This study of human gastric cancer samples was approved by the Ethics Committee of the First Hospital Affiliated to Fujian Medical University. Seventy formalin-fixed and paraffin-embeded gastric cancer tissues, and 19 samples of normal gastric tissues that were located away from the cancer were used as contorls. The samples were provied by Pathology Department and Department of Gastrointestinal Surgery Section II of the First Hospital Affiliated to Fujian Medical University. The patient group samples were conserved during the period from 2008 to 2009, and the patient group consisted of 34 men and 36 women with a median age of 63 years. None of the gastric cancer patients included in this study had received any preoperative chemotherapy or other therapy such as radiotherapy. Postoperative chemotherapy was indicated for patients who had indications of chemotherapy with intravenous infusion of 5-fluorouracil-based chemotherapy. Follow-up was done every three months for the first 2 years and annually thereafter.

Immunohistochemistry staining was performed using the standard immunoperoxidase staning procedure. In brief, serial 4-μm slices were obtained from formalin-fixed and paraffin-embedded tissues specimens. Then dewaxed in xylene, rehydrated in alcohol, and incubated with fresh 3% hydrogen peroxide (H₂O₂) for 25 min at room temperature. Sections were antigen-retrieved by boling for 15 min in citrate buffer (pH 6.0) and followed by a wash step with phosphate buffered saline (PBS). The non-specific antigen of sections were blocked with appropriate normal serum in PBS. RIP1 mouse monoclonal anti-human antibody (1:500, Abcam Biotechnology) was diluted and placed on the sections overnight in humidified boxes at 4°C. The sections were then washed with PBS for 6 min followed by an incubation with an UltraSensitive S-P kit (Maixin-Bio, Fuzhou, China) according to the manufacturer's instructions. After exposure to stable 3, 3-diaminobenzidine for 4–6 min, slides were counterstained with hematoxylin, dehydrated, and mounted. Cells with a deposition of buffy-colored granules in the sections were scored as RIP1-positive. The expression of RIP-1 was evaluated using the Image-Pro Plus 6.0 software (Media Cybernetics, Inc. USA). Five slices were randomly selected for each group and five fields were selected for each slice. The expression of RIP-1 in each group was semiquantitatively analysed using mean optical density (MOD) = the intergral optical density (IOD) / the positive area, in 25 fields, respectively.

The gastric cancer cell lines MGC, MKN-74, HGC and AGS were preserved by the Key Laboratory of the Ministry of Eduction for Gastrointestinal Cancer, Fujian Medical Universtiy, Fuzhou, Fujian, China. The MGC, MKN-74 and HGC cell lines were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) (both from Gibco, Carlsbad, CA, USA). AGS was incubated in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FBS. All the cells were incubated at 37°C under 95% air and 5% CO₂.

Semiquantitative RT-PCR was used to detect the expression of RIP-1 gene in the gastric cancer cell lines MGC, MKN-74, HGC and AGS, respectively. Total RNA was isolated from the cultured cells grown in 6-well plates using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions and quantified by UV 260/280 nm to an absorption ratio of >1.8. Total RNA (2 μg), according to the different concentrations of sample RNA, were reverse transcribed to cDNA in a final volume of 20 μl using the AVM First Strand cDNA synthesis kit (Invitrogen) following the manufacture's instructions. Additionally, they were used as cDNA template for PCR. The PCR primers used for amplification are shown in Table I. All PCRs were performed with Thermo Scientific SYBR Green qPCR kit on an Applied Biosystems StepOne Real-time PCR System. PCR conditions were: 95°C for 2 min, 95°C for 15 sec, and 60°C for 30 sec for 40 cycles. All gene transcripts were quantified by the 2^−ΔΔCt method.

Western blot analysis and detection of the blotted product were carried out as described previously (14). The following primary antibodies were used from the companies as listing, respectively: RIP-1 mouse monoclonal anti-human antibody (1:1,000), VEGF-C rabbit polyclonal anti-human antibody (1:1,000), c-jun (AP-1) monoclonal rabbit anti-human antibody (1:1,000), and p-c-jun (p-AP-1) monoclonal rabbit anti-human antibody (1:1,000) (all from Abcam Biotechnology), nuclear factor-κB (NF-κB) (p65) monoclonal mouse anti-human antibody (1:500), p-NF-κB (p-p65) monoclonal mouse anti-human antibody (1:500) (both from Cell Signaling Technology, Danvers, MA, USA), β-actin monoclonal mouse anti-human antibody (1:1,500) (from Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA).

