LINC00461, a long non-coding RNA, is important for the proliferation and migration of glioma cells

An increasing number of reports have revealed that long non-coding RNAs are important players in tumorigenesis. Here we showed that long non-coding RNA LINC00461 is highly expressed in glioma tissues compared to non-neoplastic brain tissues. The knockdown of LINC00461 suppressed cyclinD1/A/E expression which led to G0/G1 cell cycle arrest and inhibited cell proliferation in glioma cells. LINC00461 suppression also inhibited glioma cell migration and invasion. The function of LINC00461 in glioma cells is partially mediated by MAPK/ERK and PI3K/AKT signaling pathways as down-regulation of LINC00461 expression suppressed ERK1/2 and AKT activities. Moreover, LINC00461 knockdown decreased expression levels of microRNA miR-9 and flanking genes MEF2C and TMEM161B. Taken together, our results demonstrate that LINC00461 is important for glioma progression affecting cell proliferation, migration and invasion via MAPK/ERK, PI3K/AKT, and possibly other signaling pathways.

Gliomas account for the majority of malignant brain tumors in adults, causing approximately 80% of all malignant brain tumors [23]. According to the Histopathology, the World Health Organization (WHO) has divided gliomas into four grades (I-IV) [24]. Gliomas are rarely curable especially for glioblastoma (grade IV) [25]. Current standard therapy for glioblastomas includes maximal safe surgical resection, followed by concurrent radiation with temozolomide (TMZ), an oral alkylating agent [26]. But the treatment is not effective due to the high proliferation rate and aggressive invasion of glioblastomas. The median survival of patients is merely 15 months [25]. Thus, it is urgent to advance our understanding of regulatory mechanisms involved in the initiation and progression of gliomas for developing novel and effective therapeutic approaches.

Research Paper
Here, we have reported that lncRNA LINC00461 is up-regulated in human gliomas. LINC00461 is located at an intergenic region of human chromosome 5 between two protein-coding genes MEF2C (myocyte enhancer factor 2C) and TMEM161B (transmembrane protein 161B). Our study suggested that LINC00461 is important for glioma cell proliferation, migration and invasion. Furthermore, we found that LINC00461 could potentially activate MAPK/ ERK and PI3K/AKT pathways and expression levels of genes in its vicinity as well.

LINC00461 is expressed in neural stem/glioma cells
Previously, we compared transcriptomes of mouse spinal cords at E13.5 (embryonic day 13.5) with those at P0 (postnatal day 0) and identified several genes that are highly expressed at E13.5, including lncRNA C130071C03Rik. Now further studies revealed that it is specifically expressed in the ventricular zone of the mouse spinal cord at E11.5 ( Figure 1A) and E13.5 ( Figure 1B), where neural stem/precursor cells are located. At P0, its expression spreads out to the whole spinal cord ( Figure  1C). In the mouse brain, we detected its expression in the subventricular zone (SVZ) at P0 (Supplementary Figure 1A, 1B). Real-time PCR analysis showed that C130071C03Rik is highly expressed in mouse neural tissues compared to non-neural tissues ( Figure 1D).
The liftOver program was used to identify single mapped orthologous regions in genomes of diverse species. We found that the ortholog of lncRNA C130071C03Rik in humans was LINC00461. LINC00461 is transcribed from an intergenic region of human chromosome 5 between MEF2C and TMEM161B ( Figure  2A). Using in situ hybridization (ISH) technique, we demonstrated that LINC00461 transcript predominantly locates in the cytoplasm of U251 and U87MG glioma cells (Supplementary Figure 1C).
It has been reported that glioblastoma cells harbor some properties similar to those of neural stem cells. Their gene expression profiles resemble each other. Therefore we decided to find out whether this lncRNA is enriched in gliomas. Glioma datasets GSE16011 and GSE4290 showed that expression levels of LINC00461 were upregulated in gliomas (grade I-IV in GSE16011, grade II-IV in GSE4290) compared to the control ( Figure 2B). We also checked the expression of SOX2, a neural stem cell marker, and found that SOX2 mRNA expression levels were also up-regulated in gliomas ( Figure 2C). In our own Chinese glioma collection, we found that expression levels of both LINC00461 and SOX2 mRNAs were markedly higher in gliomas than those in nonneoplastic brain tissues (P < 0.001) ( Figure 2D). Pearson correlation analysis revealed significant and positive correlation between LINC00461 and SOX2 mRNAs in GSE16011 and GSE4290 datasets ( Figure 2E). Again, a positive correlation between mRNA levels of LINC00461 and SOX2 was detected in Chinese glioma samples ( Figure 2E). Up-regulation of SOX2 has been linked to the development and maintenance of gliomas. Our findings suggested that LINC00461 might be involved in the development of gliomas, regulating stem-cell like properties in gliomas.

