Rabbit antibodies against human BAMBI (HPA010866), β-catenin (C2206), TGF-β (SAB4502954) and β-actin (AV40173) were purchased from Sigma (St. Louis, MO, USA). The secondary antibodies conjugated to horseradish peroxidase against rabbit IgG (sc-2030) were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA).

Clinical samples were removed from 10 patients during tissue excision surgery at the Qilu Hospital of Shandong University (Jinan, China) and 60 patients were recruited for the 24-month survival analysis. Informed patient consent and approval from the institutional review board of the Qilu Hospital of Shandong University were obtained. The samples were paired and samples from 10 patients were subjected to semi-quantitative polymerase chain reaction (qPCR) and immunohistochemistry. The N87 human gastric adenocarcinoma cell line was obtained from the American Type Culture Collection (Manassas, VA, USA).

For comparison of biophysical properties following knockdown and overexpression of BAMBI, BAMBI interference and overexpression vectors were constructed. The recombinant expression plasmid of BAMBI small hairpin RNA (shRNA) was purchased from Santa Cruz Biotechnology, Inc. (sc-62576-SH). The recombinant expression plasmid of BAMBI was constructed. Briefly, the open reading frame of BAMBI (GenBank accession, NM012342) was cloned into the plasmid pcDNA3.1(t) (Invitrogen Life Technologies, Inc., Carlsbad, CA, USA) between the XhoI and BamHI sites to build the recombinant plasmid pcDNA3.1(t) BAMBI. The cells were transfected with pcDNA3.1(t)-BAMBI and/or BAMBI shRNA using Lipofectamine™ 2000 (Invitrogen Life Technologies) according to the manufacturer’s instructions. Following transfection and incubation for 24 h, the cells were harvested and used for the experiments described below.

The cells were randomly divided into three groups (five parallel treatments for each group), including control (non-treated group), BAMBI interference group (transfected with 1 μg BAMBI shRNA) and BAMBI overexpression group (transfected with 1 μg pcDNA3.1(t)-BAMBI).

Total RNA was extracted with TRIzol^® (Invitrogen Life Technologies) according to the manufacturer’s instructions. First-strand cDNA was reverse transcribed using SuperScript II reverse transcriptase (Life Technologies, Gaithersburg, MD, USA) from 1 μg RNA. The primers were designed by GenBank (National Institute for Biotechnology Information, Bethesda, MD, USA) (Table I). An aliquot of 20 ng cDNA was used for the 50 ml qPCR reaction. The PCR reaction was performed using the Applied Biosystems 7500 Real-Time PCR system (Life Technologies). The protocol was as follows: One cycle of 94°C for 4 min, 40 cycles of 94°C for 30 sec, 60°C for 30 sec and 72°C for 30 sec. The PCR products were assayed by dissociation curve to verify single product generation at the end of the assay. Following the PCR reaction, the relative expression levels were calculated using the SDS 1.3 software (Life Technologies) on the Applied Biosystems 7500 Real-Time PCR system.

Paraffin tissue sections were prepared and dewaxed. Then the sections were blocked in 4% evaporated milk for 30 min. Following incubation with the primary and second antibodies for 24 h and 30 min, respectively, the signals were stained by horseradish peroxidase (HRP) catalyzed TMB chromogenic staining. For cell immunohistochemistry, the cells were first cultured on the sterile glass cover slips and then fixed in 4% paraformaldehyde solution for 30 min. Subsequently, the cells were blocked in 4% evaporated milk for 30 min. Following incubation with the primay antibodies for 24 h at 4°C and fluorescein isothiocyanate-conjugated secondary antibody (Santa Cruz Biotechnology, Inc.) for 30 min. Incubation for 5 min with fluorescent dye 4′-6-diamidino-2-phenylindole was used to stain the nucleus. Then the signals were visualized by a fluorescence microscope (Axiover 200M; Carl Zeiss, Oberkochen, Germany).

Cells were lysed in radio-immunoprecipitation assay (Life Technologies) buffer on ice following centrifugation at 25,200 × g for 15 min to obtain the supernatant. The extracted protein samples were separated by 12% SDS-PAGE and transferred onto polyvinylidene fluoride membranes (Amersham Pharmacia Biotech, Picastaway, NJ, USA). The membranes were blocked in 5% skimmed milk for 30 min and subsequently incubated with primary antibodies overnight at 4°C. Following washing with phosphate-buffered saline (PBS) three times, the samples were incubated with secondary antibodies conjugated to horseradish peroxidase for 1 h at room temperature and washed with PBS three times. The membranes were then incubated with the secondary antibody. The blots were evaluated using the chemiluminescence system SuperSignal West Pico Chemiluminescent Substrate (Pierce Biotechnology, Inc., Rockford, IL, USA). Experiments were performed as three independent repeats to assess the relative protein levels.