According to the siRNA design principle, four suitable siRNA target sequences (R-1, R-2, R-3 and R-4), from human RIP-1 gene GenBank accession no. NM_003804, were synthesized as follows, respectively: R-1, GCACAAATACGAACTTCAA; R-2, GGCCAATTCCAAGTCATAT; R-3, TACAACAGAGAGGAGGAAA; and R-4, TTGTGATAATGACTTCCA. A small hairpin RNA (shRNA) of human RIP-1 in a phU6 gene transfer vector and the negative control (NC) sequence encolding an enhanced green fluorescent protein (EGFP) sequence with the puromycin resistance were constructed by Genechem Co. Ltd. (Shanghai, China). Additionally, all the plasmids were verified by DNA sequencing. The AGS and HGC cells were cultured in appropriate medium supplemented with 10% FBS. When the AGS and HGC cells were at ~90% confluency, we transfected the plasmids into these cells. We used Lipofectamine 2000 (Invitrogen) to transfect cells according to the manufacturer's instructions. We used a microscope to observe the transfection efficiency by counting the percentage of cells that were EGFP-positive.

In order to create better stable transfection for later experiments, we constructed lentiviral-mediated siRNA targeting RIP-1 vector. We constructed lentiviral vectors for human RIP-1 small hairpin RNA (shRNA) encoding a green fluorescent protein (GFP). The lentivirus vectors containing RIP-1 shRNA were constructed by ligating the digested pGCSIL-GFP and the RIP-1 shRNA PCR product and then verified by DNA sequencing. ShRNA for the negative controls was preserved in our lab. AGS and HGC cells were plated in 6-well plates with appropriate medium supplemented with 10% FBS overnight. When these cells were at approximately 35% confluent, they were used in siRNA virus infection. The multiplicity of infection of (MOI) = 50 was appropriate quantity of virus infection and 1 ml complete medium was added to 8 mg/ml polybrene and mixed with cells and incubated for 12 h at 37°C. Then, the cells were incubated in fresh complete medium containing 10% FBS for 24 h. Microscope was used to confirm the infection efficiency by observing the GFP-positive cells.

Cell proliferation was assessed by MTT assay. Three groups of cells (blank control group, NC-RNAi-LV group and R-1-RNAi-LV group) were seeded into 96-well plates at density of 10³ cells/well with 100 μl of medium supplemented with 10% FBS. The proliferative activity was determined after 1, 2, 3, 4, and 5 days by the addition of 10 μl of sterile 3-(4,5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) (5 mg/ml; Sigma) to each well. The reaction was terminated after 4 h of incubation at 37°C through the addition of 100 μl of dimethyl sulfoxide (DMSO; Sigma). The optical density (OD) value was obtained by measuring absorbance at a wavelength of 450 nm. Each well test was repeated 6 times.

Invasion assays were performed using Transwell 24-well plates with 8-μm polycarbonate membranes (BD Biosciences). Briefly, the upper side of the membranes was coated with Matrigel matrix (20 μg/well) and the membranes were then air-dried for 1 h of incubation 37°C. The lower side of the membranes was coated with 5 μg fibronectin (BD Biosciences). To test the ability of cell invasion, blank control group, NC-RNAi-LV group and R-1-RNAi-LV group gastric cancer cells (2×10⁵) in 200 μl of RPMI-1640 (DMEM for AGS cells) medium with 2.5% FBS were placed in the upper chamber. The lower chamber was filled with 700 μl RPMI-1640 (DMEM for AGS cells) medium with 10% FBS as the chemoattractant. The invasion chamber was incubated for 8 h at 37°C and 5% CO₂. The cells on the upper surface of the membrane were removed by gentle scrubbing with a cotton swab. Membranes were fixed in a stationary liquid of 95% ethanol and 5% acetic acid for 30 min and stained with crystal violet. The number of cells on the lower surface of the membrane in 5 random visual fields (×400) was then counted using a bright field light microscope. Each assay was repeated in triplicate.

