This article is mentioned in:

Introduction

Colorectal cancer (CRC) is the third and second most common cancer diagnosed in males and females, respectively, with >1.2 million new cancer cases diagnosed and >0.6 million mortalities annually (1). Although patients diagnosed with localized disease have a 5-year survival rate as high as 90.3%, this rate significantly decreases to 12.5% in patients with distant metastasis (2). Mutation, phenotypic drift, epithelial-mesenchymal transition (EMT), collective migration, resistance to apoptosis and anoikis, vascularization, and lymphangiogenesis are among the known mechanisms of metastasis (3). Thus, mechanism-based targeted therapies been on the increase and are thought to have high efficacy that inhibits a target, while having relatively fewer off-target effects and thus less non-specific toxicity. By contrast, a targeted therapeutic agent that inhibits a key pathway in a cancer cell may not completely eradicate the entire tumor, allowing residual heterogenic cancer cells to survive, followed by almost inevitable relapses (4). Accumulating evidence suggests that microRNAs (miRNAs) are important in carcinogenesis by affecting the expression of genes that regulate cancer progression (5). When specific miRNAs involved in the process of metastasis are identified, therapeutic strategies can be developed to silence oncogenes or upregulate tumor-suppressor genes.

microRNA-206, unlike other myomiR family members, has been consistently identified based on Northern blot, microarray, RNase protection assays, and RT-PCR to be highly expressed in skeletal muscle (6). Recently, the findings of a number of studies have shown that miR-206 is frequently downregulated in many human malignancies, including breast cancer (7), rhabdomyosarcoma (8), renal (9), lung (10), laryngeal (11), endometrioid (12), colorectal (13), gastric cancers (14) and melanoma (15), suggesting that miR-206 is associated with a more malignant phenotype (10,13,16). Furthermore, miR-206, in combination with miR-21, −135a, −335 and let-7a, can be used as a prognostic panel for detecting the presence of metastasis in human CRCs, with a specificity and sensitivity of 87 and 76%, respectively (13). Oncogenes, including CCND1 (15), ESR1 (17), MET (18), and NOTCH3 (19), have been identified and confirmed to be targets of miR-206, suggesting that miR-206 has a tumor-suppressor role in CRC. Of these targets, NOTCH3, identified as the third mammalian Notch receptor and expressed in proliferating neuroepithelium (20), has been reported to be frequently expressed in human CRC tissues and has been shown to play a role in the modulation of CRC cell proliferation and tumorigenic potential in xenograft models (21). Previous studies have validated the inhibitory mechanism of miR-206 through its seed match region binding to the NOTCH3 3′ untranslated region, although in the context of HeLa and C2C12 cells (19,22). However, whether the interplay between miR-206 and NOTCH3 occurring in CRC is important is unknown.

In the present study, the expression of miR-206 and NOTCH3 in CRC tissues and cell lines was analyzed. Based on an inverse association between the expression of this molecular pair, miR-206 mimics were transiently transfected into the SW480 (and the metastatic strain) and SW620 colon cancer cell lines. The results showed that upregulation of miR-206 inhibited cancer cell proliferation and migration, blocked the cell cycle, and activated apoptosis, even in SW620 cells, exhibiting a more aggressive property and involving NOTCH3 signaling, EMT, and extracellular matrix degradation. Therefore, our results confirmed miR-206 as a tumor suppressor in CRC and suggested a potential therapeutic target for clinical intervention.

Materials and methods

Patients and tissue samples

Approval from the Institutional Review Board of Huai’an First People’s Hospital was obtained for the present study. A total of 49 sequential patients who underwent surgery for the first manifestation of CRC were included. Patients who underwent preoperative radiation or chemotherapy were not included in this study. Tumors were staged according to the AJCC Cancer Staging Manual (6th edition). The primary tumors, but not the metastatic lesions, were fixed in 10% formalin, embedded in paraffin and cut (5–8 μm) for routine histopathology examination by the Department of Pathology (Huai’an First People’s Hospital, China). The tissue samples were snap-frozen and stored at −80°C for RNA and protein extraction.

