Long non-coding RNA HCG11 modulates glioma progression through cooperating with miR-496/CPEB3 axis.

OBJECTIVES: It has been widely reported that long non-coding RNAs (lncRNAs) can participate in multiple biological processes of human cancers. lncRNA HLA complex group 11 (HCG11) has been reported in human cancers as a tumour suppressor. This study focused on investigating the function and mechanism of HCG11 in glioma.
MATERIALS AND METHODS: Based on The Cancer Genome Atlas (TCGA) data set and qRT-PCR analysis, the expression pattern of HCG11 was identified in glioma samples. The mechanism associated with HCG11 downregulation was determined by mechanism experiments. Gain-of-function assays were conducted for the identification of HCG11 function in glioma progression. Mechanism investigation based on the luciferase reporter assay, RIP assay and pull-down assay was used to explore the downstream molecular mechanism of HCG11. The role of molecular pathway in the progression of glioma was analysed in accordance with the rescue assays.
RESULTS: HCG11 was expressed at low level in glioma samples compared with normal samples. FOXP1 could bind with HCG11 and transcriptionally inactivated HCG11. Overexpression of HCG11 efficiently suppressed cell proliferation, induced cell cycle arrest and promoted cell apoptosis. HCG11 was predominantly enriched in the cytoplasm of glioma cells and acted as a competing endogenous RNAs (ceRNAs) by sponging micro-496 to upregulate cytoplasmic polyadenylation element binding protein 3 (CPEB3). CEPB3 and miR-496 involved in HCG11-mediated glioma progression.
CONCLUSIONS: HCG11 inhibited glioma progression by regulating miR-496/CPEB3 axis.

INTRODUCTION

Glioma is known as one of the commonest malignant brain tumours.1, 2 So far, surgery, chemotherapy and radiotherapy are three main therapeutic methods for glioma. Glioma has become one of the leading causes of cancer‐related death.3, 4, 5 Therefore, exploring the molecular mechanism involved in tumorigenesis and progression is essential to the treatment of glioma.

Long non‐coding RNAs (lncRNAs) are acknowledged as a group of non‐coding RNAs (ncRNAs), which are longer than 200 nucleotides. lncRNAs can regulate gene expression in various manners, such as chromatin modification, mRNA splicing and post‐transcriptional modulation.6, 7, 8 Increasing studies have demonstrated that lncRNAs can act as either tumour suppressors or tumour promoters in the occurrence and development of cancers,9 including glioma.10, 11, 12 According to previous studies, lncRNA HLA complex group 11 (HCG11) was associated with the prognosis of patients with prostate cancer.13 However, the role of HCG11 in glioma is unclear. In our present study, we found that HCG11 was downregulated in glioma samples of The Cancer Genome Atlas (TCGA) database. Moreover, the expression level of HCG11 was lower in glioma tissues and cell lines. The association of HCG11 with the overall survival of glioma patients was analysed by Kaplan‐Meier method. In vitro and in vivo experiments were conducted to demonstrate the function of HCG11 in glioma cell proliferation, apoptosis and cell cycle progression. Mechanistically, lncRNAs can exert function in human cancers by acting as competing endogenous RNAs (ceRNA) to regulate microRNAs‐mRNAs axis.14, 15, 16 Mechanism investigation was conducted to demonstrate whether HCG11 exerted function in glioma in the same manner. Finally, rescue assays were conducted to demonstrate the ceRNA pathway.

MATERIALS AND METHODS

Clinical specimens

All glioma specimens and the non‐tumorous tissues used in this study were acquired and collected from glioma patients who received the surgical resections in The First Affiliated Hospital of Wenzhou Medical University. Informed consent had been signed by all patients. This study had received approval from the research ethics committee of The First Affiliated Hospital of Wenzhou Medical University. All specimens were immediately snap‐frozen in liquid nitrogen as soon as they were collected. Then, the specimens were preserved at −80°C until use. Based on the World Health Organization (WHO) Pathological Grading Standard (2016 version), glioma was admittedly classified into four grades (I‐IV). Patients enrolled in this study were all in WHO grade I (n = 51) and grade II (n = 37).