The knockdown of LINC00461 decreased cell viability of glioma cells, while had no effects on cell apoptosis
Lentivirus-mediated short hairpin RNAs (shRNAs) were applied to knockdown LINC00461 expression. 48 hours after lentivirus infection, expression levels of LINC00461 were measured by real-time PCR to determine the effect of LINC00461 shRNA. We had designed two different shRNAs. Both significantly suppressed expression levels of LINC00461 in U251 and U87MG cells ( Figure 3A, Supplementary Figure 3A) and reduced the cell viability at 2, 3, 4 and 5 days post the lentivirus treatment ( Figure 3B, Supplementary Figure 3B).

LINC00461 knockdown led to G0/G1 phase arrest and inhibited cell proliferation
To evaluate the effect of LINC00461 on cell proliferation, we used both ki67 antibody and EdU staining to measure the proliferation rate. Ki67 antibody can recognize a protein that is present during all active phases of the cell cycle, but absent in the resting stage. EdU is an analog for thymidine and can be incorporated into the newly synthesized DNA at S phase of the cell cycle. The percentage of Ki67 positive cells after www.impactjournals.com/oncotarget LINC00461 shRNA treatment was 59.50 ± 0.33% in U251 cells and 45.20 ± 0.23% in U87MG cells, while in control groups it was 79.18 ± 0.40% in U251 cells (P < 0.01) and 69.95 ± 0.26% in U87MG cells (P < 0.001) ( Figure 4A, 4B). Results were confirmed by EdU staining experiments. In LINC00461 knockdown groups, the percentage of EdU positive cells was 40.98 ± 1.68% in U251 cells and 23.23% ± 1.64% in U87MG cells, compared to 56.91 ± 2.79% (P < 0.05) in U251 cells and 38.15 ± 2.24% (P < 0.001) in U87MG cells for control groups ( Figure 4C, 4D). Our results suggested that suppressed expression of LINC00461 inhibits cell proliferation which contributes to the reduction of cell viability.
In order to explore mechanisms underlying the anti-proliferation effect of LINC00461 knockdown on glioma cells, we analyzed cell cycle distribution by flow cytometry. As showed in Figure 5, 48 hours after downregulation of LINC00461, the fraction of cells in G0/G1 phase increased to 77.86 ± 1.59% in U251 cells compared to 59.31 ± 1.07% in control cells ( Figure 5A, 5B). Similar results were obtained with U87MG cells. The fraction of cells in G0/G1 phase increased to 70.14 ± 1.13% in experimental groups, whereas the fraction in control group was 55.65±0.95% ( Figure 5A, 5B). The fraction of cells in S phase decreased significantly after LINC00461 shRNA treatment (treatment group 17.08% ± 1.55% vs. control group 30.57 ± 0.56%in U251 cells, 19.57 ± 1.52% vs. 28.79 ±2.06% in U87MG cells) ( Figure 5A, 5B).
We further measured expression levels of some proteins important for regulating the cell cycle. In LINC00461 knockdown groups, expression levels of cyclinD1, CDK4, cyclin A, cyclin E mRNAs significantly decreased in U251 and U87MG cells ( Figure 5C). Moreover, expression levels of cyclinD1 protein decreased about 50% in U251 cells (P < 0.01) and 65% in U87MG cells (P < 0.001) ( Figure 5D). These results showed that down-regulation of LINC00461 can reduce expression levels of cyclinD1, CDK4, cyclin A and cyclin E, which leads to G0/G1 phase arrest and the delay of cell cycle progression.