Immunohistochemistry

Tissue sections were deparaffinized and rehydrated, and then incubated for heat-induced antigen retrieval in 0.01 M citrate buffer (pH 6.0) at 100°C for 10 min. After cooling to room temperature, 0.6% H2O2/80% methanol was applied to the sections for 10 min at room temperature to eliminate endogenous peroxidation. After washing with D-PBS, the sections were blocked with 10% bovine serum albumin and 0.2% Triton X-100 in D-PBS at room temperature for 30 min prior to incubation with anti-NOTCH3 rabbit polyclonal antibody (AB61043a, Sangon, Shanghai, China) diluted 100-fold at 4°C overnight. After washing with PBS-T, the sections were applied with the Polink-1 HRP DAB detection system (ZSGB-BIO, Beijing, China) for color development. After the reaction was stopped by washing, the sections were counterstained with hematoxylin, dehydrated with alcohol, mounted with neutral balsam and imaged on a BX51 microscope with an E-620 digital camera (Olympus, Tokyo, Japan).

In situ hybridization

After rehydration in water, the sections proceeded to hybridization using an Enhanced Sensitive ISH detection kit (Boster, Wuhan, China) according to the manufacturer’s instructions. Briefly, the sections were treated with pepsin at 37°C for 10 min, washed with 0.5 M TBS and DEPC-treated water, pre-hybridized at 40°C for 1 h, and hybridized with 100 nM 5′-digoxigenin (DIG)-labeled LNA™ probe against hsa-miR-206 at 50°C for 8 h. After hybridization, the slides were washed with 2X, 0.5X and 0.2X SSC, sequentially. The slides were then blocked for 30 min at 37°C, and incubated with biotinylated anti-DIG antibody at 37°C for 2 h. After 0.5 M TBS washing, alkaline phosphatase (AP)-conjugated streptavidin was applied to the section for 1 h at room temperature. Immediately after washing the slides, AP substrate containing 0.2 mM levamisol was applied to the sections protected from light at 30°C for 2 h. When the desired intensity of chromogenic reaction was reached, the slides were washed with water, mounted in aqueous medium and imaged on a BX51 microscope with an E-620 digital camera (Olympus, Tokyo, Japan).

Cell lines, in vitro culture and transfection

The SW480 and SW620 human colon cancer cell lines were obtained from the Shanghai Institutes of Biological Sciences (Shanghai, China). The cells were cultured in Leibovitz’s L-15 medium (Gibco, Grand Island, NY, USA) supplemented with 10% HyClone® fetal bovine serum (Thermo Scientific, Beijing, China). The cells were maintained at 37°C and 5% CO2 in a humidified incubator to near confluence and were deprived of serum for 16 h prior to use in the experiments. Experimental data were obtained from cells passaged between 3 and 10.

For transient transfection, the cells were first cultured to reach 50–75% confluence and transfected with 100 nM miR-206 mimic or scrambled negative control (GenePharma, Shanghai, China) with Lipofectamine 2000 (Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s instructions.

Total RNA extraction and reverse transcription-quantitative PCR (RT-qPCR)

Tissue samples or cultured cells were homogenized in TRIzol® reagent (Life Technologies). Total RNA extraction was performed as described in the manufacturer’s instructions with one modification, i.e., that isopropanol precipitation was proceeded at −20°C overnight instead of at room temperature for 10 min (23). RNA concentration was measured by the NanoDrop 1000 UV/Vis spectrophotometer (Thermo Scientific, Wilmington, DE, USA).