Cell culture

All cells (two normal human astrocytes and five glioma cells) used in this study were bought from the Institute of Biochemistry and Cell Biology of the Chinese Academy of Sciences (Shanghai, China). Cells were cultured and preserved in DMEM (GIBCO‐BRL), which was mixed with 10% FBS, 100 U/mL penicillin and 100 mg/mL streptomycin in a moist air at 37°C with 5% CO2.

Glioma primary cell culture

Glioma primary culture was conducted in a human solid biopsy of patient with grade II glioma who was diagnosed at The First Affiliated Hospital of Wenzhou Medical University. To remove the adhering blood and visible necrotic portions, the fresh tumour biopsies were carefully washed. Then, the samples were sliced into small pieces (1 mm3) and washed twice with DMEM serum‐free solution. Next, the tissue specimens were incubated with 0.125% trypsin and 0.125% EDTA (pH 7.4). The ratio between the weight of glioma tissue and trypsin was 1 g/10 mL. Digestion was conducted at 37°C for about 20 minutes in a water bath via gentle stirring. The primary glioma cell was obtained by centrifugation and grown in adherent and neurosphere conditions. For adherence, cells were plated in a tissue culture flask (75 cm2), suspended in DMEM with 10% FBS. Cells were incubated at 37°C with 5% CO2.

Transfection

To overexpress or knock down of miR‐496, miR‐496 mimics, miR‐496 inhibitors and corresponding negative controls were purchased from GenePharma (Shanghai, China). In order to construct pcDNA3.1‐lncRNA HCG11 vector, the whole sequence of HCG11 was synthesized and subcloned into a pcDNA3.1 (+) vector (GenePharma, Shanghai, China). CPEB3 was silenced by using siRNA that specially targeted to CPEB3 (si‐CPEB3#1, si‐CPEB3#2, si‐CPEB3#3). miR‐496 mimics or inhibitors were transfected at a final concentration of 40 nmol/L. However, the plasmids were transfected at concentration of 2.5 μg/well in a 6‐well plate. All transfections were finished by using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). The transfection efficiency was assessed by qRT‐PCR analysis.

qRT‐PCR analysis

TRIzol solution (Invitrogen) was used to extract or isolate total RNA form glioma tissues or cells. Reverse transcription was finished by using PrimeScript™ RT Master Mix (TaKaRa, Dalian, China). To detect the expression level of miRNA, the reverse transcription of miRNAs was carried out with TIANScript M‐MLV (Tiangen, Beijing, China). qRT‐PCR was conducted on a LightCycler 480 instrument (Roche, Basel, Switzerland) using SYBR Premix Ex Taq II (TaKaRa). The conditions for thermal cycle were shown as follows: 95°C for 30 seconds followed by 40 cycles at 95°C for 5 seconds and at 60°C for 30 seconds. The specific primers of genes were bought from Invitrogen. GAPDH and U6 were separately used as the internal control. The relative expression levels of genes were calculated by using the 2−ΔΔCt method.

Cell viability and colony formation assay

Cell viability was evaluated and measured by using MTT Reagent Kit I (Roche Applied Science, Basel, Switzerland). The U87 and U251 cells were treated with pcDNA/HCG11 or pcDNA‐NC at a density of 3000 cells per well. Next, the transfected cells were cultured in 96‐well plates. The relative cell viability was detected and evaluated in accordance with the manufacturer's recommendations. To conduct colony formation assay, cells (totally 500) were put in a six‐well plate and preserved in a medium containing 10% FBS. The medium was replaced every 5 days. Two weeks later, the cells were fixed with methanol and stained with 0.1% crystal violet (Sigma, San Francisco, CA, USA). The visible colonies were calculated by our own.

EdU assay

After transfection of pcDNA/HCG11 and pcDNA‐NC for 48 hours, the proliferation abilities of U87 and U251 cells were detected by EdU cell proliferation (Ribo, Guangzhou, China). Briefly, the ClickREdU solution (Invitrogen) was added to the culture medium (1000:1). The cells which were in proliferating phase were tabbed with EdU for 2 hours. Next, the cells were washed for three times with 0.5 g/mL of PBS. Subsequently, DAPI (Invitrogen) nuclei counterstained the washed cells for 10 minutes at normal temperature in the dark place. The DAPI‐marked cells were then washed more than twice with PBS. Finally, all marked cells were analysed and measured by the flow cytometer FACSCalibur DxP (BD Biosciences, Franklin Lakes, NJ, USA).

