The tumor suppressor gene PTEN (MMAC1, TEP1) encodes a dual-specificity phosphatase and is considered a progression-associated target of genetic alterations in human gliomas. Recently, it has been reported that the introduction of wild type PTEN into glioma cells containing endogenous mutant PTEN alleles (U87MG, LN-308), but not in those which retain wild-type PTEN (LN-18, LN-229), causes growth suppression and inhibits cellular migration, spreading and focal adhesion. Here, we show that PTEN gene transfer has no effect on the chemosensitivity of the four cell lines. Further, a correlational analysis of the endogenous PTEN status of 12 human glioma cell lines with their sensitivity to seven different cancer chemotherapy drugs reveals no link between PTEN and chemosensitivity. In contrast, ectopic expression of wild type PTEN, but not the PTENG129R mutant, in PTEN-mutant gliomas markedly sensitizes these cells to irradiation and to CD95-ligand (CD95L)-induced apoptosis. PTEN-mediated facilitation of CD95L-induced apoptosis is associated with enhanced CD95L-evoked caspase 3 activity. Protein kinase B (PKB/Akt), previously shown to inhibit CD95L-induced apoptosis in nonglial COS7 cells, is inactivated by dephosphorylation. Interestingly, both PTEN-mutant U87MG and PTEN-wild-type LN-229 cells contain phosphorylated PKB constitutively. Wild-type PTEN gene transfer promotes dephosphorylation of PKB specifically in U87MG cells but not in LN-229 cells. Sensitization of U87MG cells to CD95L-apoptosis by wild-type PTEN is blocked by insulin-like growth factor-1 (IGF-1). The protection by IGF-1 is inhibited by the phosphoinositide 3-OH (PI 3) kinase inhibitor, wortmannin. Although PKB is a down-stream target of PI 3 kinase, the protection by IGF-1 was not associated with the reconstitution of PKB phosphorylation. Thus, PTEN may sensitize human malignant glioma cells to CD95L-induced apoptosis in a PI 3 kinase-dependent manner that may not require PKB phosphorylation.
PTEN is a tumor suppressor gene on human chromosome 10q23 (Li et al., 1997) that is deleted or mutated in about 20 – 40% of glioblastomas but rarely altered in low-grade gliomas (Duerr et al., 1998; Wang et al., 1997). PTEN contains the motif of protein tyrosine phophatases and functions to dephosphorylate both phophotyrosyl and phosphoseryl/threonyl residues and phosphoinositols phosphorylated on the 3′ position of the inositol ring (Li and Sun, 1997; Maehama and Dixon, 1998; Myers et al., 1997). Expression of wild-type PTEN cDNA supresses proliferation in glioma cells with mutant endogenous PTEN alleles but not in glioma cells with a wild-type PTEN background (Furnari et al., 1997). Further, adenovirus-mediated MMAC1/PTEN gene transfer suppresses tumorigenicity of U87MG glioblastoma cells in vivo (Cheney et al., 1998). PTEN has similarity to the cytoskeletal protein tensin. Exogenous expression of PTEN in fibroblasts and a glioma cell line with mutant PTEN alleles results in decreased migration by virtue of decreased integrin-mediated cell spreading and the formation of focal adhesions. These inhibitory properties of PTEN depend on an intact phosphatase domain (Tamura et al., 1998).
Malignant glioma cells show marked resistance to multiple apoptotic stimuli including radiochemotherapy. We have recently correlated the responses of a panel of 12 glioma cell lines to various chemotherapeutic drugs with the differential expression of BCL-2 family proteins and their status of p53, p21, MDM-2, p16, CDK-4 and RB (Weller et al., 1998). Unexpectedly, no molecular predictor of response to chemotherapy was identified. Therefore, other candidate molecules for the regulation of radiochemosensitivity need to be examined. Here, we investigated the effect of PTEN gene transfer in the response of four human glioma cells with different endogenous PTEN gene status to different cancer chemotherapy drugs, to irradiation, and to apoptosis induced by the cytotoxic cytokine, CD95L.