Prior to reverse transcription, RNAs were treated with DNase I (NEB, Beijing, China) according to the manufacturer’s instructions. For amplification of ACTB, NOTCH3, JAG1, HEY1, CDH2, MMP9 mRNA and RNU6-1, cDNA was generated using the ReverTra Ace-α-® Reverse Transcription kit (Toyobo, Japan) with random (dN)9 primer. According to the manufacturer’s instructions, the reaction was performed on 250 ng of total RNA in 10 μl total volume as follows: 30°C for 10 min, 42°C for 30 min, 99°C for 5 min and stored at −20°C. Concerning miR-206, primers for reverse transcription and qPCR were designed as previously described (24). A pulsed gene-specific RT reaction (46) in 10 μl total volume containing 250 ng RNA and 4 nM stem-loop primer [5′-CTCAACTGGTGTCGTGGAGTCGGCAATTCAGTTGAGCCACACAC-3′, (24)] was applied as follows: 16°C for 30 min, followed by 60 cycles at 20°C for 30 sec, 42°C for 30 sec and 50°C for 1 sec, then 99°C for 5 min and stored at −20°C.

DyNAmo™ ColorFlash SYBR-Green® qPCR kit (Thermo Scientific, Espoo, Finland) was used with a modified version relating to the volumes of reagents and cDNA in 10 μl total volume: 5 μl of 2X master mix, 1 μl RT product, and 180 nM forward and reverse primer pairs. The primers used in the qPCR were: ACTB (001101.3), forward, 5′-CACGAAACTACCTTCAACTCC-3′ and reverse, 5′-CATACTCCTGCTTGCTGATC-3′; NOTCH3 (008716.2) forward, 5′-ACAGACTGGATGGACACAGAG-3′ and reverse, 5′-GATGTCAGCAGCAACCAGATG-3′; JAG1 (000214.2) forward, 5′-TGTCGGTCTTCCAGTCTCC-3′ and reverse, 5′-CACTGCAAATGTGCTCCGTAG-3′; HEY1 (012258.3) forward, 5′-CATACGGCAGGAGGGAAAGG-3′ and reverse, 5′-AACTCGAAGCGGGTCAGAGG-3′; CDH2 (001792.3) forward, 5′-AGTCAACTGCAACCGTGTCT-3′ and reverse, 5′-AGCGTTCCTGTTCCACTCAT-3′; MMP-9 (004994.2) forward, 5′-TTCTACGGCCACTACTGTGC-3′andreverse,5′-AGAATCGCCAGTACTTCC CATC-3′;RNU6-1(004394.1)forward,5′-GCAGCACATATACTAAAATTGGAACGA-3′ and reverse, 5′-AATATGGAACGCTTCACGAATTTGC-3′; and miR-206 (MIMAT0000239) forward, 5′-AGCTCGATTAAGGTGGAATGTAAGGAAGT-3′ and reverse, 5′-CTCAACTGGTGTCGTGGAGTCGG-3′. Three-step qPCR for ACTB, NOTCH3, JAG1, HEY1, CDH2 and MMP-9 cDNA was performed as follows: 95°C for 7 min; 40 cycles of denaturation at 95°C for 20 sec, annealing at 60°C for 20 sec, and extension at 72°C for 20 sec. Two-step qPCR for RNU6-1 and miR-206 was performed as follows: 95°C for 7 min; 40 cycles of denaturation at 95°C for 10 sec and extension (RNU6-1 at 68°C, miR-206 at 59°C) for 30 sec.

RT-qPCR reactions, including RT-minus controls and no-template controls, were run in triplicate on an CFX™ 96 real-time PCR detection system (Bio-Rad, Hercules, CA, USA). The quantification cycle (Cq) was defined as the fractional cycle number at which fluorescence passes the fixed threshold determined by CFX Manager™ software ver. 1.6 (Bio-Rad). The expression level of mRNA and miRNA was calculated by the ΔCt method (25) using ACTB and RNU6-1, respectively, as a reference.