Flow cytometry

U87 and U251 cells transfected with pcDNA/HCG11 or pcDNA‐NC were harvested and collected after 48 hours for apoptosis analysis. Based on the manufacturer's recommendations, the cells were then treated with fluorescein isothiocyanate (FITC)‐annexin V and propidium iodide (PI) in dark place at normal temperature. Subsequently, the cells were analysed and identified by FACScan® at different stages.

U87 and U251 cells were transfected with necessary plasmids and harvested after 48 hours for detection of cell cycle. The cells were then stained with PI by using a Cycletest™ Plus DNA Reagent Kit (BD Biosciences). Next, they were subjected to analysis of a flow cytometer (FACScan®; BD Biosciences) and Cell Quest software (BD Biosciences). The percentages of cells in different phases were carefully calculated and compared.

In vivo tumorigenesis assay

Male BALB/C nude mice, aged 6 weeks, were commercially obtained from Shanghai SIPPR‐BK Laboratory Animal Co., Ltd. (Shanghai, China) and maintained in an SPF‐grade pathogen‐free animal laboratory. The procedures of animal experiment were approved by the Animal Research Ethics Committee of the First Affiliated Hospital of Wenzhou Medical University. Mice were randomly divided into two groups (n = 3 each group). 5 × 106 primary glioma cells transfected with pcDNA/HCG11 or NC were collected and subcutaneously injected into the left side of the mice's neck. Tumour volumes were measured every 4 days. Four weeks later, mice were killed. Tumours were excised and weighed for subsequent analysis.

Subcellular fractionation assay

RNA located in cytoplasm or nucleus was isolated by using Cytoplasmic and Nuclear RNA Purification Kit (Norgen, Thorold, ON, Canada). In short, U87 and U251 cells were harvested and lysed on ice for 5 minutes. The cells were then centrifuged at a high speed (12 000 g, 3 minutes). The cytoplasmic RNA was measured by collecting the supernatant, whereas the nuclear RNA was extracted by using the nuclear pellet.

RIP assay

Based on the experimental protocol, Thermo Fisher RIP Kit (Thermo Fisher Scientific, Waltham, Massachusetts, USA) was utilized for performing RNA immunoprecipitation. Ago2 antibodies were obtained from Abcam (Cambridge, UK). Normal mouse IgG (Abcam) was regarded as negative control for the RIP procedure. qRT‐PCR analysis examined purified RNA to verify the existence of the binding targets.

Dual‐luciferase reporter assay

To demonstrate the binding relationship between HCG11 and miR‐496, the wild‐type sequence (HCG11‐WT) and the mutant sequence of HCG11 (HCG11‐MUT) were separately inserted into the pmirGLO vector (Sangon Biotech, Shanghai, China). Cells were co‐transfected with miR‐496 mimics or miR‐NC and by using Lipofectamine 2000 (Invitrogen). The relative luciferase activities were measured with the Dual‐Luciferase Reporter Assay Kit (Promega, WI, USA) at 48 hours post‐transfection.

Chromatin immunoprecipitation (ChIP) assay

ChIP assay was carried out in glioma cells by using the EZ‐ChIP Kit (Millipore, Billerica, MA, USA). To generate different fragments, formaldehyde cross‐linked chromatin was sonicated. Then, anti‐FOXP1 and anti‐Ago2 antibodies were used to immunoprecipitate the chromatin fragments. IgG antibody was used as negative control. The precipitated chromatin DNA was analysed by qRT‐PCR analysis.

Western blot analysis

Cells were lysed in RIPA buffer (Beyotime Biotechnology, Shanghai, China) in which the protease inhibitor had been added. Protein concentration was measured with the BCA Protein Assay Kit (Thermo Fisher Scientific). Next, the protein extracts were segregated by 10% SDS‐PAGE (40 mg/lane) and transferred to PVDF membranes (Sangon Biotech). The membranes were blocked with 5% defatted milk followed by incubation for all night at 4°C with primary antibodies and with secondary antibody for 2 hours at normal temperature. The primary antibodies were shown as follows: anti‐CPEB3 (ab10883, 1:2000 dilution), anti—Bcl‐2 (ab32124, 1:2000 dilution), anti‐Bax (ab32503, 1:2000 dilution), anti—Cyclin‐D1 (ab134175, 1:2000 dilution), anti‐p21 (ab109520, 1:2000 dilution), anti—β‐catenin (ab32572, 1:5000 dilution), anti—c‐myc (ab32072, 1:1000 dilution) and anti‐GAPDH (ab8245, 1:3000 dilution). The secondary antibody was anti‐mouse IgG (ab193651) conjugated to horseradish peroxidase. All antibodies were bought from Abcam. The signals or blots were detected by the enhanced chemiluminescence detection reagent (Thermo Fisher Scientific, MA, USA). The protein levels were normalized to those of GAPDH.

