This article is mentioned in:

Introduction

Gastric cancer (GC) is one of the most common human cancers and the second leading cause of cancer-related mortality worldwide (1). The major cause of death is metastasis, which greatly hinders treatment success (2). Human gastric cancer tissue contains cancer initiating cells (CIC) (3). CIC are defined as the unique subpopulation in the tumors that possess the ability to initiate tumor growth and sustain self-renewal as well as metastatic potential (4). Elucidating molecular mechanism of formation of gastric CIC will not only help us to further understand the pathogenesis and progression of the disease, but also will offer new targets for effective therapies.

EHF/ESE-3 is a new member of the ETS transcription factors which is exclusively expressed in a subset of epithelial cells (5). Aberrent expression of EHF may affect the normal process of epithelial cell differentiation and contribute to cell transformation (5–7). Moreover, EHF may regulate epithelial growth and differentiation and have an important role in oncogenesis of epithelium-derived tumors (5,8). EHF is overexpressed in ovarian and mammary cancers and may be a predictive marker for poor survival in ovarian cancer (9). Recently, it has been reported that increased expression of EHF via gene amplification contributes to the activation of HER family signaling and associates with poor survival in gastric cancer (10).

miRNAs are regulatory, non-coding RNAs ~18–25 nucleotides in length and are expressed at specific stages of tissue development or cell differentiation, and have large-scale effects on the expression of a variety of genes at the post-transcriptional level. Through base-pairing with its targeted mRNAs, a miRNA induces RNA degradation or translational suppression of the targeted transcripts (11–15). Some miRNAs can function either as oncogenes or tumor suppressors (16–18) and expression profiling analyses have revealed characteristic miRNA signatures in certain human cancers (19–21). miR-206 was confirmed to be downregulated in gastric cancer specimens (22). Restoration of miR-206 can inhibit gastric cancer progression (22,23). However, its role in regulating formation of CIC has not been reported. In this study, we showed that miR-206 inhibits cancer initiating cells by targeting EHF in gastric cancer.

Materials and methods

Gastric cancer tissues

Gastric cancer tissues and adjacent normal tissues were obtained from Department of Gastrointestinal Surgery, Shandong Provincial Qianfoshan Hospital. All tissues were examined histologically, and pathologists confirmed the diagnosis. Medical ethics committee of Qianfoshan Hospital approved the experiments undertaken. The use of human tissue samples followed internationally recognized guidelines as well as local and national regulations. Informed consent was obtained from each individual.

Human gastric cancer cell lines

NCI-N87 and GES-1 cell lines were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). They were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (Wisent, Canada) and antibiotics (1% penicillin/streptomycin; Gibco, USA). All cell lines were grown in a humidified chamber supplemented with 5% CO2 at 37°C.

MTT assay

The proliferation of cells was assessed by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium (MTT) assay (Sigma, St. Louis, MO, USA). The MTT analysis was performed as described previously (24–29). In brief, the cells were plated in 96-well plates in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum at a density of 8×103 cells per well at 37°C in a 5% CO2 incubator for 12 h. Cells were transfected with EHF expressing plasmids or empty vectors and then were treated with fluorouracil or DMSO (10 µM) for 24 h. Then MTT (5 mg/ml) was added to the wells (20 µl per well). The plates were incubated in a cell incubator for 4 h, then the supernatant was removed and 150 µl of dimethyl sulfoxide was added to each well. After incubation for 10 min, the absorbance of each well was measured using a Synergy™ 4 (BioTek Instruments, Winooski, VT, USA) with a wavelength of 570 nm, with the reference wavelength set at 630 nm. Absorbance was directly proportional to the number of survival cells.

Colony formation

The assay was performed as described (16). For colony formation assay, cells were transfected as indicated, and then seeded in a 6-well plate. FBS (0.2 ml) was added per well on day 5. After 9-10-day incubation, plates were washed with PBS and cells were fixed with 4% paraformaldehyde and stained with 1% crystal violet. Colonies with >50 cells were manually counted.