Western blotting

Following removal of the RNA aqueous phase from TRIzol® the remaining phenol-chloroform layer was used for protein isolation according to the manufacturer’s instructions. Proteins were dissolved in 0.5% SDS/3 M urea, quantified by the BCA assay kit (CWBIO, China), and mixed with 5X SDS-PAGE sample buffer (10% SDS/5% 2-mercaptoethanol). After the samples were denatured at 95°C for 5 min, 30 μg proteins per lane were separated by 10% SDS-PAGE and transferred to Immobilon®-P membranes (Millipore, Billerica, MA, USA). The membranes were blocked in D-PBS/0.05% Tween-20 (PBST) containing 5% defatted milk for 1 h at room temperature. The blocked membranes were then probed with anti-ACTB rabbit polyclonal antibody (Proteintech, Chicago, IL, USA), diluted at 2,000-fold; anti-NOTCH3 rabbit polyclonal antibody (Sangon, China), diluted at 250-fold; or anti-Caspase-3 rabbit monoclonal antibody (Cell Signaling Technology, Boston, MA, USA) at 4°C overnight. After washing with PBST, the membranes were then incubated with HRP-conjugated goat anti-rabbit IgG antibody (Boster, China), diluted at 4,000-fold for 1 h at room temperature. After washing with PBST, the immunolabeling proteins were reacted with chemiluminescent HRP substrate (Millipore) and visualized by ChemiDoc™ XRS systems (Bio-Rad).

Cell proliferation and apoptosis assay

For the proliferation assay, the transfected cells were seeded in a flat-bottom 96-well plate at 1×104/well. After maintaining in serum-free medium for 16 h, the cells were recovered to the culture medium followed by the addition of 20 μl/well of [3-(4,5-dimethylthiazol-2-yl) 2,5 diphenyl tetrazolium bromide)] (MTT) solution (5 mg/ml in PBS) for 4 h at 37°C. The reaction was terminated by adding 200 μl dimethyl sulfoxide to each well, followed by gentle agitation for 20 min. The absorbance of each well was then measured at 490 nm using an Elx808 microplate reader (BioTek, Winooski, VT, USA).

At 36 h post-transfection, cells (5×105) were resuspended and stained using an Annexin V-FITC Apoptosis detection kit (BioBox, Nanjing, China), and then detected on a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA, USA). The apoptotic cell level was calculated by dividing FITC-stained only cells to gated cells.

Cell cycle analysis

The cells seeded in 6-well plates at 1×106cells/well were cultured and transfected as mentioned above. Following serum starvation for 16 h, the cells were digested with trypsin and neutralized with Leibovitz’s L-15 medium. Cell cycle analysis was performed using a CycleTest™ Plus DNA reagent kit (BD Biosciences). Each sample was analyzed on a FACSCalibur flow cytometer (BD Biosciences). The cell subpopulations in G0/G1, S and G2/M phases were calculated based on differences of DNA content.

Wound-healing assay

The cells seeded at 5×105 cells/well were cultured overnight in 6-well plates. The following day, 100 nM miR-206 mimic or negative control was transfected using Lipofectamine 2000. Untransfected cells were cultured in medium as blanks. After 36 h, transfected cells were grown to confluence and wounded by dragging a 0.2-ml pipette tip through the monolayer. The cells were washed using pre-warmed PBS to remove cell debris and allowed to migrate for 24 h. The rate of wound closure was expressed as a percentage of the initial scraped gap.

Cell migration assay

Cells were seeded, cultured and transfected as described above. At 36 h post-transfection, the cells collected and resuspended in Leibovitz’s L-15 medium were placed in the upper compartment at 1×105/100 μl. After 10 h of incubation, the cells on the top of the membrane were removed by swiping with a damp cotton swab. The membrane was fixed with 4% paraformaldehyde for 30 min and stained with 0.1% crystal violet for 15 min. The cells on the underside of the membrane were counted using a light microscope. For each triplicate, the number of cells in 10 fields (x200 magnification) was determined, and the counts were averaged.