Immunohistochemistry (IHC) and in situ hybridization (ISH)

IHC was performed by using anti‐CPEB3 antibody (Cell Signaling Technology) following the manufacturer's instructions. ISH was performed by using a miR‐496 probe from Exiqon (miRCURY LNA detection probe, 5′‐ and 3′‐digoxigenin (DIG) labelled) according to the manufacturer's instructions. The DIG antibody (Abcam), LSAB2 System HRP (Dako Denmark A/S, Glostrup, Denmark) and the liquid DAB Substrate Chromogen System (Dako) were used to detect the probe.

Statistical analysis

The spss 17.0 software was used for statistical analysis. All data were presented as mean ± SD of more than two independent experiments. Differences between two groups were analysed by Student's t test, while differences in multiple groups were tested by one‐way ANOVA with LSD post hoc test. Survival curves were generated by Kaplan‐Meier method (log‐rank test). Expression correlations among HCG11, miR‐496 and CPEB3 were analysed by Spearman's correlation analysis. Difference was considered statistically significant when P value is <0.05.

RESULTS

HCG11 was transcriptionally inactivated by FOXP1 in glioma and associated with prognosis

Based on the TCGA data, lncRNA HCG11 was significantly downregulated in glioma samples compared with that in corresponding normal tissues (Figure 1A). The expression level of HCG11 in paired glioma tissues and adjacent normal tissues collected from 88 patients was further determined by qRT‐PCR analysis. The result indicated that HCG11 was expressed at a low level in glioma tissues (Figure 1B). Consistently, HCG11 was expressed lower in glioma cells than that in normal astrocyte cells (Figure 1C). Next, 88 glioma patients were divided into two groups in accordance with the mean expression level of HCG11. Using Kaplan‐Meier method, we analysed the positive correlation between the expression of HCG11 and the overall survival of glioma patients (Figure 1D). These findings indicated the potential involvement of HCG11 in glioma progression. Furthermore, we detected the mechanism associated with the downregulation of HCG11. We found some potential upstream transcription factors of HCG11 from ucsc (http://genome.ucsc.edu/, date: 20180921), among which FOXP1 has been reported in human cancers as a transcription inhibitor.17, 18 Therefore, we investigated the effect of FOXP1 on HCG11 transcription. The motif structure of FOXP1 as a transcription factor was found from jaspar (http://jaspar.binf.ku.dk/cgi-bin/jaspar_db.pl), and we verified that the promoter region of HCG11 possessed the putative potential binding sites for FOXP1 (Figure 1E). ChIP assay demonstrated the affinity of FOXP1 to HCG11 promoter (Figure 1F). Luciferase activity analysis showed that FOXP1 transcriptionally inactivated HCG11 and decreased the luciferase activity of HCG11 promoter (Figure 1G), indicating the potential inhibitory effect of FOXP1 on HCG11 transcription.

Figure 1

HCG11 was transcriptionally inactivated by FOXP1 in glioma and associated with prognosis. A, The expression pattern of HCG11 in the TCGA glioma samples. B‐C, The expression of HCG11 in glioma tissues and cell lines was detected by qRT‐PCR. Corresponding normal tissues and cell lines were used as the negative controls. D, The motif structure of FOXP1 as a transcription factor. E, Kaplan‐Meier method and log‐rank test were used to analyse the relevance between HCG11 expression and the overall survival of glioma patients. F, The affinity of FOXP1 to HCG11 promoter was demonstrated by ChIP assay. G, The interaction between FOXP1 and HCG11 promoter was analysed by luciferase activity analysis. H, HCG11 was overexpressed in U87 and U251 cells by transfecting with pcDNA/HCG11. * P < 0.05, ** P < 0.01

Overexpression of HCG11 suppressed cell proliferation by inducing cell cycle arrest and promoting cell apoptosis

