Three Chinese ESCC cell lines (HKESC1, EC18 and EC109), six Japanese ESCC cell lines (KYSE30, KYSE140, KYSE180, KYSE410, KYSE510 and KYSE520; reviewed in ref. 17) and two immortalized esophageal epithelial cell lines (NE1 and NE3) were kindly provided by Professor Srivastava (Department of Pathology, University of Hong Kong). After surgical resection at Linzhou Cancer Hospital, a total of 300 primary ESCC tumor tissue samples and their paired non-tumorous tissue samples were immediately collected from the proximal resection margins. Some were quickly put into vials and stored in liquid nitrogen and other tumor tissues were routinely formalin-fixed and paraffin-embedded. Before operative treatment, no patients recruited in the study received chemotherapy or radiotherapy and the follow-up period was 60 months. Samples used in the present study were approved by the Individual Institutional Committees for Ethical Review of Research Involving Human Subjects.

Through the application of TRIzol (Invitrogen, Carlsbad, CA, USA), total RNA was extracted from the cell lines and frozen ESCC tissues. Advantage RT for PCR kit (Clontech, Mountain View, CA, USA) was employed to conduct reverse transcription of total RNA (2 μg) and cDNA was subjected to PCR for 30 cycles of amplification with the following pairs of primers: SEMA3B Fw, 5′-CCAGTGCCAAGAGGCGGTTC and SEMA3B Rv, 5′-AGCACCTGGGTGTGGGCTGT. The GAPDH gene served as a control with the pair of primers: GAPDH Fw, 5′-CGGGAAGCTTGTCATCAATGG and GAPDH Rv, 5′-GGCAGTGATGGCATGGACTG. In order to analyze the relative grey value, the mRNA expression was quantified by Quantity One software, during which GAPDH was regarded as an internal control.

As previously described, tissue microarrays (TMA) containing 300 pairs of primary ESCC tissue samples and their corresponding non-tumorous tissues were constructed (19). In short, tissue sections with a thickness of 5 μm were cut from tissue microarray blocks and mounted on microscope slides. By adopting the standard streptavidin-biotin-peroxidase complex method, immunohistochemistry (IHC) was performed (19). TMA slides were deparaffinized in xylene, rehydrated through a graded alcohol series and incubated with 3% hydrogen peroxide. For antigen retrieval, TMA slides were boiled in 10 mM sodium citrate buffer (pH 6.0) for 15 min with the help of a pressure cooker. Blocked by 10% normal rabbit serum at room temperature for 30 min, the slides were then incubated with primary anti-SEMA3B monoclonal antibody (1:50 dilution; Lifespan) at 4°C overnight. At a concentration of 1:75, the TMA sections were incubated with biotinylated goat anti-rabbit immunoglobulin for 30 min at 37°C. Through the application of ImmPRESS™ Universal Antibody kit, primary antibody staining was visualized with NovaRed (both from Vector Laboratories, Burlingame, CA, USA) as a substrate. In the next step, the sections were counterstained with hematoxylin, dehydrated and mounted. As previously described, an immunoreactivity scoring (IRS) system was applied (19). The percentage of SEMA3B-positive cells was scored as 0, <5%, negative; 1, 5–25%, sporadic; 2, >25–50%, focal; 3, >50%, diffuse; and the intensity of SEMA3B-positive staining was scored as 0, negative; 1, weak; 2, moderate; and 3, strong. Both the percentage of positive cells and cell staining intensity were determined in a double-blinded manner. According to the following formula, the total score was determined: Staining index = intensity × positive rate. In the present study, a staining index of ≤4 was considered to be downregulated expression and a staining index of >4 was considered as normal expression.

In order to test the function and mechanism underlying growth inhibition by SEMA3B, SEMA3B was PCR amplified, sequence-verified, cloned into pcDNA3.1(+) vector (Invitrogen) and transfected into ESCC, KYSE30 cells which lacked SEMA3B expression. A stable SEMA3B-expressing clone was selected and SEMA3B cDNA was re-sequenced. Blank vector-transfected KYSE30 cells (Vec-30) were used as control. RT-PCR and western blot analysis were employed to detect SEMA3B expression. Taking GAPDH as an internal control, mRNA and protein expression was quantified by Quantity One software, to analyze the relative grey value.

In order to test the tumor-suppressive function of the SEMA3B gene, foci formation assay was performed. Into a 6-well plate, 1×10³ SEMA3B-30 and Vec-30 cells were planted. After being cultured for 10 days, surviving colonies (>50 cells/colony) were counted with Giemsa staining and triplicate independent experiments were carried out. Moreover, cell growth rates of the SEMA3B-30 and Vec-30 cells were detected by MTT assay. At a density of 1×10³/well, the cells were seeded in a 96-well plate. According to the manufacturer's instructions, the cell growth rate was aassessed using cell proliferation MTT kit (Sigma). The experiment was conducted in triplicate.

