All experimental procedures were approved by the Institutional Review Board of Wuhan University of Science and Technology (Wuhan, China). Human gastric cancer tissues (n=65) and matched adjacent normal tissues were obtained from patients who underwent surgery at Tianyou Hospital of Wuhan University of Science and Technology between June 2007 and March 2011. None of the patients had received any treatment prior to resection. Written informed consent was obtained from the patients. The clinicopathological data were collected from medical records.

HEK293 human embryonic kidney cells, GES-1 immortalized normal gastric mucosa cells, and AGS, SGC7901, MKN28, MGC803 and HGC27 gastric cancer cells were purchased from the American Type Culture Collection (Rockville, MD, USA) and cultured according to the manufacturer’s instructions. All transfections were performed using Lipofectamine 2000 (Invitrogen Life Technologies, Carlsbad, CA, USA).

Total RNA was isolated from the tissues or cells using TRIzol reagent (Invitrogen Life Technologies) according to the manufacturer’s instructions. miR-183 expression levels were measured using a mirVana qRT-PCR miRNA Detection kit (Ambion Inc., Austin, TX, USA). U6 small RNA served as an internal reference. Bmi-1 mRNA expression levels were measured using a SYBR Premix Ex Taq™ kit (Takara Bio, Inc., Shiga, Japan), and β-actin expression levels served as an endogenous control. RT-qPCR was conducted using an Applied Biosystems 7,900 Fast Real-Time PCR system (Applied Biosystems, Foster City, CA, USA). cDNA was synthesized from 200 ng of total RNA, using a high-capacity cDNA Reverse Transcription Kit (Applied Biosystems) in a total volume of 20 μl per reaction. The sequences of the primers were as follows: Forward, 5′-GCTTCAAGATGGCCGCTTG-3′ and reverse, 5′-TTCTCGTTGT TCGATGCATTTC-3′ for Bmi-1; and forward, 5′-TGGATCAGCAAGCAGGAGTA-3′ and reverse, 5′-TCGGCCACATTGTGAACTTT-3′. Quantitative PCR was performed as follows: heating at 95°C for 5 min, followed by 95°C for 15 sec, 60C for 15 sec and 72C for 32 sec for 40 cycles. The data were analyzed using the 2^−ΔΔCt method.

Total cell lysate was extracted from the cells and tissues using RIPA reagent (Beyotime Institute of Biotechnology, Shanghai, China). Proteins were separated using SDS-PAGE and transferred to polyvinylidene difluoride membranes (Sterlitech Corporation, Kent, WA, USA). The membranes were then blocked with 5% skimmed milk in Tris-buffered saline with Tween 20 buffer (TBST; 10 mmol/l Tris-HCl, 150 mmol/l NaCl and 0.1% Tween 20; pH 8.0) for 1 h at room temperature then incubated with primary antibodies, including monoclonal anti-human antibodies Bmi-1 (1:500, Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) and polyclonal rabbit antibody GAPDH (1:1,000; Sigma-Aldrich, St. Louis, MO, USA) overnight at 4°C. Following three washes with TBST, the membranes were incubated with a goat anti-rabbit secondary antibody (Boster Systems, Inc., Pleasanton, CA, USA) for 1 h. The bands were then visualized using the enhanced chemiluminescence system (Amersham Pharmacia Biotech, Amersham, UK) following the manufacturer’s instructions.

The hsa-miR-183 mimic and negative control (NC) oligonucleotides were obtained from Guangzhou Ribobio Co., Ltd. (Guangzhou, China). SGC7901 and AGS cells were plated in a six-well plate the day prior to transfection. The cells were transfected with hsa-miR-183 mimic or NC (50 nmol/l) using Lipofectamine 2000. After 24 h, the cells were collected to perform in vitro assays.

At 24 h post-transfection, the cells were seeded in 96-well plates (2×10³ cells/well). The viability of the cells was examined by MTT assay (Sigma-Aldrich), conducted daily for five days.

For the colony formation assay, 500 transfected cells were placed in a six-well plate and cultured for 14 days using RPMI 1640 medium (Gibco-BRL, Carlsbad, CA, USA) containing 10% fetal bovine serum. Cell colonies were fixed with methanol and then stained with 0.1% crystal violet. Colonies were observed using a microscope and colonies containing >50 cells were counted (IX71, Olympus Corporation, Tokyo, Japan).

The cell invasive capacity was assessed with specialized Transwell chambers (8-μm pore; BD Biosciences, Franklin Lakes, NJ, USA). The transfected cells (5×10⁴ cells/well) were added to the upper chamber of the inserts, which was coated with a Matrigel mix; 500 μl fetal bovine serum was added to the lower chamber as a chemoattractant. After 24 h, any non-invading cells on the upper surface were removed with swabs and the cells that had migrated to the lower side of the membrane were fixed with methanol, stained with 0.1% crystal violet and air-dried. Images of the cells were then captured. The number of invading cells was evaluated in five fields using microscopy. The mean of triplicate assays for each experimental condition was analyzed.