To investigate the role of HCG11 in tumorigenesis of glioma, we designed and performed gain‐of‐function assays in two glioma cells which possessed the lowest expression level of HCG11. Before the functional assays, HCG11 was overexpressed in U87 and U251 cells by transfecting with pcDNA/HCG11. The empty vector was used as negative control (NC). The overexpression efficiency was examined with qRT‐PCR (Figure 1H). According to the results of MTT, colony formation and EdU assays, overexpression of HCG11 had negative effect on the glioma cell proliferation (Figure 2A‐C). Then, we applied flow cytometry analysis to detect whether HCG11 regulated cell proliferation by affecting cell cycle and cell apoptosis. As illustrated in Figure 2D,E, HCG11 overexpression induced cell apoptosis and led to cell cycle arrest. Furthermore, we measured the levels of proteins associated with cell cycle (Cyclin D1 and p21) and cell apoptosis (Bcl‐2 and Bax). As presented in Figure 2F, overexpression of HCG11 increased the protein levels of Bax and p21, while decreased the protein levels of Bcl‐2 and Cyclin D1, further demonstrating the effects of HCG11 on cell cycle progression and apoptosis.

Figure 2

Overexpression of HCG11 suppressed cell proliferation by inducing cell cycle arrest and promoting cell apoptosis. A‐C, MTT, colony formation assay and EdU proliferation test were separately conducted in glioma cells transfected with pcDNA/HCG11 or empty vector (NC). D, E, Flow cytometry analyses were applied to examine cell apoptosis rate and cycle distribution in cells transfected with pcDNA/HCG11 or NC. F, Western blot was used to measure the levels of proteins which are closely related to apoptosis (Bcl‐2 and Bax) and cell cycle progression (Cyclin D1 and p21) in cells transfected with pcDNA/HCG11 or NC. * P < 0.05, ** P < 0.01

Overexpression of HCG11 suppressed primary glioma cell proliferation

We repeated the functional assays in primary glioma cells to verify the function of HCG11. Similarly, HCG11 was overexpressed in primary glioma cell (Figure 3A). The cell proliferation of primary glioma cell was efficiently suppressed by the overexpression of HCG11 (Figure 3B‐D). Consistently, the cell cycle was arrested at G0/G1 phase and cell apoptosis was promoted in primary glioma cell transfected with pcDNA/HCG11 (Figure 3E,F).

Figure 3

Overexpression of HCG11 suppressed primary glioma cell proliferation. A, Overexpression efficiency of HCG11 in primary glioma cell. B‐D, Proliferative ability of primary glioma cell transfected with pcDNA/HCG11 or NC. E, F, Cell cycle distribution and apoptosis rate in cells transfected with pcDNA/HCG11 or NC. * P < 0.05, ** P < 0.01

HCG11 acted as a ceRNA in glioma by sponging miR‐496

Mechanistically, lncRNAs can regulate gene expression at post‐transcriptional level by acting as a ceRNA.19, 20 Based on the experimental result of subcellular fractionation assay, HCG11 expression was enriched in the cytoplasm of glioma cells (Figure 4A). Therefore, we investigated whether HCG11 exerted ceRNA function in glioma. Subsequently, we found out 16 miRNAs which harboured the binding sites with HCG11 from starBase v2.0 (http://starbase.sysu.edu.cn/, date: 20180425). Then, we overexpressed HCG11 to detect the expression levels of all those 16 miRNAs. The results suggested that only three miRNAs (miR‐590‐3p, miR‐144‐3p and miR‐496) were obviously downregulated (Figure 4B). Furthermore, the expression levels of these three miRNAs in glioma tissues were detected by comparing with corresponding normal tissues. Among which only miR‐496 presented the significant upregulation in glioma tissues (Figure 4C). The negative expression association between miR‐496 and HCG11 in glioma tissues was analysed by Spearman's correlation analysis (Figure 4D). Furthermore, we analysed the correlation between miR‐496 expression and the overall survival of glioma patients. Similarly, all patient samples were divided into two groups in accordance with the mean expression level of miR‐496. Kaplan‐Meier analysis revealed the negative correlation between miR‐496 expression and the overall survival rate of glioma patients (Figure 4E). Moreover, patients in HCG11 high/miR‐496 low group had the highest overall survival rate, while patients in HCG11 low/miR‐496 high group had the lowest overall survival, indicating HCG11 and miR‐496 potentially synergistically affected the overall survival of glioma patients. Then, we obtained the putative binding sites between HCG11 and miR‐496 (Figure 4F). Dual‐luciferase reporter assay revealed that miR‐496 mimics efficiently decreased the luciferase activity of wild‐type HCG11 vector (HCG11‐WT) but not that of mutant‐type HCG11 vector (HCG11‐MUT; Figure 4G). According to the result of RIP assay, both HCG11 and miR‐496 were enriched in Ago2‐containing beads (Figure 4H). Therefore, we confirmed the interaction between HCG11 and miR‐496 in glioma cells.