Statistical analysis

Data were presented as mean ± SD. A Shapiro-Wilk normality test was used to evaluate the distribution of quantitative observations. Skew distributional data expressed as median (interquartile range) were analyzed using non-parametric tests. The Wilcoxon matched-pairs signed-rank test was used to compare between tumors and matched adjacent normal tissues. A comparison between tissue samples with different stage was analyzed with the Mann-Whitney test. The correlation between miR-206 and NOTCH3 protein expression was analyzed using Spearman’s rank correlation test. The in vitro cell experiment was repeated at ≥3 times. Two-factor analysis of variance was performed using two-way ANOVA. Data were analyzed using Stata/SE version 11.2 (StataCorp, TX, USA) and plotted in Prism® ver. 5 (GraphPad, College Station, TX, USA). P<0.05 was considered statistically significant.

Results

microRNA-206 expression is opposite to NOTCH3 in primary CRCs

According to the occurrence of metastasis, patients were enrolled in the following categories: stage I and II with non-metastatic disease (n=28); and stage III and IV with metastatic disease (n=21). Only primary tumors and matched adjacent normal tissues, but not the metastatic lesions, were analyzed in the present study. A significant difference was noted in the miR-206 expression of patients with stage I–II or stage III–IV disease between tumors and normal tissues (stage I–II: P<0.0001; stage III–IV: P=0.0001; Fig. 1A), while there were no obvious differences between stage I–II and stage III–IV patients in normal tissues or tumors (normal tissues, P=0.4057; tumors, P=0.2577). By contrast, NOTCH3 protein was expressed in opposition to miR-206, and was significantly upregulated in tumors and minimally detected in normal tissues (stage I–II, P<0.0001; stage III–IV, P=0.0001; Fig. 1B). However, patients with advanced stages had a higher level of NOTCH3 in tumors and normal tissues (normal tissues, P=0.0306; tumors, P=0.0068). Based on these data, we further examined tissue distributions by ISH and immunohistochemistry, respectively. microRNA-206 expression was weak-to-moderate in normal mucosa, but was absent in tumor tissues. By comparison, NOTCH3 was predominantly localized in the cytoplasm and nucleus of tumor cells with strong staining, but was rarely visible in normal mucosa (Fig. 1C). Apparently, there was an inverse association between miR-206 levels and NOTCH3 expression among the 49 CRC patients (Spearman’s rho =−0.786, P<0.0001; Fig. 1D).

Figure 1 miR-206 expression is opposite to NOTCH3 in primary colorectal cancers. (A) The levels of miR-206 were significantly lower in tumors (T) than in normal tissues (N) of patients with Stage I–II (n=28, P<0.0001) and Stage III–IV (n=21, P=0.0001). (B) The levels of NOTCH3 protein were significantly higher in tumors (T) than in normal tissues (N) of patients with Stage I–II (n=28, P<0.0001) and Stage III–IV (n=21, P=0.0001). (C) The tissue sections were examined for miR-206 by in situ hybridization and NOTCH3 by immunohistochemistry. Scale bars, 50 μm. (D) miR-206 expression in the 49 colorectal patients was reverse to the NOTCH3 protein expression (Spearman’s r =−0.786, P<0.0001). The levels of miR-206 were quantified by RT-qPCR and normalized to RNU6-1. The relative levels of NOTCH3 were determined by western blotting and expressed as the ratio to ACTB.

Upregulation of miR-206 alters NOTCH3 expression in colon cancer cells

A previous study had identified NOTCH3 as a target of miR-206, resulting in apoptosis and migration in HeLa cells (19). To investigate the effect of miR-206 targeting NOTCH3 in CRC, we introduced a colon cancer cell line (SW480) and a metastatic strain (SW620) for subsequent experiments. Following transfection of cells with miR-206 mimic (100 nM), NOTCH3 mRNA and protein levels were significantly reduced in SW480 (mRNA, P<0.05; protein, P<0.001) and SW620 (mRNA, P<0.05; protein, P<0.001), which compared to the medium group (Fig. 2). As expected, treatment with scrambled negative control under the same conditions had little effect on NOTCH3 mRNA and protein expression in the two cell lines (all tests, P>0.05). Repression of NOTCH3 expression by miR-206 was more notable on protein translation than mRNA transcription. These findings suggested an inverse association between miR-206 and NOTCH3 expression in CRC.