PTEN gene transfer induces growth suppression but not apoptosis in glioma cell lines
PTEN mutations in a set of glioma cell lines have recently been characterized (Furnari et al., 1997). Here, we sought to examine the effect of ectopic PTEN expression on the sensitivity of glioma cells to chemotherapeutics, irradiation or CD95L. We introduced exogenous PTEN genes into four glioma cell lines. LN-18 and LN-229 are wild-type for PTEN and show PTEN protein expression on immunoblots (Figure 1a). U87MG and LN-308 harbor different mutations of the PTEN gene (Furnari et al., 1997). U87MG cells, which carry a deletion of codons 55 – 70 (exon 3) within the tension/auxilin homology region, display a slightly smaller endogenous PTEN protein signal and a correspondingly broader band when full length PTEN alleles are exogenously expressed (Figure 1a). Expression of the ectopic PTEN alleles was verified by immunoblotting for the HA-tag (Figure 1b). Parental and control-transfected LN-308 cells do not show a band corresponding to PTEN due to a truncating mutation in the PTEN gene (Figure 1a, LN-308 pBP), but display a PTEN protein signal when transfected with PTEN or PTENG129R. Introduction of wild-type PTEN but not PTENG129R resulted in marked growth suppression in U87MG and LN-308 cells whereas the growth of LN-18 and LN-229 cells was unaltered by wild-type PTEN (Figure 1c). To differentiate whether the observed effects were due to growth suppression or induction of apoptosis in the wild-type PTEN-transfected U87MG and LN-308 cells, we determined the degree of DNA fragmentation after mock transfection or expression of wild-type PTEN or PTENG129R. There was no increase levels of DNA fragmention above baseline after transfection with either plasmid. In contrast, as a positive control, treatment with teniposide (VM26) resulted in a DNA fragmentation of up to 23% (Figure 1d).
Ectopic PTEN expression does not alter glioma cell sensitivity to chemotherapeutics but enhances radiosensitivity and CD95L-induced apoptosis
Next, we asked whether PTEN may play a role in regulating the chemosensitivity of malignant glioma cell lines. We assessed the modulation by the PTEN status of the cytotoxic and antiproliferative effects of five different chemotherapeutics, vincristine, cytarabine, (VM26), cisplatin and BCNU. Neither transfection with PTEN-HA nor with the mutant allele nor with the pBP control vector resulted in an altered response of any of the cell lines to any of the drugs, as assessed at the level of the EC50 values for growth inhibition in 72 h continous exposure assays. As an example, the data for vincristine are shown in Figure 2 (left panel). Further, we pooled our previously published data on the endogenous PTEN status of 12 human glioma cell lines (Furnari et al., 1997) with their sensitivity to five different cancer chemotherapy drugs (Weller et al., 1998) and determined that there is no difference in the EC50 value for either BCNU, vincristine, cytarabine, teniposide (VM26), and cisplatin in acute cytotoxicity assays between PTEN-wild-type and PTEN-mutant glioma cell lines (data not shown).
In contrast, ectopic expression of wild-type PTEN in cell lines mutant for PTEN (U87MG, LN-308) resulted in a strong sensitization to irradiation, notably in the clinically relevant dose range of 1 – 2 Gy (Figure 2, middle panel). The dose of irradiation required to reduce the cell counts to 37% was lowered by wild-type PTEN from 3.9 to 1.2 Gy in U87MG cells and from 4.5 to 0.8 Gy in LN308 cells. This effect was specific in that wild-type PTEN gene transfer did not alter the sensitivity to irradiation in PTEN wild-type cell lines (LN-18, LN-229) and in that the PTENG129R mutant failed to alter the radiosensitivity of any of the four cell lines.
Next, we treated the glioma cells with CD95L and cycloheximide (CHX), a protein synthesis inhibitor that facilitates CD95L-induced apoptosis (Weller et al., 1994). Interestingly, U87MG and LN-308 glioma cells forced to express wild-type PTEN showed a strong increase of apoptosis compared with vector controls or PTENG129R-expressing cells (Figure 2, right panel). The EC50 values for CD95L-induced apoptosis were shifted by wild-type PTEN from 70 U/ml to 20 U/ml in U87MG cells and from >130 U/ml to 40 U/ml in LN-308 cells. Similar to the results for irradiation shown in Figure 2, wild-type PTEN did not alter the sensitivity of LN-18 and LN-229 to CD95L, and mutant PTENG129R had no effect on CD95L-induced apoptosis.
To evaluate whether the augmentation of CD95L-induced apoptosis was due to facilitation of apoptosis upstream or downstream of caspase 3 activation, we transfected the cell lines with the various PTEN vectors and the determined the caspase 3 activity evoked by CD95L/CHX exposure. Figure 3 shows that there was significantly more caspase 3 activity in response to CD95L/CHX treatment in wild-type PTEN-transfected U87MG and LN-308 cells than in PTENG129R-transfected cells or vector control cells, suggesting that PTEN facilitates the CD95 signaling cascade upstream of caspase 3 activation.
