Human glioblastomaastrocytoma, epithelial-like cell line U-87MG was purchased from ATCC (www.atcc.org). These cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 μg/ml streptomycin (all from Gibco Invitrogen, Carlsbad, CA, www.invitrogen.com). Neurospheres were formed in DMEM containing Nutrient Mixture F-12 (DMEM/F12, Gibco Invitrogen) supplemented with 20 ng/ml basic fibroblast growth factor (Gibco Invitrogen), 20 ng/ml recombinant human epidermal growth factor (R&D Systems, Minneapolis, MN, http://www.rndsystems.com), and 20 μl/ml N2-supplement (Gibco Invitrogen). A PTEN-expressing plasmid was constructed by cloning PCR-amplified PTEN cDNA into pLPCX (Clontech, Mountain View, CA, http://www.clontech.com). The PTEN-expressing plasmid was transfected into U-87MG cells with Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. U-87MG cells that stably express PTEN, by chromosomal integration of the PTEN-expressing plasmid, were selected in DMEM supplemented with 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 500 μg/ml puromycin (Gibco Invitrogen) for 2 to 3 weeks.

Total-RNA was extracted using TRIzol (Invitrogen) and 2–5 μg total-RNA was reverse-transcribed using the SuperScriptII™ First-Strand Synthesis System (Invitrogen) according to the manufacturer’s instructions. Real-time RT-PCR was carried out on cDNAs using the QuantiTect SYBR-Green PCR Kit (Qiagen, Valencia, CA). Reactions were conducted in triplicate using the Exicycler™96 Real-Time Quantitative Thermal Block (Bioneer, Korea, http://www.bioneer.co.kr). For quantification, the expression of target genes was normalized against that of the glyceraldehyde 3-phosphate dehydrogenase gene (GAPDH). The PCR primer pairs (sense, antisense) were as follows: CD133 (cag agt aca acg cca aac ca, aaa tca cga tga ggg tca gc) and GAPDH (ggg tgt gaa cca tga gaa, gtc ttc tgg gtg gca gtg at).

Cells were washed twice with cold phosphate-buffered saline (PBS), lysed with tissue lysis buffer (20 mM Tris base pH 7.4, 137 mM NaCl, 2 mM EDTA, 1% Triton X-100, 25 mM β-glycerophosphate, 2 mM sodium pyrophosphate, 10% glycerol, 1 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride and 1 mM benzamidine) and clarified by centrifugation at 12,000 x g for 10 min. Whole-cell extracts were prepared and 20–50 μg protein was resolved by SDS-PAGE and blotted using antibodies against PTEN (sc-6817-R, Santa Cruz Biotechnology), Akt (sc-8312, Santa Cruz Biotechnology), p-Akt (#4058, Cell Signaling Technology), p-Stat3 (#9131, Cell Signaling Technology), Stat3 (sc-482, Santa Cruz Biotechnology) and β-actin (sc-47778, Santa Cruz Biotechnology). Immunoreactivity was detected by enhanced chemiluminescence (Amersham Biosciences).

Cells were fixed with 4% paraformaldehyde in PBS for 20 min and permeabilized with 0.1% Triton X-100 in PBS for 15 min. After treatment with 1% goat serum for 1 h, cells were incubated with primary antibodies against goat anti-human p-Stat3 or CD133 (1:100 dilution) at 4°C overnight. Cells were washed with PBS and then incubated with Alexa 488-labeled anti-goat IgG secondary antibody (1:300; Molecular Probes, Eugene, OR) for 2 h.

The SA-β-gal staining method was used to analyze senescence. Cells were washed twice with PBS (pH 7.2), fixed with 2% formaldehyde and 0.2% glutaraldehyde in PBS (pH 7.2), and washed in PBS (pH 7.2) supplemented with 1 mM MgCl₂. Cells were stained in X-gal solution [1 mg/ml X-gal (Boehringer Ingelheim, Ridgefield, CT, http://us.boehringer-ingelheim.com/)], 0.12 mM K₃Fe[CN]₆, 0.12 mM K₄Fe[CN]₆, 1 mM MgCl₂ in PBS at pH 6.0) at 37°C overnight and then examined under microscope (CKX41, Olympus, Japan, http://www.olympus-global.com/).

For scratch tests, cells were plated in 6-well plates at 90–100% confluency. A scratch was made at the center of the well with a 200-μl pipette tip. Cell migration was observed under microscope 24 h after wounding. This experiment was repeated three times.