Western blotting

Protein extracts were resolved through SDS-polyacrylamide gel electrophoresis, transferred to PVDF membranes (Bio-Rad, Berkeley, CA, USA), probed with antibodies against EHF, ALDH1, CD44, CD133, c-MET, PAX3, cyclin D2 and CDK4 or β-actin (Abcam, Cambridge, MA, USA) and then with secondary antibodies (Abcam).

Sphere formation assay

Cells (103/ml) in serum-free RPMI-1640/1 mM Na-pyruvate were seeded on 0.5% agar precoated 6-well plates. After 10 days, half the medium was exchanged every third day. Single spheres were picked out and counted.

Quantitative reverse transcription PCR detection for miR-206

It was performed as described previously (30).

Methods of bioinformatics

The analysis of potential microRNA target sites was by the prediction algorithms - miRanda (http://www.microrna.org/).

Immunofluorescence staining

The staining was performed as described (31). Cells were stained for immunofluorescence on coverslips. After fixation and permeabilization, the cells were incubated with primary antibodies against EHF (ab172730; 1:200 dilution; Abcam) and then incubated with the secondary antibodies. The coverslips were counterstained with 4′, 6- diamidino-2-phenyl indole and imaged under a confocal microscope TCS SP5 (Lecia, Solms, Germany).

Statistical analysis

Data are presented as mean ± SEM. Student's t-test (two-tailed) was used to compare two groups. Spearman correlation was used to analyze correlation between miR-206 and EHF. P<0.05 was considered significant.

Results

EHF promotes formation of CIC phenotypes and clonogenic ability in gastric cancer NCI-N87 cells

To identify the role of EHF, we tested whether EHF expressing plasmids could stably express EHF protein in NCI-N87 cells. The results showed that EHF protein could be significantly increased by EHF expressing plasmids in the cells (Fig. 1B). In order to identify whether EHF can affect CIC traits in NCI-N87 cells, we performed sphere forming assay to assess the capacity of CIC or CIC-like cell self renewal in NCI-N87 cells. Sphere forming assay showed EHF overexpressing cells formed much bigger spheres after 14 days of culture as compared with control cells, indicating markedly increased CIC traits by EHF (Fig. 1A). To identify whether EHF can regulate ALDH1, CD133 and CD44 protein expression, we performed western blotting in NCI-N87 cells transfected with EHF expressing plasmids and empty vectors. The results showed that ALDH1, CD44 and CD133 protein are upregulated in NCI-N87 cells transfected with EHF expressing plasmids (Fig. 1B).

Figure 1. EHF promotes formation of CIC phenotypes in gastric cancer NCI-N87 cells. (A) Sphere growth for NCI-N87 cells transfected with EHF expressing plasmids and empty vectors (mock). n=3. (B) Western blotting for EHF, ALDH1, CD44 and CD133 in NCI-N87 cells transfected with EHF expressing plasmids and empty vectors (mock). β-actin was a loading control. n=3.

Clonogenic ability was increased by EHF in gastric cancer NCI-N87 cells

To determine whether cells with elevated stem-like cell characteristics could have increased clonogenic ability in NCI-N87 cells, we performed clonogenic assay. We found that clonogenic ability was significantly increased in NCI-N87 cells transfected with EHF expressing plasmids compared with NCI-N87 cells transfected with empty vectors (Fig. 2).

Figure 2. Clonogenic ability is increased by EHF in gastric cancer NCI-N87 cells. Colony formation assay for NCI-N87 cells transfected with EHF expressing plasmids and empty vectors (mock). n=3.

Silencing EHF inhibits formation of CIC phenotypes and clonogenic ability in gastric cancer NCI-N87 cells

To identify the role of sh-EHF, we tested whether sh-EHF plasmids could stably downregulate EHF protein in NCI-N87 cells. The results showed that EHF protein could be significantly decreased by sh-EHF plasmids in the cells (Fig. 3B). In order to identify whether sh-EHF can affect CIC traits in NCI-N87 cells, we performed sphere forming assay to assess the capacity of CIC or CIC-like cell self renewal in NCI-N87 cells. Sphere forming assay showed silencing of EHF formed much smaller spheres after 14 days of culture as compared with control cells transfected with scrambles, indicating markedly decreased CIC traits by sh-EHF (Fig. 3A). To identify whether EHF can regulate ALDH1, CD133 and CD44 protein expression, we performed western blotting in NCI-N87 cells transfected with sh-EHF plasmids and scrambles. The results showed that ALDH1, CD44 and CD133 protein are downregulated in NCI-N87 cells transfected with sh-EHF plasmids (Fig. 3B).

