LncRNA MEG3 inhibits cell proliferation and induces apoptosis in laryngeal cancer via miR‐23a/APAF‐1 axis

Abstract Long non‐coding RNA (LncRNA) MEG3 serves a regulatory role in the progression of several types of cancer, but the role of MEG3 in laryngeal cancer is still unknown. The aim of this study was to explore the regulatory role and mechanism of MEG3 in laryngeal cancer. MEG3 expression in 50 laryngeal cancer tissue samples was detected by reverse transcription‐quantitative polymerase chain reaction (RT‐qPCR). The effects of MEG3 overexpression on laryngeal cancer cells were investigated in vitro and in vivo. The mechanism of competitive endogenous RNA (ceRNA) was validated through luciferase reporter assay, RT‐qPCR and Western blotting. MEG3 was down‐regulated in laryngeal cancer tissues, and the low MEG3 expression was associated with advanced clinical stage. Additionally, MEG3 overexpression inhibited the proliferation and induced the apoptosis of laryngeal cancer cells in vitro and in vivo. Particularly, MEG3 bound to miR‐23a specifically and a reciprocal negative regulation existed between miR‐23a and MEG3. Moreover, MEG3 up‐regulated apoptotic protease activating factor‐1 (APAF‐1), a known miR‐23a's target, thereby leading to the activation of caspase‐9 and caspase‐3. Meanwhile, these activated effects were rescued by miR‐23a overexpression. In conclusion, the present study demonstrated that MEG3 functions as a novel tumour suppressive LncRNA in laryngeal cancer for the first time. Furthermore, MEG3 may act as a ceRNA to regulate APAF‐1 expression by competitive binding to miR‐23a, thereby regulating the progression of laryngeal cancer.

gene family, are dysregulated in laryngeal cancer tissues. 6,7 More interestingly, several LncRNAs, such as HOTAIR, 8 NEAT1 9 and H19, 10 function as oncogenes to promote the tumorigenesis of laryngeal cancer. Therefore, identifying the role of LncRNAs in laryngeal cancer may bring a novel insight into the occurrence and development of laryngeal cancer.
Maternally expressed gene 3 (MEG3), located in human chromosome 14q32.3 region, has been identified as a LncRNA because its transcript lacks a significant open reading frame. MEG3 is found to be down-regulated in cancer tissues and lost in some cancer cells. 11 Moreover, MEG3 is characterized as a tumour suppressor as growing evidence that overexpression of MEG3 inhibits tumour cell growth, invasion and metastasis, induces tumour cell apoptosis and enhances tumour cell chemosensitivity. 12 However, the role of MEG3 in laryngeal cancer remains unknown. Furthermore, recent study has demonstrated that MEG3 could modulate endogenous miRNAs available for binding to their target mRNAs acting as miRNA sponges, also called competing endogenous RNA (ceRNA), thereby leading to the repression of these mRNAs. 13 For instance, MEG3 could modulate Bcl-2 by competitively binding to miR-181 to regulate gastric carcinogenesis. 14 Additionally, we predicted that MEG3 could bind to miR-23a specifically using an online bioinformatics tool (LncBase Predicted v.2) 15 and our previous study has demonstrated that miR-23a promotes the laryngeal cancer cell proliferation and induces the apoptosis by targeting apoptotic protease activating factor-1 (APAF-1). 16 Therefore, we have been suggested that MEG3 regulates APAF-1 expression by functioning as a ceRNA to competitively bind to miR-23a, thereby modulating the progression of laryngeal cancer.
In the present study, the result showed that the expression of MEG3 was down-regulated in laryngeal cancer tissues.
Overexpression of MEG3 inhibited the growth, proliferation and induced the apoptosis of laryngeal cancer cell in vitro and in vivo.

| Colony formation assay
After 24 hours of transfection, cells (3 × 10 3 cells/well) were seeded into 6-well plates and incubated in RPMI 1640 (GIBCO) with 10% FBS. After 9 days, colonies were harvested and fixed with methanol for 30 minutes, stained with haematoxylin for 20 minutes and photographed by a digital camera (Nikon).

| Cell apoptosis measured by flow cytometry
The cell apoptosis was measured by flow cytometry using Annexin V-FITC/PI Apoptosis Detection Kit (KeyGEN Biotech) according to the manufacturer's instructions. Data collection and analysis were performed using flow cytometry analyser (FACSCalibur).

| Hoechst staining assay
Cells were cultured in 24-well plates that were plated with glass

| In vivo tumorigenicity assay
The animal experiment was approved by the Animal Ethics Tumour growth was examined every three days with callipers, and tumour volumes were calculated according to the following formula: ½ × length × width 2 . After 15 days of injection, the mice were killed under sodium pentobarbital anaesthesia and the xenograft tumours were resected and weighed.

