In this study, we randomly selected 62 esophageal cancer patients who underwent potentially curative surgery without preoperative chemotherapy or radiotherapy between February 26 and March 16, 2006 in Anyang Tumor Hospital, Henan, China. Among them, 36 were men and 26 were women, ranging from 38-75 years of age with a mean age of 52.6 years. Overall, 22 cases were poorly differentiated, 25 were moderately differentiated, and 15 were well differentiated. For the lymphatic node metastasis, 20 cases were positive and 42 cases were negative. Seven cases were classified as T₁ and T₂, and 55 as T₃ and T₄ in terms of T stages. Surgically removed specimens were routinely fixed in buffered formalin and embedded in paraffin blocks for clinical diagnosis and reclassification for this study. The normal esophageal squamous tissues were taken from mucosae 3 cm away from carcinomas, and the adjacent atypical hyperplastic epithelium 2 cm away from the carcinomas. The final pathological diagnosis was based on the result of histological examination.

Immunohistochemical analysis for E-cadherin and N-cadherin was performed on 5-&mgr;m sections made from tissue microarray blocks. The Envision Plus detection system (Dako, Carpinteria, CA, USA) was used for the immunostaining. The sections were deparaffinized in xylene and then were microwaved in 10 mmol/L citrate buffer (pH 6.0) to unmask the epitopes. Endogenous peroxidase activity was blocked by incubation with 0.03% hydrogen peroxide in methanol for 5 min. Slides were incubated with mouse anti-human E-cadherin antibody (microwaving retrieval in citrate buffer at 1:100 concentration, Abcam) and mouse anti-human N-cadherin antibody (microwaving retrieval in low pH buffer at 1:100 concentration; Abcam), respectively. Then, polyperoxidase anti-mouse IgG (Beijing Zhongshan Golden Bridge Biotechnology Co. Ltd., China) was added for incubation for another 30 min followed by gentle rinsing with washing buffer for three times. Thereafter, the sections were stained for 5 min with 3,3-diaminobenzidine (DAB) (Beijing Zhongshan Golden Bridge Biotechnology), counterstained by hematoxylin, dehydrated, and mounted in Diatex. Expression of E-cadherin and N-cadherin in breast cancer was used as a positive control, while the same concentration of PBS was applied as a negative control for E-cadherin and N-cadherin.

For both E-cadherin and N-cadherin, only membrane and cytoplasm staining was considered as positive. The intensity and percentage of immunostained carcinoma cells were all taken into consideration according to the previously published method with modification[6]. Briefly, the extent of positivity was scored as 0 when no positive cell was observed; 1 when the percentage of positive cells was < 30%; 2 when it was 30%-60%; and 3 when it was > 60%. The intensity was scored as 0 when no positive cells were identified; 1, weak; 2, moderate; and 3, strong staining. Multiplying the extent by intensity gave the following immunohistochemical staining grades as 0, 1, 2, 3, 4, 6 and 9. For statistical analyses, grades 0, 1 and 2 were considered as negatively stained, and grades > 2 were considered as positively stained.

PT67 packaging cells were seeded onto a six-well plate at 1 × 10⁵ cells per well and incubated for 24-48 h. The cells were allowed to grow to 60%-70% confluence and then rinsed with fresh DMEM medium. The control vector pEGFP-MSCVneo and recombinant retroviral vector pMSCVneo/N-cadherin plasmids (kindly provided by Dr. Ma Jie, Chinese Academy of Medical Sciences, China) (The former contained the enhanced green fluorescent protein and neomycin resistance genes, the latter contained the N-cadherin shRNA, U6-promotor, enhanced green fluorescent protein and neomycin resistance genes) were transfected into the packaging cell line PT67 by lipofectamine 2000 (Invitrogen Corp, UK) and incubated for 6 h following the manufacturer’s instructions. The mixture was replaced with fresh medium to stop transfection and the transfected PT67 cells were further cultured with G418 (1000 mg/L) selecting medium for 2 wk. Viral supernatant of the drug-resistant clones was filtered through a 0.45-&mgr;m filter and condensed by ultracentrifugation at low temperature, then preserved at -80°C. The supernatant, from the fresh PT67 packaging cells containing recombinant retrovirus, was used to infect NIH3T3 cells to determine viral titer as previously described[7]. The highest titer clones were selected for further experiments.

