The present study was approved by the Medical Ethics Committee of Shanghai Zhongshan Hospital, Fudan University (Shanghai, China). Prior written informed consent was obtained from each patient. A total of 10 patients with gastric cancer (6 male, 4 female; age range, 39–58 years) were included in the present study. The specimens were handled and made anonymous according to the ethical and legal standards. Paired tissue specimens (tumor and adjacent normal mucosa) from 10 patients with gastric cancer were obtained and histologically confirmed by a pathologist at Shanghai Zhongshan Hospital. The samples were derived from patients who had not received adjuvant treatment, including radiotherapy or chemotherapy prior to surgery in order to eliminate potential treatment-induced changes to gene expression profiles. Following excision, tissue samples were immediately snap frozen in liquid nitrogen and stored at −80°C until RNA extraction.

Immunohistochemical staining was performed for RASSF8 to evaluate immunoreactivity in the aforementioned tissue samples (21) using the Elivision™ Plus Two-Step system (Fuzhou Maixin Biotech Co., Ltd., Fuzhou, China). Paraffin-embedded tissue sections (5-µm thick) were prepared. Slides were deparaffinized, rehydrated and subjected to microwave heat antigen retrieval in 0.01 M citrate buffer (pH 6.0) for 20–25 min at 25°C. Following blocking of endogenous peroxidase activity with 3% H₂O₂ for 15 min at 25°C, the sections were incubated with a primary antibody directed against RASSF8 (dilution, 1:300; cat. no. ab214324; Abcam, Cambridge, MA, USA) overnight at 4°C. Following washing with PBS for 3 times (10 min each, 25°C), the specimens were incubated with F(ab')2-goat anti-rabbit immunoglobulin G (IgG) (H+L) secondary antibody (dilution, 1:200; cat. no. Q-11401MP; Thermo Fisher Scientific Inc., Waltham, MA, USA) at 37°C for 1 h and subsequently washed with PBS three times. Immunoreactivity was visualized using the chromogen, 3,3′diaminobenzidine (Fuzhou Maixin Biotech Co., Ltd.). Slides were then counterstained with 0.1% hematoxylin solution for 5 min at 25°C.

The immunostaining frequency for each tumor was scored as follows: 0 (<10%); 1 (10–30%); 2 (31–60%); and 3 (>61%). Staining intensity was documented as: 0, no immunostaining; 1, weak; 2, moderate and 3, strong. Total immunostaining score results were calculated from the multiplication of both parameters. Samples were scored totally as follows: A scale of 0 (−, total immunostaining score, 0); 1+ (score range, 1–2); 2+ (++, total immunostaining score, 3~4); and 3+ (+++, total immunostaining score, 6~7). Immunostaining was assessed by an experienced pathologist who was blinded to the clinical data of the patients. RASSF8 expression was determined by assessing the percentage and intensity of stained tumor cells. For RASSF8 protein, immunostainings were scored as strong (2++ and 3+++), weak or negative (1+ and 0) according to the rate of labeled tumor cells and the membrane staining intensity.

The human AGS, BGC-823, HGC-27, MKN-45 and SGC-7901 gastric cancer cell lines, GES-1 normal gastric cell line, and HEK 293T cells were purchased from the Cell Bank of Chinese Academy of Science (Shanghai, China). AGS, BGC-823, MKN-45, SGC-7901 and HEK 293T cells were cultured in Dulbecco's modified Eagle's medium (DMEM). HGC-27 and GES-1 were cultured in RPMI-1640. In each case, the medium was supplemented with 10% fetal bovine serum (FBS) all from Thermo Fisher Scientific, Inc. (Gibco) and 1% penicillin and streptomycin (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany). Cells were maintained at 37°C in a humidified atmosphere consisting of 5% CO₂.

Total RNA was extracted from cell lines or frozen tissues using TRIzol reagent and miRNeasy mini kit (Invitrogen; Thermo Fisher Scientific, Inc.). The RT-qPCR was performed using All-in-One™ miRNA qRT-PCR Detection kit (GeneCopoeia Inc., Rockville, MD, USA) according to the manufacturer's protocol. The iQ-5 system (Bio-Rad Laboratories, Inc., Hercules, CA, USA) was used to monitor the PCR in real-time. The PCR cycling profile was as follows: Denaturation at 95°C for 2 min; followed by 40 cycles of annealing at 95°C for 5 sec; and extension at 60°C for 35 sec. The average cycle threshold (Cq), from triplicate assays, was used for further calculations. Relative expression levels were normalized to the control (GES-1, control cell line; normal gastric mucosal tissue, control tissue). The endogenous U6 snRNA was chosen as the internal control. The 2^−ΔΔCq method (22) was used to quantify the relative amount of miR-377/RASSF8, where ΔΔCq=[(CqmiR-377/RASSF8)-CqU6RNA]-(Cqcontrol-CqU6RNA).

