The human gastric adenocarcinoma cell lines SGC-7901 (moderately differentiated) and MKN-45 (poorly differentiated) and the normal human gastric epithelial cell line GES-1 were obtained from the Cell Bank of Xiangya Central Laboratory, Central South University (Changsha, Hunan, China). The human gastric carcinoma cell lines NCI-N87 (well differentiated) and MGC-803 (poorly differentiated) were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China).

MGC-803, MKN-45 and SGC-7901 cells were maintained in RPMI-1640 medium. NCI-N87 and GES-1 cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; HyClone-Pierce, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS; Invitrogen, Carlsbad, CA, USA), penicillin (100 U/ml; Sigma Chemicals, St. Louis, MO, USA), and streptomycin (100 μg/ml; Sigma Chemicals). All cells were cultured in a humidified incubator with 5% CO₂ at 37°C.

Total cellular RNA was isolated from cells of each cell line using TRIzol reagent (Invitrogen) and purified with an RNeasy Mini kit (Qiagen, Hilden, Germany). After being labeled using the miRCURY Array Power Labeling kit (Exiqon, Vedbaek, Denmark), these RNA samples were used as probes to hybridize with the miRCURY microRNA Array kit (Exiqon) in a Hybridization Chamber II (Ambion, Austin, TX, USA) in accordance with the manufacturer’s instructions. Hybridized miRNA microarrays were then scanned using a GenePix 4000B microarray scanner (Axon Instruments, Inc., Union City, CA, USA), and the data were analyzed using GenePix Pro software v6.0 (Axon). Differential miRNA expression (fold change >1.50 or ratio <0.667) and cluster analysis were performed using MeV software (v4.6, TIGR). Some of these differentially expressed miRNAs were validated using qRT-PCR.

Gastric cancer tissue samples were obtained from 23 patients who underwent partial gastrectomy between March and November 2011 at the First Affiliated Hospital and Second Affiliated Hospital of the University of South China (Hengyang, China). Patients included 12 men aged 60.9±9 years and 11 women aged 52.8±9.5 years. These patients did not receive any radiotherapy and chemotherapy before surgery.

Gastric cancer tissues and matched non-tumor tissues were obtained from surgical resections and were each divided into two parts: one for pathological diagnosis and another for extracting total RNA. Pathological diagnoses were performed independently by two pathologists in accordance with the WHO standard for pathological diagnosis of gastric cancer. Nine of these cases were histologically diagnosed as well/moderately differentiated gastric cancers, while 14 were diagnosed as poorly differentiated gastric cancers.

The present study was approved by our institutional review boards of the First and Second Affiliated Hospitals of the University of South China. Each patient provided informed consent before participation in the study.

Total cellular RNA was isolated from the gastric cell lines and tissues using TRIzol reagent (Invitrogen) and then reverse transcribed to cDNA using a Hairpin-it miRNA qPCR quantitation kit (Shanghai GenePharma Co., Ltd., Shanghai, China) in accordance with the instructions. The PCR primers were: hsa-miR-7 (to generate an 82-bp PCR product), 5′-CCACGTTGGAAGACTAGTGATTT-3′ and 5′-TATGGTTGTTCTGCTCTCTGTCTC-3′; human epidermal growth factor receptor (EGFR; to generate a 93-bp PCR product), 5′-AAAGAATACCATGCAGAAGGAGG-3′ and 5′-GACATCACTCTGGTGGGTATAGA-3′; U6 (as an internal control, to generate a 70-bp PCR product), 5′-ATTGGAA CGATACAGAGAAGATT-3′ and 5′-GGAACGCTTCACGA ATTTG-3′, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH, an internal control, to generate a 113-bp PCR product), 5′-CATGAGAAGTATGACAACAGCCT-3′ and 5′-AGTCCTTCCACGATACCAAAGT-3′. All of the primers were obtained from Shanghai GenePharma. The qRT-PCR conditions consisted of an initial 95°C for 3 min; and 40 cycles of 95°C for 30 sec and 62°C for 40 sec. To generate the melting curves, the PCR amplification was carried out by slowly heating at 0.1°C/sec increments from 60 to 94°C.

