A total of 53 brain tumors including 10 WHO grade III and 43 WHO grade IV gliomas was examined. Patient characteristics are listed in Table I. Seven adjacent normal human brain tissues were used as the control. The study was approved by the Hunan Normal University Human Ethics Committee, and informed consent was obtained from all patients. The immunohistochemistry was analyzed on formalin-fixed and paraffin-embedded samples. Endogenous peroxidase was blocked with 3% H₂O₂ in methanol for 10 min. The mouse monoclonal anti-Eps8 antibody (BD Biosciences, CA, USA) at a 1:500 dilution or mouse control IgG was added at 4°C overnight. The sections were incubated with HRP-conjugated goat anti-mouse secondary antibody (Sigma-Aldrich Corp., St. Louis, MO, USA) for 30 min and then with 3,3-diaminobenzidine (DAB)/H₂O₂ for 5 min. The sections were counterstained with hematoxylin, mounted and photographed using an optical microscope. The percentage of tumor cells stained was scored as: 0 (no cell staining), + or 1 (≤30%), ++ or 2 (31–60%), and +++ or 3 (61–100%). Some samples with a staining between two score values were given 0.5.

The human neuroblastoma cell line SH-SY5Y was cultured in a 1:1 mixture of F12 and Eagle’s minimum essential medium (EMEM) supplemented with non-essential amino acids, glutamine, sodium pyruvate, penicillin, streptomycin, and 10% fetal bovine serum (FBS). HEK293FT and 293T cells, as well as human glioma A172, U251, and SHG-44 cells were cultured in complete Dulbecco’s modified Eagle’s medium (DMEM) containing penicillin, streptomycin and 10% fetal bovine serum. Cells were maintained in a humidified atmosphere of 5% CO₂ and 95% air at 37°C.

To generate the pEGFP-C3-Eps8 plasmid, full-length human Eps8 was released by EcoRI and XhoI digestion of a Myc-tagged human Eps8 expression vector pCMV-Myc-Eps8 (20) and subcloned into the EcoRI and SalI sites of vector pEGFP-C3. U251 cells were transfected with pEGFP-C3-Eps8 or pEGFP-C3 control plasmids using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). Stably transfected cells were isolated by G418 screening (21,22). The transfection efficiency was detected using a fluorescence microscope (Zeiss Axioskop 2, LLC, USA).

Samples were denatured in sample buffer [2% sodium dodecyl sulfate (SDS), 10% glycerol, 2.5% β-mercaptoethanol, and bromophenyl blue] and heated to 100°C for 5 min before gel electrophoresis. Samples were then separated on a 10% SDS-PAGE gel and transferred to a polyvinylidene difluoride (PVDF) membrane. The blots were probed with rabbit polyclonal anti-GFP antibody, mouse monoclonal antibodies against β-catenin, β-actin, Akt, phosphorylated Akt, ERK and phosphorylated ERK (Santa Cruz Biotechnology, Santa Cruz, CA, USA). HRP-conjugated goat anti-rabbit and goat anti-mouse secondary antibodies (Sigma) were used for detection. The signal was visualized with SuperSignal West Femto Maximum Sensitivity Substrate (Pierce Biotechnology, Rockford, IL, USA).

Four siRNA sequences targeting Eps8 (NM_004447.5) were from the positions (siRNA1, 93–111; siRNA2, 246–264; siRNA3, 678–696; siRNA4, 1974–1992) relative to the start codon and synthesized by Shanghai Integrated Biotech Solutions Co. Ltd., China. After testing knockdown efficiencies, stem-loop DNA oligonucleotides were synthesized by Shanghai GeneChem Co. Ltd. (shRNA1 sense, 5′-CCG GAC GGA CAG AGA ACA TGG TTC ACT CGA GTG AAC CAT GTT CTC TGT CCG TTT TTT G-3′; antisense, 5′-AAT TCA AAA AAC GGA CAG AGA ACA TGG TTC ACT CGA GTG AAC CAT GTT CTC TGT CCG T-3′; shRNA2 sense, 5′-CCG GAC CAC TGT TGA TGA TGG AAT ACT CGA G TA TTC CAT CAT CAA CAG TGG TTT TTG-3′; antisense, 5′-AAT TCA AAA AAC CAC TGT TGA TGA TGG AAT ACT CGA GTA TTC CAT CAT CAA CAG TGG T-3′) and cloned into the lentivirus-based RNAi vector pGCSIL-GFP. A nontargeting stem-loop DNA was also inserted into the pGCSIL-GFP vector as a negative control (NC). Lentiviral particles were prepared as described previously (23). SHG-44 cells were infected with Eps8-RNAi-lentivirus or NC-GFP-lentivirus and detected on day 6 by a fluorescence microscope and western blot analysis.

For cell survival assays, 5,000 cells stably expressing GFP-Eps8 or GFP and parental U251 cells were plated in triplicate in 6-well plates in complete medium containing 400 μg/ml G418. For SHG-44 cells, 5,000 cells expressing Eps8-RNAi-LV or NC-GFP-LV were plated in triplicate in 6-well plates. After 8–10 days, cell numbers were counted with a hemocytometer after trypan blue staining of viable cells.

For the cell viability assay, 3,000 cells were seeded in octuplicate in 96-well plates. On days 1, 4, 8 and 10 or days 1, 3, 5 and 8, cells were treated with 1 mg/ml 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenytetrazolium bromide (MTT) at 37°C for 4 h. Then 100 μl dimethylsulfoxide (DMSO) was added to the medium to dissolve the formazan crystals. The absorbency value at 490 nm in each well was obtained using a spectrophotometer (UV-2102C, Changsha, China).

