metastasis of EC.

To investigate the role of reduced CD63 expression in the migration of EC, the esophageal cancer cells TE-1 were transfected with CD63-silencing RNA. Western blot analysis revealed that CD63 protein expression was markedly reduced by the siRNA transfection (Fig. 4A). Matrigel invasion assays were performed to explore the effect of CD63 on the invasiveness of TE-1 cells. As presented in Fig. 4B and C, the ability of TE-1 cells to invade through the Matrigel and membrane was increased by CD63 knockdown when compared with the control cells or the empty plasmid transfected cells. The CD63 siRNA transfected TE-1 cells exhibited a ~2-fold higher invasiveness than the control cells. In vitro wound healing assays were performed to detect the motility of TE-1 cells, which is another important characteristic of metastatic cells. As presented in Fig. 4D and E, CD63 knockdown noticeably reduced the wound area, suggesting that CD63 knockdown enhanced the migration capability of esophageal cancer.

To explore the mechanisms underlying the effect of CD63 on the invasiveness of EC, TE-1 cells were transfected with CD63 siRNA or empty plasmid. After 48 h of incubation, western blotting assays were used to detect the expression of EMT-associated proteins and MMPs. E-cadherin was established as the epithelial maker, and vimentin as the mesenchymal maker. As presented in Fig. 5, TE-1 cells transfected with CD63 siRNA exhibited downregulation of E-cadherin expression and upregulation of vimentin when compared with the control cells or the empty plasmid transfected cells. The matrix metalloproteinase MMP-2 and MMP-9 expression was increased by CD63 knockdown. These results suggest that CD63 knockdown regulates the invasiveness of esophageal cancer cells via promoting EMT.