Male four to six-week-old athymic BALB/c nu/nu mice were obtained from the Shanghai SLAC Laboratory Animals Co., (Shanghai, China) and kept in the Experimental Center of the Fujian Medical University. All the mice were handled according to the recommendations of the institutional guidelines and were approved by the Ethics Committee of the Medical Faculty of the Fujian Medical University. Blank control group, NC-RNAi-LV group and R-1-RNAi-LV group were randomly divided into three groups using 15 mice (five mice/group) for the subcutaneous gastric cancer nude mouse model. Three groups of different gastric cancer cells (10⁶) in 100 μl serum-free medium were seeded subcutaneously into the upper flank region of the nude mice. Tumor growth was monitored and measured in two dimensions (width and length respectively) every 3 days. The size of tumor growth was estimated using the formula = width² (mm²) × length (mm)/2 (the width and length are the shortest and longest diameters of the subcutaneous tumor). After 5 weeks, nude mice were sacrificed and the tumors were collected and weighed immediately. Then all the samples were fixed in 10% formalin solution for the immunohistochemical analysis.

The statistical analysis used GraphPad Prism 5 software. Data were analyzed by the one-way ANOVA or Student's t-test. The data were expressed as the means ± standard deviation (SD). Survival curves were calculated by the Kaplan-Meier method and compared by using the log-rank test. A P-value <0.05 was considered to indicate a statistically significant difference.

We evaluated the expression of RIP-1 in 70 clinical samples from gastric cancer patient and 19 normal gastric tissues distant from the cancer site using immunohistochemistry. In the normal gastric tissues, low levels of RIP-1 expression were detected (Fig. 1A); but, in the gastric cancer tissues high levels of expression of RIP-1 were found (Fig. 1B). The MOD of RIP-1 in the normal gastric tissues (0.1478±0.1346) was markedly lower than gastric cancer tissues (0.3576±0.0961; P<0.001; Table II). Thus, RIP-1 was overexpressed in the gastric cancer tissues. Importantly, the RIP-1 immunoreactivity was positive at the site of invasion, but little or no immunoreactivity was detected at the parts of interstitial substance (Fig. 1C and D). This suggested that RIP-1 overexpression may promote gastric cancer invasion.

Table III shows the relationship between RIP-1 expression and select clinicopathological parameters. RIP-1 expression did not vary significantly with age, gender, tumor size and histological grade; however, there was a significant relationship between RIP-1 overexpression and clinical stage (I–II vs. III–IV; P=0.031), lymph node metastasis (no vs. positive; P=0.036) and Helicobacter pylori (negative vs. positive; P=0.0312).

The expression levels of RIP-1 were positively correlated with the prognosis of gastric cancer patients. The study samples were split into two groups according to the MOD expression levels: a low RIP-1 expression levels group (<0.3011, which is the average MOD values of 70 gastric cancer patients of tissues) and a high RIP-1 expression levels group (≥0.3011). Additionally, the patients with high expression of RIP-1 had very discouraging prognosis; however, the patients with low expression of RIP-1 had a better prognosis. Thus, the survival and prognosis of the low expression of RIP-1 group was significantly longer than the high expression of RIP-1 group (P<0.05) (Fig. 2).

We used qRT-PCR and western blot technology to detect the expression of RIP-1 in human gastric cancer cell lines (MGC, MKN-74, HGC, and AGS). RIP-1 mRNA was detected in all four cell lines by qRT-PCR. In addition, expression of RIP-1 protein in all four cell lines was confirmed. The expression quantity of RIP-1 mRNA and protein in HGC and AGS cells was higher compared with the MGC and MKN-74 cells (Fig. 3). Thus, we used the HGC and AGS cells to further evaluate the role of RIP-1 in gastric cancer.

