C6 glioma cells were purchased from the Shanghai Institute of Biochemistry and Cell Biology (Shanghai, China). The cells were cultured in the serum containing medium composed of Dulbecco’s modified Eagle’s medium (DMEM) with high glucose and 10% fetal bovine serum. Cells that were in the exponential growth phase were then collected and transplanted to a new culture flask with an equal volume of serum-free neural stem cell (NSC) medium containing DMEM/F12 with 2% B27, epidermal growth factor (EGF, 10 ng/ml) and basic fibroblast growth factor (bFGF, 10 ng/ml). Cells were incubated at 37°C with 95% air, 5% CO₂ and 100% humidity. Two weeks later, floating primary tumor spheres were collected. The sphere cells were harvested, dissociated into single cells, and plated into a 96-well plate for the subsphere-forming assay by limiting dilution as described previously (25). In brief, the cells in single-cell suspension were diluted to one cell per microliter and plated at one to two cells per well. Cells were fed with 200 μl serum-free media with growth factors, changing half of the medium every 5 days. After plating, the cells were observed, and only wells containing a single cell were considered.

Subspheres from one mother cell were characterized by immunocytochemistry against anti-CD133 antibody (1:400, Roche, Germany) or anti-nestin antibody (1:1,000, Chemicon) as described previously. In brief, the subspheres were incubated with primary antibodies: rat anti-CD133, mouse monoclonal anti-nestin overnight at 4°C, followed by incubation with Alexa Fluor 568-conjugated goat anti-rat (1:1,000, Molecular Probes), 488-conjugated goat anti-mouse (1:800, invitrogen). Cell nuclei were counter-stained with Hoechst 33342 for 30 min at room temperature (RT). Immunopositive cells were observed using a fluorescent microscope.

Four self-complementary hairpin DNA oligos targeting rat VEGF mRNA were designed by using Dharmagon siDESIGN Center algorithrm (http://www.dhar-macon.com/DesignCenter/DesignCenterPage.aspx) as follows: VEGF shRNA1 is 5′-TGCGGATCAAACCTCACC AAA-3′; VEGF shRNA2 is 5′-GAGCGGAGAAAGCATTTG TTT-3′; VEGF shRNA3 is 5′-GCGAGGCAGCTTGAGTTA AAC-3′; VEGF shRNA4 is 5′-GCCTCTGAAACCATGAAC TTT-3′, including a 9-nt loop sequences. Also, a mock sequence that lacks any similarity to the rat sequence was designed and synthetized. Lentiviral vectors for rat VEGF shRNA encoding a green fluorescent protein (GFP) sequence and a puromycin resistance gene was constructed by GenePharma (Shanghai, China), named as pGLVH1/GFP/PURO. DNA oligos were annealed and inserted in the rat H1 promoter site of the PGLVH1-GFP vector. Then, the RNA interference (RNAi) cassette was cloned into the latter vector and lentiviral vectors expressing shRNA were constructed. They were then confirmed by DNA sequencing identification.

Lentiviral vector DNAs and packaging vectors (pHelper1.0, pHelper2.0) (Qiagen) were then transfected into 293T cells. After transfection, the cells were incubated at 32°C to increase viral titer. Forty-eight hours later, the supernatant containing the retroviral particles was collected, filtered through the 0.45-μm low protein binding syringe filter and the titer of lentivirus was determined.

C6GSCs were maintained in above mentioned medium and were plated into 6-well plates at 4×10⁵ cells/well. Twenty-four hours later, the cells were transfected with viral supernatants in the presence of polybrene (6 μg/ml final concentration) for 12 h at a multiplicity of infection (MOI) of 10 and 20, then replaced with a fresh medium containing 5 μg/ml puromycin. After 72 h, the cells had been transfected for the subsequent study.

