A total of 40 GC patients were recruited at the Sun Yat-sen Memorial Hospital of Sun Yat-sen University (Guangzhou, Guangdong, China), and tumor and corresponding non-tumor control tissues were collected. After washing with sterile phosphate-buffered saline (PBS), all tissue samples were immediately stored at −80°C until further study. This study was approved by the Ethics Committee of Sun Yat-sen Memorial Hospital, and written informed consent was obtained from all patients.

The GC cell line SNU-1 was obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA), and GC cell line BGC823 was purchased from Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences. Both SNU-1 and BGC823 cells were maintained in RPMI-1640 medium with 100 U/ml penicillin, streptomycin and 10% fetal bovine serum (FBS, cat. no. 100–106; Gemini Bio Products, West Sacramento, CA, USA) at 37°C and 5% CO₂.

Negative control siRNA (NC) and two siRNAs against human ANGPTL4 (siRNA1, siRNA2) were designed and obtained from GenePharma (Shanghai, China). The sequence of siRNA1 was 5′-GGUGACUCUUGGCUCUGCC-3′ (sense) and 5′-GGUGACUCUUGGCUCUGCC-3′ (antisense). The sequence of siRNA2 was 5′-AGGGAAUCUUCUGGAAGAC-3′ (sense) and 5′-GUCUUCCAGAAGAUUCCCU-3′ (antisense). According to the manufacturer's instructions, SNU-1 or BGC823 cells were transfected with 40 nmol/l of ANGPTL4 siRNA1, siRNA2 or NC siRNA using Lipofectamine^® 3000 (Invitrogen, Carlsbad, CA, USA). qRT-PCR and western blot analysis were used to detect the knockdown efficacy of the ANGPTL4 siRNAs.

Total RNA was extracted from the SNU-1 and BGC823 cells using TRIzol reagent (Invitrogen), and 5 µg RNA was used for synthesizing cDNA using the Revert Aid First Strand cDNA Synthesis kit (Invitrogen, UK) according to the manufacturer's instructions. Quantitative real-time PCR was performed using a 7900 HT Fast system (Applied Biosystems, Life Technologies, Foster City, CA, USA). During the course of the reaction, the Cq value was obtained, and the results were analyzed using 2^−∆∆Cq calculation. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an endogenous housekeeping gene. The sequence of human ANGPTL4 PCR primers was 5′-GGCGAGTTCTGGCTGGGTCT-3′ (sense) and 5′-TGGCCGTTGAGGTTGGAATG-3′ (antisense). The sequences of the GAPDH primers were: 5′-ATGTCGTGGAGTCTACTGGC-3′ (forward) and 5′-TGACCTTGCCCACAGCCTTG-3′ (reverse).

Total proteins of the human GC cells were prepared using RIPA buffer (cat. no. R3792). Cell lysis was preformed on ice using ultrasonication. Cell debris was centrifuged (13,000 rmp, 20 min, 4°C) and the supernatant was collected. The concentrations of proteins in each sample were detected using a protein assay reagent (Bio-Rad Laboratories, Inc., Hercules, CA, USA). SDS-PAGE (10%) was used to separate total proteins (40 µg). The separated proteins were then transferred onto Immun-Blot^® PVDF membranes (cat. no. 162-0177; Bio-Rad Laboratories). The membranes were blocked with 5% skim milk in TBS for 2 h at room temperature, and incubated with the primary antibody overnight at 4°C. Then the membranes were incubated with secondary horseradish peroxidase antibody (HRP, cat. no. 074-1806; KPL Co., USA). The results were analyzed using a chemiluminescence substrate kit and an ECL system (both from Amersham Pharmacia Biotech Inc., Piscataway, NJ, USA). The primary antibodies were ANGPTL4 (1:1,000; AF3485; R&D Systems Inc., Minneapolis, MA, USA) and GAPDH (1:200; sc-51631; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA).

Cell migration and cell invasion assays were performed using Transwell chambers (8-µm pore size; BD Biosciences, San Jose, CA, USA) with or without Matrigel coating (BD Biosciences). Briefly, SNU-1 or BGC823 cells (0.2 ml, 2×10⁵ for the invasion assay; 1×10⁵ for the migration assay) were seeded in the upper chamber, and the lower chamber was filled with 10% FBS containing medium (0.6 ml). After 24 h, cells on the upper chamber were removed, and migrated cells were fixed and stained with 5% crystal violet. The cells were captured and counted under a microscope.

After transfection, SNU-1 or BGC823 cells were seeded into a 96-well plate at a concentration of 2×10⁴ cells per well. Incubated for 24 h, cells were then washed with PBS and 20 µl of MTT solution was added to each well. After at least 30 min of incubation at 37°C, 150 µl dimethyl sulfoxide (DMSO) was added to each well. The plate was then shaken at room temperature for ≤10 min, and absorbance was determined at 490 nm by a microplate reader (Sunrise™; Tecan Group Ltd., Switzerland).