The RNAi effect of silencing RIP-1 gene in the HGC and AGS cells was assessed. The DNA sequencing results verified that RIP-1 siRNA plasmid construction was successful. To determine the level of RIP-1 expression levels 72 h after transfection in HGC and AGS cells we used qRT-PCR. Expression of the siRNA plasmids vectors (NC/R-1/R-2/R-3/R-4/siRNA) enhanced green fluorescent protein (EGFP) 72 h after transfection, as shown by fluorescence microscopy, in the HGC and AGS cells (Fig. 4A and B). qRT-PCR revealed a decrease in RIP-1 mRNA in siRNA-transfected cells. The R-1/siRNA plasmid vector resulted in a higher suppression of the levels of RIP-1 mRNA expression than other vectors (R-2/siRNA, R-3/siRNA, and R-4/siRNA), while non-transfected and NC/siRNA vectors had no effect on the levels of RIP-1 mRNA expression in the HGC and AGS cells (Fig. 4C and D). Because semi-quantitative analysis by qRT-PCR showed that there was marked inhibition between R-1/siRNA and non-transfected siRNA groups in the HGC and AGS cells, we selected R-1/siRNA sequence to knock down the RIP-1 in our further studies.

We used lentiviral-mediated R-1/siRNA targeting RIP-1 vector to improve the stable transfection for the next experiment. Transfection efficiency was quantified by counting the cells under a fluorescent microscope 72 h after infection, and the efficiency of HGC and AGS cells were >80% (Fig. 5A–D). The qRT-PCR showed that the expression of RIP-1 mRNA was significantly inhibited after infection in HGC and AGS cells. Compared with the control group and NC-RNAi-LV group, the R-1-RNAi-LV group showed relatively lower expression of RIP-1 mRNA in HGC and AGS cells (Fig. 5E and F). In accordance with mRNA, western blotting revealed that RIP-1 protein expression was suppressed in the R-1-RNAi-LV group compared with the control group and NC-RNAi-LV group in HGC and AGS cells. There were distinct densitometric differences between R-1-RNAi-LV group and control group and NC-RNAi-LV group in HGC and AGS cells (Fig. 5J, H, J and K).

To experimentally address the biological function of RIP-1 in gastric cancer cells, we analyzed the effect of lentivirus-mediated RIP-1 siRNA on HGC and AGS cell proliferation and invasion by MTT assay and cell invasion assay, respectively. MTT experiment results showed that compared to control group and NC-RNAi-LV group, the cell proliferation of the R-1-RNAi-LV group was slower in HGC and AGS cells (Fig. 6). The cell viability of R-1-RNAi-LV group was decreased when compared with control group in both gastric cancer cell lines. The results from cell in vitro invasion indicated that the numbers of cells that penetrated the basal membrane from the R-1-RNAi-LV group was significantly less than the blank control group, which was similar to the number of cells from the NC-RNAi-LV group in HGC and AGS cells (Fig. 7).

To investigate the mechanisms of RIP-1 silencing responsible for the decrease in proliferation and invasion, we assessed the changes in levels of NF-κB (p65), p-NF-κB (p-p65), c-jun (AP-1), p-c-jun (p-AP-1) and VEGF-C protein and VEGF-C mRNA; these factors are vital to the growth and invasion of cancer cells. Western blot analysis revealed that NF-κB (p65), p-NF-κB (p-p65), c-jun (AP-1), p-c-jun (p-AP-1) and VEGF-C protein levels in the R-1-RNAi-LV cells decreased compared with the control NC-RNAi-LV and in HGC cells (Fig. 8). The changes in VEGF-C mRNA expression in the three groups were the same as the levels of VEGF-C protein in HGC cells (Fig. 8F).

To address the impact of RIP-1 expression on gastric cancer tumor biology in vivo, three groups (HGC-control, HGC-NC-RNAi-LV, and HGC-R-1-RNAi-LV groups) of HGC cells were subcutaneously xenografted onto nude mice. Nude mice were all successfully established in the subcutaneous mouse model. None of the mice had ascites or metastasis to the liver, lungs, or lymph nodes at the time of autopsy. At the end of the 5-week experimental period, the size of the subcutaneous xenograft tumor showed significant differences between the HGC-R-1-RNAi-LV and HGC-NC-RNAi-LV groups, which was similar to the HGC-control group. We found that xenograft growth was markedly slower from the second week until the day on which the mice were sacrificed (Fig. 9A and B). Because RIP-1 expression was reduced in the HGC-R-1-RNAi-LV cells, we used immunohistochemistry to analyze the expression of RIP-1 in the nude mice. Compared with the HGC-control and HGC-NC-RNAi-LV groups, the levels of RIP-1 expression were significantly reduced in the HGC-R-1-RNAi-LV group (Fig. 9C and D).