Figure 2 Upregulation of miR-206 alters NOTCH3 expressions in colon cancer cells. Following transfection of cells with scramble normal control or miR-206 mimic (100 nM), (A) NOTCH3 mRNA and (B) protein levels were significantly reduced in SW480 and SW620 cells. Data are shown as mean ± SD of triplicate determinations. *P<0.05 vs. medium, #P<0.001 vs. medium.

miR-206 inhibits cancer cell proliferation and migration, and activates apoptosis

In view of the aforementioned data, recovery of miR-206 levels attenuated NOTCH3 protein expression, making miR-206 a potential target for therapeutic intervention. Thus, the effects of miR-206 on cell growth behavior require elucidation. Based on the MTT assay, cell viability following transfection with a miR-206 mimic (100 nM) was significantly reduced when compared to the medium group (SW480, P<0.01; SW620, P<0.01; Fig. 3A). Further examination of the cell cycle (Table I) showed that the percentage of transfected cells in the G0/G1 phase (SW480, P<0.01; SW620, P<0.01) was significantly increased, accompanied by a reduction in the S (SW480, P<0.05; SW620, P<0.05) and G2/M (SW480, P<0.05; SW620, P<0.05) phases in the two cell lines compared to the medium group. The cell cycle data obtained were consistent with the proliferation results. In addition, miR-206 mimic transfection increased the apoptotic cell number (SW480, P<0.001; SW620, P<0.001; Fig. 3B) and upregulated caspase-3 content (SW480, P<0.001; SW620, P<0.01; Fig. 3C) in the two cell lines.

Figure 3 miR-206 inhibits cancer cell proliferation and activates apoptosis. Following transfection of cells with scramble normal control or miR-206 mimic (100 nM). (A) The cell viability measured by MTT assay was significantly reduced. (B) Apoptotic cells measured by flow cytometry. (C) Caspase-3 expression measured by western blotting was significantly increased. Data are shown as mean ± SD of triplicate determinations. *P<0.01 vs. medium, #P<0.001 vs. medium.

Table I Effect of miR-206 on the cell cycle of colon cancer cells.

Table I

Effect of miR-206 on the cell cycle of colon cancer cells.

	SW480			SW620
Phase	Medium	Control	Mimic	Medium	Control	Mimic
G0/G1	70.79±1.82	71.31±1.74	77.53±2.32b	69.60±2.09	70.79±1.90	76.00±1.98b
S	24.49±2.39	24.25±2.42	19.42±2.94a	25.40±1.48	24.41±1.39	20.83±1.24a
G2/M	4.72±0.60	4.44±0.85	3.06±0.63a	5.00±0.84	4.80±0.90	3.17±0.90a

{ label (or @symbol) needed for fn[@id='tfn1-or-33-03-1402'] } Cells seeded in six-well plate at 1×10⁶/well were cultured and transfected with scrambled negative control or miR-206 mimic (100 nM). The cell cycle distribution was detected by flow cytometry and the proportions of cells within G₀/G₁, S, G₂/M phases were determined. Data are shown as mean ± SD of triplicate determinations.

a P<0.05 vs. medium,

b P<0.01 vs. medium.

To determine whether miR-206 affects migration in CRC, the wound-healing and Boyden chamber assays were used to perform cell motility measurements. Cells transfected with miR-206 mimic presented a significant reduction in the percentage of wound closure in SW480 and SW620 cells compared to the medium group (SW480, P<0.001; SW620, P<0.001; Fig. 4A–C). Similar decreases in cell migration in the two cell lines transfected with miR-206 mimic were also observed using the Boyden chamber assay compared to the medium group (SW480, P<0.001; SW620, P<0.001; Fig. 4D).