Figure 3. Silencing EHF inhibits formation of CIC phenotypes and clonogenic ability in gastric cancer NCI-N87 cells. (A) Sphere growth for NCI-N87 cells transfected with sh-EHF expressing plasmids and scrambles. n=3. (B) Western blotting for EHF, ALDH1, CD44 and CD133 in NCI-N87 cells transfected with sh-EHF expressing plasmids and scrambles. β-actin was a loading control. n=3.

Clonogenic ability was attenuated by silencing EHF in gastric cancer NCI-N87 cells

To further determine whether silencing EHF could affect clonogenic ability in NCI-N87 cells, we performed a clonogenic assay. We found that clonogenic ability was significantly decreased in NCI-N87 cells transfected with sh-EHF plasmids compared with NCI-N87 cells transfected with scrambels (Fig. 4).

Figure 4. Clonogenic ability is attenuated by silencing EHF in gastric cancer NCI-N87 cells. Colony formation assay for NCI-N87 cells transfected with sh-EHF expressing plasmids and scrambles. n=3.

EHF promotes fluorouracil-resistance in gastric cancer NCI-N87 cells

To further identify whether EHF can affect fluorouracil efficacy in NCI-N87 cells, we transfected NCI-N87 cells with EHF expressing plasmids. Then we performed MTT assay in the cells transfected with EHF expressing plasmids and empty vectors. The results showed that overexpressing EHF could transform fluorouracil sensitive NCI-N87 cells to fluorouracil-resistant cells (Fig. 5), suggesting that its overexpression promoted fluorouracil-resistance.

Figure 5. EHF promotes fluorouracil-resistance in gastric cancer NCI-N87 cells. MTT for cell viability in NCI-N87 cells. NCI-N87 cells transfected with EHF expressing plasmids and empty vectors (mock) were untreated or treated with fluorouracil.

miR-206 inhibits EHF expression in immortalized human gastric epithelial mucosa GES-1 cells and gastric cancer NCI-N87 cells

Having demonstrated that EHF is associated with formation of CICs phenotypes and increased clonogenic ability in NCI-N87 cells, next we studied the mechanisms regulating EHF expression in GES-1 cells and NCI-N87 cells. MicroRNAs (miRs) are a class of small non-coding RNAs (~22 nucleotides) and negatively regulate protein-coding gene expression by targeting mRNA degradation or translation inhibition (12,32).

To further confirm whether EHF could be regulated by microRNA, we used the common prediction algorithm - miRanda (http://www.microrna.org/microrna/home.do) to analyze 3′UTR of EHF. A dozen of microRNAs were found by the algorithm. However, we are interested in miR-206, because it has been reported that miR-206 is significantly down-regulated in gastric carcinoma (22,23). Target sites on 3′UTR of EHF are shown in Fig. 6A and it carries the identical sequence in the human, mouse (M.m.) and rat (R.n.) mRNA orthologues (Fig. 6A).

Figure 6. miR-206 inhibits EHF expression in gastric cancer NCI-N87 cells. (A) Schematic of predicted miR-206-binding sites in the 3′UTR of EHF mRNA by miRanda. Upper panel, target sites between human EHF 3′UTR and miR-206. Lower panel, sequence of the miR-206 binding site on EHF 3′UTR of human (H.s.), mouse (M.m.), and rat (R.n.). (B) Real-time PCR for miR-206 in NCI-N87 cells transfected with pre-miR-206 and control miR. U6 was a loading control. n=3. (C) Real-time PCR for miR-206 in GES-1 cells transfected with pre-miR-206 and control miR. U6 was a loading control. n=3. (D) Real-time RT-PCR for EHF in NCI-N87 cells and GES-1 cells infected with pre-miR-206 or control miR. GAPDH was a loading control. n=3. (E) Luciferase activity of wild-type (UTR-WT) or mutant (UTR-mut.) EHF 3′UTR reporter gene in NCI-N87 cells transfected with pre-miR-206 or control miR. (F) Western blotting for EHF in NCI-N87 cells and GES-1 cells transfected with pre-miR-206 and control miR. β-actin was a loading control. n=3. (G) Migration and invasion assay for NCI-N87 cells transfected with pre-miR-206 and control miR. n=3.

