Thirty pairs of human glioma and adjacent normal tissues were obtained from patients who underwent surgical resection in Yulin Hospital of Traditional Chinese Medicine. The specimens were preserved in liquid nitrogen after removal and stored at −80°C until RNA extraction. The study was performed in accordance with the Helsinki declaration and was approved by the Human Ethics Committee/Institutional Review Board of Yulin Hospital of Traditional Chinese Medicine.

The human glioma cells U-251 and M059J were purchased from American Type Culture Collection (Manassas, VA, USA). The normal astrocyte cells were preserved in clinical laboratory of Yulin Hospital of Traditional Chinese Medicine and were primarily isolated from neuronal tissues of mouse embryos as previously described with appropriate modification (19,20). Cell culture bottles used here were pre-coated with 50 ng/1 ml Poly-L-lysine hydrobromide (Sigma) overnight in a sterile environment, aiming to improve the adhesion and accelerate the proliferation of cells (21,22). The SPF athymic nude mice used here were provided by the experimental animal center of the Yulin Hospital of Traditional Chinese Medicine. All the cell lines were maintained routinely in RPMI-1640 media (Gibco, cat. no. 11875-093) supplemented with 10% fetal bovine serum (Life Technologies, Inc., Grand Island, NY, USA). All cells were grown at 37°C in a humidified 5% CO₂ atmosphere.

Recombinant lentiviral vector carrying LncR-MEG3 or MEG3-shRNA were constructed according to previous studies (23). U-251 cells were transfected with recombinant lentiviral vector or an empty lentiviral vector control and then were selected according to the protocol. Up- or downregulation of MEG3 was detected through qRT-PCR. The miR-93 mimic or inhibitor and corresponding negative control were designed and synthesized by GeneChem (Shanghai, China). These mimics, inhibitor and negative control (NC) were transfected into U-251 cells using Lipofectamine 2000 (Invitrogen, Burlington, ON, Canada) according to the manufacturer's instructions. Cells were harvested 48 h after transfection for further experiments.

Total RNA from glioma tissues and cell lines was harvested using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. Total RNA was eluted with RNase-free water and stored at −80°C. RNAs were reversed and transcribed into cDNAs using the RT-PCR kit purchased from Takara according to the manufacturer's protocol. SYBR Premix Ex Taq (Takara) was used to detect the expression of MEG3 and miR-93 according to the manufacturer's protocol. The RT-PCR primers for MEG3 and miR-93 were purchased from GeneCopoeia (San Diego, CA, USA). The specific primers were as follows: MEG3 (forward: 5′-CCTGCTGCCCATCTACACCTC-3′; reverse: 5′-CCTCTTCATCCTTTGCCATCCTGG-3′); miR-93 (forward: 5′-AGGCCCAAAGTGCTGTTCGT-3′; reverse: 5′-GTGCAGGGTCCGAGGT-3′). GAPDH and U6 snRNA were used as the internal controls of the mRNA or miRNA, respectively. Fold change of MEG3 or miR-93 was calculated by the equation 2^−∆∆Ct (24).

Cell proliferation was assayed using the Cell Counting Kit-8 (CCK-8, Dojindo Laboratories, Tokyo, Japan) according to the manufacturer's protocol. Firstly, U-251 cells were pretreated with lncRNA-MEG3 or/and miR-93 mimic or mimic control or left untreated. After 2 days, a total of approximately 5×10³ transfected cells were seeded onto 96-well plates incubated for 1, 2, 3, 4, 5 days. Cells were then incubated with CCK-8 solution for 2 h at 37°C. The absorbance was measured at 450 nm using multifunctional microplate reader SpectraMax M5 (Molecular Devises, Sunnyvale, CA, USA) at indicated time points. All experiments were repeated at least three times. The cell proliferation trends were depicted according to the absorbance at each time point.

After treated as previously described, cells were double-labeled with Annexin V-fluorescein isothiocyanate (FITC) and PI apoptosis detection kits (Annexin V-FITC Apoptosis Detection kit, eBioscience) according to the manufacturer's protocol, and were analyzed using a BD FACSCalibur flow cytometer (BD Biosciences, San Jose, CA, USA) equipped with cell quest software (BD Biosciences). The apoptosis rate was evaluated for further analysis. The experiments were performed in triplicate.