To construct the miR-7 expression vector, we synthesized a mature miR-7 sequence (5′-TGGAAGACTAG TGATTTTGTTGT-3′) and inserted these double-stranded oligonucleotides into a linearized PcDNATM6.2-GW/EmGFP-miR vector (Invitrogen). After amplification, we sequenced this vector to confirm successful cloning. We also constructed a negative control vector using non-sense sequences (5′-AAATGTACTGCGCGTGGAGAC-3′) to insert into a pcDNA 6.2-GW/miR vector. For gene transfection, we grew MGC-803 cells and transfected the above vectors into this cell line using GBfectene-Elite (Genebank Biosciences, Suzhou, China) in accordance with the manufacturer’s instructions.

The following experimental groups were created: untreated, transfection reagent only, negative control vector and miR-7 vector. The transfection rate was confirmed using an inverted fluorescence microscope and qRT-PCR 48 h after transfection.

Cells in each group were suspended and inoculated into 96-wells at 1,000 cells/well. After reaching 80% confluency, the cells were transfected with plasmids (0.05 μg/well) using GBfectene-Elite in accordance with the manufacturer’s instructions, and cultured for 72 h. At the end of the experiment, 10 μl of the CCK-8 solution (Yiyuan Biotechnology, Guangzhou, China) was added into each well and the cells were further incubated at 37°C in a humidified incubator with 5% CO₂ for 1 h. Optical densities were measured at an absorbance of 450 nm using an ELX800 enzyme-linked immunoassay analyzer (BioTek, Winooski, VT, USA). The average and standard deviation of each group were calculated and the data are presented in a histogram.

A Transwell chamber (PengBo Biotech, Changsha, China) was used to detect gastric cancer cell migration and invasion capacity. For invasion, the membrane in the Transwell chamber was coated with Matrigel basement membrane matrix (BD BioCoat, Bedford, MA, USA). The miR-7-transfected or control vector-transfected cells were suspended at 1.0×10⁶/ml in DMEM containing 1.0% BSA. One hundred microliters was added into the upper chamber of the Transwell, and 500 μl DMEM containing 20% FBS was added into the lower chamber. The cells were incubated for 48 h. At the end of the experiments, the cells on the upper side of the membrane were carefully wiped out, and the cells that had migrated or invaded into the lower side of the membrane were fixed with 4% paraformaldehyde and stained with 4′,6-diamidino-2-phenylindole (DAPI). Images were then captured, and the number of cells was counted in 5 random fields for each group and summarized as mean ± standard deviation (SD) for statistical analysis.

Cells were washed 3 times with phosphate-buffered saline (PBS) and lysed in RIPA buffer (Beyotime Institute of Biotechnology, Shanghai, China) to extract the total cellular protein. Equal amounts of protein samples were resolved via 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto a polyvinylidene difluoride (PVDF) membrane (Beyotime Institute of Biotechnology). After blocking with a blocking buffer (GenePharma) for 1 h, the membrane was incubated with a primary antibody against EGFR (Proteintech Group, Chicago, IL, USA) or GADPH (Sigma-Aldrich) at a dilution of 1:500 at 4°C overnight. The next day the membranes were washed with PBS thrice and then further incubated with a horseradish peroxidase-conjugated secondary antibody (Jackson ImmunoResearch, West Grove, PA, USA) at a dilution of 1:1,000 at 37°C for 1 h. After washing with PBS thrice, positive protein bands were visualized using SuperSignal West Pico chemiluminescence substrates (Thermo Fisher Scientific, Rockford, IL, USA). Image analysis was performed with Alpha Imager 2200 (UVP, Upland, CA, USA).

To assess the direct binding of miR-7 to EFGR cDNA, we performed a luciferase assay. In brief, the cells were co-transfected with EGFR 3′UTR-pmirGLO plasmid and hsa-mir-7 mimic vector or the negative control plasmid hsa-miR-NC using Lipofectamine 2000 (Invitrogen) in accordance with the manufacturer’s instructions. Twenty-four or 48 h after transfection, the cells were prepared for the luciferase assay using the Dual-luciferase reporter assay system (Promega, Madison, WI, USA) in accordance with the kit instructions. Luciferase activity was detected with a GloMax™20/20 Luminometer (Promega). The relative luciferase activity was defined as the ratio of firefly luciferase activity to Renilla luciferase activity. Δ luciferase activity was used for statistical analysis of data, where Δ luciferase activity = (Firefly/Renilla) sample/(Firefly/Renilla) control.