All of the experimental animals were manipulated in accordance with the guidelines of the CHA University Institutional Animal Care and Use Committee (IACUC090012). Adult male Sprague-Dawley rats (n=11) weighing 260–280 g (Orient, Seoul, Korea) were used in this experiment. One million ordinary U-87 MG cells (called WT thereafter) or U-87 MG cells overexpressing PTEN (called PTEN OE thereafter) in a volume of 2 μl were stereotaxically implanted into one hemisphere of the brain (n=5 WT and n=6 PTEN OE) using the following coordinates: AP, +1.0 mm; ML, −3.0 mm and DV, −5.0 mm. Animals were euthanized at 5 weeks following implantation and brains were taken for histological analyses. H&E staining was carried out on 10 μm-thick brain sections, whereas immunohistochemical staining was carried out on 40 μm-thick sections using the following antibodies: p-Stat3 (1:100, Cell Signaling), PTEN (1:100, Santa Cruz Biotechnology), human-specific nuclei (hNu) (1:200, Chemicon), goat anti-mouse IgG-conjugated Alexa-555 (1:200, Molecular Probes) and goat anti-rabbit IgG-conjugated Alexa-488 (1:200, Molecular Probes).

Graphical data are presented as means ± SD. Each experiment was performed at least three times and subjected to statistical analysis. Statistically significant differences between two groups were determined using Student’s t-test. A p<0.05 was considered significant. Statistical analysis was performed using the SAS statistical package vs. 9.13 (SAS Inc., Cary, NC, http://www.sas.com/).

The PTEN pathway is known to control normal stem cell maintenance, self-renewal and migration. PTEN loss confers increased self-renewal capacity to normal stem cells/progenitors, which can cause the development of cancer stem cells and ultimately tumorigenesis. Expression of the GSC tumor marker CD133 was used to examine the effect of PTEN expression on the GSC population. First, the PTEN-deficient U-87MG cells (wild-type, WT) were engineered to stably express PTEN (PTEN-overexpressing, PTEN OE) (Fig. 1A). Fig. 1B shows that the overexpression of PTEN significantly reduced the levels of CD133 RNA and protein, suggesting that PTEN expression resulted in the depletion of GSCs. As confirmation, the loss of further GSC characteristics was analyzed. Cell migration ability is one of the prominent characteristics of glioblastoma cancer stem-like cells (30), and, as expected, PTEN OE cells showed defective migration ability compared to WT (Fig. 1C). In addition, the ability to form neurospheres, another feature of GSCs, was also significantly diminished in PTEN OE cells (Fig. 1D). Taken together, these results suggest that PTEN expression clearly reduced the population of GSCs among U-87MG cells.

Negative regulation by PTEN of PI3K activity and the PI3K/Akt pathway is of critical importance for cell proliferation. To further analyze the tumor suppressive mechanism of PTEN, the effect of PTEN expression on U-87MG cell growth was examined. As expected, cell counting and MTT assay revealed that PTEN expression reduced the proliferation of U-87MG cells (Fig. 2A and B). Cellular senescence, which can be induced by various stimuli, can function as a critical barrier for cancer development and significantly restrict cancer progression. Since proliferation is directly linked to cellular senescence, the effect of PTEN expression on the induction of senescence in U-87MG cells was analyzed. Compared with the WT cells, the number of cells positive for β-gal staining was significantly increased in PTEN OE at each passage (Fig. 2C). Addition of bpV to the culture medium to inhibit PTEN activity reduced the population of β-gal-positive cells in PTEN OE, suggesting that cellular senescence is induced by PTEN activity (Fig. 2C). Treatment with bpV consistently rescued the growth retardation of PTEN OE (Fig. 2D). Taken together, these results indicate that the restoration of PTEN activity in glioblastoma cells markedly inhibits cell proliferation and leads to senescence.

PTEN inhibits the PI3K/Akt signaling pathway by catalyzing the dephosphorylation of PIP3, while the loss of PTEN induces the activation of the PI3K/Akt cascade, which leads to the stimulation of cell growth and survival (7,8). The LIFRβ-Stat3 signaling pathway is also regulated by PTEN and functions as a brain tumor suppressor in astrocytes (24). In contrast, Stat3 activation has been described in a subset of glioblastoma, indicating that Stat3 plays distinct roles in cell transformation depending on the oncogenic environment (23,31). To examine whether PTEN expression can affect Stat3 and Akt signaling pathways, the phosphorylation level of Stat3 and Akt in PTEN OE was compared with that in WT. In response to PTEN expression, the level of phospho-Stat3 (p-Stat3) and phospho-Akt (p-Akt) was markedly reduced (Fig. 3A). Dephosphorylation of Stat3 in PTEN OE was confirmed by immunocytochemical staining (Fig. 3B). Consistent with the result shown in Fig. 1B, the expression level of CD133 was also significantly decreased in PTEN-expressing cells (Fig. 3B). To confirm that the dephosphorylation of Stat3 and Akt was induced by PTEN, the phosphorylation status of STAT3 and AKT was examined in PTEN OE following treatment with bpV for 24 h. As shown in Fig. 3C, chemical inhibition of PTEN activity enhanced Stat3 and Akt phosphorylation, suggesting that Stat3 and Akt signaling is directly regulated by PTEN.