In an attempt to identify the role of miR-206 in regulating EHF expression in GES-1 cells and NCI-N87 cells, we transfected GES-1 cells and NCI-N87 cells with pre-miR-206 and control miR. After transfection, miR-206 expression was detected by real-time PCR and the results showed that miR-206 was significantly increased by pre-miR-206 in the cells (Fig. 6B and C). miR-206 overexpression did not cause degradation of EHF mRNA (Fig. 6D), however, it reduced the activity of a luciferase reporter gene fused to the wild-type EHF 3′UTR (Fig. 6E), indicating that miR-206 targets EHF through translational inhibition.

The action of miR-206 on EHF depends on the presence of miR-206 binding sites on the 3′UTR of EHF, because the activity of a luciferase reporter with a mutant 3′UTR of EHF was not reduced by expression of miR-206 (Fig. 6E). To support the results, we performed western blotting to detect EHF protein in NCI-N87 and GES-1 cells. We observed an evident reduction for EHF protein in miR-206-overexpressing cells (Fig. 6F). In order to detect whether EHF and miR-206 can affect abilities of migration and invasion, we performed migration and invasion assay in NCI-N87 cells. Evidently, re-expression of miR-206 completely abrogated EHF-induced cell motility and invasiveness (Fig. 6G), suggesting that this EHF is indeed a functionally important target of miR-206.

miR-206 inhibits formation of CIC phenotypes and clonogenic ability in gastric cancer NCI-N87 cells

In order to identify whether pre-miR-206 can affect CIC traits in NCI-N87 cells, we performed sphere forming assay to assess the capacity of CIC or CIC-like cell self renewal in NCI-N87 cells. Sphere forming assay showed miR-206 overexpressing cells formed much smaller spheres after 14 days of culture as compared with control cells, indicating markedly decreased CIC traits by miR-206 (Fig. 7A). To identify whether miR-206 can regulate ALDH1, CD133 and CD44 protein expression, we performed western blotting in NCI-N87 cells transfected with pre-miR-206 and control miR. The results showed that ALDH1, CD44 and CD133 protein are downregulated in NCI-N87 cells transfected with pre-miR-206 (Fig. 7B).

Figure 7. miR-206 inhibits formation of CIC phenotype in gastric cancer NCI-N87 cells. (A) Sphere growth for NCI-N87 cells transfected with pre-miR-206 and control miR (mock). n=3. (B) Western blotting for ALDH1, CD44 and CD133 in NCI-N87 cells transfected with miR-206 expressing plasmids and control miR (mock). β-actin was a loading control. n=3.

Clonogenic ability was attenuated by miR-206 in gastric cancer NCI-N87 cells

To further determine whether overexpressing miR-206 could affect clonogenic ability in NCI-N87 cells, we performed a clonogenic assay. We found that clonogenic ability was significantly decreased in NCI-N87 cells transfected with pre-miR-206 compared with NCI-N87 cells transfected with control miR (Fig. 8).

Figure 8. Clonogenic ability was attenuated by miR-206 in gastric cancer NCI-N87 cells. Colony formation assay for NCI-N87 cells transfected with pre-miR-206 and control miR. n=3.

miR-206 is inversely correlated with EHF mRNA expression in gastric cancer tissues

Furthermore, we detected the endogenous mRNA level of EHF in GC tissues. The results showed EHF is upregulated (Fig. 9A) and miR-206 is downregulated (Fig. 9B) and we observed a negative correlation between miR-206 mRNA levels and EHF mRNA expression in GC tissue samples (n=28) (Fig. 9C, ρ=−0.76).