Wild-type PTEN mediates dephosphorylation of PKB in PTEN-mutant glioma cells
Given the phosphatase activity of PTEN, we next asked whether the observed facilitation of radiation- and CD95L-induced apoptosis in U87MG and LN-308 cells was linked to the dephosphorylation of specific molecules. PKB, a serine/threonine protein kinase, has been shown to inhibit CD95L-mediated apoptosis in COS7 cells (Häusler et al., 1998). Further several growth factors like PDGF, EGF or IGF-1 convert PKB from an inactive to an active form in neuronal cells by phosphorylation (Dudek et al., 1997). Therefore, we stimulated mock- or PTEN-transfected U87MG and LN-229 cells with control medium or IGF-1 (25 ng/ml) for 10 min under serum-free conditions, performed immunoprecipitation from whole cell lysates using an anti-PKB antibody, and subjected the lysates to SDS – PAGE and anti-phosphothreonine immunoblotting. Immunoblotting for total PKB protein levels was done in parallel. PKB was expressed in both cell lines, irrespective of endogenous or exogenous PTEN status (Figure 4a). Further, the levels of PKB protein were unaffected by IGF-1.
Next, we examined the state of threonine phosphorylation in both cell lines, depending on the expressed transgene and on the absence or presence of IGF-1. Interestingly, ectopic expression of wild-type PTEN resulted in a loss of phophorylated PKB in U87MG. No such effect was seen with PTENG129R in U87MG cells, and no modulation of PKB phosphorylation was in LN-229 cells with either plasmid. However, IGF-1 did not alter the amounts of phosphorylated PKB under any of these conditions (Figure 4b). Given the failure of IGF-1 to prevent the dephosphorylation of PKB by PTEN, we expected that IGF-1 would also fail to modulate the biological effects of PTEN. This, however, was not the case. As shown in Figure 5 (lower panel), IGF completely prevented the PTEN-mediated sensitization of apoptosis in U87MG cells, without altering induction of apoptosis in mock-transfected U87MG cells or PTEN-transfected LN-229 cells. To determine whether this effect was mediated through PI 3-kinase, we co-exposed the cells to an inhibitor of PI 3-kinase, wortmannin (100 nM), and IGF-1. Wortmannin eliminated the effect of IGF-1 on U87MG expressing wild-type PTEN (Figure 5, lower panel), indicating a role for PI 3-kinase in IGF-1 signaling. There was no effect of wortmannin on CD95L-induced apoptosis in the other sublines (Figure 5).
PTEN is a mutational target in a variety of human glioma, breast, prostate, melanoma and kidney tumor specimens or cell lines (Liu et al., 1997; Rhei et al., 1997; Teng et al., 1997). PTEN has features of a protein tyrosine phophatase which dephosphorylates both phophotyrosyl and phosphoseryl/threonyl residues (Li and Sun, 1997). Expression of wild-type PTEN in glioma cells lacking wild-type PTEN function results in growth suppression (Furnari et al., 1997) and inhibition of migration, spreading and focal adhesion formation (Tamura et al., 1998).
In the present study, we transferred mutant or wild-type PTEN alleles into glioma cell lines with mutant (U87MG, LN-308) or wild-type (LN-18, LN-229) PTEN alleles and observed the growth inhibitory effect of wild-type PTEN in PTEN-mutant cells. Expression of the mutant form of PTEN did not interfere with cell growth, indicating that this mutant was not functional in this paradigm and that there was no dominant-inhibitory activity towards wild-type PTEN activity. Next, we monitored the modulation of chemosensitivity and sensitivity to radiotherapy and CD95L-induced apoptosis. We found that the sensitivity to vincristine, cytarabine, VM26, cisplatin and BCNU was virtually unaltered by wild-type PTEN in all four cell lines, irrespective of their PTEN status. Further, a correlation of the endogenous PTEN status of 12 human glioma cell lines (data from: Furnari et al. (1997)) with their sensitivity to seven different cancer chemotherapy drugs, including BCNU, vincristine, cytarabine, VM26, doxorubicin, camptothecin and β-lapachone (data from: Weller et al., (1998)) revealed no correlation between PTEN and chemosensitivity.
