Gastric cancer is the fifth most common malignant tumor in the world and the third leading cause of cancer death worldwide. The incidence of gastric cancer and the resulting mortality have decreased worldwide, especially in developed countries, primarily because of better living conditions and improvements in diagnosis and treatment. However, gastric cancer remains a challenge in East Asia, with high incidence and mortality rates persisting.1,2 Gastric cancer is difficult to cure unless it is found at an early stage because few symptoms are manifested during the early stage, and the disease usually is advanced when the diagnosis is determined.

The epithelial–mesenchymal transition (EMT) is a process through which epithelial cells lose their cell polarity and cell–cell adhesion and gain migratory and invasive properties to become mesenchymal cells.3 The EMT, reported to play important roles in the progression and metastasis of cancer, is associated with a poor prognosis.4 E-cadherin is required for the maintenance of cell adhesion, and lack of E-cadherin expression is important for the EMT. Downregulation of E-cadherin expression due to mutation, deletion, CpG hypermethylation, or SNAI-mediated transcriptional repression of the CDH-1 gene, which encodes E-cadherin, leads to the EMT in gastric cancer.5–7

MicroRNAs (miRNAs) are small noncoding RNAs of 20–25 nucleotides that bind to the 3′ untranslated region (UTR) of multiple-target mRNAs, enhancing their degradation and inhibiting their translation. Reports show an association of MiRNAs with a variety of diseases, including cancer.8,9 Recent studies have shown that miRNAs regulate not only proliferation, differentiation, and migration,10 but also the EMT by suppressing EMT-related transcription factors in cancer cells.11,12

Peritoneal metastasis, the most frequent pattern of metastasis, has been shown to correlate with poor prognosis in advanced gastric cancer.13,14 Some studies have shown that the EMT plays a crucial role in the formation of peritoneal metastases by gastric cancer cells.15,16 Therefore, it is necessary to elucidate the epigenetic mechanisms of the EMT to improve early diagnosis and treatment of peritoneal metastases.

Recent studies have shown that miR-506 controls the EMT by inhibiting the expression of SNAI2 and PRRX1 and that aberrant low expression of miR-506 is associated with a poor prognosis in ovarian and breast cancers.17,18 However, the importance of miR-506 expression as a prognostic factor for the EMT and peritoneal metastasis has not been studied in gastric cancer to date. Therefore, the current study investigated the role of miR-506 in the survival of Japanese patients with gastric cancer and analyzed the function of miR-506 in the EMT in gastric cancer cell lines.

Materials and Methods

Patients

This study enrolled 141 patients with gastric cancer who underwent gastrectomy at Kyushu University Beppu Hospital and affiliated hospitals between 1989 and 2009. Written informed consent was obtained from all the patients in accordance with the guidelines approved by the Institutional Research Board. This study was conducted under the supervision of the ethical board of Kyushu University and affiliated hospitals. Detailed information is described in the Supplementary Material.

Cell Lines

The human gastric cancer cell lines MKN7 and MKN45 were obtained from the Riken Bioresource Center (Tsukuba, Japan) and maintained in RPMI 1640 medium containing 10 % fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin sulfate. Cells were cultured at 37 °C in a humidified atmosphere containing 5 % carbon dioxide (CO2).

Transfection with miR-506 Precursor (pre-miR-506)

Cells were transfected with either pre-miR-506 or pre-miR-negative control (Ambion, Austin, TX, USA) using Lipofectamine RNAiMAX (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s instructions.

Preparation of RNA for Reverse-Transcription Polymerase Chain Reaction

Total RNA was isolated using a modified acid-guanidine-phenol–chloroform procedure, as described previously.19 Complementary DNA (cDNA) was synthesized from 8 μg total RNA using random hexamer primers and Moloney murine leukemia virus (M-MLV) reverse transcriptase (Invitrogen Life Technologies).

Evaluation of Gene and miRNA Expression in Clinical Samples

Quantitative reverse transcription polymerase chain reaction (qRT-PCR) was performed in a LightCycler 480 instrument (Roche Applied Science, Basel, Switzerland) using a LightCycler 480 Probes Master kit (Roche Applied Science). The detailed protocol and the primer sequences used in this procedure are described in the Supplementary Material.