For the luciferase assays, the potential miR-183 binding site in the 3′-UTR of Bmi-1 mRNA was predicted by TargetScan (www.targetscan.org) and miRanda (www.microrna.org). Wild-type (wt) and mutant (mt) Bmi-1 mRNA 3′-UTRs were synthesized and cloned into the Xba I site of a pGL3 basic vector (Promega Corporation, Madison, WI, USA) downstream of the luciferase stop codon, and the resulting vectors were termed pGL3-wt-Bmi-1 and pGL3-mt-Bmi-1, respectively. The HEK293 cells were cultured in 24-well plates and co-transfected with pGL3-Control (0.4 mg), pGL3-wt-Bmi-1 (0.4 mg) or pGL3-mt-Bmi-1 (0.4 mg), plus pRL-TK luciferase reporters (25 ng/well) and pcDNA-miR-183 (20 nm) or pcDNA-miR-NC (20 nm) using Lipofectamine 2000. After 48 h, the cells were collected and luciferase activity was assessed using a Dual-Luciferase Reporter Assay kit (Promega Corporation).

The data are expressed as the mean ± standard deviation. Statistical significance was analyzed using Student’s t-test (two-tailed) or the χ² test. All statistical analyses were performed using GraphPad Prism 5.0 (GraphPad Software, Inc., La Jolla, CA, USA) and SPSS 13.0 (SPSS, Inc., Chicago, IL, USA) software packages. The Kaplan-Meier method with log-rank test was used to analyze the prognostic significance. P<0.05 was considered to indicate a statistically significant difference.

miR-183 expression levels were measured in the gastric cancer tissues and cell lines. miR-183 was significantly downregulated in the gastric cancer cell lines compared with the GES-1 normal gastric epithelial cell line (Fig. 1A). Of all six gastric cancer cell lines, SGC7901 exhibited the lowest levels of miR-183 expression, whereas MGC803 possessed the highest levels of miR-183 expression.

To examine the clinicopathological significance of miR-183 expression in gastric cancer patients, miR-183 expression levels were measured in 65 gastric cancer tissue samples and adjacent normal tissue specimens. miR-183 expression levels were significantly reduced in the tumor tissues compared with the normal tissues (P<0.001; Fig. 1B). Furthermore, miR-183 expression levels were significantly reduced in the tissues with distant metastasis compared with the tissues without distant metastasis (P<0.001; Fig. 1C). Subsequently, the patients were divided into two groups, as determined by the median miR-183 expression levels in each patient. Low miR-183 expression levels were significantly associated with increased tumor size (P=0.003), the presence of distant metastasis (P=0.018) and increased tumor-node-metastasis stage (P=0.019; Table I). In addition, patients with low miR-183 levels had significantly lower survival rates compared with patients with high miR-183 levels (P=0.020; Fig. 1D).

As miR-183 is downregulated in gastric cancer tissues and cell lines, miR-183 may function as a tumor suppressor in gastric cancer. SGC7901 and AGS cells were transfected with miR-183 mimics to overexpress miR-183. The efficiency of transfection was confirmed by RT-qPCR (Fig. 2A). As shown in Fig. 2B, the ectopic expression of miR-183 significantly inhibited SGC7901 and AGS cell viability (P<0.05). In addition, ectopic miR-183 expression significantly reduced the numbers of SGC7901 and AGS cell colonies that formed compared with the negative control (both P<0.05; Fig. 2C). To assess the effect of miR-183 on the invasive abilities of gastric cancer cells, a Transwell assay was performed. Overexpression of miR-183 significantly suppressed the invasive ability of the SGC7901 and AGS cells (both P<0.05; Fig. 2D).

To analyze the underlying mechanism of miR-183 in gastric cancer, TargetScan (www.targetscan.org) and miRanda (www.microrna.org) were used to search for potential targets of miR-183 in gastric cancer. Bmi-1 was predicted as a potential target of miR-183 by TargetScan and miRanda. The 3′-UTR of Bmi-1 contains the binding site for the seed region of miR-183 (Fig. 3A). To investigate whether miR-183 regulates Bmi-1 expression in the cellular environment, the SGC7901 and AGS cells were transfected with miR-183 mimics to overexpress miR-183. The ectopic expression of miR-183 significantly reduced the Bmi-1 mRNA and protein levels in the SGC7901 and AGS cells (P=0.027; Fig. 3B and C). To further confirm that Bmi-1 is the direct target of miR-183 in gastric cancer, human Bmi-1 3′-UTR, containing either the wt or mt miR-183 binding sequence was cloned downstream of the firefly luciferase reporter gene in the pGL3 vector. HEK293 cells were transfected with either of the two types of reporter plasmid, plus either miR-183 mimics or NC vector. The luciferase activity of the reporter in the vector containing the Bmi-1 3′-UTR wt was significantly reduced by miR-183 mimics (P=0.009); however, the luciferase activity of the Bmi-1 3′-UTR mt was not significantly affected (Fig. 3D).