In contrast, irradiation-induced cell death was significantly enhanced by wild-type PTEN gene transfer in glioma cell lines lacking wild-type PTEN (U87MG, LN-308) but not in cell lines retaining wild-type PTEN (LN-18, LN-229) (Figure 2, middle panel). The phosphatase domain was required for radiosensitization since the PTENG129R mutant had no such effect. We have previously characterized CD95L-induced apoptosis as a novel approach of immunochemotherapy to malignant glioma (Roth et al., 1997; Weller et al., 1994). Here, we showed that wild-type PTEN specifically sensitized the PTEN-mutant glioma cell lines, U87MG and LN-308), to CD95L-induced apoptosis (Figure 2). This sensitization operated upstream of caspase 3 activation (Figure 3). These observations led us to compare data on PTEN status, CD95 expression and sensitivity to CD95L in an extended series of glioma cell lines (Table 1). A comparison of the three wild-type PTEN cell lines, LN-18, LN-229, and LN-428, with the eight PTEN mutant cell lines showed that lower levels of CD95 expression were more likely to transmit a death signal when PTEN is wild-type than with mutant PTEN. This was particularly true for the cell lines U138MG, LN-319 and U373MG which were rather resistant to CD95L despite strong CD95 expression.
Activation of PI 3-kinase has recently been shown to inhibit CD95L-induced apoptosis in COS7 cells (Häusler et al., 1998). This effect was hypothesized to be mediated by PKB, a serine/threonine protein kinase that is phophorylated at two major sites, Thr308 in the kinase domain and Ser473 in the carboxy-terminus, and activated thereby (Downward, 1998). PKB provides a survival signal that protects cells from apoptosis induced by various stresses, and the gene has been shown to be amplified in many human tumors (Marte and Downward, 1997). Recently, it has been shown that PTEN acts as phosphoinositide 3-phosphatase that regulates phosphatidylinositol 3,4,5 triphosphate, PI(3,4,5)P3, levels (Maehama and Dixon, 1998). PI(3,4,5)P3 in turn has been shown to directly activate PKB by Thr308 phosphorylation. PKB may act via the p70 S6 kinase to inhibit glycogen synthase kinase-3, the BCL-2-related BAD protein, and other proteins (Marte and Downward, 1997; Crowder and Freeman, 1998; Downward, 1998; Dudek et al., 1997). Thus, PKB is a potential downstream target of PI(3,4,5)P3.
Both PTEN-mutant U87MG cells and PTEN-wild-type LN-229 cells expressed the PKB protein (Figure 3a). Both cell lines contained significant amounts of phosphorylated (active) PKB. The transfer of mutant or wild-type PTEN alleles into U87MG or LN-229 cells did not alter the levels of PKB protein. However, expression of wild-type PTEN in PTEN mutant U87MG resulted in a striking loss of phosphorylated PKB, suggesting that intact PTEN can dephosphorylate PKB on threonine residues and that loss of PTEN activity resulted in a constitutive activation of PKB. Reversal of this phenotype promoted enhanced sensitivity to CD95L-induced apoptosis (Figures 2 and 3). Interestingly, the presence of phosphorylated PKB in untransfected PTEN wild-type LN-229 cells suggested that there were as yet undefined pathways to escape from the PTEN-mediated dephosphorylation of PKB in glioma cells.
In growth factor dependent paradigms of cell survival, PKB is activated through PI 3-kinase (Crowder and Freeman, 1998). IGF/IGF-receptor interaction is one physiological trigger of PKB activation through PI 3-kinase in a variety of cells. We found that IGF abrogated the PTEN-mediated sensitization of glioma cells towards CD95L-induced apoptosis. This action was blocked by wortmannin, suggesting that PI 3-kinase was required for the effect of IGF. Since IGF exerted its effect without preventing the dephosphorylation induced by PTEN, PKB phosphorylation may be dispensable for the cytoprotective effect of IGF in this paradigm. Thus, antagonism of specific actions of wild-type PTEN which promote stimulus-evoked apoptosis in glioma cells might underlie some of effects in the cancerogenesis of gliomas attributed to IGF (Baserga, 1995).
Materials and methods
Transfections were performed using the calcium phosphate technique. Briefly, 10 μg of control pBP (pBP), of mutant PTENG129R-HA (PTENG129R), or of pBP-PTEN-HA (PTEN) (Furnari et al., 1997) were used per 100 mm dish containing cells seeded at 7×105 cells per dish 24 h before transfection. Transfections were terminated at 6 h by rinsing the cells with PBS and supplementing serum-containing medium. At 24 h post transfection, the cells were split at 1 : 3 dilution and maintained for 7 days in puromycin-containing medium (500 ng/ml) at 2% fetal calf serum.
Immunoblot analysis and immunoprecipitation
Immunodetection of the HA-tagged PTEN was done as described previously (Furnari et al., 1997) with an anti-HA antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and chemiluminescence (Amersham, Braunschweig, Germany). For detection of PTEN protein, an anti-PTEN antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) was used.