LINC00461 knockdown inhibited cell migration and invasion
We tried to find out whether down-regulation of LINC00461 could affect cell migration and invasion, which are key determinants for tumor aggressiveness and metastasis. Wound-healing assay was performed to investigate the effect of LINC00461 on cell migration. At the time points of 12 h and 24 h after wound scraping, slow migration in LINC00461 knockdown group was observed in both U251 and U87MG cells ( Figure 6A, 6B, Supplementary Figure 3C). Transwell assay was also used to further verify the effect of LINC00461 on migratory and invasive properties of tumor cells. After LINC00461 The expression of LINC00461 positively correlated with that of SOX2 in GSE16011, GSE4290, and CBSS. Each sample has been measured three times. Data are presented as mean ± SEM. * , P < 0.05; ** , P< 0.01; *** shRNA treatment, numbers of cells underwent migration and invasion were markedly reduced in U251 (P < 0.01 in migration assay and P < 0.05 in invasion assay) and U87MG cells (P < 0.01 in migration assay and P < 0.05 in invasion assay) ( Figure 6C, 6D). These data showed that knockdown of LINC00461 reduces migratory and invasive properties of glioma cells.

LINC00461 knockdown suppressed expression levels of markers for stem cells and EMT
Our data have showed that LINC00461 is highly expressed in neural stem cells and increases cell proliferation and migration, which are important features of stem cells. We decided to find out whether LINC00461 regulates "stemness" of glioma cells. We chose some classic stem cell markers, such as SOX2, Nestin and CD44. The result showed that LINC00461 knockdown significantly reduced mRNA levels of stem cell markers: SOX2 and CD44 in U251 cells, Nestin and CD44 in U87MG cells ( Figure 7C).
It is well known that cancer cells usually undergo EMT to gain migratory and invasive properties. So we analyzed expression levels of EMT markers as well. Realtime PCR results showed that LINC00461 knockdown suppressed mRNA levels of EMT markers, such as N-cadherin and ZEB1 ( Figure 7B), in U251 and U87MG cells.

Potential mechanisms for LINC00461 promoting glioma cell proliferation, migration and invasion
MAPK/ERK pathway is well known for its roles in oncogenesis. Phosphatidylinositol 3-kinase (PI3K)/ AKT signaling pathway also plays a major role in cell proliferation, migration and invasion, especially in glioma cells. In order to find out the mechanism underlying the regulation of LINC00461 on glioma cells, we studied effects of LINC00461 on different signaling pathways. Expression levels of total and phosphorylated AKT and ERK1/2 proteins were measured in glioma cells by Western blot technique. Results showed that LINC00461 knockdown significantly suppressed phosphorylation of ERK1/2 proteins (P < 0.001) in both U251 and U87MG cells ( Figure 7A, 7B). Moreover, down-regulation of LINC00461 also decreased expression levels of total AKT (P < 0.01) and phosphorylated AKT proteins (P < 0.001) in both U251 and U87MG cells ( Figure 7A, 7B). These results indicated that LINC00461 might promote glioma cell proliferation, migration, and invasion through MAPK/ ERK and PI3K/AKT signaling pathways.

