mRNA and miRNA expression microarray data from 180 samples were downloaded from the Gene Expression Omnibus website (http://www.ncbi.nlm.nih.gov/geo/; accession no. GSE4290). The data were compiled from 23 non-tumor, 26 astrocytoma (7 grade II, 19 grade III), 50 oligodendroglioma (38 grade II, 12 grade III) and 81 GBM samples. The tumor sample expression profile, including HOTAIR expression data, was also downloaded. HA1800 human astrocytes and A172 glioma cells were purchased from the Cell Resource Center of Shanghai Institute of Life Sciences (Shanghai, China). The cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% fetal calf serum (Gibco; Thermo Fisher Scientific, Inc.), penicillin (100 U/ml) and streptomycin (100 mg/ml) at 37°C in 5% CO₂.

Total RNA was extracted from the cells using TRIzol Reagent (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's instructions. Briefly, 1 ml TRIzol per 5×10⁵ cells was added to cells, prior to adding 0.2 ml of chloroform per 1 ml TRIzol. The mixture was mixed vigorously by hand and allowed to stand for 2–3 min at room temperature. The mixture was then centrifuged at 10,000 × g for 10 min at 4°C. The upper clear phase was transferred to a fresh tube and 0.5 ml isopropanol per 1 ml of the clear phase was added, which was mixed vigorously by rapid shaking and left to stand for 10 min. The precipitated RNA was collected by centrifugation at 10,000 × g for 10 min at 4°C and then carefully decanting/pipetting the supernatant. The RNA precipitate was washed once with 70% ethanol, dissolved in 25 µl RNase free water, and then stored at −80°C. The concentration of the Recombinant DNase I RNase-free (Takara Biotechnology Co., Ltd., Dalian, China) used to treat the RNA sample was 5 units/µl.

cDNA was synthesized using the HiFi-MMLV cDNA kit (Beijing ComWin Biotech Co., Ltd., Beijing, China) and qPCR was conducted using the UltraSYBR Mixture (Beijing ComWin Biotech Co., Ltd.). Briefly, 5 µg purified RNA sample was mixed with Primer Mix, dNTP Mix, DTT, RT-buffer, HiFi-MMLV and RNase-free water using a pulled pipette (total volume, 20 µl). All qRT-PCR reactions were run in a StepOnePlus™ Real-Time PCR System (Applied Biosystems; Thermo Fisher Scientific, Inc.). The mixture was then incubated at 42°C for 30–50 min and then 85°C for 5 min. The RT products were quickly centrifuged and stored at −20°C. No cDNA was used as a negative control. To amplify hsa-miR-148b-3p cDNA, specific RT primers were used based on its sequence, and the U6 RT primer for was the same as the U6 reverse PCR primer. The hsa-miR-148b-3p RT primer was 5′-GTTGGCTCTGGTGCAGGGTCCGAGGTATTCGCACCAGAGCCAACACAAAG-3′. PCR primers were as follows: Forward, 5′-GGCACCACACCTTCTACAAT-3′ and reverse, 5′-GTGGTGGTGAAGCTGTAGCC-3′ for the β-actin gene; forward, 5′-CAGTGGGGAACTCTGACTCG-3′ and reverse, 5′-GTGCCTGGTGCTCTCTTACC-3′ for the HOTAIR gene; forward, 5′-CGGTCAGTGCATCACAGAA-3′ and reverse, 5′-GTGCAGGGTCCGAGGT-3′ for hsa-miR-148b-3p; and forward, 5′-CTCGCTTCGGCAGCACA-3′ and reverse, 5′-AACGCTTCACGAATTTGCGT-3′ for U6. All primers were synthesized by GenScript (Nanjing, China). The relative fold change in mRNA expression level was calculated using the 2^−ΔΔCq method (40).

siRNA HOTAIR (siHOTAIR), siRNA negative control (siNC), miR-148b-3p mimic and miR-148b-3p inhibitor were synthesized by Biomics Biotechnologies Co., Ltd. (Nantong, China). The sequences were as follows: Sense, 5′-CCACAUGAACGCCCAGAGAUU-3′ and antisense, 5′-AAUCUCUGGGCGUUCAUGUGG-3′ for si-HOTAIR (41); and sense, 5′-UUCUCCGAACGUGUCACGUTT-3′ and antisense, 5′-ACGUGACACGUUCGGAGAATT-3′ for si-NC (22). Cells were grown tuntil they reached 10⁵ in number in 6-well plates prior to transfection. One day before transfection, the control group cells were plated at the same density as the si-HOTAIR and si-NC groups. The cells were transfected using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's instructions.