Total protein was extracted from related glioma tissue/cells and 20 µg of isolated protein was separated by SDS-PAGE and transferred onto a PVDF membrane (Millipore, Billerica, MA, USA). The membranes were blocked in PBS with 0.1% Tween-20 containing 5% non-fat milk for 2 h at room temperature, and then were incubated with the primary antibodies: anti-Ki67, anti-PCNA, anti-caspase-3, anti-caspase-9, anti-P13K, anti-p-P13K, anti-AKT, anti-p-AKT, anti-GAPDH (Abcam, Cambridge, UK) and the corresponding HRP-conjugated secondary antibodies, followed by detection and visualization using a ChemiDoc XRS imaging system and analysis software (Bio-Rad, San Francisco, CA, USA). U6 and GAPDH (Abcam) were used as endogenous references.

Northern blot analysis was performed as previously described (25). The expression levels of miR-93 in gliomas samples, adjacent normal tissues, gliomas cell lines, and normal astrocyte cell line were determined by northern blot assay.

The Luc-MEG3-WT and Luc-MRG3-MUT were constructed as follows. The 3′-UTR of the MEG3 gene, which contains two putative miR-93 targeting sites, was amplified by chemical synthesis and inserted into the luciferase reporter vector (pGL4.74). U-251 cells were seeded onto 6-well culture plates in DMEM medium containing 10% fetal bovine serum and incubated overnight. Cells were co-transfected with 0.1 µg Luc-MEG3-WT or Luc-MRG3-MUT, together with 40 nM miR-93 mimic or 40 nM negative control for 24 h. Luciferase activity assays were detected by a dual-luciferase reporter system according to the manufacturer (Promega, E2920).

Formalin fixed paraffin-embedded gliomas were cut with a microtome into 5-µm paraffin sections. Antigen retrieval was carried out in heated 10 mM citrate buffer of pH 6.0 for 10 min at 96-98°C. Slides were incubated with primary antibodies against Ki-67, PCNA and AKT (Boster Bioengineering, Wuhan, China). Corresponding mouse horseradish peroxidase (HRP)-conjugated secondary antibody was added for 1 h at room temperature. Cells were counterstained with 10 mg/ml DAPI. Sections were subsequently incubated with the cell and tissue staining kit HRP-DAB system (R&D Systems, Minneapolis, MN, USA), according to the manufacturer's instructions.

Specific pathogen-free (SPF) athymic nude mice (male, six to eight weeks of age) were housed and manipulated according to the protocols approved by the experimental animal center of the Yulin Hospital of Traditional Chinese Medicine. For investigating tumorigenicity of MEG3 in vivo, xenograft mouse model was created by subcutaneous injection of 1×10⁷ U-251 cells transfected with lncRNA-MEG3 or control fragment to SPF nude mice. After the development of a palpable tumor, the tumor volume was monitored every 6 days and assessed by measuring the 2 perpendicular dimensions using a caliper and the formula (a × b²)/2, where a is the larger and b is the smaller dimension of the tumor. At 30 days after inoculation, the mice were sacrificed and tumor weights were assessed. Tumors from each mouse were randomly selected for immunohistochemical (IHC) analysis. All the animal experiments were performed according to relevant national and international guidelines and were approved by the animal experimental ethics committee.

The significance of differences between 2 groups was estimated using the Student's t-test. Data are shown as mean ± SD of at least three independent experiments performed in triplicate. All of the P-values were 2-sided and differences were considered statistically significant at P<0.05.

To determine whether MEG3 was associated with the development of glioma, the expression of MEG3 in glioma tissues and cell lines was firstly evaluated through qPCR. As shown in Fig. 1A, the level of MEG3 in tumor tissues was obviously lower than that in normal tissues (P<0.01). Besides, the expression of MEG3 in corresponding cell lines (U-251/MO59J) was detected by qRT-PCR. The level of MEG3 in normal glioma cell lines was significantly decreased compared with normal human astrocyte cell line (P<0.05, P<0.01, Fig. 1B). These results suggest that MEG3 level is reduced in glioma.

To elucidate the role of MEG3 in regulating the proliferation of glioma cells, two representative glioma cell lines U-251 were transfected with lncRNA-MEG3 or control fragment. Expression of MEG3 was significantly increased in U-251 cells treated with lncRNA-MEG3 (P<0.001, Fig. 2A). Then, upregulated MEG3 largely suppressed the proliferation of U-251 cells (P<0.05, Fig. 2B and C). To convince the results, the expression of proliferation marker proteins Ki67 and PCNA was evaluated through western blotting. Results showed that the level of Ki67 and PCNA was significantly suppressed in U-251 cells transfected with lncRNA-MEG3 (Fig. 2D). Flow cytometric analysis indicated that apoptosis rate was obviously elevated with the overexpression of MEG3 (P<0.001, Fig. 2E and F). Similarly, the expression of apoptosis-related proteins (caspase-3 and caspase-9) was enhanced under the treatment of lncRNA-MEG3. These results indicate that overexpressed MEG3 suppresses cell growth and induces cell apoptosis in glioma.