DISCUSSION
We have identified a mouse lncRNA C130071C03Rik which is highly expressed in mouse neural tissues. Our studies showed that it is specifically expressed in mouse neural stem cells during development. Its human homolog, LINC00461, is up-regulated in human glioma tissues. We performed a series of experiments and found that LINC00461 plays multiple roles in glioma cells, affecting cell proliferation, migration, and invasion. Our results also suggested that it has little effect on cell apoptosis. The knockdown of LINC00461 caused cells being arrested during G1/S transition with concomitant decreased cyclinD1/A/E expression. We also found that LINC00461 is involved in MAPK/ERK and PI3K/AKT signaling pathways and regulates genes in its vicinity.
There are very few studies related to this lncRNA. It has been reported by Oliver et al. in 2015. They found that Visc-1 (aka. C130071C03Rik) was expressed in the cortical plate and the ventricular and subventricular zones of the telencephalon as well as the rostral and caudal interneuron migratory streams at E16.5 [27]. Their data suggested its role in cell proliferation and migration, which is consistent with our findings. LINC00461 has also been reported to be up-regulated in several types of cancers. For example, Pastori et al. have reported LINC00461 as one of the top 100 most up-regulated lncRNAs in glioblastomas [28]. Crea et al. also showed that LINC00461 is upregulated in metastatic prostate cancers [29].
In our study, we found that mouse ortholog of LINC00461 is expressed in the ventricular zone of mouse spinal cord at E11.5 and E13.5. It is also highly enriched in gliomas. Further studies revealed significant and positive correlation between LINC00461 and SOX2 mRNA in gliomas. SOX2 activity has been found to be associated with the maintenance of the undifferentiated state of cancer stem cells [30]. These results suggested that LINC00461 may regulate "stem-like" properties of neural/ cancer stem cells.
LINC00461 plays multiple functional roles in glioma cells. One of them is cell cycle regulation. The loss of the regulatory control of the cell cycle, which leads to an unrestrained cell proliferation, is a hallmark of cancer cells [31]. Our data suggested that LINC00461 knockdown could induce the G1/S transition arrest which leads to the decreased viability of glioma cells. Further, we found that suppression of LINC00461 resulted in decreased expression levels of cyclinD1/E/A and CDK4. CyclinD1 plays pivotal roles in G1/S transition and is frequently over-expressed in a large number of cancers including gliomas. Its high expression is associated with the tumor malignancy and poor prognosis. CyclinD1 binds to CDK4 and CDK6 (cyclin-dependent kinase 4/6) to phosphorylate retinoblastoma protein and activate the transcription factor E2F-1. E2F-1 initiates the transcription of some cell cycle regulators, such as cyclin E and cyclin A, to drive the transition from G1 to S phase [32].
In most epithelial carcinomas, cancer cells usually undergo the epithelial-mesenchymal transition (EMT) to gain invasiveness and start metastasis. During the EMT process, several EMT markers or inducers, including ZEB1 and SNAIL, are activated to promote the "stemness" of cancer cells, thus induce cancer invasion [33]. In our study, we found that reduced LINC00461 expression reduced expression levels of some EMT markers (N-cadherin and ZEB1) and several other factors related to glioma "stemness" (CD44, SOX2, Nestin). Therefore, LINC00461 may participate in glioma progression via promoting EMT.
Current studies have found lncRNAs play critical roles in the progression of gliomas. HOTAIR (HOX transcript antisense intergenic RNA) has been reported to be up-regulated in gliomas. Its knockdown inhibits cell proliferation, migration, invasion, promotes apoptosis and cell cycle arrest by up-regulating miR-326 in human glioma cells [34]. Wang et al. have reported that lncRNA CRNDE (colorectal neoplasia differentially expressed) is up-regulated in gliomas and promotes glioma cell growth and invasion through mTOR signaling pathway [21]. Shi et al. have found that lncRNA H19 is closely correlated with the severity of tumors. H19 transcript can produce a miRNA (miR-675), which directly regulates CDH13 (Cadherin 13), thereby modulating glioma cell invasion [35].