Construction of Reporter Plasmids and Evaluation of Luciferase Reporter Activity

To construct a luciferase reporter plasmid, most of the length of the SNAI2 3′UTR, which contained the miR-506 binding sites, was subcloned into the pmirGlo Dual-luciferase miRNA Target Expression Vector (Promega, Madison, WI, USA) located 5′ to the firefly luciferase. Nucleotide sequences of the constructed plasmids were confirmed by DNA-sequencing analysis. Detailed information is provided in the Supplementary Material.

Immunoblotting Analysis

Total cellular protein was extracted from MKN7 and MKN45 cells 48 h after transfection with pre-miR-506. Total protein (40 μg) was extracted from MKN cells and electroblotted as previously described.20 Detailed information is provided in the Supplementary Material.

Immunohistochemical Analysis

Levels of E-cadherin and SNAI2 protein expression were measured by immunohistochemical analysis in 39 pathologic tissue samples available from 141 cases analyzed by RT-PCR. Formalin-fixed, paraffin-embedded tissue sections corresponding to the samples used for mRNA expression analysis were analyzed. Detailed information is provided in the Supplementary Material.

Cell Proliferation and Cell Migration Analysis

Cell proliferation was evaluated by MTT assay using a Cell Proliferation Kit 1 (Roche Applied Science) according to the manufacturer’s instructions. Migration assays were conducted using the BD Falcon HTS Fluoro Block Insert (BD Biosciences, San Jose, CA, USA). Detailed information is provided in the Supplementary Material.

Statistical Analysis

Continuous variables are expressed as means ± standard deviations. Data were analyzed using JMP 9 software (JMP, Cary, NC, USA). Overall survival rates were calculated according to the Kaplan–Meier method, and the log-rank test was applied to compare the survival curves. Multivariate analysis for the survival was performed on the basis of the Cox proportional hazards model. The relationship between groups was analyzed using the Chi square test and Fisher’s test. Continuous variables between two groups were analyzed using Student’s t test after experiments had been repeated at least three times. A probability level of 0.05 was chosen for statistical significance.

Results

Clinicopathologic Significance of miR-506 Expression in Gastric Cancer

Expression of miR-506 was examined in 141 tumors by qRT-PCR to investigate the clinical significance of miR-506 in gastric cancer. Cases were subdivided into two groups [a low-expression group (n = 85) and a high-expression group (n = 56)] according to the level of miR-506 expression. These groups were classified using the minimum P value approach, which is a comprehensive method for finding the optimal risk separation cutoff point in continuous measurements.21 Clinicopathologic factors then were analyzed between the two groups (Supplementary Table 1). The group with low miR-506 expression contained significantly more poorly differentiated grades than the group with high miR-506 expression (P = 0.015). The patients with low miR-506 expression divided by the median value of miR-506 also tended to have a lower survival rate than those with high expression (P = 0.051, data not shown).

With regard to overall survival, the patients with low miR-506 expression had a significantly lower survival rate than those with high expression (P = 0.0160; Fig. 1a). We also performed subgroup analysis of the patients without peritoneum metastasis (n = 118). The patients with low miR-506 expression also had a significantly lower survival rate in this subgroup (P = 0.0096; Fig. 1b). Uni- and multivariate analyses for overall survival showed that miR-506 expression was an independent and significant prognostic factor (relative risk 1.78; 95 % confidence interval, 1.00–3.30; P = 0.049; Table 1).

Fig. 1
figure 1

The low expression of miR-506 was significantly correlated with poor overall survival. Kaplan–Meier overall survival curves according to miR-506 expression levels are shown. a The overall survival of patients with low miR-506 expression (n = 85) was significantly lower than that of patients with high expression (n = 56; P = 0.0160, log-rank test). b The overall survival of patients with low miR-506 expression (n = 71) was significantly lower than that of patients with high expression (n = 47) among the patients without peritoneum metastasis (P = 0.0096, log-rank test)

Full size image
Table 1 Uni- and multivariate analyses of clinicopathologic features for overall survival (Cox proportional regression model)
Full size table

Regulation of SNAI2 Expression by miR-506 in Gastric Cancer

We next explored the potential target genes of miR-506 in gastric cancer. Using TargetScan, an online tool available at http://www.targetscan.org/vert_50/, we identified a potential miR-506 binding site in the 3′UTR of the transcript encoding SNAI2 (Supplementary Fig. 1). A luciferase reporter assay was conducted for direct investigation of binding and repression between miR-506 and SNAI2. Transient cotransfection of MKN7 and MKN45 cells with the reporter plasmid and pre-miR-506 significantly reduced luciferase activity compared with the negative control cells (P < 0.05; Fig. 2a). These data indicated that SNAI2 mRNA is a direct functional target of miR-506 in gastric cancer.