| Luciferase activity assay
The fragment of MEG3 containing the predicted miR-23a binding sites was amplified by PCR and thereby constructing pcDNA-MEG3 Wt (GenePharma Co Ltd). The pcDNA-MEG3 Mut was synthesized by mutating the target sequences in MEG3 (GenePharma Co Ltd).
The cells were seeded in 24-well plate and transfected with GV272-MEG3-Wt and GV272-MEG3-Mut together with miR-23a mimic or mimic NC. At 48 hours post-transfection, luciferase activity was analysed using the Dual-Glo Luciferase Assay System (Promega) according to the manufacturer's instructions.

| Western blotting
The tissues and cells were smashed and homogenized with RIPA

| Statistical analysis
The SPSS 17.0 software was used for statistical analysis. Data were expressed as the mean ± SEM. Difference between two groups was evaluated using Student's t test. Difference within three or more groups was evaluated using one-way analysis of variance (ANOVA), and if the difference was significant, multiple comparison analysis was further performed using Fisher's least significant difference (LSD) test. All P values < .05 were considered statistically significant.

| Expression of MEG3 in laryngeal cancer tissues
To explore the expression pattern of MEG3 in laryngeal cancer tissues, 50 pairs of laryngeal cancer samples were analysed by RT-qPCR. The result showed that MEG3 expression was significantly down-regulated in laryngeal cancer tissues compared with adjacent normal tissues. Specifically, MEG3 expression was decreased in 74% (37 out of 50) detected sample ( Figure 1A,B). Moreover, we analysed the association between MEG3 expression and clinical features. The findings indicated that the low MEG3 expression was associated with advanced clinical stage of laryngeal cancer patients ( Figure 1C).
However, there was no statistical difference in the MEG3 expression between lymph node metastasis group and non-lymph node metastasis group ( Figure 1D) as well as between well differentiation group and moderate/poor differentiation group ( Figure 1E).

| MEG3 regulated the proliferation and apoptosis of laryngeal cancer cells
To

| MEG3 regulated the growth, proliferation and apoptosis of laryngeal cancer cells in vivo
We further investigated the effect of MEG3 on the growth, proliferation and apoptosis of laryngeal cancer xenograft tumour in nude mice. The results showed that the ratio of tumour growth was decreased in pcDNA3.1-MEG3 group as evidenced by a notable decrease in tumour weight and volume ( Figure 4A). Moreover, the IHC results showed that Ki67 protein expression, a cellular marker for proliferation, was also decreased in pcDNA3.1-MEG3 group ( Figure 4B).
Meanwhile, the expression of cleaved caspase-3, a marker for apoptosis, was increased in pcDNA3.1-MEG3 group measured by IHC and Western blotting ( Figure 4B,C). Taken together, these results suggest that MEG3 inhibits the growth and proliferation and induces the apoptosis of laryngeal cancer cells in vivo.

| Reciprocal repression of miR-23a and MEG3
To confirm whether miR-23a binds to MEG3 specifically, a luciferase reporter assay was employed. The result of luciferase reporter assay showed that cotransfection with miR-23a mimic and MEG3 luciferase reporter significantly decreased the luciferase activity in Hep-2 cells as compared with cotransfection with mimic NC and MEG3 luciferase reporter; meanwhile, the luciferase activity did not decrease by cotransfection with miR-23a mimic and MEG3 mutant luciferase reporter ( Figure 5B). This result provided a direct evidence that miR-23a binds to MEG3 specifically. Moreover, we found that overexpression of MEG3 reduced miR-23a expression in Hep-2 and AMC-HN-8 cells ( Figure 5C). Conversely, overexpression of miR-23a could also decrease MEG3 expression in Hep-2 and AMC-HN-8 cells ( Figure 5D). In addition, MEG3 expression negatively correlated with miR-23a level in laryngeal cancer samples ( Figure 5E). All these data demonstrate that a reciprocal negative regulation exists between miR-23a and MEG3.

| MEG3 regulated APAF-1 expression via miR-23a
As APAF-1 is a target of miR-23a 16 and miR-23a has be demonstrated to bind to MEG3 specifically through luciferase reporter assay, the present study have been suggested that MEG3 regulated APAF-1 expression via sponging miR-23a. As expected, overexpression of MEG3 increased the mRNA ( Figure 6A) and protein expression of APAF-1, and subsequently led to an increase in the protein level of cleaved caspase-9 and cleaved caspase-3 in Hep-2 ( Figure 6B) and AMC-HN-8 cells ( Figure 6C). However, the activation of APAF-1, cleaved caspase-9 and cleaved caspase-3 by MEG3 was reversed by miR-23a overexpression (cotransfection with MEG3 and miR-23a mimic). These data suggest that the regulation of APAF-1 by MEG3 is mediated by miR-23a.