Total RNA was extracted from 1×10⁶ transfected cells in each group using TRIzol reagent (Invitrogen, Carlsbad, CA, USA), and 2 μg of total RNA was reverse transcribed into cDNA with oligo dT primers using Omniscript reverse transcriptase (Qiagen, Hilden, Germany) according to the manufacture’s protocol. Hot start PCR was performed to analyze expression levels with the following primers. VEGF: sense 5′-GACCCTGGTGGAC ATCTTCCAGGA-3′ and antisense 5′-GGTGAGAGGTCTAG TTCCCGA-3′. GAPDH: sense 5′-TGATGACATCAAGAAG GTGGTGAAG-3′ and antisense 5′-TCCTTGGAGGCCAT GTGGGCCAT-3′.

The PCR products were subjected to agarose gel electrophoresis followed by staining with ethidium bromide. The band optical density of VEGF, relative to that of GAPDH (internal control) was quantified by image analysis system (Leica Q550I W, Cambridge, UK).

Western blot analysis was performed for whole cell lysate. Aliquots of total protein (50 μg per lane) were electrophoresed on a 12% SDS-polyacrylamide gradient gel and transferred to nitrocellulose membranes (Millipore). Washed in rinse buffer at RT and incubated in blocking buffer (5% fat-free milk in rinse buffer) for 30 min, the membranes were incubated for 2 h at RT with antibody. Further washed with rinse buffer, the membranes were incubated with 1:1,000 diluted HRP-conjugated secondary antibody (Santa Cruz) for 2 h at RT, followed by developing with enhanced chemiluminescence reagents (Amershame, Little Chalfont Buckinghamshire, UK). In addition, β-actin was used as a reference protein. The optical densities were analyzed by using ImageMaster™2D Platinum (Version 5.0, Amersham Biosciences, Piscataway, NJ, USA).

The non-transfected C6GSCs (control group), mock transfected C6GSCs (mock group) and lentiviral vector-VEGF shRNA1 transfected C6GSCs (silence group) were cultured in the above mentioned medium. Photomicrographs were taken by an inverted phase contrast microscope and the size of cell spheres was analyzed. At 7 days in vitro (DIV), some cell spheres of three groups were dissociated into single-cell suspension mechanically by trypsin digestion. Collected cells were washed once in cold phosphate-buffered saline (PBS). The cells were fixed in 70% ethanol, treated with 100 mg/l RNase at 37°C for 30 min and stained with 50 mg/l propidium iodide (Sigma) for 30 min. The cells were analyzed using flow cytometry (Epics XL; Beckman Coulter, Fullerton CA, USA).

C6GSCs of three groups (1×10⁵ cells/ml) were plated onto poly-D-lysine coated cover slips in 24-well plates containing serum-free medium, respectively. After 7 DIV, cells were fixed with 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer (PB) for 1 h, blocked in 5% goat serum for 30 min at RT and then incubated for 12 h at RT with primary antibodies. After washing three times in PBS, the cells were incubated in secondary antibodies for 6 h at RT. Antibodies were as follows: the primary monoclonal antibodies were against mouse microtubule associated proteins-2 (anti-MAP-2, 1:200, Chemicon), mouse glial fibrillary acidic protein (anti-GFAP, 1:200, Sigma), rat anti-nestin (1:2,000, Millipore). Secondary antibodies were Alexa Fluor 568-conjugated (red) goat anti-mouse IgG (1:500, Invitrogen) or FITC-conjugated (green) goat anti-rat IgG (1:200, Millipore). Cell nuclei were counter-stained with Hoechst 33342 for 30 min at RT. Immunopositive cells were observed using fluorescent microscopy.

C6GSCs of three groups (1×10⁵ cells/ml) were added to the upper chamber of the Transwell. The lower chamber was filled with DMEM/F12 medium. After incubation for 12 h at 37°C, the cells were fixed with 4% PFA for 30 min and stained with Hoechst 33342 for 30 min at RT. The filters were then rinsed thoroughly in distilled water and checked by bright field microscopy to ensure that the cells were adherent and had migrated. The non-migrating cells were then removed from the upper surface (inside) of the Transwell with a wet cotton swab.