In this study, we have found that LINC00461 regulates miRNA miR-9, which is the mature form of miR-9-2 gene. The sequence of miR-9-2 is included in the sequence of LINC00461. Knockdown of LINC00461 significantly suppressed expression levels of miR-9. Wu et al. have reported that expression levels of miR-9 in gliomas are significantly higher than those in nonneoplastic brain tissues [36]. The overall survival time of glioma patients with high miR-9 expression levels is obviously shorter than that with low miR-9 expression levels [36]. Schraivogel et al. found that miR-9 promotes neurosphere-like formation of glioblastoma stem cells by targeting a tumor suppressor CAMTA1 (calmodulinbinding transcription activator 1) [37]. It has also been reported that miR-9 can promote glioma cell migration by repression of NF1 (neurofibromin1) [38]. Based on these studies and our own work, we concluded that LINC00461 might regulate glioma progression via miR-9.
Studies have shown that miR-9 could target multiple signaling pathways [39]. In ovarian cancer and gastric cancer, miR-9 has been shown to target NF-κB pathway leading to inhibition of cell proliferation and metastasis [40,41]. On the other hand, miR-9 overexpression has been shown to enhance metastasis in esophageal squamous cell carcinoma by targeting E-cadherin [42]. Zhang et.al have found that overexpression miR-9 could block the phosphorylation of AKT (Ser473) and ERK [42]. In our study, we found that overexpression or inhibitor of miR-9 has no effects on AKT and ERK in glioma cells (Supplementary Figure 4). Further studies are needed to decipher the mechanism of miR-9 regulation on gliomas.
Recent studies have found that long intergenic non-coding RNAs (lincRNAs) can regulate expression levels of protein-coding genes in their vicinity [43]. We found that LINC00461 is important for the expression of neighbor genes, TMEM161B and MEF2C, possibly via cis-interaction. Expression levels of both genes decreased after the treatment of LINC00461 shRNA. The function of TMEM161B is still not clear, while MEF2C has been reported to play an anti-apoptotic role, protecting the differentiating cells from death during neurogenesis [44], and promote neuronal synapse formation [45,46]. Li et al. showed that Mef2c knockout in Nestinexpressing neural stem/progenitor cells (NSCs) impaired neuronal differentiation in vivo, resulting in aberrant cell compaction and smaller somal size. However, the proliferation and survival of NSCs were not affected [47]. Still no reports have been published about effects of MEF2C on gliomas.
Previous studies have shown that several signaling pathways are involved in gliomas [48]. Among them, MAPK/ERK and PI3K/AKT are two frequently activated pathways [49,50]. ERK1/2 (Extracellular signal-regulated kinases 1 and 2) belong to the mitogen-activated protein kinase (MAPK) family and play crucial roles in a wide range of biological processes, including cell proliferation, differentiation, and motility. They have been reported to be activated in glioblastomas [51]. Meanwhile, elevated AKT phosphorylation levels have been observed in 84% cases of glioblastomas [52]. The PI3K/AKT signaling pathway plays fundamental roles in regulating cellular processes such as cell proliferation, survival, and migration [53]. Our studies found that LINC00461 positively affects both MAPK/ERK and PI3K/AKT pathways. Knockdown of LINC00461 can suppress not only phosphorylation levels of ERK1/2 and AKT proteins, but also total AKT protein levels. Dual activation of MAPK/ERK and PI3K/AKT signaling pathways has been reported in many types of cancers. In thyroid cancers, both MAPK/ERK and PI3K/ AKT pathways are reported to be activated, which play fundamental roles in promoting cell growth, survival and are associated with tumorigenesis [54,55]. Lv et al. found that CXCR4 knockdown repressed glioma cell migration via deactivation of both ERK and PI3K/AKT pathways [56]. Our laboratory previously has found that TCF3 (transcription factor 3) can activate both MAPK/ERK and PI3K/AKT signaling pathways to suppress cell apoptosis and promote cell migration [57]. These data suggested that LINC00461 might promote glioma cell proliferation, migration, and invasion via both MAPK/ERK and PI3K/ AKT signaling pathways (Figure 8).
It is worthwhile to note that cyclinD1 can be regulated by a number of signaling pathways, including PI3K/AKT and ERK1/2 pathways. PI3K/AKT pathway can promote cyclinD1 expression and positively regulates G1/S cell cycle progression through deactivation of GSK3bata [58]. In addition, persistent activation of ERK1/2 is required to pass the G1 restriction point in fibroblasts [59,60]. Our data suggested that LINC00461 can activate both MAPK/ERK and PI3K/AKT signaling pathways which lead to cyclinD1 activation and cell proliferation ( Figure  8).
In summary, our studies suggested that lncRNA LINC00461 plays important roles in gliomas via recruiting multiple pathways. Based on these studies, LINC00461 could be a potential target in glioma diagnosis and treatment.

Human tissues
Glioma gene expression data were downloaded from the Gene Expression Omnibus Web site (http:// www.ncbi.nlm.gov/geo/, GSE16011 and GSE4290).  were immediately frozen in liquid nitrogen after resection and kept in liquid nitrogen for storage. Chinese glioma specimens included 9 samples of grade II, 5 samples of grade III and 5 samples of grade IV. Five cases of nonneoplastic brain samples were collected from patients with cerebral trauma.

Cell culture
U251 and U87MG human glioma cells were maintained in alpha-modification of eagle's medium (α-MEM) or Dulbecco's Modified Eagle Medium (DMEM) with high glucose (Hyclone, USA) and supplemented with 10% fetal bovine serum (Sigma, USA). Cells were kept in an incubator at 37°C, a minimum relative humidity of 95%, and atmosphere of 5% CO 2 in air.

RNA extraction, reverse-transcription and realtime PCR
Total RNA from tissues or cells was extracted with Trizol reagent (Invitrogen, USA) under the manufacturer's instruction. RNA concentration was measured by NanoDrop-2000 (Thermo Fisher, USA). 2.5 μg of total RNA was used for cDNA synthesis with RevertAid First Strand cDNA Synthesis kit (Thermo Fisher, USA). Expression levels of mature miR-9 were detected by the Bulge-LoopTM miRNA q-RT-PCR primer kits (RiboBio, Guangzhou, China). Roche light Cycler 480 mix (Roche, Mannheim, Germany) was employed to perform real-time PCR on a CFX 96 Real-Time PCR Detection System (Biorad, USA). GAPDH mRNA or U6 snRNA was used as the internal control. Sequences of real-time PCR primers were listed in Table 1.

Western blotting analysis
48 h after lentivirus infection, total protein was extracted from cells by RIPA cell lysis buffer with proteinase and phosphatase inhibitors. The supernatant of lysates was collected and electrophoresed in 10%-12% SDS-PAGE gel, then transferred onto PVDF membranes. Membranes were blocked in Tris-buffered saline plus 0.1% Tween-20 (TBST) with 5% non-fat dry milk or bovine serum albumin (BSA) at room temperature for 1h, followed by primary antibody incubation at 4°C overnight. On the next day after washing with TBST, membranes were incubated with secondary antibodies labeled with HRP at room temperature for 1 h. Then proteins were detected by ECL

CCK-8 assay
The CCK-8 cell counting kit (Zoman, Beijing, China) was used to determine the cell viability. U251 or U87MG cells were infected with lentivirus for 48 h. Then, cells were seeded onto 96-well plates at the density of 2 x 10 3 cells per well for measurement at time points of 1day (d), 2d, 3d, 4d and 5d after lentivirus infection. At each time point, 10μl CCK-8 solution was added into each well. After 1 h incubation at 37°C, optical density (OD) at 450 nm was measured by a microplate reader (EXL800 from BioTek, Winooski, Vermont, USA).

