TRIzol reagent and Lipofectamine 2000 were obtained from Invitrogen (Carlsbad, CA, USA). Horseradish peroxidase (HRP) Affinipure goat anti-mouse/rabbit IgG (H+L) and SP9000 were purchased from Zhongshan Golden Bridge Biotechnology Co., Ltd. (Beijing, China). FITC Affinipure goat anti-mouse IgG was from Jackson ImmunoResearch Co. The pCDNA3.1 vector and pCDNA3.1/EphA2 were obtained from Invitrogen Co. The rabbit anti-EphA2 antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA, USA).

Eighty-two cases of surgically resected primary gastric carcinoma tissues, paired adjacent mucosa (2–5 cm from the margin of the gastric carcinoma) and normal mucosa at the surgical margin (at least 5 cm from the margin of the gastric carcinoma and histologically proven) tissues collected between January and July 2007 at the Inpatient Department of the Fourth Hospital of Hebei Medical University were used in this study. These gastric carcinoma tissues included 30 cases of well-differentiated gastric carcinoma and 52 cases of moderately and poorly differentiated gastric carcinoma [superficial muscle invasive (n=20), deep muscle invasive (n=62), lymph node metastasis positive (n=51), lymph node metastasis negative (n=31)]. Thirty cases were randomly chosen to detect EphA2 mRNA and protein expression [9 cases of well differentiated tissues and 21 cases of moderately and poorly differentiated tissues; superficial muscle invasive (n=8), deep muscle invasive (n=22), lymph node metastasis positive (n=21), lymph node metastasis negative (n=9)]. Experimental protocols were approved by the Institutional Human Care and Use Committee of Hebei Medical University.

Human gastric cancer cells (MGC803) were grown in humidified air with 5% CO₂ in an incubator at 37°C in RPMI-1640 supplemented with 10% fetal bovine serum, 100 μg/ml streptomycin and 100 U/ml penicillin. In order to investigate the effect of GA on cell proliferation, the cells were divided into 20, 40, 200, 400 and 2,000 nmol/l groups. In order to explore the mechanism of GA on apoptosis of MGC803 cells, the cells were divided into six groups: control, GA (200 nmol/l), GA + pYr-3.1-shNC, GA + pYr-3.1-shNC; GA + pcDNA3.1-EphA2-IRES-EGFP, and GA + pcDNA3.1-NC-IRES-EGFP; and were collected at 48 h.

The tissues were fixed in 4% formaldehyde. Antigen recovery was performed using a microwave. The sections were incubated with primary antibodies against EphA2 (1:100) overnight at 4°C. On the following day, the sections were incubated with polyperoxidase anti-mouse IgG at 37°C, and finally stained with diaminobenzidine. The sections were imaged with an Olympus microscope. Positive staining of EphA2 protein was located in the cytoplasm or/and membrane of epithelial cells. The results were divided into 4 grades as previously described (6).

The cells were fixed in a 4% paraformaldehyde solution and permeabilized with 0.1% Triton X-100. Following incubation with anti-EphA2 (1:100; Santa Cruz) antibodies and the secondary antibody conjugated with HRP, the slices were stained with diaminobenzidine using the method introduced by Feng et al (7).

Total RNA was extracted from gastric cancer tissues or MGC830 cells with TRIzol reagent according to the instructions of the manufacturer. Total RNA (2 μg) was reverse transcribed into cDNA by AMV reverse transcriptase at 42°C for 1 h and then heated to 94°C for 5 min in a total reaction volume of 20 μl. The PCR condition used for EphA2 and the internal reference GAPDH was 94°C for 5 min followed by 33 cycles of 94°C for 45 sec, 60°C for 45 sec and 72°C for 45 sec, and 72°C for 7 min. The specific primers were as follows: EphA2 sense, 5′-CCAAGTTCGCTGACATCGT-3 and anti-sense, 5′-GCCATGAAGTGCTCCGTAT-3′; GAPDH sense, 5′-ACCACAGTCCATGCCATCAC-3′ and antisense, 5′-TCC ACCACCCTGTTGCTGTA-3′. The expected PCR products were 196 and 486 bp for EphA2 and GAPDH, respectively. The amplicons were analyzed by electrophoresis, imaged using UVI gel imaging system and quantified using Gel-Pro Analyzer 3.1 software. Expression levels of EphA2 were normalized to the internal reference GAPDH.

Total protein extraction from the collected tissues and cultured cells was performed as described previously (8). The protein extracts were separated by 10% SDS-PAGE and then transferred to PVDF membranes. The membranes were incubated overnight at 4°C with anti-EphA2 (1:100; Santa Cruz) and β-actin (1:200; Santa Cruz) antibodies. Subsequently, the membranes were incubated with HRP-conjugated goat anti-rabbit IgG (1:5,000) and then exposed to X-ray film using an enhanced chemiluminescence system (Pierce Biotechnology, Rockford, IL, USA). EphA2 protein expression was quantified by comparison with β-actin.

MGC803 cells in a logarithmic phase were inoculated into a 96-well plate with 1×10⁵ cells in each well. Cell viability at 24 h was examined using the MTT method. OD_490nm values were detected in six duplicate wells and their averages were used to plot the growth curve and calculate the growth inhibition rate of each treatment using the following formula: Growth inhibition rate (IR) = OD_490nm of the control group − OD_490nm of the treatment group/OD_490nm of the control group × 100%.