Statistical analyses were performed using the statistical software SPSS 16.0 (SPSS, Inc., Chicago, IL, USA). All data are expressed as mean ± standard deviation. Differences between groups were analyzed with one-way ANOVA and the least significant difference test; P<0.05 was considered to indicate a statistically significant result.

We performed a microRNA microarray assay to identify differentially expressed miRNAs in gastric cancer cell lines (MGC-803, MKN-45, SGC-7901 and NCI-N87) that varied in the degrees of tumor differentiation, as well as in the normal gastric mucosal epithelial cell line GES-1 (Fig. 1A). In these gastric cancer cell lines, we found 14 miRNAs that were overexpressed relative to the normal gastric mucosal GES-1 cells (ratio >1.5 between tumor and normal cells). These overexpressed miRNAs were: hsa-miR-23a, hsa-miR-27b, hsa-miR-1908, hsa-miR-26b, hsa-miR-3607-3p, hsa-miR-20a*, hsa-miR-21, hsa-miR-122*, hsa-miR-193a-3p, hsa-miR-26a, hsa-miR-29c, hsa-miR-16, hsa-miR-30b and hsa-miR-3647-3p. In contrast, 19 miRNAs were reduced by 1.5-fold or more (i.e., ratio <0.667 between tumor and normal cells). The miRNAs that were underexpressed were: hsa-miR-1247, hsa-miR-20b*, hsa-miR-4301, hsa-miR-1299, hsa-miR-484, hsa-miR-885-5p, hsa-miR-1265, hsa-miR-589, hsa-miR-550a, let-7i*, hsa-miR-2355-3p, hsa-miR-4284, let-7i, hsa-miR-7, hsa-miR-1260, hsa-miR-622, hsa-miR-185, hsa-miR-378 and hsa-miR-378c.

The miRNA hsa-miR-7 was downregulated in all four gastric cancer cell lines comprising well-, moderately and poorly differentiated gastric cancer cells, relative to normal GES-1 cells. Thus, we chose hsa-miR-7 for further investigation. The underexpression of hsa-miR-7 in the NCI-N87, SGC-7901, MKN-45 and MGC-803 cell lines was confirmed via qRT-PCR (Fig. 1B).

We then assessed miR-7 expression in gastric cancer tissue specimens, and found that relative to the normal tissues, expression of miR-7 was decreased in 86.9% of gastric cancer tissues. Levels of miR-7 expression were significantly lower in moderately and poorly differentiated cancer tissues when compared with that in the well-differentiated gastric cancer tissues (Fig. 1C and D).

To further assess the role of miR-7 in gastric cancer, we constructed a miR-7 expression vector for gene transfection and found that the transfection efficiency of the miR-7 plasmid and negative control groups was >80% in MGC803 cells after a 48-h transfection, via observation under a fluorescence microscope. The qRT-PCR data showed that levels of miR-7 expression were significantly greater 48 h after miR-7 plasmid transfection compared with the negative control cells (Fig. 2A). Restoration of miR-7 expression reduced gastric cancer cell viability (Fig. 2B) compared with the controls.

We used a Transwell chamber to detect the effects of miR-7 on the invasion and migration abilities of MGC803 gastric cancer cells. The miR-7-transfected cells exhibited significantly reduced invasion and migration abilities (33±11 and 59±34, respectively) compared with the negative control cells (135±37 and 160±11), transfection reagent only cells (191±12 and 183±23), and the parental cells (149±39 and 182±20; Fig. 2C and D).

To explore miR-7 targeting (binding) genes, we performed a bioinformatic analysis using TargetScan, PicTar and miRanda software and found that EGFR could be a target gene (Fig. 3A). To confirm whether miR-7 regulates EGFR expression by binding to the EFGR 3′UTR, we performed a dual luciferase reporter assay. The results showed that 24 and 48 h after co-transfection of EGFR 3′UTR-pmirGLO plasmid and hsa-miR-7, fluorescent activity was significantly reduced when compared to that of the miR-NC control group (Fig. 3D).

We then performed qRT-PCR and western blot analysis and found that levels of EGFR mRNA and protein were significantly less in the miR-7-transfected cells when compared with the levels in the negative control, transfection reagent only cells and the parental cells (Fig. 3B and C). These data suggest that miR-7-induced inhibition of gastric cancer viability and invasion acts through degradation of EGFR mRNA and a decrease in EGFR protein expression.