Fig. 2
figure 2

In gastric cancer cell lines and samples from patients with gastric cancer, miR-506 suppressed the expression of SNAI2 and subsequently increased the expression of E-cadherin. a Luciferase assays demonstrated that miR-506 repressed its target in MKN7 cells (left) and MKN45 cells (right) (P < 0.05). Relative luciferase level = [(sample Luc/sample Renilla)/(control Luc/control Renilla)]. Luc, raw firefly luciferase activity; Renilla, internal transfection control Renilla activity. The error bar represents the standard deviation (SD) from three replicates. b Expression of SNAI2/GAPDH as measured by quantitative real-time polymerase chain reaction (PCR) analysis in MKN7 cells (left) and MKN45 cells (right) after transfection with pre-miR-506. c Expression of E-cadherin transcripts and protein was measured by quantitative real-time PCR analysis and Western blot analysis, respectively, in MKN7 cells transfected with pre-miR-506. Protein expression was normalized to the expression of β-actin. d The group with high miR-506 expression exhibited significantly lower SNAI2 expression than the group with low expression (P = 0.009). Error bars represent the SD from three replicates. NC, pre-miR negative control; 30 nmol, pre-miR-506 30 nmol; 50 nmol, pre-miR-506 50 nmol

Full size image

Endogenous miR-506 expression then was measured in three gastric cancer cell lines. Cell lines with high endogenous miR-506 expression (MKN-1 and NUGC-3) showed significantly lower SNAI2 expression than cells with low endogenous miR-506 expression (MKN-7) (Supplementary Fig. 2). Next, we sought to confirm that miR-506 mediated the expression of SNAI2 in two gastric cancer cell lines (MKN7 and MKN45) using pre-miR-506. We confirmed that miR-506 expression in cells transfected with pre-miR-506 was significantly higher than in cells transfected with pre-miR-negative control using qRT-PCR (P < 0.05; Supplementary Fig. 3). Moreover, SNAI2 expression was significantly suppressed in MKN7 cells exposed to pre-miR-506 (P < 0.05; Fig. 2b). In MKN45 cells transfected with pre-miR-506, the expression level of SNAI2 tended to be suppressed, but this difference was not statistically significant due to the low basal expression of this target (Fig. 2b).

To investigate the function of miR-506 in the EMT, we examined the expression level of the epithelial marker, E-cadherin, after overexpression of miR-506 in gastric cancer cells. As shown in Fig. 2c, MKN7 cells transfected with pre-miR-506 exhibited significantly increased expression of E-cadherin mRNA and protein.

Moreover, in clinical samples, the group with high miR-506 mRNA expression exhibited significantly lower SNAI2 expression than the group with low miR-506 expression (P = 0.009; Fig. 2d). We explored the protein expression of SNAI2 in 39 gastric cancer patients. Immunohistochemical analysis showed that SNAI2 protein expression inversely correlated with miR-506 mRNA expression, and the correlation was statistically significant (P = 0.006, Supplementary Table 2). Expression of SNAI2 was inversely correlated with E-cadherin expression in identical lesions of resected gastric cancer samples, as shown in Fig. 3. These data directly demonstrated that miR-506 controlled the expression of SNAI2 through binding to its 3′UTR and induced the EMT in gastric cancer.

Fig. 3
figure 3

Immunohistochemical analysis of clinical samples. Clinical samples showed an inverse correlation between the expression of SNAI2 and the expression of E-cadherin, as examined by immunohistochemical staining. Scale bar: 50 μm

Full size image

Suppression of Cell Proliferation and Migration in Gastric Cancer by miR-506

Next, we evaluated the role of miR-506 in determining the malignant potential of gastric cancer. Proliferation assays were conducted with gastric cancer cells transfected with pre-miR-506 and with negative control cells. The findings showed that miR-506 significantly suppressed the proliferation of both gastric cancer cell lines (P < 0.05; Fig. 4a). The expression of miR-506 also significantly inhibited the migratory capacity of the cells compared with that of control cells (P < 0.05; Fig. 4b). These data demonstrated that impaired expression of miR-506 promoted the malignant potential of gastric cancer.