Having known that miRNA always works by binding with the 3′ untranslated region (3′UTR) of corresponding mRNAs. To explore the relationship between lncRNA-MEG3 and miR-93, complementary sites of miR-93 in MEG3 RNA were firstly predicted through bioinformatics analysis (Fig. 3A). The significantly increased expression of miR-93 was found in tumor tissues compared with normal tissues (P<0.01, Fig. 3B). The result was further convinced through northern blotting in tumor tissues and normal tissues (Fig. 3C). Besides that, the expression of miR-93 was found upregulated in glioma cell lines (U-251/MO591) compared with human common astrocyte through qRT-RCR and northern blotting (P<0.05, P<0.01, Fig. 3D and E). The increased level of miR-93 was decreased by adding lncRNA-MEG3 into U-251 cells transfected with miR-93 mimic. Similarly, the downregulated level of miR-93 was increased by adding MEG3-shRNA into U-251 cells transfected with miR-93 inhibitor (P<0.01, P<0.001, Fig. 3F and G). The targeting relationship between MEG3 and miR-93 was further identified through Luciferase activity assay. Luciferase reporter assays showed that relative luciferase activity was obviously decreased by overexpressed miR-93 in U-251 cells transfected with MEG3 WT (P<0.001, Fig. 3H). The negative correlation between MEG3 and miR-93 was further detected through relative expression analysis from 30 gliomas samples (Fig. 3I). All the results above illustrate the fact that miR-93 is a direct target of MEG3.

miR-93 is found effective in various cancers, but few studies clarified how it works in glioma. Cell proliferation of U-251 cells was largely increased in miR-93 mimic group detected through CCK8 assay (P<0.01, Fig. 4A). Besides that, the expression of proliferation marker proteins Ki67 and PCNA was raised in U-251 cells treated with miR-93 mimic (Fig. 4B). Flow cytometric analysis demonstrated that cell apoptosis rate was significantly suppressed by miR-93 mimic (P<0.01, Fig. 4C and D). Western blot assay further proved this opinion. The expression level of apoptosis-related proteins (caspase-3 and caspase-9) was largely reduced in U-251 cells treated with miR-93 mimic (Fig. 4E). Integrated these results, we concluded that miR-93 promoted the growth of glioma cells.

To further explore the signal pathway underlying the disincentive role in the proliferation of glioma, U-251 cells were transfected with lncRNA-MEG3 and/or miR-93 mimic or mimic control. The expression of P13K/AKT and phosphorylated p-P13K/p-AKT was detected through western blot. The result exhibited that PI3K/AKT pathway was activated by miR-93 mimic but was restrained by lncRNA-MEG3 (Fig. 5A and B). The level of p-P13K and p-AKT was strongly decreased by adding lncRNA-MEG3 into U-251 cells transfected with miR-93 mimic (P<0.001, Fig. 5C and D). Moreover, the subcellular localization of AKT was measured. Compared with the mimic control group, the cytomembrane staining of AKT was reduced in lncRNA-MEG3 group but was promoted in miR-93 mimic group. Then, increased cytoplasmic and cytomembrane staining of AKT was decreased by co-transfecting MEG3 in U-251 cells pretreated with miR-93 mimic (Fig. 5E). The results above suggest that MEG3 restrains the activation of PI3K/AKT pathway by reducing cytomembrane translocation of AKT.

Having identified the role of MEG3 in glioma proliferation in vitro, further research was conducted to investigate the effects of MEG3 in vivo. Glioma xenograft mouse model was created by subcutaneous injection of U-251 cells pretreated with lncRNA-MEG3 or control fragment to SPF nude mice. Tumor formation and tumor volume were effectively suppressed by MEG3 compared with the control group (Fig. 6A and B). Upregulated expression of MEG3 was detected in MEG3 model mice (P<0.001, Fig. 6C). The expression of proliferation marker proteins Ki67 and PCNA was cut down in MEG3 model mice (Fig. 6D). Relative expression of miR-93 was suppressed by MEG3 detected through qRT-PCR and northern blotting, respectively (P<0.001, Fig. 6E). The expression of P13K and AKT was elevated but the level of p-P13K and p-AKT was reduced in MEG3 model mice (Fig. 6F). Results above indicate that MEG3 suppresses tumor growth in vivo.