The tissues and the cell pellets treated by GA were fixed in 70% ethanol, and stained with 500 μg/ml propidium iodide (Sigma) containing 3 Kunitz units of RNase and 1% fetal bovine serum in PBS. Then, the stained cells were analyzed in an Epics-XLII flow cytometer (Beckman Coulter, Miami, FL, USA) and the number of the cells in each phase of the cell cycle was calculated using the ModFit LT cell cycle analysis program according to the manufacturer’s instructions. The proliferation index (PI) represents the number of proliferating cells and was calculated using the following equation: PI = (S + G₂M)/(S + G₂M + G₀/G₁) × 100%.

The sh-EphA2 vector and negative control sh-Scramble vector were designed and produced by Invitrogen. The expression vector of EphA2 (pcDNA3.1-EphA2-IRES-EGFP) and the control vector (pcDNA3.1-IRES-EGFP) were designed and produced by the Beijing Fungenome Technology Co. (Beijing, China). Transient transfection of MGC803 cells was carried out using Lipofectamine 2000 according to the manufacturer’s instructions. At ~80% confluency, the cells were transfected with 2.0 μg of vector DNA with 4 μl of Lipofectamine 2000 in 2 ml of serum-free RPMI-1640 medium. Six hours after the transfection, the cells were incubated for 48 h with normal RPMI-1640 medium containing 10% fetal bovine serum. Then, 200 nmol/l GA was added to the medium, and the cells were cultured for 48 h. Subsequently, EphA2 expression, PI, cell cycle distribution and apoptosis were determined.

Data are expressed as mean ± standard variation and were analyzed using SPSS 15.0 statistical software package. Differences between samples were tested using single factor analysis of variance and LSD method for multiple comparisons. A p-value <0.05 was considered as a significant difference. Before comparison, data homogeneity of variance was first examined using the F test. In the case of heterogeneity of variance, the approximate variance F test/Welch method was used.

As shown in Fig. 1A and Table I, positive expression of EphA2 protein was noted in the cytoplasm and/or membrane and was higher in the carcinoma tissues than that in the adjacent and normal tissues (P<0.01). Moreover, the positive expression rate was higher in moderately and poorly differentiated tissues than that in the well-differentiated tissues (P<0.05). In addition, the expression of EphA2 in tumors with deep infiltration was higher than that in tumors with superficial layer invasion (P<0.05). Higher expression was noted in cases with lymph node metastasis than in non-lymph node metastasis (P<0.05, Tables II and III). There was no difference in EphA2 mRNA among the tissue groups as shown by RT-PCR (P>0.05, Fig. 1B), while the results of the western blot analysis were identical with IHC and the EphA2 protein expression was increased in the gastric carcinoma tissues.

According to the results of IHC, the carcinoma cases were divided into two groups: EphA2 high-expression group (++~+++) and EphA2 low-expression group (−~+). The apoptosis rate was lower in the EphA2 high-expression group than that in the low-expression groups (P=0.018) and PI had a contrasting trend (P=0.002) as shown by FCM (Fig. 2 and Table IV).

In order to explore the precise effect of EphA2 on the cell proliferation and apoptosis of gastric cancer cells, the MGC803 cell line was used, which overexpressed the EphA2 protein. In addition, RNAi was used to knock down the expression of EphA2. As shown in Fig. 3A and B, MGC803 cells transfected with the sh-EphA2 vector showed low EphA2 mRNA and protein expression, particularly in the S3 vector group. On the contrary, no change in EphA2 protein expression was found between cells transfected with the blank control vector and the untransfected MGC803 cells.

FCM revealed that the number of cells in the S phase decreased and the number of cells in the G₀/G₁ phase was increased. Therefore, the PI decreased following transfection of the MGC803 cells with the sh-EphA2 vector (Table V and Fig. 3C). Furthermore, silencing of EphA2 expression induced cell apoptosis (Table VI and Fig. 3D).

The MTT results showed that GA inhibited the proliferation of MGC803 cells significantly in a time-dependent and dose-dependent manner and the highest inhibition rate reached 77.69±0.91%; IC₅₀ was 558.94 nmol/l after 48 h of GA treatment (Fig. 4A).

FCM analysis showed that the percentage of cells in the S phase was increased while the percentage of cells in the G₂/M phase was decreased in the GA-stimulated cells compared with the control group, and the percentage of cells in the S phase was 36.35±3.04% in the control group, 45.00±1.41% in the middle-dose group, and 45.25±3.04% in the high-dose group, respectively. At same time, the percentage of cells in the G₀/G₁ phase increased (Fig. 4B and Table VII). As shown in Table VIII, GA induced cell apoptosis of MGC803 cells in a dose-dependent manner (P<0.01). As shown in Fig. 4C, GA treatment inhibited the expression of EphA2 protein.

In order to investigate whether GA induces MGC803 cell apoptosis through downregulation of EphA2 protein, transient transfection experiments were performed. Compared with the control group, pcDNA3.1-EphA2-IRES-EGFP effectively upregulated the EphA2 protein and mRNA levels in the MGC803 cells by 2.3 and 2.5-fold. As detected by FCM, GA increased the apoptosis ratio of MGC803 cells; however, the apoptosis ratio in the GA+EphA2 vector group was decreased compared with the GA group. Importantly, knockdown of EphA2 expression increased the apoptosis ratio induced by GA in the MGC803 cells (Fig. 5).