Figure 9. miR-206 is inversely correlated with EHF mRNA expression in gastric cancer tissues. (A) Real-time RT-PCR for EHF in gastric cancer and adjacent normal tissues. GAPDH was a loading control. n=3. (B) Real-time PCR for miR-206 in gastric cancer and adjacent normal tissues. U6 was a loading control. n=3. (C) Spearman correlation analysis for miR-206 and EHF mRNA in gastric cancer tissues.

miR-206 inhibits c-MET, PAX3, cyclin D2, and CDK4 protein in NCI-N87 cells

In order to detect whether miR-206 can inhibit c-MET, PAX3, cyclin D2, and CDK4 protein expression, we transfected NCI-N87 cells with pre-miR-206 and control miR. Then western blotting was performed to detect the proteins. The results showed that c-MET, PAX3, cyclin D2 and CDK4 protein were downregulated in NCI-N87 cells transfected with pre-miR-206 (Fig. 10A).

Figure 10. miR-206 inhibits c-MET, PAX3, cyclin D2, and CDK4 protein in NCI-N87 cells. (A) Western blotting for c-MET, PAX3, cyclin D2, and CDK4 protein in NCI-N87 cells transfected with pre-miR-206 and control miR (mock). (B) miR-206 regulates invasion, migration, proliferation, CICs and metastasis by targeting c-MET, PAX3, EHF, cyclin D2, and CDK4 protein in gastric cancer.

Discussion

In this study, we first provided strong evidence supporting miR-206 inhibits cancer initiating cells (CIC) by targeting EHF in gastric cancer. Oncogenic activities of EHF have been reported in gastric cancer (10). EHF was frequently overexpressed and amplified in gastric cancers compared with matched non-cancerous gastric tissues (10). Consistent with previous reports (10), we found that EHF mRNA expression was upregulated in gastric cancer tissues. EHF amplification or overexpression was significantly associated with poor clinical outcomes and may be used as a potential prognostic marker for gastric cancer patients (10). CIC were found to possess the ability to sustain tumor self-renewal, initiate tumor progression, and possibly also contribute to metastasis and poor prognosis in gastric cancer (33). We showed that EHF could promote formation of CIC in gastric cancer NCI-N87 cells. Clonogenic ability was significantly increased with formation of CIC in various cancers (34–36). In line with the reports (34–36), we observed that overexpressing EHF promoted abilities of colony formation while it induced formation of CIC.

5-Fluorouracil chemotherapy is the first treatment of choice for advanced gastric cancer; however its effectiveness is limited by drug resistance. Emerging evidence suggests that the existence of CIC contributes to 5-Fluorouracil resistance (37,38). Our results showed that EHF overexpression promoted 5-fluorouracil resistance. In this study, we only used MTT to analyze 5-fluorouracil resistance. Apoptosis assay and cell cycle analysis can be performed to further confirm the roles of EHF in NCI-N87 cells.

miR-206 can inhibit progression by downregulating c-MET, PAX3, cyclin D2 and CDK4 in gastric cancer (Fig. 10B) (22,23,39). Its expression is downregulated in gastric cancer. miR-206 was downregulated in gastric cancer cells especially in high metastatic cell lines and its expression was significantly decreased in metastatic lymph nodes, compared with their corresponding primary tumor samples (23). We found that miR-206 is downregulated in gastric cancer, compared with adjacent normal tissues. Its restoration inhibited CIC traits in gastric cancer cells (Fig. 10B). In addition, we identified that EHF is one of target genes of miR-206 and it plays an important role in the network regulated by miR-206. miR-206 inhibits the expression of PAX3, cyclin D2, c-Met and cycle-related proteins CDK4 in gastric cancer cells. Consistent with previous report, we found that PAX3, cyclin D2, c-Met and CDK4 were downregulated by miR-206 in gastric cancer cells (Fig. 10B). Elucidating miR-206-mediated molecular mechanism responsible for CIC traits will help us to further understand the pathogenesis and progression of the disease and offer new targets for effective therapies.

Acknowledgements

This study was supported by Natural Science Foundation of Shandong Province: ZR2015HL081; Science and Technology Development Project of Shandong Province: 2011YD21035; Natural Science Foundation of Shandong Province: ZR2015HL070.

References