RASSF8 overexpression lentivirus was purchased from Shanghai R&S Biotechnology Co., Ltd. (Shanghai, China). BGC-823 cells were seeded into 3.5-cm dishes (1×10⁶ cells/dish) 24 h prior to lentivirus infection. The next day, lentivirus was added into dishes with a multiplicity of infection of 4 to infect cells. The infection efficiency was detected using fluorescence microscopy analysis of green fluorescence protein 48 h after infection, and was only accepted when the efficiency was >90%.

The miR-377 mimic and negative control (miR-NC) were purchased from Shanghai GenePharma, Co., Ltd. (Shanghai, China). The transfection of miRNAs (50 nM) was performed using X-tremeGENE (Roche Applied Science, Rotkreuz, Switzerland) according to the manufacturer's protocol. The 3′-untranslated region (UTR) was PCR-amplified from BGC-823 genomic DNA and cloned downstream of the luciferase gene in the pGL vector (Promega Corporation, Madison, WI, USA).

Total protein from cells was extracted according to the protocol of the Whole Protein Extraction kit (KGP2100; Nanjing KeyGen Biotech Co., Ltd., Nanjing, China). Equivalent quantities (40 µg protein/lane) of protein were separated by 10% SDS-PAGE and transferred to polyvinylidene difluoride membranes. Membranes were blocked with 10% skimmed milk (10% bovine serum albumin (KGY00810, Nanjing KeyGen Biotech Co., Ltd) for phosphorylated-protein), and then incubated with the appropriate primary antibody overnight at 4°C. Membranes were subsequently washed with Tris-Buffered saline-Tween 20 and incubated with the corresponding horseradish peroxidase-conjugated AffiniPure goat anti-rabbit IgG (H+L) secondary antibody (dilution, 1:200; cat. no. 111-035-003; Jackson ImmunoResearch Inc., West Grove, PA, USA) at 25°C for 1 h. Bound secondary antibody was visualized using an enhanced chemiluminescence system (Pierce Biotechnology, Rockford, IL, USA). The primary antibodies used were as follows: Anti-GAPDH (1:1,000; cat. no. 2118; Cell Signal Technology, Inc., Danvers, MA, USA); and anti-RASSF8 (1:2,000; cat. no. ab126110; Abcam). The results were normalized to GAPDH to correct for loading.

A total of 1×10⁴ BGC-823 cells/well were plated in 96-well plates and cultured for 1–6 days in normal conditions. Cell viability was assessed using Cell Counting kit-8 (Dojindo, Kumamoto, Japan). A total of 10 µl of cell counting assay kit-8 (CCK-8) solution was added daily to each well. After treatment with CCK-8 at 37°C for 2 h, the absorbance at 450 nm was measured using a microplate reader.

The transfected cells were plated in 6-well plates at a density of 1×10³ cells/well and incubated in culture medium (DMEM containing 10% FBS) for 10 days without any disturbance. The cells were fixed with 70% (v/v) ethanol for 15 min at 25°C and stained with 0.5% crystal violet for 1 h at 37°C. Visible colonies were counted in 10 different fields and the mean value was calculated.

The cells were seeded into 6-well plates at a density of 1×10⁵ cells/well after transfection for 24 h, and maintained in DMEM/RPMI-1640 containing 10% FBS. After being cultured for 24, 4 h and 72 h, the cells were harvested and fixed in 70% (v/v) ice-cold ethanol for 15 min at 4°C, then stored at −20°C overnight. The fixed cells were washed twice with PBS (1 ml) and stained with propidium iodide (PI) solution (Invitrogen; Thermo Fisher Scientific, Inc.) (10 µg/ml) including RNase (0.5 mg/ml) for 30 min at 4°C in the dark. Flow cytometry was then performed using a FACSCalibur cytometer (BD Biosciences, Franklin Lakes, NJ, USA) to examine the cell cycle distribution. Experiments were performed in triplicate.

