The present study was approved by the Ethical Committee of The Third Xiangya Hospital (Changsha, China). A total of 78 gastric cancer tissues, as well as the adjacent non-tumor tissues were obtained from The Third Xiangya Hospital between January 2010 and March 2011. These patients with gastric cancer included 45 male and 33 female aged between 43–83 years, with a mean of 67.4 years. Written informed consent was obtained from all patients. All tissue samples were confirmed using histopathological evaluation and stored at −80°C until further use. The clinical information for patients with gastric cancer is summarized in Table I.

Total RNA was extracted using the TRIzol^® reagent (Thermo Fisher Scientific, Inc., Waltham, MA, USA), which was then converted to cDNA using a Taqman^® miRNA Reverse Transcription kit (Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. The expression of miR was determined using a Hairpin-it miRNAs qPCR Quantitation kit (Shanghai GenePharma Co, Ltd., Shanghai, China). U6 was used as the internal reference. The expression of mRNA was examined using a SYBR^® Green qPCR Assay kit (Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. GAPDH was used as the internal reference for mRNA. The thermocycling conditions were initial denaturation at 95°C for 10 min, then 40 cycles at 95°C for 15 sec and annealing/elongation at 60°C for 15 sec. The relative expression was analyzed by the 2^−ΔΔCq method (22). The primers for miR-23a and U6, which was used as the internal reference for microRNAs, were purchased from Fulgene (Guangzhou, China; cat nos. HmiRQP0345 and HmiRQP9001, respectively). The primer sequences were as follows: miR-23a forward, 5′-CCTACTGTCGTCCCAAGACCT-3′ and reverse, 5′-GGGGCTCGTGCAGAAGAAT-3′; U6 forward, 5′-ACAACTTTGGTATCGTGGAAGG-3′ and reverse, 5′-GCCATCACGCCACAGTTTC-3′. GAPDH forward, 5′-GGAGCGAGATCCCTCCAAAAT-3′ and reverse, 5′-GGCTGTTGTCATACTTCTCATGG-3′.

Human gastric mucosa epithelial line GES-1, and gastric cancer lines HGC-27, BGC-823, SGC-7901 and AGS were obtained from the Cell Bank of Central South University (Changsha, China). The cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS; both from Thermo Fisher Scientific, Inc.) at 37°C in a humidified incubator containing 5% CO₂. Cell transfection was conducted using Lipofectamine^® 2000 (Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. AGS cells were transfected with 100 nM of scrambled miR mimic (miR-NC; Guangzhou Fulengen Co., Ltd., Guangzhou, China.), miR-23a mimic (miR-23a; Guangzhou Fulengen Co., Ltd.), miR-23a inhibitor (Fulengen), negative control inhibitor (NC inhibitor; Guangzhou Fulengen Co., Ltd.), co-transfected with miR-23a inhibitor and non-specific siRNA (miR-23a in+siNC; OriGene Technologies, Inc., Beijing, China.) or co-transfected with miR-23a inhibitor and SPRY2-specific siRNA (miR-23a in+siSPRY2; OriGene Technologies, Inc.). Following transfection for 48 h, the following experiments were performed.

Cells were lysed in RIPA buffer (Thermo Fisher Scientific, Inc.), and the protein was extracted and quantified using a BCA Protein Assay kit (Pierce; Thermo Fisher Scientific, Inc.). Proteins (50 µg) were separated with 10% SDS-PAGE and transferred onto a polyvinylidene difluoride (PVDF) membrane (Thermo Fisher Scientific, Inc.). The membrane was incubated with TBS with Tween^® 20 containing 5% non-fat milk at room temperature for 3 h, and then with rabbit anti-mouse SPRY2 (dilution, 1:50; cat no. ab85670), total-ERK (dilution, 1:100; cat no. ab196883), phospho-ERK (dilution, 1:100; cat no. ab214362) or GAPDH antibodies (dilution, 1:100; cat no. ab9485; all from Abcam, Cambridge, MA, USA) at room temperature for 3 h. Following washing with PBS with Tween 20 three times, the PVDF membrane was incubated with horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (dilution, 1:10,000; cat no. ab6721; Abcam) at room temperature for 1 h. Chemiluminescence detection was performed using a Pierce™ Fast Western Blot kit (Pierce; Thermo Fisher Scientific, Inc.). The relative protein expression was analyzed using Image-Pro Plus software version 6.0 (Media Cybernetics, Inc., Rockville, MD, USA), represented as the density ratio vs. GAPDH.