EC9706 cells were plated in a six-well plate at 5 × 10⁴ cells/mL and cultured with 5% CO₂ at 37°C. Twenty-four hours later, the cells were exposed to 2 mL viral supernatant at a consecutive multiplicity of infection once for 12 h, in the presence of 8 &mgr;g/mL polybrene (Sigma). Subsequently, viral supernatant was replaced with fresh medium. After being incubated for another 24 h, the infected EC9706 cells were screened with G418 (600 mg/L) (Sigma) selecting medium for 2 wk. Drug-resistant clones were expanded with G418 (300 mg/L) (Sigma) selecting medium. Two days after culturing, the expression of EGFP in infected EC9706 cells was observed under fluorescence microscope.

Total RNA and protein isolation was performed using the Macherey-Nagel total RNA and protein isolation kit according to the user manual. About 5 × 10⁶ EC9706 cells transfected with N-cadherin RNAi, control vector, or the untreated cells were collected and lysed. Through the NucleoSpin RNA/Protein column, RNA and DNA were bound to the column and protein was contained in the flow-through. After digestion of DNA, total RNA was isolated by washing the column. Protein was isolated from the flow-through and incubated for 3 min at 98°C for dissolving and denaturation, and stored at -20°C until used. All of the preparation and handling steps of RNA took place in a laminar flow hood under RNase-free conditions.

RNA quality and quantity were determined by absorbance readings at 260 and 280 nm with the Nano Drop (ND-1000) spectrophotometer. RNA integrity was tested by PCR amplification of the GAPDH gene. Reverse transcription of RNA was performed using Transcriptor First Strand cDNA Synthesis Kit (Roche). cDNA was synthesized from 5 &mgr;g total RNA isolated from EC9706 cells transfected with N-cadherin siRNA, control vector, or the untreated cells according to the manufacturer’s handbook.

The primer pairs and hydrolysis probes for N-cadherin, E-cadherin, matrix metalloproteinase-9 (MMP-9) and GAPDH were designed by Universal ProbeLibrary Assay Design Center (Roche). All primer sequences listed in Table 1 were synthesized by Shanghai Sangon (China). The hydrolysis probes were designed and synthesized by Roche Diagnostics.

Real-time PCR was performed with the ABI Prism 7500 Sequence Detection System (ABI) in a total volume of 20 &mgr;L in glasscapillaries containing 2 &mgr;L of cDNA, 0.5 &mgr;mol/L of each primer, 0.1 &mgr;mol/L of hydrolysis probe and 4 &mgr;L of LightCycler TaqMan Master Mix (Roche Diagnostics). PCR reaction was initiated with a 12-min denaturation at 95°C and terminated with a 30-s cooling step at 40°C. The cycling protocol consisted of denaturation at 95°C for 10 s, annealing at 54°C for 10 s, and extension at 72°C for 10 s, and was cycled 45 times. Fluorescence detection was performed at the end of each extension step. The housekeeping genes GAPDH and DEPC-H₂O were set as internal control and negative control, respectively.

Twenty microliters of protein samples were separated on a 10% SDS-acrylamide gel (Bio-Rad) for 1 h at 150 V, and the proteins were transferred to nitrocellulose membrane (Whatman). After blocking in 5% fat-free milk, the membrane was treated with the dilution of the primary antibody overnight at 4°C and the dilution of the secondary IgG-horseradish peroxidase (HRP) conjugated antibody for 1 h at room temperature. All dilutions were in PBS containing 5% Blotto (Santa Cruz) and 0.1% Tween-20. The stained membranes were visualized by enhanced chemiluminescence reaction using the ECL Plus (GE Healthcare). Western blot experiments were repeated at least three times on every sample, with similar results.

Matrigel-coated filter inserts with 8-&mgr;m pores that fit into 24-well invasion chambers were obtained from Becton Dickinson. EC9706 cells transfected with N-cadherin RNAi, control vector, or the untreated cells were detached from the tissue culture plates, washed, resuspended in conditioned medium (10⁶ cells/mL), and then added to the upper compartment of the invasion chamber with or without plasmin (1.8 mg). Conditioned medium (500 &mgr;L) was added to the lower compartment of the invasion chamber. The Matrigel invasion chambers were incubated at 37°C for 24 h in 5% CO₂. After incubation, the filter inserts were removed from the wells, and the cells on the upper side of the filter were removed using cotton swabs. The filters were fixed, mounted, and stained according to the manufacturer’s instructions. The cells that invaded through the Matrigel were counted on the underside of the filter. Three to five invasion chambers were used for each experimental condition. The values obtained were calculated by averaging the total number of cells from three filters.