Figure 4

HCG11 acted as a ceRNA in glioma by sponging miR‐496. A, The cytoplasmic localization of HCG11 in glioma cells was identified by subcellular fractionation assay. GAPDH (cytoplasmic control), U6 (nuclear control). B, The expression levels of 16 miRNAs were examined in response to HCG11 overexpression. C, The expression levels of three miRNAs in paired glioma tissues and corresponding normal tissues. D, The negative expression correlation between HCG11 and miR‐496 in glioma tissues was analysed. E, The correlation between the expressions of miR‐496 and HCG11 and the overall survival of glioma patients. F, The binding sequence between miR‐496 and HCG11 was displayed. G, Luciferase reporter assay was carried out to validate the interaction of HCG11 and miR‐496. H, RIP assay was performed to demonstrate the combination between HCG11 and miR‐496. * P < 0.05, ** P < 0.01

miR‐496 exerted pro‐oncogenic functions in glioma by affecting cell proliferation, cell cycle progression and cell apoptosis

We designed loss‐of‐function assays to investigate the role of miR‐496 in glioma. At first, miR‐496 was silenced in U87 and U251 cells by miR‐496 inhibitors (20 and 40 nmol/L). miR‐NC was used as the negative control. It was found that the expression level of miR‐496 was efficiently decreased at a concentration of 40 nmol/L (Figure 5A). Moreover, MTT assay was conducted to detect whether cell viability was affected by miR‐496 inhibitors in a dose‐dependent manner. Interestingly, cell viability was most efficiently suppressed when cells were treated with 40 nmol/L of inhibitors (Figure 5B). Therefore, we chose this concentration for subsequent experiments. As presented in Figure 5C,D, the decreased cell proliferation was observed in cells transfected with miR‐496 inhibitors. The role of miR‐496 in regulating cell cycle progression and apoptosis was analysed by flow cytometry analysis. As a result, downregulation of miR‐496 accelerated cell apoptosis (Figure 5E) and repressed cell cycle progression (Figure 5F). The expression levels of proteins correlated with cell cycle progression and apoptosis were measured in response to miR‐496 inhibitors. The result was consistent with that of flow cytometry analysis (Figure 5G). Moreover, we overexpressed miR‐496 in normal human astrocyte (Figure 5H) and found no significant change in the cell viability (Figure 5I). All these experimental results suggested that miR‐496 exerted pro‐oncogenic functions in glioma.

Figure 5

miR‐496 exerted pro‐oncogenic functions in glioma by affecting cell proliferation, cell cycle progression and cell apoptosis. A, The expression level of miR‐496 was decreased by miR‐496 inhibitors (20 and 40 nmol/L). B, Cell viability was measured in cells treated with miR‐NC and miR‐496 inhibitors (20 and 40 nmol/L). C, D, Colony formation assay and EdU proliferation assay were conducted to detect the impact of miR‐496 inhibitors on cell proliferation. E, F, Cell apoptosis rate and cycle progress in cells transfected with miR‐NC or miR‐496 inhibitors. G, The levels of proteins which are involved in cell cycle progression and apoptosis were tested in indicated cells. H, miR‐496 was overexpressed in normal astrocyte NHA. I, The effect of miR‐496 mimics on cell viability of NHA was detected by MTT assay. * P < 0.05, ** P < 0.01; n.s: no significance

HCG11 enhanced the expression level of CPEB3 in glioma cells by competitively binding miR‐496