Apoptosis rate of GC cells after the transfection of ANGPTL4 siRNA was measured by flow cytometry using Annexin V/PI double staining (BD Biosciences, Franklin Lakes, NJ, USA). The cells were seeded in a 24-well plate at a concentration of 2×10⁴ per well. Following overnight incubation, cells were washed at least twice with cold phosphate-buffered saline (PBS). Cells were re-suspended with 0.5 ml binding buffer supplemented with 5 µl PI and 5 µl FITC-labeled Annexin V and incubated at room temperature for 10 min (protected from light). Apoptotic SNU-1 or BGC823 cells were analyzed in triplicates and the analysis was repeated three times independently by flow cytometry on a FACScan (Beckman Instruments, Fullerton, CA, USA).

Paraffin-embedded tissues were sliced into 4-µm sections, and then the sections were treated with xylene and ethanol to remove paraffin, and were rehydrated in PBS. All sections were immersed in 0.01 M sodium citrate buffer (pH 6.0) at 95°C for ≤20 min for antigen retrieval. The sections were then treated with 3% H₂O₂ for 10 min to block the endogenous peroxidase activity, and incubated with 5% normal goat serum to block the non-specific binding. Samples were then incubated overnight at 4°C with primary rabbit antibody against human ANGPTL4 (R&D Systems). After washing with PBS, the sections were sequentially incubated with goat anti-rabbit HRP, and the reaction signals were visualized by the diaminobenzidine reaction (DAB; Dako Ltd.) and sections were counterstained with hematoxylin.

After transfection with ANGPTL4 siRNA, SNU-1 or BGC823 cells (2×10⁴) were seeded in a 96-well plate and incubated overnight at 37°C. Then cells were treated with various concentrations (0, 2.5, 5.0, 10.0, 20.0 and 40.0 µg/ml) of cisplatin (Sigma, Santa Clara, CA, USA). MTT assay was applied to examine cell viability.

Statistical analyses were conducted using SPSS 19.0 software (SPSS, Inc., Chicago, IL, USA). All data are expressed as the mean ± standard deviation (SD). The Student's t-test was carried out to compare the level of cell proliferation, apoptosis, migration, invasion and cisplatin resistance among groups, and P<0.05 was considered to be indicative of statistical significance.

To study the roles of ANGPTL4 in GC pathogenesis, protein expression of ANGPLT4 was analyzed by IHC in tumor samples and adjacent normal specimens of 40 patients with GC (I, II, III and IV stage). The results indicated that the expression of ANGPTL4 was significantly increased in the GC tissues compared with that noted in the normal tissues. In addition, ANGPTL4 expression levels were associated with the clinical phase of GC (Fig. 1).

To investigate the role of ANGPTL4 in human GC, two siRNAs targeting ANGPTL4 were designed, and SNU-1 and BGC823 cells were transfected. The knockdown efficiency of those two siRNAs was validated by detecting the mRNA expression level using qRT-PCR assay. Results from qRT-PCR showed that, compared to NC, the mRNA expression of ANGPTL4 in SNU-1 and BGC823 cells was extremely decreased after siRNA1 and siRNA2 transfection (P<0.01 and P<0.001, respectively; Fig. 2A and B). We further assessed the ANGPTL4 protein level in the transfected SNU-1 and BGC823 cells, and the results indicated that ANGPTL4 protein was also downregulated (Fig. 2C and D). These results demonstrated that the siRNAs were effective in silencing ANGPTL4 expression.

In order to explore the biological function of ANGPTL4 in human GC, we silenced the ANGPTL4 expression by siRNA in SNU-1 and BGC823 cells. Cell proliferation and apoptosis were then detected. MTT assay indicated that knockdown of ANGPTL4 significantly inhibited the cell proliferation in both SNU-1 and BGC823 cells, compared with NC (P<0.05, P<0.01 and P<0.001, respecively; Fig. 3). Annexin V/PI staining and flow cytometry assay were employed to further determine whether cell apoptosis could be affected by ANGPTL4. We found that the percentage of apoptosis was significantly upregulated after the silencing of ANGPTL4 in both the SNU-1 (P<0.001, Fig. 4A) and BGC823 cells (P<0.001, Fig. 4B). This indicated that knockdown of ANGPTL4 inhibited cell proliferation and promoted cell apoptosis in human GC cell lines.

Migration and invasion are two important factors involved in the progression of cancer cells. To determine whether ANGPTL4 is involved in the migration and invasion of human GC cells, we performed Transwell assays. The results showed that cell migration and invasion were significantly reduced after transfection of SNU-1 and BGC823 cells by siRNA1 or siRNA2 compared to the blank control (P<0.001, Fig. 5). These results confirm that knockdown of ANGPTL4 inhibited human GC cell migration and invasion.

Whether the expression of ANGPTL4 affects the sensitivity of human GC cells to cisplatin remains unexplored. To ascertain the effects, after transfection with ANGPTL4 siRNA and stabilization, human GC cells were treated with various concentrations (0, 2.5, 5.0, 10.0, 20.0 and 40.0 µg/ml) of cisplatin for 24 h. The results of MTT assay revealed that knockdown of ANGPTL4 significantly decreased the viability of the SNU-1 (P<0.05, P<0.01, P<0.001, Fig. 6A) and BGC823 cells (P<0.05, P<0.01, P<0.001, Fig. 6B) in response to cisplatin in a dose-dependent manner. These results obviously indicated that ANGPTL4 knockdown enhanced the sensitivity of human GC cells to cisplatin.