The effect of HOTAIR downregulation on the viability of A172 cells was assessed using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Briefly, cells were trypsinized and seeded at a density of 1×10⁴ cells/well in 96-well plates immediately after siRNA transfection. Then, 10 µl MTT solution (5 mg/ml) was added and the plates were incubated for an additional 4 h at 37°C. Following removal of the medium, formazan crystals were dissolved in 150 µl dimethyl sulfoxide. The absorbance of the MTT formazan was measured at 550 nm using a SpectraMax M3 microplate reader (Molecular Devices, LLC, Sunnyvale, CA, USA). Experiments were repeated three times using 6 wells for each treatment to ensure the reproducibility of results.

Cells were separately harvested 0, 24, 48 and 72 h after siRNA transfection, fixed in 70% ethanol, and stained with propidium iodide (Nanjing KeyGen Biotech Co. Ltd., Nanjing, China) containing RNase A (1 mg/ml; Takara Biotechnology Co., Ltd.) for 30 min at 37°C. Subsequently, 500 µl of cells was filtered through 200-µm mesh sieves, and the cell cycle profiles were assayed using a Guava easyCyte 8 Flow Cytometer (EMD Millipore, Billerica, MA, USA).

Transwell inserts (diameter, 6.5 mm) with a pore size of 8 µM (Corning Incorporated, Corning, NY, USA) were coated with Matrigel (100 µg/well; BD Biosciences, San Jose, CA, USA) and placed into the wells of 24-well culture plates. Following transfection with siRNA for 48 h, 1×10⁴ cells were transferred into the top of the invasion chambers in serum-free DMEM, and DMEM containing 20% fetal calf serum was added to the lower chambers. After 24 h of incubation at 37°C, non-invasive cells were removed with a cotton swab, and the invading cells were fixed with 4% paraformaldehyde for 15 min, stained with Giemsa for 20 min at room temperature and observed under an ECLIPSE Ti S microscope (Nikon, Tokyo, Japan). Experiments were independently repeated three times.

To verify whether miR-148b-3p regulates HOTAIR as a direct target, a predicted binding site (12 bp) for miR-148b-3p was identified using the online software program starBase v2.0 (42). The human HOTAIR fragment containing the putative binding sites for miR-148b-3p was synthesized, annealed and inserted into the NotI and XbaI restriction sites of the pmirGLO luciferase reporter vector (Promega, Madison, WI, USA), downstream of the luciferase gene, to generate the recombinant vectors pmirGLO-wild-type (WT) and pmirGLO-mutant (MUT). For the pmirGLO-MUT construct, 12 mismatches were introduced into the HOTAIR sequence, producing a change of GATGCATTTTCTGTGCACTGGtoGATGCACCTCCCACATGTCAG; therefore, a human HOTAIR fragment containing mutated binding sites was synthesized. The constructs were validated by Sanger sequencing by Sangon Biotech Co., Ltd. (Shanghai, China).

For the luciferase reporter assay, A172 cells were co-transfected with miRNA (miR-148b-3p mimics or miR-148-3p mimic negative control; Biomics Biotechnologies Co., Ltd., Nantong, China) and reporter vectors (pmirGLO-WT or pmirGLO-MUT) using Lipofectamine 2000. Luciferase activity was assayed 48 h after transfection using a Dual-Luciferase Reporter Assay system (Beyotime Institute of Biotechnology, Haimen, China). The values were normalized to those obtained for miRNA negative control transfection. All transfection experiments were performed in triplicate.

All statistical analyses were performed using SPSS software version 18.0 (SPSS, Inc., Chicago, IL, USA). Data are expressed as means ± standard deviation of experiments performed in triplicate. HOTAIR expression in normal tissues and tumor tissues was compared using analysis of variance (ANOVA) followed by Student-Newman-Keuls analysis. GraphPad Prism version 5.0 (GraphPad Software, La Jolla, CA, USA) was used for the graphing. Statistical analysis on other experiments was performed using one-way ANOVA and the Student-Newman-Keuls test for multiple comparisons, and unpaired Student's t-tests for comparisons between groups. P<0.05 was considered to indicate a statistically significant difference.