A total of 2×10⁵lentivirus-infectedcells BGC-823 cells/well were plated in 6-well plates, and then transfected with miR-NC and miR-377 mimics for 48 h. The transfected cells were harvested by brief trypsinization and centrifugation (170 × g) for 5 min at 4°C, washed in ice-cold PBS and resuspended in PBS containing Annexin V-fluorescein isothiocyanate. PI (Thermo Fisher Scientific, Inc.) was then added, and the cells were incubated for 30 min at 20°C. A total of 1×10⁴ cells/sample were acquired using a FACScan flow cytometer and the proportions of labeled cells were analyzed using Paint-A-Gate software (BD Paint-A-Gate Pro for Windows; version 649728; BD Biosciences).

The upper surface of a polycarbonate membrane in a Transwell filter insert was coated with Matrigel (10 µg/µl; 30 µl). After 30 min of incubation at 37°C, the Matrigel solidified and served as the extracellular matrix for tumor cell invasion analyses. The cells were harvested in 100 µl serum-free growth medium and seeded into the upper compartment of the chamber. BGC-823 cells or lentivirus-infected cells were seeded into the upper chambers at the density of 8×10⁵ cells in 100 µl serum-free DMEM. The lower chambers were filled with 500 µl RPMI-1640 containing 10% FBS. After 24 h of incubation, non-invading cells on the top of the membrane were removed by scraping. Invaded cells on the bottom of the membrane were fixed with 4% paraformaldehyde for 15 min at 25°C, followed by staining with 0.5% crystal violet for 20 min at 25°C. The invaded cells on the membrane were then counted in three random microscopic fields/well, using ImageJ software (k 1.45; National Institutes of Health, Bethesda, MD, USA) (23). Invasion studies were repeated three times.

TargetScan (www.targetscan.org/vert_71) and Starbase (starbase.sysu.edu.cn) were used to predict the targeting association between miR-377 and RASSF8.

A total of 1×10⁵ HEK 293 T cells/well were seeded into 24-well culture plates in RPMI-1640 containing 10% FBS without antibiotics. After 24 h, the cells were cotransfected with miR-NC, miR-377 mimics and the 3′-UTR of RASSF8 responsive firefly luciferase reporter plasmid and Renilla luciferase plasmid (R&S Biotechnology Co., Ltd., Shanghai, China) and transfection reagent (POLO deliverer TM 3000 Transfection Reagent POLO3000, CT001; R&S Biotechnology Co., Ltd.) according to the manufacturer's protocol. Cells were incubated for 24 h under normal conditions, then cell lysates were prepared and luciferase activities were measured using the Dual-Luciferase Reporter Assay system (Promega Corporation). Firefly luciferase activity was normalized to the activity of Renilla luciferase.

Results are presented as the means ± standard deviation of three independent samples. A comparison of the level of RASSF/miR-377 expression between gastric cancer and adjacent normal tissue was performed using the Wilcoxon signed-rank test. Significant differences in the mean values were evaluated using the Student's unpaired t-test. Where multiple comparisons were required, analysis was performed using one-way analysis of variance with Bonferroni correction. P<0.05 was considered to indicate a statistically significant difference.

The overall expression levels of RASSF8 and miR-377 human gastric cancer tissues, and cell lines were determined using RT-qPCR. As presented in Fig. 1A, levels of RASSF8 in the human GES-1 normal cells were significantly higher compared with the levels of RASSF8 in BGC-823, AGS, MKN-45, HGC-27 and SGC-7901. In addition, RASSF8 expression in tumor tissues demonstrated significantly attenuated levels compared with the corresponding normal tissues, with a mean1.6-fold decrease (Fig. 1B). Furthermore, levels of miR-377 in the BGC-823, AGS and MKN-45 cell lines were significantly higher compared with that in the normal GES-1 cell line (P<0.01; Fig. 1C), which was consistent with a previous report (24). However, no significant differences were identified among the levels of miR-377 in the HGC-27, SGC-7901 and normal GES-1 cell lines in the present study. Furthermore, a significantly increased expression level of miR-377 between the tumor and normal groups was identified, with a mean 20.4-fold increase (P<0.05; Fig. 1D).