AGS cells in 100 µl DMEM containing 0.5 g/l MTT (Thermo Fisher Scientific, Inc.) were seeded in a 96-well plate at a density of 1×10⁴ cells/well. Following incubation at 37°C for 0, 24, 48 or 72 h, 50 µl DMSO (Thermo Fisher Scientific, Inc.) was added. Following incubation at 37°C for 10 min, the wavelength of 570 nm for each sample was measured using the Tecan Infinite^® M200 plate reader (Tecan Group, Ltd., Mannedorf, Switzerland).

A wound healing assay was conducted to determine the cell migratory capacity. AGS cells were cultured to full confluence, and wounds of ~1 mm width were created with a plastic scriber. Then, cells were washed once using PBS. Prior to migration, AGS cells were photographed under an inverted microscope (magnification, ×40; Olympus Corporation, Tokyo, Japan) as the negative control. Following being cultured at 37°C with 5% CO₂ for 48 h, AGS cells were observed under an inverted microscope (magnification, ×40; Olympus Corporation). This experiment was repeated for 3 times.

A Cell Invasion Assay kit (Chemicon International, Inc., Temecula, CA, USA), with an adhesion matrix supplied in the kit, was used to perform the Transwell assay, according to the manufacturer's guidelines. An AGS cell suspension was prepared in 300 µl DMEM containing 50,000 AGS cells and was seeded in the upper chamber. Then, 500 µl DMEM with 10% FBS was added into the lower chamber. Following incubation at 37°C for 24 h, the non-invading AGS cells on the upper face of the membrane were scraped, while the invading AGS cells on the lower face of the membrane were stained with 0.5% crystal violet at room temperature for 20 min and counted under an inverted microscope (magnification, ×200).

The putative target genes of miR-23a were predicted using Targetscan software (www.targetscan.org/), in accordance with the manufacturer's protocol.

The mutant type (MT) 3′-UTR of SPRY2A was constructed using a Quik-Change Site-Directed Mutagenesis kit (Agilent Technologies, Inc., Santa Clara, CA, USA), according to the manufacturer's protocol. The wild type (WT) or MT 3′-UTR of SPRY2 were then inserted into the psiCHECK2 vector (Promega Corporation, Madison, WI, USA). For the luciferase reporter gene assay, AGS cells were transfected using Lipofectamine^® 2000 (Invitrogen, Thermo Fisher Scientific, Inc.) with 100 nM of WT SPRY2-3′-UTR or MT SPRY2-3′-UTR plasmid, with or without 100 nM of miR-23a mimics, respectively. The luciferase activity was then measured at 48 h following transfection using the Dual-Luciferase Reporter assay system (Promega Corporation) on the Lmax Multiwell luminometer (Molecular Devices, LLC, Sunnyvale, CA, USA). Renilla luciferase activity was normalized to firefly luciferase activity.

The data are expressed as the mean ± standard deviation following at least three independent experiments. Statistical analysis of differences was performed via one-way analysis of variance and Turkey's post-hoc test for multiple groups or Student's t-test for two groups, using the SPSS version 20.0 software package (IBM Corp., Armonk, NY, USA). P<0.05 was considered to indicate a statistically significant difference. A χ² test was performed for Table I, with a log-rank test conducted on the Kaplan-Meier survival curve.

In the present study, RT-qPCR was conducted first to examine the expression levels of miR-23a in gastric cancer. The data indicated that miR-23a was significantly upregulated in gastric cancer tissues compared with adjacent non-tumor tissues (Fig. 1A). Further investigation revealed that the increased expression of miR-23a was significantly associated with node metastasis and advanced clinical stage (Table I). The results indicated that miR-23a may contribute to gastric cancer growth and metastasis. Furthermore, the patients with gastric cancer with high miR-23a expression demonstrated significantly reduced survival time, compared with those with low miR-23a levels (Fig. 1B); therefore, miR-23a may be used as a potential predicator for the prognosis of patients with gastric cancer.