Since the expression of PTEN in U-87MG cells were shown to modulate Stat3 and Akt signals, we next examined whether Stat3 signals are regulated by Akt pathway in U-87MG. To do this, U-87MG cells were treated with PI3K inhibitor LY294002 and, unexpectedly, Stat3 was only transiently dephosphorylated and rapidly restored to its original level within 60 min after LY294002-mediated inhibition of Akt activity (Fig. 4A). This result suggests that Stat3 signal is not a downstream target of Akt pathway in U-87MG. Since PTEN expression resulted in the depletion of the GSC population and caused growth retardation and senescence of U-87MG cells, we speculate that these phenomena might be induced by the disturbance of Akt and Stat3 signaling. To test this hypothesis, Akt activity in U-87MG cells was inhibited by treatment with LY294002 and then the extent of senescence, cell growth and neurosphere formation was examined. As expected, LY294002 treatment specifically induced senescence and severely inhibited cell growth (Fig. 4B and C). In addition, neurosphere formation was completely blocked by the treatment of LY294002 (Fig. 4D). Taken together, these results suggest that the senescence, growth retardation, and depletion of GSCs in U-87MG cells are, at least in part, mediated by the Akt signaling pathway.

Constitutive activation of Stat3 is frequently detected in clinical samples from a wide range of human carcinomas and the established cancer cell lines, suggesting that Stat3 activity is crucial to cell survival and growth (32,33). In addition, the JAK/Stat3 signaling pathway is required for the growth of stem-like cancer cells in various tumors (34,35). For these reasons, we decided to examine the effect of inhibition in Stat3 signaling on on cell growth, senescence, and neuro-sphere formation of U-87MG cells. We found that Stat3 was completely dephosphorylated by treatment with 2 μM WB1066, while phosphorylation of AKT was not affected (Fig. 5A). Interestingly, we observed that cellular senescence, which can be estimated by β-gal staining, was not directly induced by inhibiting the Stat3 signaling pathway, although it caused significant growth retardation (Fig. 5B and C). Treatment with WB1066 completely inhibited the ability of U-87MG to form neurospheres, suggesting that the Stat3 signaling pathway is required for maintaining the cancer stem-like cell population (Fig. 5D). These results demonstrate that Stat3 signaling regulates proliferation and maintenance of the cancer stem-like cell population but is not involved in the senescence of U-87MG cells. Since PTEN expression in U-87MG disrupts Akt and Stat3 signals, and that growth retardation, senescence and depletion of cancer stem cells are caused by PTEN expression in U-87MG, we speculate that PTEN induces growth retardation and cancer stem cell depletion of U-87MG in conjunction with both Stat3 andAkt signaling pathways, whereas the senescence response of U-87MG cells upon PTEN expression is mainly regulated by the Akt signaling pathway.

To address whether PTEN is involved in the proliferation of GSCs, PTEN OE or as controls, WT U-87MG cells were implanted into the brain of 8-week-old rats. Strikingly, no tumor-like structures were detected in any of the PTEN OE-implanted group (n=6), whereas robust tumor-like structures were formed in all of the WT-implanted group (n=5) (Fig. 6A and B). Histological analysis of the WT-implanted brains indicated that the tumor-like structures consisted mainly of undifferentiated glioblastoma cells (Fig. 6A; WT 2). In contrast, such cells were not detected in the PTEN OE-implanted brains (Fig. 6B; PTEN OE 4). Furthermore, immunohistochemical staining revealed that many human-specific nuclear antigen (hNu)-positive cells were detected in the WT-implanted brain sections, indicating that U-87MG glioblastoma cells had proliferated extensively after implantation (Fig. 6B; WT). p-Stat3-positive cells were also common in the area of tumor formation (Fig. 6B; WT). In contrast, in the PTEN OE-implanted brains, only a few hNu-positive cells were detected, suggesting that most of the PTEN OE U-87MG glioblastoma cells were not maintained after implantation (Fig. 6B; PTEN OE in the bottom panel). No p-Stat3-positive cells were detected either in the PTEN OE-implanted brains (Fig. 6B; PTEN OE in the bottom panel). Taken together, these results strongly suggest that PTEN negatively regulates the proliferation of GSCs in U-87MG glioblastoma cells, presumably through the inhibition of the Stat3 pathway.

From our study, we found that analysis of the signaling pathways underlying the tumor suppressor function of PTEN in U-87MG cells can provide experimental evidence how the loss of PTEN promotes pathogenesis of glial tumors. In particular, identification of the Stat3 and Akt signaling pathways as downstream targets of PTEN provides a mechanism by which PTEN modulates cell proliferation and senescence and maintenance of the GSC population. Accumulating evidence supports the concept that GSCs, which have stem cell-like properties, are the major cause of the occurrence and maintenance of glioblastoma. In this regard, our results will provide important experimental basis for developing therapeutics to treat glioblastoma in the future.