Images were analyzed by Leica Qwin image processing and analysis software (Leica imaging system, Cambridge, UK). Statistical analysis was performed using statistics package for social science 16.0 (SPSS 16.0). Data are presented as mean ± standard deviation (M ± SD). Statistical comparisons were performed using one-way analysis of variance (ANOVA) and differences at P<0.05 were considered statistically significant.

The C6 cells were cultured in serum-free NSC medium, and many formed floating primary tumor spheres. Subspheres from one mother cell were CD133 and nestin double-positive. The nestin fluorescence intensity of the C6GSC sphere of the silence group was lower than that of the other two groups (Fig. 1).

In order to exclude an off-target silencing effect mediated by specific shRNA, we employed 4 different sequences of VEGF shRNA. Lentiviral infection efficiency was calculated by GFP expression under a fluorescence microscope. The infection efficiency of lentiviral vector-shRNA was 98.02±1.37% when MOI was 1:20 (Fig. 2). To evaluate the inhibition of VEGF mRNA expression, semi-quantitative RT-PCR was performed 72 h after infection, the VEGF mRNA expression in lentiviral vector-VEGF shRNA1 transfected C6GSCs was reduced by ∼75%, as compared with the non-transfected and mock transfected ones (P<0.05). In addition, no difference was observed among the non-transfected cells, the mock-infected cells, and VEGF shRNA2, 3 and 4 transfected cells (Fig. 3), indicating that the corresponding mRNA sequence for VEGF shRNA1, not VEGF shRNA2, 3 and 4 is specific RNAi target. Western blot analysis was also performed 72 h after infection. The VEGF protein expression demonstrated a significant reduction in lentiviral vector-VEGF shRNA1 transfected C6GSCs, as compared with non-transfected cells and mock transfected cells (P<0.05) (Fig. 4), suggesting that VEGF shRNA1 strongly blocked VEGF expression, whereas no obvious inhibition of VEGF protein was observed in VEGF shRNA2, 3 and 4 transfected cells.

The C6GSCs of 3 groups were cultured under the same conditions outlined in Materials and methods. The appearance of the sphere of three kinds of stem cells is similar at the beginning, showing a small and uniform cell sphere (Fig. 5A–C). In the following days in culture, nontransfected C6GSCs and mock transfected C6GSC spheres grew significantly faster and reached a larger size than lentiviral vector-VEGF shRNA1 transfected C6GSC spheres (Fig. 5D–I). At 7 DIV, the proliferation of lentiviral vector-VEGF shRNA1 transfected C6GSCs reached the peak. On the contrary, the other two groups proliferated continually and the diameter of a cell sphere was up to ∼236.6820.43 and 240.8520.52 μm (Fig. 5J). Flow cytometry also indicated that proliferation index of lentiviral vector-VEGF shRNA1 transfected C6GSCs were lower than that of the other two groups (Fig. 6).

Differentiation culture of three kinds of stem cells was performed in the conditions outlined in Materials and methods. After 7 DIV, three kinds of stem cells exhibited multi-directional differentiation, all of them differentiated into MAP-2-positive neurons and GFAP-positive astrocytes. MAP-2-positive neurons and GFAP-positive astrocytes derived from lentiviral vector-VEGF shRNA1 transfected C6GSCs were nestin-negative, cell morphology was good, showing extensive processes and rounded bodies. MAP-2-positive neurons and GFAP-positive astrocytes derived from the other two groups were nestin-positive, cell morphology was less distinct, the differentiated cells were poorly differentiated (Figs. 7 and 8).

Cell invasion is a characteristic feature of carcinoma cells. To investigate the role of VEGF in tumor cell invasion, we measured the ability of C6GSCs to invade through Matrigel-coated filters in a Transwell chamber assay. After transfected with VEGF shRNA1, the invasive cells were obviously decreased compared with the other two groups. However, there were no significant differences between the other two groups (Fig. 9).