RASSF8 was immunohistochemically stained in the tissue sections of gastric cancer and their corresponding adjacent non-cancerous mucosa. It was demonstrated that the RASSF8 protein was abundantly expressed in the upper glandular layer of the superficial epithelium, while expression of RASSF8 protein was significantly downregulated in gastric cancer tissue compared with normal gastric mucosa (Table I; Fig. 2; P<0.05).

To examine the role of RASSF8/miR-377 in tumorigenesis of gastric cancer, the effect of RASSF8 overexpression and miR-377 on the proliferation of BGC-823 gastric cancer cell lines was examined. Following infection with RASSF8 overexpression lentivirus the cells were transfected with miR-377 mimic or the miR377-NC. The RT-qPCR results demonstrated that RASSF8 was significantly increased in the cells transfected with RASSF8 overexpression lentivirus (Fig. 3A). The CCK-8 assay demonstrated that the overexpression of RASSF8 significantly suppressed the proliferation of BGC-823 cells, whereas miR-377 attenuated the effect of RASSF8 on cell proliferation (Fig. 3B). The proliferation of the RASSF8 overexpression with miR377 group remained significantly lower compared with that of the negative control group, in which cells were transfected by negative control miRNA. The colony formation assay was performed to further confirm the effect of RASSF8/miR-377 on gastric cancer cell proliferation, and the data indicated that the overexpression of RASSF8 significantly decreased colony numbers in BGC-823 cell cultures, which was not reversed by treatment with miR-377, which confirmed the results of CCK-8 assay (Fig. 3C). Furthermore, the cell cycle distribution was assessed by flow cytometry. As presented in Fig. 3D, the overexpression ofRASSF8decreased the proportion of cells in the G₁phase at 24, 48 and 72 h, compared with the control (BGC-823 cells without transfection) or negative control. Additionally, overexpression of RASSF8 increased the proportion of cells in the S phase at 24 and 48 h, compared with the control or negative control, and at 72 h, compared with the control, as well slightly decreased at 72 h compared with the negative control and led to cell cycle arrests in BGC-823 cells. Similarly, miR-377 attenuated rather than reversed the effect of RASSF8 on cell cycle. Collectively, these results demonstrated that RASSF8 inhibits gastric cancer cell growth.

As presented in Fig. 4, RASSF8 overexpression significantly increased the rate of apoptotic cells compared with the control (P<0.01). Furthermore, no significant differences were identified between the RASSF8 and RASSF8 combined with miR-377 groups regarding the rate of apoptotic cells.

To investigate the effect of RASSF8 ontheinvasivecapability of gastric cancer cells, Transwell assays were performed in BGC-823 cells. Fig. 5A demonstrates the changes in theinvasive capacity of BGC-823 cells. BGC-823 cells infected with RASSF8 overexpression lentivirus revealed significantly decreased invasive capabilities compared with NC lentivirus infected-BGC-823 cells, with an 80% reduced ability to invade through Matrigel membranes (Fig. 5B). Compared with the RASSF8 overexpression group, no significant differences were observed in the invasive capabilities of BGC-823 cells transfected with RASSF8 and miR-377. Collectively, these results suggest that RASSF8 decreased BGC-823 cell invasion capability and miR-377 could not reverse the effect of RASSF8 on BGC-823 cell invasion.

To gain insight into the biological implications of RASSF8/miR-377 on gastric cancer tumorigenesis, TargetScan, miRanda and DIANA were used to identify putative human protein-coding gene targets of miR-377. The tumor suppressor gene RASSF8 was predicted to have miR-377-binding elements in its 3′-UTRs with high fidelity scores (Fig. 6A). To assess whether the predicted miR-377-binding sites in the 3′-UTR of the target gene was responsible for miR-377 regulation, the 3′-UTR regions downstream of a luciferase reporter gene were cloned and transfected into the BGC-823 control, RASSF8 overexpression with miR-377, and RASSF8 overexpression with miR-377-NC groups. The luciferase activity of RASSF8-overexpressed cells was significantly increased compared with the control. miR-377 significantly suppressed the luciferase activity of RASSF8-overexpressed cells, which was not observed following transfection with miR-377-NC (Fig. 6B). However, miR-377 attenuated but not reversed the luciferase activity of cells transfected with RASSF8 overexpression compared with control, which suggested that miR-377 could decreased but not inhibit the effect of overexpression of RASSF8 compared with control or negative control.