Cell invasion assays were performed in 24-well FluoroBlok cell culture inserts (BD Biosciences, San Jose, CA, USA) with an 8-μm-pore size polyethylene terephthalate membrane (BD Biosciences). The insert was coated with 200 μl Matrigel matrix (1 μg/μl, BD Biosciences) and incubated at 4°C overnight. Following starvation for 6 h in serum-free Dulbecco’s modified Eagle’s medium (DMEM; Life Technologies), cells were harvested from one subconfluent 10-cm dish by incubation with cell dissociation buffer (Life Technologies), centrifuged at 448 × g for 5 min and resuspended in DMEM. The cells (1×10⁵ in 500 μl DMEM) were seeded onto the insert and 250 μl DMEM with 10% FBS was added into the lower chamber of the transwells. Following incubation for 18 h at 37°C, the medium inside the insert was removed and the insert was placed in a fresh 24-well plate. The invaded cells at the reverse side of the insert were labeled with the fluorescent dye Calcein AM (4 μM in PBS; BD Biosciences) for 1 h at 37°C. Fluorescence was detected at 494 nm/517 nm (excitation/emission wavelength) using a Beckman DU-800 (Beckman Coulter, Inc., Carlsbad, CA, USA) ultraviolet spectrophotometer.

All data are presented as the mean ± standard error of the mean of independent experiments. One-way analysis of variance was used to determine differences among groups. The normality and constant variance for experimental data were analyzed by Levene’s test. Data that did not have homogenous variance underwent logarithmic transformation to meet the requirements of analysis of variance. P≤0.05 was considered to indicate a statistically significant difference. If P-values exceeded the critical value (P<0.05), the Newman-Keul’s post hoc test was performed to compare the groups. The statistical analyses were performed each month. Kaplan-Meier curves were assembled in Prism version 6 (GraphPad software, San Diego, CA, USA). Mean survival times were estimated from the Kaplan-Meier curves. Comparisons between survival curves were made using the log-rank test.

To the best of our knowledge, the present study was the first to assess the expression of BAMBI in gastric cancer. The expression levels of BAMBI mRNA in tumor lesions of patients with gastric cancer were assessed using qPCR. Out of 50 tumor samples, 47 had a higher expression level of BAMBI than the respective adjacent normal tissue (Fig. 1A). In addition, the results showed that the upregulated expression levels of BAMBI were correlated with distal metastasis and recurrence of disease (Table 2). The expression of BAMBI protein was upregulated in gastric tumor tissue rather than normal tissue (Fig. 1B). Kaplan-Meier survival curves (Fig. 1C) showed that BAMBI expression was associated with overall and disease-free survival. Individuals with low expression levels of BAMBI mRNA (relative expression of 1–2) showed a significantly higher overall survival and disease-free survival rate than individuals with high expression levels of BAMBI mRNA (relative expression of 3–4).

To investigate the effect of BAMBI knockdown on human gastric cells, N87 cells were transfected with the BAMBI interference vector, and the resulting characteristics were assessed. Following transfection with shBAMBI, the expression of BAMBI decreased significantly (Fig. 2A). In addition, a time-dependent effect was revealed following transfection, with mRNA expression levels reaching a minimum 96 h post transfection.

Knockdown of BAMBI also had a negative effect on the cell growth and invasion ability of N87 cells (Fig. 2B–D). Similarly, the effect on cell viability was observed to be time dependent. The lowest cell viability was found 96 h post transfection. Thus, 96 h of incubation following transfection were the optimal conditions in the present study. Subsequently, the invasion and migration abilities were calculated and showed a significant decrease 96 h post transfection with the BAMBI interference vector (P<0.05).

The mechanism underlying the BAMBI expression in gastric cancer and decreased cell migration following BAMBI knockdown was investigated. The effects of shBAMBI transfection on tumor migration factors, including β-catenin and TGF-β, were confirmed. Under the optimal treatment conditions, 96 h posttransfection, the transcription of β-catenin and TGF-β expression levels were significantly decreased (P<0.05; Fig. 3A). Western blot analysis and immunofluorescence also suggested a decreased expression of β-catenin and TGF-β following knockdown of BAMBI (Fig. 3B and C).