Figure 4 miR-206 reduces cancer cell migration. Following transfection with scrambled normal control or miR-206 mimic (100 nM), the cell migration was significantly decreased in the (A–C) wound healing assay or (D) Boyden chamber assay. Data are shown as mean ± SD of triplicate determinations. *P<0.001 vs. medium.

microRNA-206 affects NOTCH targets and metastasis-associated genes

We determined whether NOTCH transcriptional targets and metastasis-associated genes alter their expression following miR-206 mimic transfection. As shown in Fig. 5, the mRNA levels of two NOTCH transcriptional targets (JAG1 and HEY1) and two metastasis-associated genes (CDH2 and MMP9) were significantly downregulated 24 h post-transfection with miR-206 mimic in SW480 and SW620 cells. However, the basal expression of these genes, with the exception of MMP-9, had a higher level in SW620 cells, which may be relevant to its more aggressive property than SW480 cells.

Figure 5 miR-206 affects NOTCH targets and metastasis-associated genes. Following cell transfection with scrambled normal control or miR-206 mimic (100 nM), the mRNA levels of JAG1, HEY1, CDH2 and MMP-9 were detected by RT-qPCR. Data are shown as mean ± SD of triplicate determinations. *P<0.05 vs. medium, #P<0.01 vs. medium.

Discussion

Accumulating studies acknowledge that miR-206 is a tumor suppressor that is downregulated in various tumors (7–13,15,26) and is involved in cancer metastasis (13,27). NOTCH3, a verified target of miR-206 (19), has also been reported to be frequently expressed in human CRC samples, and has a capacity in the modulation of CRC cell proliferation and tumorigenic potential in xenograft models (21). Therefore, we investigated the tumor suppressive and metastatic effects of miR-206 and its target (NOTCH3) in CRC. First, we evaluated the expression in two CRC cohorts grouped by the onset of metastasis. An inverse association between the expression of this molecular pair among the total CRC patients supported the negative regulation of miR-206 targeting NOTCH3. miR-206 levels were further reduced with CRC progression. However, the two CRC cohorts did not differ in tumor tissue or normal mucosal miR-206 expression. This finding invalidates miR-206 as a prognostic marker used independently for metastasis, as it was combined into a 5-miR panel for predicting metastasis (13).

In CRC tissues, miR-206 was expressed at an extremely low level and was scarcely probed by ISH. Even in normal mucosa, miR-206 expression was minimal compared to tissue-specific expression in skeletal muscle. It is plausible that miR-206 is not identified in high-throughput profiling due to its low abundance in intestinal tissue (28–30). This low-abundance miRNA may regulate key components, the expression of which is markedly restricted under normal circumstances. As one such component, NOTCH3 was shown to have less expression in normal mucosa and overexpression in CRC tissue in the present study, which coincided with findings of other studies (21,31). In addition, only one mutation of the NOTCH3 gene was detected in 48 CRC patients (32), suggesting the gene mutation may not be important in tumorigenesis. A deficiency in miR-206 in CRC may in turn make NOTCH3, one of its targets, lose control with intensified upregulation. A high NOTCH3 expression has been shown to be strongly correlated with metastasis in CRC (31), and may be an independent prognostic factor for hepatocellular carcinoma as well (33). Compared to a decrease of the overexpressed NOTCH3 by introducing an exogenic antagonist, rescuing the reduced miR-206 in keeping with the physiologic requirements may be more rational. Following transfection with miR-206 mimics in colon cancer cells, the mRNA and protein levels of NOTCH3 were all downregulated. The same phenomenon has also been observed in HeLa (19) and HepG2 (34) cells. This finding supports the hypothesis that miRNAs are involved in gene regulation via post-transcriptional silencing on its targets and interfering with the transcription of target genes in an indirect manner. Initially, studies on miR-206 mostly focused in muscle tissue, which was due to the high tissue-specificity of miR-206 screened in high-throughput analyses (6). Occasionally, a significant expression of miR-206 was detected in normal murine skin (24,35) and HPV8-mediated skin tumors (36). Since the first evidence for the post-transcriptional regulation of ESR1 by miR-206 in MCF-7 breast cancer cells, miR-206 has been denoted as a tumor suppressor that is downregulated in many types of cancer, and a number of target genes, including oncogenes and tumor-suppressor genes, have been identified (37). Our results confirmed the repressive effect of miR-206 on NOTCH3 expression at the mRNA and protein levels in CRC, which also suggests that miR-206 coordinates molecular networks in cancer.