| D ISCUSS I ON
Maternally expressed gene 3 is an imprinted gene commonly expressed in several normal tissues, such as pituitary and cerebellum. 11 However, it is usually down-regulated in non-small-cell lung cancer, 17 breast cancer, 18 hepatocellular carcinoma, 19 gastric cancer, 14 cervical cancer, 20 oesophageal cancer, 21 prostate cancer, 22 nasopharyngeal carcinoma 23 and gliomas, 24 even lost in pituitary tumours, meningiomas and myelomas. 25 The low expression of MEG3 The effects of MEG3 on the growth of laryngeal cancer cells in vivo. Hep-2 cells transfected with empty vector or MEG3 plasmid were injected into nude mice (n = 4/group). A, Tumour volumes were measured every 3 d after injection. At 15 d after injection, xenograft tumours were harvested and weighed. B, Immunohistochemistry assay was performed to assess the protein levels of Ki67 and cleaved caspase-3. Magnification: ×400, scale bars: 50 μm. C, Western blotting was employed to measure protein level of cleaved caspase-3 in xenograft tumour tissues. The data were obtained from three independent experiments and presented as mean ± SEM, ** P < .01 in cancer cells and tissues is likely to be related to aberrant promoter hypermethylation. 21,25 Like most of previous researches, we found that the expression of MEG3 was generally down-regulated in laryngeal cancer tissues. Moreover, the decreased expression of MEG3 was associated with advanced clinical stage, which was similar to Zhuo et al finding in breast cancer 19 and Tian et al finding in osteosarcoma. 26 Taken together, our findings suggest that MEG3 may serve as a novel biomarker for the progress of laryngeal cancer.
Maternally expressed gene 3 is initially found to be a growth suppressor in human cancer cells including HeLa, MCF-7 and H4. 11 Growing evidence has also demonstrated that MEG3 characterizes as a tumour suppressor in different types of cancer as evidence that overexpression of MEG3 could suppress the proliferation and induce apoptosis of cancer cells. 14,17-25 However, little is known about the role of MEG3 in laryngeal cancer up to date. In the present study, we found that overexpression of MEG3 inhibited the proliferation and induced the apoptosis in human laryngeal cancer cells and the similar evidence was also provided in vivo. These results were consistent with previous findings in non-small-cell lung cancer 17 and prostate cancer, 22 and further suggest that MEG3 may serve as a novel tumour suppressive LncRNA to regulate laryngeal carcinogenesis. Recent studies have illuminated that LncRNAs can function as a kind of ceRNA to interact with miRNAs and further regulate the expression of target mRNAs through sharing miRNA response elements (MREs). 27 Thus, LncRNA-miRNA-mRNA forms a novel regulatory network at post-transcription level. 28  Apoptotic protease activating factor-1, a tumour suppressor, is a major target of miR-23a verified in various types of cancer, including laryngeal cancer. 16,32,33 In this study, we found that overexpression of MEG3 remarkably increased APAF-1 expression. Several studies have demonstrated that APAF-1 is the adaptor molecule which in the presence of cytosolic cytochrome c and dATP interacts with pro-caspase-9, resulting in the sequential cleavage and activity of caspase-9 and caspase-3, followed by apoptosis. 34, 35 Perkins C et al have further demonstrated that overexpression of APAF-1 promoted the oligomerization of pro-caspase-9, thereby enhancing the cleavage and activation of caspase-9 and caspase-3. 36 Our data also support this by showing that the levels of cleaved caspase-9 and caspase-3 were enhanced along with an increasing in APAF-1 expression by MEG3 overexpression. These results imply that MEG3 can activate mitochondrial apoptosis-related pathway, and at least in part explain why MEG3 can induce apoptosis and inhibit the growth and proliferation of laryngeal cancer cells. Meanwhile, we found that the activations of APAF-1 and its downstream caspases induced by MEG3 were rescued by overexpression of miR-23a, which implicated that the regulation of APAF-1 by MEG3 is mediated by miR-23a.
Taken together, these results support the conclusion that MEG3, miR-23a and APAF-1 form a ceRNA regulatory network in the progression of laryngeal cancer.
However, we must acknowledge that there existed two limitations in this study. First, although we have already demonstrated that the activations of APAF-1 and its downstream caspases induced by MEG3 were rescued by miR-23a, it is still not known whether the phenotype in laryngeal carcinoma induced by MEG3 was rescued by miR-23a. Thus, more studies should be conducted to confirm what extent the phenotype in laryngeal cancer induced by MEG3 was rescued by miR-23a. Second, besides APAF-1, miR-23a is found to target many apoptosis-related genes, such as PDCD4, 37 BCL2 38 and Fas. 39 Therefore, there exists the possibility that the apoptosis rescued by miR-23a overexpression may attribute to the regulation of these apoptosis-related genes. However, there is not yet enough evidence in our study to confirm the possibility and we must acknowledge that this is also a limitation of our present study.
In conclusion, the present study demonstrates that MEG3 acts as a novel tumour suppressive LncRNA in laryngeal cancer for the first time. Furthermore, MEG3 acts as a ceRNA to regulate APAF-1 expression via competitively binding to miR-23a, thereby regulating the progression of laryngeal cancer. Undoubtedly, identifying the role and mechanism of MEG3 in laryngeal cancer will bring a novel insight into the progression of laryngeal cancer. Ka-Fai To for kindly giving us MEG3 expression plasmid and empty vectors.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest.