The human U251, U87, LN229, U87 and A172 glioblastoma cell lines were purchased from the Institute of Biochemistry and Cell Biology, Chinese Academy of Science. All cells were maintained in a 37°C, 5% CO₂ incubator in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (Invitrogen, Carlsbad, CA, USA).

Human glioma samples were obtained from the Department of Neurosurgery, Sir Run Run Shaw Hospital, Medical College, Zhejiang University after informed consent from adult patients diagnosed with glioblastoma, freshly resected during surgery and immediately frozen in liquid nitrogen for subsequent total RNA extraction.

Total RNA was isolated from cultured cells, human glioma specimens, or NAT (normal adjacent tissue)s using TRIzol® reagent (Invitrogen) according to the manufacturer’s instructions. Quantitative real-time polymerase chain reaction (qRT-PCR) was performed in triplicate in ABI 7500HT fast real-time PCR system (Applied Biosystems, Foster City, CA, USA) and normalized with U6 and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) endogenous control. Total RNA from NATs was used as a control. miR-92a levels were measured with the TaqMan microRNA assay kit, and endogenous mRNA levels of Bim were detected using SYBR-Green PCR Master Mix kit in accordance with the manufacturer’s instructions (Applied Biosystems).

miR-92a: 5′-UAUUGCACUUGUCCCGGCCUGU-3′, AS-miR-92a: 5′-ACAGGCCGGGACAAGUGCAAUA-3′ and scrambled control oligonucleotides were purchased from Shanghai GenePharma Co., Ltd. (Shanghai, China). Bim expression plasmid was preserved in our laboratory. miRNA sequences/Bim expression plasmid were transfected into cultured cells using Lipofectamine 2000 reagent (Invitrogen) following the manufacturer’s instructions.

The MTT assay was used to determine relative cell growth as follows. U251 and U87 cells were plated at 10⁴ cells/well in 96-well plates with 6 replicate wells for each condition, transfected with oligonucleotides, and assayed 48-h post-transfection. Cell growth assay was performed by MTT (Sigma, St. Louis, MO, USA) as previously described (18). The cell viability was determined at 540-nm absorbance using an enzyme-linked immunosorbent assay plate reader. All data points represent the mean of a minimum of 6-wells.

Forty-eight hours after transfection, apoptosis in cultured cells was evaluated with Annexin V labeling, caspase-3/7 activity and mitochondrial membrane potential. For the Annexin V assay, an Annexin V-FITC labeled Apoptosis Detection kit (Abcam) was used according to the manufacturer’s protocol. Caspase-3/7 activity was measured using caspase-Glo-3/7 reagent (Promega, Madison, WI, USA). Mitochondrial membrane potential was determined with cationic dye JC-1 (5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzi-midazolylcarbocyanine-chloride/C25H27Cl3N4) staining (19). Briefly the cells were harvested and first stained with PI. Following wash twice with PBS, the cells were incubated with 10 mg/ml JC-1 for 20 min at room temperature and then analyzed with FACSCalibur to detect green fluorescence at excitation/emission wavelengths of 485/530 nm and red fluorescence at excitation/emission wavelengths of 485/590 nm. TUNEL assay was used to detect the apoptosis in tumor specimens and was performed as previously described (20).

Western blot analysis, miRNA-LNA in situ hybridization and immunohistochemistry were performed as previously described (21). For reporter assay, the pGL3-WT-Bim-3′UTR-Luc reporter (WT Bim) was created by the ligation of Bim 3′UTR PCR products into the XbaI site of the pGL3 control vector (Promega). The pGL3-MUT-Bim-3′UTR-Luc reporter (MUT Bim) was generated from pGL3-WT-Bim-3′UTR-Luc by replacing the binding site of miR-92a with restriction enzyme cutting site CGGATCCG. For the reporter assay, cells were cultured in 96-well plates and transfected with WT Bim/MUT Bim and AS-miR-92a. Luciferase activity was measured 48 h after transfection with the Dual-luciferase reporter assay system. The Renilla luciferase activity was utilized as an internal control.

U251 glioma cells were subcutaneously injected to 5-week-old female nude mice (Cancer Institute of The Chinese Academy of Medical Science). When the tumor volume reached 100 mm³, the mice were randomly divided into two groups (10 mice per group) which were treated with 200 pmol scramble oligo or AS-miR-92a in 10 μl Lipofectamine 2000 through local injection of xenograft tumor in multiple sites. The treatment was performed once every 2 days for 14 days. The tumor volume was measured with a caliper every 2 days, using the formula: volume = length × width²/2.

All tests were done using SPSS Graduate Pack 11.0 statistical software (SPSS Inc., Chicago, IL, USA). Statistical evaluation for data analysis was determined by one-way ANOVAs, the Student’s t-test and the Chi-square test. Differences with P<0.05 were considered statistically significant.

To explore the expression of miR-92a in human glioma samples and cell lines, we performed the TaqMan-based real-time stem-loop RT-PCR analyses. Our data showed that miR-92a was obviously upregulated in glioma cell lines vs. NATs (Fig. 1A). Similar results were observed in the glioma tissues (Fig. 1B). Taken together, our results demonstrated that miR-92a was abnormally overexpressed both in human glioma samples and cell lines.

To examine biological significance of miR-92a in glioma, U251 and U87 cells were treated with AS-miR-92a and scramble oligonucleotides (Fig. 1B). First of all, real-time PCR showed that AS-miR-92a reduced miR-92a levels by 81% in U251 cells and 74% in U87 cells (P<0.05) (Fig. 2A). MTT assay showed that proliferation was inhibited in the miR-92a abrogation glioma cells, which was the most apparent on Day 3 (Fig. 2B). We further explored effects of AS-miR-92a on apoptosis. Annexin V labeling revealed that knockdown of miR-92a significantly increased cell apoptosis compared to the cells treated with scramble oligonucleotide and control group (Fig. 2C). Moreover, western blot assay displayed that pro-apoptotic protein Bax and cleaved-caspase-3 expression were significantly upregulated while Bcl-2 expression was downregulated in As-miR-92a group (Fig. 2F). In addition, caspase-3/7 activity was also considerably elevated in miR-92a deleted cells (Fig. 2E). Since collapse of the mitochondrial membrane potential is one of the early events in apoptosis (22), we next examined if miR-92a regulates mitochondrial membrane potential. The cells with suppression of miR-92a were stained with cationic dye JC-1. FACSCalibur analysis showed that the mitochondrial membrane potential was largely damaged when miR-92a were downregulated (Fig. 2D). These findings indicate that miR-92a plays an important role in regulation of cell apoptosis.

Based on the analysis of databases miRanda (http://www.cbio.mskcc.org/mirnaviewer), PicTar (http://www.pictar.bio.nyu.edu), and TargetScan (http://www.targetscan.org), we predicted that Bim may be a target gene for miR-92a (Fig. 3A). To determine whether Bim is regulated by miR-92a, we knocked-down miR-92a in U251 and U87 cells, and evaluated Bim protein levels at 48 h post-transfection. And western blot analysis showed that AS-miR-92a obviously increased Bim protein levels in U251 and U87 glioma cells (Fig. 3B).

To assess whether Bim is a direct target of miR-92a, we created WT Bim and MUT Bim plasmids, which were co-transfected with AS-miR-92a or scrambled oligonucleotide respectively into glioma cells for 48 h, followed by measurement of luciferase activity in transfected cells. Our results showed that the reporter plasmid with wild-type 3′UTR of Bim caused a significant increase in luciferase activity in cells transfected with AS-miR-92a, while MUT reporter plasmid produced no change in luciferase activity (Fig. 3C). Taken together, these data indicated that miR-92a directly modulates Bim expression by binding to 3′UTR of Bim.

We next assessed the importance of Bim in miR-92a-mediated cell survival. Since A172 cells had relatively low expression of miR-92a, we transfected AS-miR-92a or scramble oligonucleotide 24 h after transfection with Bim expression plasmid lacking 3′UTR. The proliferation and apoptosis were assessed in co-transfected A172 cells, which showed that ectopic expression of Bim abrogated miR-92a effects on cell proliferation and apoptosis (Fig. 4A-C). Subsequently, western blot results demonstrated the upregulation of Bim protein levels was significantly prevented by miR-92a (Fig. 4D). Taken together, these results suggested that Bim is a critical target in miR-92a-mediated glioma apoptosis.

Since miR-92a are frequently elevated in glioblastoma and play an important role in cell survival, we further examined the effects of knockdown of miR-92a on glioma growth in vivo. As shown in Fig. 5A, AS-miR-92a significantly reduced tumor growth compared with scramble group (P<0.05). LNA-ISH analysis confirmed that miR-92a levels were considerably reduced in AS-miR-92a group (Fig. 5B). TUNEL assay analysis of xenograft tumor taken at 28 days after treatment revealed much more apoptosis in AS-miR-92a group when compared to tumors from scramble groups (Fig. 5B). In addition, Ki-67 staining shows that AS-miR-92a treated tumors had a lower proliferation index compared with the control groups (Fig. 5B). Immunohistochemical stanining analysis revealed that Bim levels were upregulated in AS-miR-92a group (Fig. 5B), confirming the data in vitro that BIM as a direct target of miR-92a. Additionally, Bax expression was increased, whereas Bcl-2 expression was decreased in xenograft tumor sections (Fig. 5B). These findings further indicate that miR-92a targets Bim and that AS-miR-92a could be therapeutic means for glioblastoma intervention.

Having demonstrated Bim as a major target of miR-92a, we further investigated the correlation of between miR-92a and Bim expression in gliomas. We examined 96 human glioma specimens with LNA-ISH and immunohistochemical staining. Representative images of miR-92a and Bim are shown in Fig. 6A. Upregulation of miR-92a was detected in 44 gliomas (Fig. 6B). Of the 44 tumors with elevated miR-92a, 36 (81.8%) had low levels of Bim (P<0.05). Thirty-nine of 52 (75%) specimens with downregulated miR-92a presented high levels of Bim. In addition, we found that miR-92a expression increased significantly in high grade gliomas compared with low grade gliomas.