The liquid colony formation was carried out. 1,000 cells were seeded in triplicate in 6-well plates, and cells were grown for 2 weeks in a humidified incubator at 37°C with 5% CO₂. Colonies were fixed with methanol, stained with Giemsa (BBI International, Cardiff, UK) and photographed with a digital camera (Olympus, Tokyo, Japan). Only colonies containing over 30 cells were counted.

Data are presented as the means ± SD from at least three independent experiments. The significance of difference was assessed using the Student’s t-test. Values of p<0.05 were considered statistically significant.

The expression level of Eps8 was examined in 10 WHO grade III, 43 WHO grade IV gliomas and 7 adjacent normal human brain tissues by immunohistochemical analysis using mouse monoclonal Eps8-specific antibody. We found that Eps8 was completely localized in the cytoplasm (Fig. 1A). Eps8 expression was detected in 30 (56.6%) of the 53 brain tumors with strong staining (3+), 20 (37.7%) of the 53 gliomas were moderately stained (2+), and 3 (5.7%) were weakly stained or negative for Eps8 expression (+/0), which showed Eps8 was significantly highly expressed in the gliomas (P<0.001; Student’s t-test). A complete loss of Eps8 expression was observed in normal brain tissues (Fig. 1B). Therefore, Eps8 expression was markedly increased in the human glioma tissues when compared with expression in the normal brain tissues.

We next analyzed the expression of Eps8 protein in three human glioma cell lines (A172, U251 and SHG-44) and human neuroblastoma cell line SH-SY5Y. Higher expression of Eps8 protein was evident in the SHG-44 cells when compared with that in the A172 and U251 cells. SH-SY5Y cells demonstrated a slight level of Eps8 protein (Fig. 1C). Thus, the level of Eps8 expression differed in the two glioma cell lines. U251 and SHG-44 cell lines were further selected to overexpress and knockdown Eps8 protein, respectively.

To further investigate the role of Eps8 in human glioma cells, we constructed the pEGFP-C3/Eps8 plasmid and established U251 cell lines stably overexpressing GFP/Eps8 and GFP by G418 screening (Fig. 2A). The level of Eps8 overexpression was demonstrated by western blotting. As shown in Fig. 2B, U251-GFP and U251-GFP/Eps8 cells were found to express the target proteins at high levels.

We next examined whether Eps8 overexpression in U251 glioma cells promotes cell proliferation. Cells were plated in triplicate in 6-well plates, and the cell number was counted on day 10. We found that the overexpression of Eps8 in U251 cells resulted in an increased cellular growth compared with the controls (Fig. 2C). Similarly, we tested the cell viability using MTT assay, Eps8 overexpression resulted in a striking increase in the number of viable cells (Fig. 2D). Furthermore, the liquid colony formation assay demonstrated a great increase in the numbers of colonies of the Eps8-overexpressing U251 cells (Fig. 2E). Taken together, these data suggest that Eps8 contributes to glioma cell survival and proliferation in vitro.

The above results indicate that Eps8 significantly promotes U251 cellular proliferation. To gain further supporting evidence, we knocked down Eps8 expression using shRNA and determined whether shRNA-mediated Eps8 knockdown inhibits the proliferative effects. Four siRNA sequences targeting different regions of Eps8 mRNA were constructed and their abilities to knock down Eps8 expression in transfected HEK293FT cells were demonstrated by western blotting. As shown in Fig. 3, the four sequences knocked down the Eps8 gene to different levels, especially siRNA1 and siRNA2.

Consequently, siRNA1 and siRNA2 were designed to construct the lentivirus-based RNAi vector pGCSIL-Eps8. Packaging plasmids and pGCSIL-Eps8 were cotransfected into the 293T cells. The fluorescence intensity was markedly increased 4 days after transfection and the transfection efficiency was nearly 90% in the SHG-44 cells (Fig. 4A). Meanwhile, the expression of Eps8 proteins was detected by western blot analysis. Eps8 shRNA1 and shRNA2 significantly suppressed the expression of Eps8 protein when compared with the negative control shRNA (Fig. 4B).

Given that Eps8 is a critical regulator of glioma cell proliferation, we examined the effect of Eps8 on SHG-44 glioma cell proliferation. As shown in Figs. 4C and D, Eps8 knockdown markedly inhibited the proliferation of SHG-44 cells, whereas control shRNA exhibited no effect. The colony formation assays showed that deletion of Eps8 expression produced a significant decrease in the clonogenicity of SHG-44 cells, while control shRNA produced no effect (Fig. 4E). These data clearly indicate that Eps8 knockdown reduces glioma cell proliferation.

Since Eps8 regulates many genes involved in the development of human cancers, we determined to investigate whether Eps8 influences these target genes in glioma cell lines and tissues. As shown in Fig. 5A, Eps8 overexpression had no effect on ERK and Akt protein levels, while it increased the levels of phosphorylated ERK and Akt expression. Moreover, the level of β-catenin was concomitantly enhanced in Eps8-transfected cells. Likely, Eps8 knockdown decreased the expression of phosphorylated ERK, phosphorylated Akt and β-catenin, but not ERK or Akt (Fig. 5B).

We then analyzed the expression of Eps8 and target genes in normal brain tissues and gliomas. Eps8 had a significantly higher expression in high grade gliomas (WHO III-IV) than in low grade gliomas (WHO I-II) and normal brain tissues. In addition, the expression of Eps8 revealed broadly paralleled expression of target genes in the gliomas and normal brain tissues, particularly phosphorylated ERK and phosphorylated Akt. However, the levels of β-catenin were partially correlated with Eps8 expression in certain samples (Fig. 5C). Taken together, these data show that Eps8 regulates cell proliferation and survival, at least in part, by affecting phosphorylated ERK and Akt/β-catenin activities.