χ² test and Spearman rank correlation coefficient analysis were used to assess the univariate association between the immunohistochemical status and the clinicopathological characteristics. Results were expressed as mean ± SD. Statistical analysis was made using one way ANOVA or paired-samples t test of SPSS 11.0. P < 0.05 was considered statistically significant.

For E-cadherin, positive immunostaining was observed on the membrane of cancer cells and the intercellular junctions. It was strongly expressed in the normal esophageal squamous tissues (95.2%), moderately expressed in the adjacent atypical hyperplastic epithelium (71.0%), and weakly expressed in the ESCC tissues (40.3%) (Figure 1A, B and Table 2). Contrary to the E-cadherin, N-cadherin, which existed in the cytoplasm, was strongly expressed in the ESCC (75.8%) and moderately expressed in the adjacent atypical hyperplastic epithelium (61.3%), but weakly expressed in the normal esophageal squamous tissues (29.0%) (Figure 1C, D and Table 2). The correlations between the clinicopathological features and the expressions of E-cadherin and N-cadherin in the primary tumors are summarized in Table 2. Higher level of N-cadherin expression was significantly associated with higher histological grade, deeper invasion and more lymph node metastasis, while E-cadherin expression was associated with totally opposite sides.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Expression of E-cadherin and N-cadherin in ESCC tissues and normal tissues. A: Strong expression (yellowish brown) of E-cadherin on the membrane of normal esophageal epithelial cells (PV, × 200); B: Weak expression (yellow) of E-cadherin on the membrane of ESCC cells (PV, × 200); C: Weak expression (yellow) of N-cadherin in the cytoplasm of normal esophageal epithelial cells (PV, × 200); D: Strong expression (yellowish brown) of N-cadherin in the cytoplasm of ESCC cells (PV, × 200).

In order to know whether the expression of E-cadherin and N-cadherin in these tumors was associated, a crosstable analysis was performed (Table 3), which showed that their expression was significantly negatively correlated.

After G418 selection for 30 d, the stable colonies of pEGFP-MSCVneo plasmid (control vector) with viral titer 1 × 10⁷ cfu/L and colonies of pMSCVneo/N-cadherin (RNAi vector) with viral titer 3 × 10⁷ cfu/L were picked up to infect EC9706 cells (Figure 2A).

Open in New Tab   Full Size Figure   Download Figure

Figure 2 EGFP expression in PT67 cells and EC9706 cells after pMSCVneo/N-cadherin transfection. A: Selected by 1000 mg/L G418 for 15 d and 300 mg/L G418 for 15 d, the PT-67 cells transfected with pMSCVneo/N-cadherin plasmid expressed EGFP stably. The figure was taken under fluorescence microscope (× 400, at 488 nm); B: Selected by 600 mg/L G418 for 14 d and 300 mg/L G418 for 10 d, the EC9706 cells infected with pMSCVneo/N-cadherin viral supernatant expressed EGFP stably. The figure was taken under fluorescence microscope (× 400, at 488 nm).

EC9706 cells were infected with the concentrated viral supernatant and selected with G418 as described above. After screening with G418 for 2 wk, the isolated G418-resistant clones (Figure 2B) were transferred into larger culture vessels to expand for further experiments.

To investigate the expression phenotypes of E-cadherin and MMP-9 when N-cadherin was down-regulated, real-time PCR and Western blotting were employed to examine the expressions of the N-cadherin, E-cadherin and MMP-9 in ESCC cell line EC9706. Transfection of N-cadherin RNAi lowered N-cadherin mRNA level (0.397 ± 0.013) to less than 40%, compared with the untreated cells (1.000 ± 0.016) (Figure 3), and the N-cadherin protein (0.342 ± 0.006) to 50%, compared with the untreated cells (0.679 ± 0.004) (Figure 4).