Microarray analysis was conducted to explore the underlying molecular mechanism of HCG11. The results indicated that overexpression of HCG11 led to the upregulation of 197 mRNAs while led to the downregulation of 303 mRNAs (Figure 6A). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed the potential association between HCG11 and Wnt signalling pathway (Figure 6B). Among top ten upregulated mRNAs shown in Figure 6A, CPEB3 was predicted to be a potential target of miR‐496 (Figure S1). Therefore, we chose it for further study. Western blot analysis showed that overexpression of HCG11 decreased the protein levels of β‐catenin, Cyclin D1 and c‐myc (Wnt pathway factors) but increased the level of CPEB3 (Figure 6C). CPEB3 was downregulated in glioma tissues (Figure 6D), which was opposite to miR‐496 but was consistent with HCG11 (Figure 6E). The binding sequence between miR‐496 and CPEB3 was predicted and is illustrated in Figure 6F. According to the result of dual‐luciferase reporter assay, the luciferase activity of the wild‐type CPEB3 (CPEB3‐WT) but not that of mutant‐type CPEB3 (CPEB3‐MUT) was reduced by miR‐496 mimics (Figure 6G), suggesting the interaction between miR‐496 and CPEB3. Furthermore, we found that the luciferase activity of CPEB3‐WT decreased by miR‐496 mimics in HEK293T cell was partially rescued by co‐transfecting with pcDNA‐HCG11 (Figure 6H). Finally, both mRNA and protein levels of CPEB3 were increased in response to the downregulation of miR‐496 (Figure 6I), indicating the negative regulatory effect of miR‐496 on the expression of CPEB3.

Figure 6

HCG11 enhanced the expression level of CPEB3 in glioma cells by competitively binding miR‐496. A, microarray analysis was applied to screen the mRNAs which can be regulated by HCG11. B, KEGG pathway analysis revealed the signalling pathways potentially involved in HCG11‐mediated functions. C, The levels of CPEB3 and the Wnt signalling factors were tested in cells transfected with pcDNA/HCG11 or NC. D, E, CPEB3 was downregulated in glioma tissues, which was opposite with miR‐496 but consistent with HCG11. F, The binding sites of miR‐496 to CPEB3 3′UTR. G, H, Dual‐luciferase reporter assay was applied to demonstrate the combination among HCG11, miR‐496 and CPEB3. I, qRT‐PCR and Western blot analysis were utilized to detect the mRNA and protein level of CPEB3 in response to the downregulation of miR‐496. * P < 0.05, ** P < 0.01

miR‐496 and CPEB3 involved in HCG11‐mediated cell proliferation and apoptosis

Finally, rescue assays were performed in U87 cells. For rescue assays, CPEB3 was silenced by specific siRNAs (si‐CPEB3#1, si‐CPEB3#2, si‐CPEB3#3). Obviously, si‐CPEB3#1 exhibited the best knockdown efficiency (Figure 7A). Therefore, we chose it for subsequent assays. Cell proliferation assays demonstrated that decreased cell proliferation caused by HCG11 overexpression was partially rescued by the upregulation of miR‐496 or the knockdown of CPEB3 (Figure 7B‐D). Similarly, flow cytometry analysis also demonstrated that HCG11‐induced cell cycle arrest and cell apoptosis were rescued by miR‐496 mimics or si‐CPEB3#1 (Figure 7E,F). Consistently, the levels of proteins associated with cell cycle and cell apoptosis mediated by HCG11 overexpression were recovered by miR‐496 mimics or si‐CPEB3#1 (Figure 7G). Collectively, HCG11 inhibited glioma progression through modulating miR‐496/CPEB3 axis.

Figure 7

miR‐496 and CPEB3 involved in HCG11‐mediated cell proliferation and apoptosis. A, Knockdown efficiency of CPEB3 in U87 cell. B‐D, The effect of miR‐496 mimics or si‐CPEB3 on HCG11‐mediated cell proliferation was detected. E‐G, The effects of miR‐496 mimics or si‐CPEB3 on HCG11‐mediated cell cycle progression or apoptosis were analysed. * P < 0.05, ** P < 0.01

Overexpression of HCG11 inhibited glioma cell growth in vivo

In vivo experiment was conducted to verify the role of HCG11 in tumorigenesis. As shown in Figure 8A, tumours derived from HCG11‐overexpressed primary glioma cell were obviously smaller than those derived from cell transfected with empty vector. Consistently, the tumour volume and tumour weight in pcDNA/HCG11 group were smaller than those in NC group (Figure 8B,C). The expression of miR‐496 was reduced in pcDNA/HCG11 groups compared with NC groups (Figure 8D). Through immunohistochemistry assay, we found that CPEB3 expression was increased in pcDNA/HCG11 groups compared with NC groups (Figure 8E). As shown in Figure 8,F,G, we found the same results for the expression of miR‐496 and CPEB3. Based on all these results, we confirmed that HCG11 played tumour‐suppressive role in glioma.