First, the expression pattern of HOTAIR was analyzed through whole genome gene expression profiling of 157 glioma and 23 normal tissue samples from NCBI GEO data (accession no. GSE4290). As shown in Fig. 1A, HOTAIR expression was significantly higher in carcinoma tissues compared with in normal tissues (P=0.010), and grade IV and III tumor tissues both demonstrated a significant increase in HOTAIR transcription levels compared with grade II tumor tissues (P=0.047). However, no significant difference in HOTAIR expression was observed between grade IV and III samples. Next, HOTAIR expression was detected in HA1800 human astrocytes and A172 glioma cells by RT-qPCR. The HOTAIR mRNA level was significantly increased in A172 cells compared with HA1800 cells (P<0.001; Fig. 1B). These findings indicate that HOTAIR may have an important role in glioma progression.

To further investigate the oncogenic role of HOTAIR in glioma pathogenesis, A172 cell lines were transfected with si-HOTAIR. Successful transfection was confirmed by RT-qPCR (Fig. 2A). Cells were separately collected at 0, 24, 48 and 72 h post-transfection, and cell growth ability was determined using the MTT assay. Silencing of HOTAIR expression significantly reduced growth rates compared with the si-NC and untransfected cell groups at 48 (P=0.012) and 72 h (P<0.001) post-transfection (Fig. 2B). Additionally, FCM was performed following transfection to assess cell cycle distribution. The results showed that the cell population in the G1 phase was increased but the S phase population was decreased after HOTAIR gene silencing compared with the results observed for the si-NC cells (Fig. 2C), further indicating that knockdown of HOTAIR expression may suppress cancer cell proliferation. In addition, the effect of HOTAIR knockdown on A172 cell invasion was investigated using a Matrigel invasion assay. The in vitro Matrigel invasion assay revealed that the invasiveness of A172 cells transfected with si-HOTAIR was significantly suppressed compared with the control and si-NC cells (P=0.007; Fig. 2D).

RT-qPCR revealed that the expression level of miR-148b-3p was significantly lower in A172 cells compared with HA1800 cells (P<0.001; Fig. 3A). To determine whether miR-148b-3p is able to suppress HOTAIR expression, miR-148b-3p mimic or inhibitor were transfected into A172 cells. As shown in Fig. 3B, the miR-148b-3p mimic reduced HOTAIR expression in a dose-dependent manner, decreasing it by ~62% 72 h after transfection (P<0.001). By contrast, the miR-148b-3p inhibitor increased the level of HOTAIR in a dose-dependent manner (Fig. 3C). Furthermore, si-HOTAIR transfection significantly increased the expression of miR-148b-3p in A172 cells at all time points (P<0.001; Fig. 3D), indicating that there is a strong inverse correlation between HOTAIR and miR-148b-3p expression levels.

To verify whether miR-148b-3p regulates HOTAIR as a direct target, a predicted binding site for miR-148b-3p was identified using the online software program starBase v2.0 (42) (Fig. 3E) an wild-type or mutant miR-148b-3p target binding sequences were cloned into the pmirGLO luciferase vector. Following co-transfection with the miR-148b-3p mimic in A172 cells, a Dual-Luciferase Assay was performed determine the luciferase activity. The data shows that cells co-transfected with the constructs containing the pmirGLO-WT and miR-148b-3p mimic had significantly lower luciferase activity compared with that of those transfected with pmirGLO-MUT and miR-148b-3p mimic (P=0.007; Fig. 3F). All the results indicate that miR-148b-3p suppresses HOTAIR by binding to HOTAIR in a sequence-specific manner.

To detect whether the aggressiveness of A172 cells could be restored by miR-148b-3p, an MTT assay, FCM and a Transwell invasion assay were performed following transfection with the miR-148b-3p mimic. Transient transfection with the miR-148b-3p mimic decreased the proliferation of A172 cells by ~15% at 72 h compared with the mimic negative control group (P=0.002; Fig. 4A). The miR-148b-3p mimic also increased the cell population in the G1 phase and decreased the S-phase population (Fig. 4B), demonstrating that miR-148b-3p blocks the cell cycle of A172 cells. Furthermore, the Transwell invasion assay results clearly revealed that the miR-148b-3p mimic decreased cell invasion to ~70% compared with the untreated and miR-148b-3p mimic negative controls (P=0.002; Fig. 4C).