Open in New Tab   Full Size Figure   Download Figure

Figure 3 E-cadherin, N-cadherin and MMP-9 gene transcripts detected by real-time PCR. 1: Positive control; 2: Untreated EC9706 cells; 3: EC9706 cells with control vector; 4: EC9706 cells with N-cadherin RNAi. ^aP < 0.05 vs 2 and 3.

Open in New Tab   Full Size Figure   Download Figure

Figure 4 E-cadherin, N-cadherin and MMP-9 proteins detected by Western bloting. 1: Positive control; 2: Untreated EC9706 cells; 3: EC9706 cells with control vector; 4: EC9706 cells with N-cadherin RNAi. E-cadherin proteins were 0.247 ± 0.010, 0.252 ± 0.087, 0.249 ± 0.07 and 0.250 ± 0.006, respectively, from lane 1 to lane 4. P > 0.05. The N-cadherin proteins were 0.681 ± 0.003, 0.679 ± 0.004, 0.653 ± 0.009 and 0.342 ± 0.006, respectively, from lane 1 to lane 4. P < 0.05. The MMP-9 proteins were 0.624 ± 0.011, 0.628 ± 0.010, 0.623 ± 0.009 and 0.282 ± 0.010, respectively, from lane 1 to lane 4. P < 0.05.

The levels of MMP-9 mRNA and protein lowered by about 50% in the N-cadherin depleted cells compared with the untreated cells (Figures 3 and 4). The MMP-9 mRNA reduced from 1.241 ± 0.023 in the untreated cells to 0.566 ± 0.016 in N-cadherin RNAi cells, and the protein reduced from 0.628 ± 0.010 to 0.282 ± 0.010. However, it seems that N-cadherin shRNA did not affect the levels of E-cadherin mRNA and protein (Figures 3 and 4).

N-cadherin level was elevated with the malignancy in the ESCC tissues in a previous experiment. When N-cadherin was down-regulated, the migration of the cells was also reduced, compared with the untreated cells. Compared with the untreated EC9706 cells (Figure 5A) and the cells with control vector (Figure 5B), the N-cadherin negative-cells (Figure 5C) migration across the membranes decreased dramatically.

Open in New Tab   Full Size Figure   Download Figure

Figure 5 Migration assay in transwell chamber. A: Untreated EC9706 cells; B: EC9706 cells with control vector; C: EC9706 cells with N-cadherin RNAi. They were 123.40 ± 8.234, 126.00 ± 10.295 and 49.60 ± 6.804, respectively, from A to C. P < 0.05. (HE, × 200).

Among the members of cadherin family, E-cadherin and N-cadherin have been extensively studied for their biological activities and associations with cancer cell invasion. Latest research on prostate cancer and breast cancer has proved that the up-regulated N-cadherin plays even more important roles in cell progression and metastasis.

The shift in expression from E- to N-cadherin and their mutually exclusive expression pattern in invasive tumors strongly reflects dedifferentiation from an epithelial to a mesenchymal phenotype, often associated with an increased invasive state. This “cadherin switch” has been observed, especially in those where mild and non-progressive cells transformed into more invasive phenotypes. Therefore, many studies have focused on the exact underlying mechanism involved in this cadherin switch.

In the current study, the expression of N-cadherin and E-cadherin was first examined in esophageal squamous cell carcinoma (ESCC) specimens and the results revealed that increased expression of N-cadherin and decreased expression of E-cadherin were related to invasion, differentiation, and lymph node metastasis, the roles of N-cadherin in the invasiveness of ESCC were first investigated in the EC9706 cell line transfected by retroviral-mediated N-cadherin RNAi and the results revealed that N-cadherin knock-down significantly decreased the invasiveness of EC9706 cells.

This study has indicated that N-cadherin is an important factor in the invasiveness of ESCC and N-cadherin may serve as a potential molecular target for biotherapy of ESCC.

Cadherins are a class of type-1 transmembrane proteins. They are calcium-dependent cell-cell adhesion glycoproteins. They play important roles in tissue formation and maintenance during embryonic development, and in the induction and maintenance of normal architecture and function in adult tissues.

The authors examined the expression pattern of N-cadherin and E-cadherin. They demonstrated that N-cadherin is an important factor in the invasiveness of ESCC and it may serve as a potential molecular target for biotherapy of ESCC.