Wound healing assay
U251 and U87MG cells were seeded in 6-well plates at a density of 2 x 10 5 cells per well. 48 h after lentivirus infection, the cell layer was wounded by a sterile 10μl pipette tip. Cells were washed three times with PBS to remove floating cells. Then cells were allowed to grow for 24 hours. The wound closure was observed by a light microscope at x100 magnification. Images of three random fields were captured at 0 h, 12 h, and 24 h time points post wounding.

Cell migration and invasion assays
Twenty-four-well plates and Transwell chambers (pore size at 8μm) (corning, New York, USA) were used for cell migration and invasion experiments. For the migration assay, 48 h after lentivirus infection, 200 μl of serum-free medium containing 1 x 10 5 cells was added to the upper chamber, and 600 μl of medium with 20% serum was added to the lower chamber as a chemoattractant. Cells were incubated for another 20 h at 37°C before measurement. For the invasion assay, the upper chamber was pretreated with 50 μl matrigel (BD Bioscience, San Jose, USA) for 4 h at 37°C. Then, 200 μl of serum-free medium containing 2 x 10 5 cells was added to the upper chamber, while 600μl medium with 20% serum to the lower chamber. Cells were incubated for another 30 h at 37°C. After the incubation, cells in top chambers were removed by a cotton swab. Cells on the lower surface of the insert were fixed in 4% PFA for 15 min, and stained with 0.1% crystal violet for 30min. An optical microscope was used to visualize stained cells in five random fields.

EdU assay
EdU assay kit (Ribobio, Guangzhou, China) was used to detect cell proliferation. 6 x 10 3 cells were cultured for each well in 24-well plates. 48 h after lentivirus infection, 200 μl of 50 μM EdU solution was added into each well. Cells were cultured for another 2 h. Afterward, cells were fixed with 4% PFA for 30 min, and incubated in 100 μl PBS with 0.5% tritonX-100 for 10 min. Then cells were incubated with Apollo solution (Ribobio, Guangzhou, China) for 30 min, followed by DAPI staining for nuclei. EdU-labeled cells were counted manually in ten random fields from each well.

Immunofluorescence
Cells were fixed with 4% PFA and washed with PBST. Then cells were blocked in PBS with 5% normal goat serum at room temperature for 1h. Samples were then incubated with the primary antibody against Ki67 (No. 14-5698, 1:200, eBioscience, San Diego, USA) at 4°C for overnight. On the next day after washing with PBS, slides were incubated with the secondary antibody (No. A23340, 1:500, Abbkine, USA) at room temperature for 1 h. Afterward, slides were mounted with anti-fade mounting medium (with DAPI) and observed under an epifluorescence microscope.

Apoptosis analysis
Annexin V-PE /7-AAD apoptosis detection kit (Keygen Biotech, Nanjing, China) was used to analyze cell apoptosis. Cells were washed with ice-cold PBS and stained with annexin-V and 7-AAD. Apoptotic cells were quantified by a dual-color flow cytometry machine (FACS Aria III, BD Biosciences, San Jose, CA, USA). Ten thousand cells were counted per sample. Data were analyzed by Flowjo 7.6 software.

Cell cycle analysis
The content of DNA was detected by a cell cycle detection kit (Keygen biotech, Nanjing, China). 48 h after lentivirus infection, U251 and U87MG cells were harvested in ice-cold PBS, then fixed with 70% cold ethyl alcohol at 4°C for overnight. Fixed cells were incubated with RNAse A for 30 min and stained with propidium iodide (PI) for 30 min in the darkness. Stained cells were measured by flow cytometry (FACS Aria III, BD Biosciences, San Jose, CA, USA). Data were analyzed with a cell-cycle program (Modfit) to calculate percentages of cells at G0/G1, G2/M and S phases.

Statistical analysis
All experiments were performed for at least three times independently. All data were presented as mean ± SEM. SPSS 19.0 and Prism 5.0 statistical softwares were employed for the analysis. Corresponding significance levels were indicated in each figure.

CONFLICTS OF INTEREST
The authors declared that they have no competing interests.