Overexpressed miR-206 has been shown to attenuate colon cancer cell proliferation and migration in vitro. Such inhibitory effects of miR-206 are evident in other cancer models in which growth-associated targets, such as CCNC (15), CCND1 (15), CCND2 (14), CDC42 (38), CDK4 (15), MET (18), and VEGF (11), have been identified. The increase in cell number in the G0/G1 phase caused by miR-206 mimic transfection suggests that cyclin genes are involved in the regulatory networks of miR-206, and probably account for the downregulation of NOTCH3 at the mRNA level. Therefore, underexpression of miR-206 in CRC may contribute to oncogenesis by dysregulating the cell-cycle drivers.

In contrast to Notch1, Notch2 and Notch4, Notch3 does not appear to be essential for embryonic development as Notch3-null mice are viable and fertile, but have impaired maturation of vascular smooth muscle cells (39). NOTCH3 activation also protects against apoptosis through CBF1/RBP-Jκ independently, but cross-talks to the ERK/MAPK (40) or PI3-kinase/AKT (41) pathway, which emphasizes its indispensable role in maintaining cell viability. Thus, the downregulation of NOTCH3 following miR-206 mimic transfection increases apoptosis and caspase-3 content. Furthermore, MMP-2 and MMP-9, are crucial in tumor invasion and metastasis because of specificity for type IV collagen (the principal component of the basement membrane) (42), have been shown to be regulated by NOTCH3 via the ERK1/2 pathway, and strongly correlate with hepatocellular carcinoma metastasis (33). The invasion and migration of MDA-MB-231 breast cancer cells in vitro were inhibited with the downregulation of MMP-9 via miR-206 targeting CDC42 (38). Similarly, CDH2 (N-cadherin), mediating cell-cell interaction to prevent cell apoptosis by activating the PI3-kinase/AKT pathway (43), was downregulated following miR-206 mimic transfection in the current study. CDH2 was upregulated in dedifferentiated, invasive prostate carcinomas (44) and has been reported to participate in TNF-α-induced EMT in CRC cells (45). The levels of expression of JAG1, HEY1, and CDH2 mRNA were higher in SW620 than SW480 cells, which supports the metastatic property of SW620 compared to its primary strain (SW480). Nevertheless, the upregulation of miR-206-induced tumor suppressive effects was similar in the 2 cell lines, suggesting miR-206 plays a fundamental role in the tumorigenesis of CRC involving multiple signaling pathways.

Consistent with previous reports, our results support the hypothesis that expressed miR-206 can decrease cell proliferation and migration, block the cell cycle, and activate apoptosis in CRC cells. The tumor suppressive capacity of miR-206 has a similar effect on CRC cells with different metastatic potential and may be explained by its direct NOTCH3 signaling inhibition and indirect cross-talk with other signaling pathways involving CDH2 and MMP-9. These results emphasize the potential therapeutic strategy for CRC by rescuing miR-206, which is downregulated in CRC. Future studies may continue to evaluate the effects of miR-206 on tumor growth in vivo, and to understand how it affects the pathogenesis of CRC, which may reveal feasible therapies for this disease.

Acknowledgements

The authors thank Dr Yuan Mu (Department of Clinical Laboratory, Nanjing Children’s Hospital, Nanjing Medical University) for the technical assistance and helpful discussions.

References