For the cell migration assays, SEMA3B-30 or Vec-30 cells were cultured in a 6-well plate until confluent and the cell layer was wound using a sterile tip. After being incubated for 36 h, the cells were photographed under a phase-contrast microscope. The experiment was conducted in triplicate. For the invasion assay, SEMA3B-30 or Vec-30 cells were starved of serum-free medium for 24 h before the assay. After suspension in 500-μl serum-free medium, the cells (5×10⁴) were loaded onto the upper compartment of an invasion chamber coated with Matrigel (BD Biosciences). However, the lower compartment was filled with 500-μl complete medium as chemoattractant. After 48 h, the invasive cells were fixed, stained and counted, and independent experiments were carried out in triplicate.

SEMA3B-30 or Vec-30 cells (1×10⁶) were cultured in RPMI-1640 medium containing 10% fetal bovine serum (FBS). When the cells reached 70% confluency, the serum was withdrawn from the culture medium. After 72 h, 10% FBS was added to the medium for an additional 12 h. The cells were fixed in 70% ethanol and stained with propidium iodide. In addition, DNA content was analyzed by applying Cytomics FC (Beckman Coulter).

SEMA3B-30 and Vec-30 cell lysates were obtained. Followed by protein transfer to polyvinylidene fluoride membranes (PVDF), equal amounts of protein from each sample were diluted with loading buffer, denatured and separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). After being incubated in a blocking solution containing 5% non-fat milk powder in Tris-buffered saline and Tween-20 (TBST) buffer (10 mM Tris-HCl, pH 8.0, 150 mM NaCl and 0.1% Tween-20) for 1 h at room temperature, the membranes were immunoblotted overnight with primary monoclonal antibodies for GAPDH, p53, p21, cyclin D1, total Akt and Akt (Ser473; Cell Signaling Technology) at a dilution of 1:1,000 at 4°C and then incubated with the secondary antibody (1:1,000 dilution) for 2 h at room temperature. Enhanced chemiluminescence detection system was adopted to detect the protein antibody complex. The protein expression was quantified by Quantity One software, so as to analyze relative grey value, during which GAPDH served as an internal control.

SPSS standard version 16.0 was used to carry out the statistical analysis. Correlations between SEMA3B expressions and clinicopathological characteristics were assessed by χ² or Fisher's exact tests. In addition, disease-specific survival (DSS) was calculated from the date of diagnosis to the date of cancer-related death or last follow-up. According to the Kaplan-Meier method, survival curves were generated and statistical analysis was performed by making use of the log-rank test. However, the Cox proportional hazards regression model was employed to identify independent prognostic factors. A P-value of <0.05 was considered to indicate a statistically significant result.

Our previous study showed that SEMA3B was significantly downregulated in ESCC by cDNA microarray (16), thus the expression of SEMA3B at the mRNA and protein levels was explored in the present study. The expression of SEMA3B at the mRNA level was assessed by semi-quantitative RT-PCR in 60 primary ESCC tumors as well as their paired non-tumorous tissues and 9 ESCC cell lines. The results showed that the downregulation of SEMA3B was detected in 34/60 (56.7%) ESCC tumor tissues in comparison with their paired non-tumorous tissues. Ratios of SEMA3B gene/GAPDH expression for ESCC tissues and their paired non-tumor tissues were 0.19±0.11 and 0.99±0.10, respectively (Fig. 1A). In comparison with the non-tumor tissues, lower expression of SEMA3B (P<0.05) was found in the ESCC tissues. Meanwhile, the downregulation of the expression of SEMA3B was detected in 6/9 (66.7%) (EC18, HYESC1, KYSE30, KYS180, KYS410 and KYS510) ESCC cell lines. Ratios of SEMA3B gene/GAPDH expression for ESCC cell lines and immortalized esophageal epithelial cell lines were 0.39±0.12 and 0.71±0.09, respectively (P<0.05; Fig. 1B). Furthermore, the quantity of samples was expanded to verify its expression at the protein level. A consistent result was detected in the expression of SEMA3B at the protein level by IHC through application of TMA. Informative results were obtained from 222 pairs of ESCCs. Non-informative samples including lost samples, unrepresentative samples, samples with too few tumor cells, and samples with inappropriate staining were not involved in data analysis. Downregulation of SEMA3B at the protein level was found in 125/222 (56.3%) informative ESCC cases (Fig. 1C).