Fig. 4
figure 4

Overexpression of miR-506 suppressed the proliferation and migration of gastric cancer cells. As shown, miR-506 significantly suppressed the a proliferation and b migratory capacity of gastric cancer cells compared with control cells (P < 0.05). Error bars represent the standard deviation from three replicates. NC, pre-miR negative control; 30 nmol, pre-miR-506 30 nmol; 50 nmol, pre-miR-506 50 nmol

Full size image

Discussion

In this study, the expression of miR-506 was inversely correlated with the expression of SNAI2 in clinical samples, and gastric cancer cell lines with miR-506 overexpression exhibited decreased expression of SNAI2 and increased expression of E-cadherin. Moreover, the study also provided direct evidence that miR-506 suppresses SNAI2 and that overexpression of miR-506 significantly suppresses cell migration. This is the first report to show that miR-506 controls the EMT by inhibiting SNAI2 in gastric cancer.

Previous studies have shown that the EMT plays an important role in peritoneal metastasis.15,16 In this study, however, we did not observe a relationship between the expression of miR-506 and the incidence of peritoneal metastasis. The data indicated that although reduced expression of miR-506 induces the EMT, other factors besides epigenetics are essential to the formation of peritoneal metastasis.22,23

Previous studies have shown that low expression of miR-506 is associated with a poor prognosis in serous ovarian and breast cancers.17,18 Consistent with this, our study initially showed that low expression of miR-506 was an independent prognostic factor in gastric cancer patients. The low expression of miR-506 was significantly correlated with poorer differentiation and more invasive properties. However, no significant relationships were found between the expression of miR-506 and lymph node metastasis or peritoneal metastasis. The group with low miR-506 expression contained more undifferentiated histopathologic grades, which are characterized by more malignant, mesenchymal-like cells with low E-cadherin expression.24,25 This indicates that miR-506 is essential to maintain the differentiation of cancer cells and that loss of miR-506 expression could lead to a poor prognosis.

In addition to its functions in the EMT, recent findings have shown that miR-506 acts as a tumor suppressor. In lung cancer, ectopic expression of miR-506 suppresses cell viability to induce the accumulation of reactive oxygen species (ROS),26 and miR-506 inhibits tumor growth by targeting the hedgehog pathway transcription factor Gli3 and the CDK4/6-FOXM1 axis in cervical cancer and ovarian cancer.27,28 In our study, overexpression of miR-506 by transfection with pre-miR-506 significantly suppressed cell proliferation in gastric cancer cell lines, suggesting that miR-506 may act as a tumor-suppressor miRNA to prevent gastric cancer progression.

Many miRNAs have been shown to control the EMT in various cancers.29 For example, the miR-200 family, miR-30a, and others have been shown to control the EMT by repressing the expression of EMT-related target proteins, such as ZEB or Vimentin. 12,30 It is possible that miR-506 indirectly controls the EMT by pathways other than the miR-506-SNAI2 axis. However, only one practical axis exists between miR-506 and SNAI2 because other EMT-related genes, such as SNAI1, ZEB1, and ZEB2, do not have binding sites for miR-506 in their 3′UTRs.

Few studies have described the mechanisms through which miR-506 expression is controlled in highly malignant cancers. One study showed that nuclear factor-κB (NF-κB) plays an important role in the EMT.31 Additionally, NF-κB has been shown to suppress the expression of miR-506 by binding to its promoter region in breast cancer.18 Interestingly, genomic sequences upstream of miR-506 may be putative p53-response elements. Consistent with this, p53 has been shown to promote the expression of miR-506 in lung cancer.26 In the current study, miR-506 was associated with proliferation and migration ability but was not correlated with invation and p53 expression (data not shown). Therefore, further study is warranted to investigate another mechanism, such as amplification, deletion, or methylation, that controls the expression of miR-506.

In conclusion, this study showed that miR-506 directly controlled the EMT by regulating SNAI2 and was an independent prognostic factor in Japanese gastric cancer patients. Moreover, our data supported the conclusion that miR-506 may act as a tumor suppressor in the context of gastric cancer. Because miRNAs have recently been recognized as potential therapeutic agents or targets in various diseases and because systemic delivery of synthetic miRNA has been shown to inhibit the growth of tumors,32,33 we hypothesize that miR-506 could be a potential therapeutic molecule in the treatment of gastric cancer. Further studies investigating the potential applications and significance of miR-506 are ongoing.