Figure 8

Overexpression of HCG11 inhibited glioma cell growth in vivo. A, The size of tumours derived from HCG11‐overexpressed primary glioma cell or that derived from cell transfected with empty vector (NC). B, C, Tumour volume and tumour weight were measured and calculated. D, E, The expression of miR‐496 and CPEB3 in the nude mice was detected by in situ hybridization and immunohistochemistry assays. F, G, The expression of miR‐496 and CPEB3 in the nude mice was detected by qRT‐PCR and Western blot. * P < 0.05, ** P < 0.01

DISCUSSION

More and more lncRNAs have been proved to be biomarkers in glioma progression. For instance, lncRNA SNHG16 plays oncogenic role in glioma via miR‐4518/PRMT5 axis21; lncRNA PVT1 modulates the malignant behaviours of glioma cells via targeting miR‐190a‐5p/miR‐488‐3p 22; lncRNA H19 facilitates the glioma progression via regulating miR‐152.23 All these studies demonstrated the crucial role of lncRNAs in regulating gene expression and biological processes in human cancers. Our present study focused on investigating the function of lncRNA HCG11 in glioma. Downregulation of HCG11 was identified in both glioma tissues and cell lines. Moreover, we analysed the positive correlation between HCG11 expression and the overall survival of glioma patients. Functionally, we found that HCG11 overexpression efficiently suppressed cell proliferation, induced cell cycle arrest and facilitated cell apoptosis. Furthermore, we demonstrated that overexpression of HCG11 inhibited cell growth in vivo. All these experimental data suggested that HCG11 played tumour‐suppressive role in glioma.

MicroRNAs (miRNAs) are another subgroup of ncRNAs, which are shorter than 22 nt. They have been proved to be pivotal regulators in various biological processes.24, 25 Similarly, the role of miRNAs has been widely reported in glioma.26, 27, 28, 29 Mechanistically, miRNAs can interact with lncRNAs, thereby releasing the downstream targets.30, 31 In our present study, by applying bioinformatics analysis, we found 16 miRNAs potentially bind with HCG11, among which only miR‐496 was negatively regulated by HCG11 and was significantly upregulated in glioma tissues. In contrast to HCG11, miR‐496 expression had negative correlation between the overall survival rates of glioma patients. More importantly, we found that HCG11 could interact with miR‐496 in glioma cells. The function of miR‐496 in glioma cellular processes was validated by loss‐of‐function assays. Interestingly, downregulation of miR‐496 led to the consistent results with HCG11 overexpression. Therefore, we identified the pro‐oncogenic role of miR‐496 in glioma.

The potential downstream target mRNAs of HCG11 and miR‐496 were found out by using microarray analysis and bioinformatics analysis. KEGG pathway analysis and Western blotting revealed that HCG11 possibly inactivated Wnt signalling pathway in glioma. Previously, CPEB3 has been demonstrated to be a tumour suppressor in human malignant tumours.32, 33, 34 Here, we demonstrated that CPEB3 competed with HCG11 to bind with miR‐496. Finally, rescue assays were designed and conducted in U87 cells. The results demonstrated the function of HCG11‐miR‐496‐CPEB3 axis in glioma progression. In conclusion, HCG11 acts as a ceRNA to affect glioma progression via modulation of miR‐496/CPEB3 axis. This study provided a new regulatory network in glioma and might contribute to find potential therapeutic targets for glioma.

CONFLICT OF INTEREST

None.

AUTHOR CONTRIBUTIONS

All authors participated in project design. The experiments were conducted by Yangzong Chen and Chunchun Bao. All other authors contributed to the experimental procedures. Xiuxing Zhang, Xinshi Lin, Hongou Huang and Zhiqiang Wang were responsible for material collection. Yangzong Chen was responsible for article writing. All authors gave advice to this manuscript.

Click here for additional data file.