Afterwards, we examined the clinical association between the expression of SEMA3B protein and the clinicopathological characteristics of the ESCC cases. The analysis demonstrated that downregulation of SEMA3B protein was significantly correlated with lymph node metastasis (P=0.000) and advanced clinical stage (P=0.001; Table I), but was not correlated with gender, age, cell differentiation and general classification. The univariate survival analysis revealed that downregulation of SEMA3B, poor differentiation, status quo of lymph node metastasis and advanced tumor-node-metastasis (TNM) stage were significantly correlated with poor overall survival (P<0.05; Table II). All these variables which showed statistical significance in the univariate analysis were further examined by multivariate analysis (Table III). The results revealed that SEMA3B downregulation was an independent risk factor for overall patient survival (P=0.017), tumor cell differentiation (P=0.005) and TNM stage (P=0.005). The results of Kaplan-Meier analysis demonstrated that ESCC patients with SEMA3B downregulation (median survival time, 19 months) had a shorter disease-specific survival (DSS) than patients with normal SEMA3B expression (median survival time, 30 months; P=0.017; Fig. 2).

The pcDNA3.1(+)-SEMA3B plasmid was constructed and its sequence was subsequently confirmed (Fig. 3A). Both the two plasmids, pcDNA3.1(+)-SEMA3B and blank vector were transfected into KYSE30 cells separately. In order to detect the transfection efficiency of the plasmids in the SEMA3B-30 cells, the SEMA3B gene and protein expression in the SEMA3B-30 cells were confirmed by RT-PCR and western blot analysis. Ratios for SEMA3B gene/GAPDH gene expression for the KYSE30, Vec-30 and SEMA3B-30 cells were 0.0068±0.0047, 0.0084±0.0069 and 0.3766±0.0574, respectively (Fig. 3B; P<0.05). Rates for SEMA3B protein/GAPDH protein expression were 0.2321±0.0510, 0.2691±0.0333 and 0.4460±0.0343, respectively (Fig. 3C; P<0.05). In comparison with the KYSE30 and blank vector-transfected KYSE30 (Vec-30) cells, the expression of SEMA3B was higher in the SEMA3B-30 cells.

The tumor-suppressive function of SEMA3B was assessed by foci formation and cell growth assays. The mean value of the colony formation quantity of the SEMA3B-30 cells was 55.5±10.60, which was lower than that of the Vec-30 cells (220±46.32). The foci formation assay showed that the efficiency of foci formation was highly inhibited (P<0.05) in the SEMA3B-30 cells in comparison with the Vec-30 cells (Fig. 4A). As confirmed in the cell growth assay, the cell proliferation rate in the SEMA3B-30 cells was significantly inhibited by SEMA3B (P<0.05) in comparison with the Vec-30 cells (Fig. 4B) after five successive days of the measurement of OD values.

The TMA results indicated that the downregulation of the protein expression of SEMA3B was significantly correlated with lymph node metastasis, thus the effects by SEMA3B on cell migration and invasion were studied by wound-healing and cell invasion assays. The wound-healing assay indicated that SEMA3B inhibited cell mobility (Fig. 4C). The Matrigel invasion assay also showed that the number of migrating SEMA3B-30 cells (75±30.61) was lower than that of the Vec-30 cells (768.33±94.73; Fig. 4D). SEMA3B inhibited ESCC cell invasiveness, as the number of invasive SEMA3B-30 cells was significantly decreased in comparison with the Vec-30 cells (P<0.05).

To explore the mechanism underlying the growth inhibition of SEMA3B, flow cytometry was used to compare the cell cycle distribution in the SEMA3B-30 cells and the control Vec-30 cells. After 3 days of serum starvation followed by addition of 10% serum for 12 h, significant G1/S phase arrest was found in the SEMA3B-30 cells in comparison with the control Vec-30 cells. The percentage of SEMA3B-30 cells in the G1 phase was 86.35%, while this percentage in the Vec-30 cells was 76.553%. The percentage of SEMA3B-30 cells in the S phase was significantly decreased in comparison with that in the control Vec-30 cells (4.812 and 10.500%; P<0.05; Fig. 5A).

To elucidate the potential molecular mechanism of SEMA3B in cell cycle arrest, the effects of SEMA3B were examined on several key cell cycle regulators including p21, p53, cyclin D1, and Akt from the phosphatidylinositol 3-kinase/Akt pathway. Total Akt and phosphorylated Akt (Ser473) proteins were determined by western blotting with isotype-specific antibodies. According to Fig. 5B, downregulation of phosphorylated Akt at Ser473 was significantly downregulated by SEMA3B in the SEMA3B-30 cells in comparison with that in the Vec-30 cells (P<0.05), while the total Akt level was not changed (P>0.05). Furthermore, the expression of p21, p53 and cyclin D1 in the SEMA3B-30 cells was found to be increased in comparison with the Vec-30 cells (P<0.05; Fig. 5B).