The miR-23a levels in gastric cancer cell lines, including HGC-27, BGC-823, SGC-7901 and AGS cells, were further examined. The data indicated that miR-23a was also significantly upregulated in gastric cancer cell lines compared with normal gastric epithelial GES-1 cells (Fig. 2A). AGS cells were used to conduct in vitro experiments. As miR-23a was demonstrated to be significantly upregulated in gastric cancer, AGS cells were transfected with a miR-23a inhibitor to knockdown its expression. Following transfection, the miR-23a levels were significantly reduced in the miR-23a inhibitor group compared with the NC inhibitor group (Fig. 2B). The MTT assay data further demonstrated that knockdown of miR-23a caused a significant decrease in AGS cell proliferation after 48 h (Fig. 2C). The results of a wound healing assay and Transwell assay demonstrated that inhibition of miR-23a significantly suppressed AGS cell migration and invasion (Fig. 2D and E). According to this data, it was indicated that miR-23a serves a promoter role in gastric cancer growth and metastasis, and may be used as a potential therapeutic target.

Bioinformatics analysis was conducted to predict the target of miR-23a, which indicated SPRY2 to be a potential target (Fig. 3A). To further confirm this target association, the luciferase reporter gene plasmid containing the WT or MT of SPRY2-3′-UTR (Fig. 3B) was constructed. Following this, a luciferase reporter assay in AGS cells was performed. The luciferase activity was significantly reduced following co-transfection with the miR-23a mimics and WT 3′-UTR of SPRY2 plasmid, but unaltered following co-transfection with the miR-23a mimics and MT SPRY2-3′-UTR plasmid (Fig. 3C). This data demonstrated that miR-23a directly binds to the 3′-UTR of SPRY2 mRNA in gastric cancer AGS cells.

Following this, the expression levels of SPRY2 in gastric cancer tissues and cell lines were investigated. As depicted in Fig. 3D, the mRNA levels of SPRY2 were significantly reduced in gastric cancer tissues compared with normal adjacent tissues. Similarly, the mRNA and protein levels of SPRY2 were downregulated in gastric cancer cell lines (Fig. 3E and F).

The effects of miR-23a on the protein expression of SPRY2 in AGS cells were studied. The data indicated that knockdown of miR-23a enhanced the mRNA and protein expression of SPRY2 in AGS cells (Fig. 4A and B). To further confirm this data, AGS cells were transfected with the miR-23a mimic to increase its expression. Following transfection, the miR-23a levels were upregulated in the miR-23a group compared with the miR-NC group (Fig. 4C). The upregulation of miR-23a reduced the mRNA and protein levels of SPRY2 (Fig. 4D and E); therefore, miR-23a negatively regulates the expression of SPRY2 in AGS cells.

Based on the aforementioned data, it was speculated that SPRY2 may be involved in the miR-23a-mediated AGS cells. To clarify this speculation, AGS cells were co-transfected with the miR-23a inhibitor and SPRY2-specific siRNA plasmid, or co-transfected with the miR-23a inhibitor and non-specific siRNA in the control group. Following transfection, the mRNA and protein levels of SPRY2 were significantly downregulated in the miR-23a in+siSPRY2 group compared with the miR-23a in+siNC group (Fig. 5A and B). Furthermore, it was demonstrated that the proliferation, migration and invasion of AGS cells were significantly upregulated in the miR-23a inhibitor+SPRY2 siRNA group compared with the miR-23a inhibitor+NC siRNA group (Fig. 5C-E). The results indicated that the suppressive effect of miR-23a knockdown on AGS cells may occur via the upregulation of SPRY2.

SPRY2 has been reported to have inhibitory effects on the activity of ERK signaling, which is essential for cancer cell growth and metastasis (23). The activity of ERK signaling was thus studied. As depicted in Fig. 6A, knockdown of miR-23a reduced the phosphorylation levels of the ERK protein; however, the phosphorylation levels of the ERK protein were increased in the miR-23a in+siSPRY2 group compared with the miR-23a in+siNC group (Fig. 6B); therefore, it is suggested that miR-23a promotes the activity of ERK signaling via targeting SPRY2.