PKB was detected using cell lysates prepared in the presence of a phosphatase inhibitor (sodium vanadate, Na3VO4, 10 mM), titrated acidic (yellow) with 3 M HCl) with standard lysis buffer. Immunoprecipitation was started by adding α-PKB antibody (PKB/Akt-1, Santa Cruz) (2 – 4 μg/sample) and incubating the reaction for at 4 h at 4°C. According to standard protocols (Santa Cruz Biotechnology 6/1996), protein A/G coupled agarose was added, followed by another 1 – 3 h of incubation at 4°C. The precipitate was centrifuged at 6000 – 8000 r.p.m. and washed carefully six times with sodium vanadate-enriched lysis buffer. Laemmli buffer was added and proteins seperated by standard SDS – PAGE. Immunoblotting was done with a monoclonal anti-phosphothreonine antibody (Santa Cruz) for the detection of threonine phosphorylation. In parallel, an aliquot of the same sample was probed with the anti-PKB antibody to verify precipitation as above. Detection was done using chemiluminescence (Amersham, Braunschweig, Germany).
Viability and apoptosis studies
Cell growth and survival were assessed by crystal violet staining (Weller et al., 1997). Proliferation was assessed after transfection and selection with puromycin by counting of viable cells after 7 days. Results were normalized in terms of percentage of vector transfection with the vector control set 100% in each case (Furnari et al., 1997). CD95L-containing supernatant was obtained from murine CD95L-transfected N2A murine neuroblastoma cells (Roth et al., 1997). Apoptotic cell death was assessed morphologically and measured by quantitative assessment of DNA fragmentation (Weller et al., 1997). All experiments were performed in 2% fetal calf serum containing medium, except for the survival studies with IGF-1. In the studies serum free media were applied before adding the growth factor.
Caspase 3 activity was measured by conversion of the fluorescent substrate, DEVD – AMC (Schulz et al., 1997).
All experiments were performed in triplicate and repeated at least once. Data from representative experiments are presented and expressed as means and SEM. Significance was calculated using the unpaired t-test.
epidermal growth factor
insulin-like growth factor-1
protein kinase B
- PI (3,4,5)P3:
- PI 3-kinase:
phosphoinositide 3-OH kinase
platelet derived growth factor
Baserga R. . 1995 Cancer Res. 55: 249–252.
Cheney I, Johnson D, Vaillancourt M-T, Avanzini J, Morimoto A, Demers G, Wills K, Shabram P, Bolen J, Tavtigian S and Bookstein R. . 1998 Cancer Res. 58: 2331–2334.
Crowder R and Freeman R. . 1998 J. Neurosci. 18: 2933–2943.
Downward J. . 1998 Curr. Biol. 10: 262–267.
Dudek H, Datta S, Franke T, Birnbaum M, Yao R, Cooper G, Segal R, Kaplan D and Greenberg M. . 1997 Science 275: 661–665.
Duerr E-M, Rollbrocker B, Hayashi Y, Peters N, Meyer-Puttlitz B, Louis D, Schramm J, Wiestler O, Parsons R, Eng C and Deimling v A. . 1998 Oncogene 16: 2259–2264.
Furnari F, Lin H, Huang H-J and Cavenee W. . 1997 Proc. Natl. Acad. Sci. USA 94: 12479–12484.
Häusler P, Papoff G, Eramo A, Reif K, Cantrell D and Ruberti G. . 1998 Eur. J. Immunol. 28: 57–69.
Li D-M and Sun H. . 1997 Cancer Res. 57: 2124–2129.
Li J, Yen C, Liaw D, Podsypanina K, Bose S, Wang S, Puc J, Miliaresis C, Rodgers L, McCombie R, Bigner S, Giovanella B, Ittmann M, Tycko B, Hibshoosh H, Wigler M and Parsons R. . 1997 Science 275: 1943–1947.
Liu W, James C, Frederick L, Alderete B and Jenkins R. . 1997 Cancer Res. 57: 5254–5257.
Maehama T and Dixon J. . 1998 J. Biol. Chem. 273: 13375–13378.
Marte B and Downward J. . 1997 Trends Biochem. Sci. 22: 355–358.
Myers MP, Stolarov JP, Eng C, Li J, Wang SI, Wigler MH, Parsons R and Tonks NK. . 1997 Proc. Natl. Acad. Sci. USA 94: 9052–9057.
Rhei E, Kang L, Bolomolniy F, Federici M, Borgen P and Boyd J. . 1997 Cancer Res. 57: 3657–3759.
Rieger J, Naumann U, Glaser T, Ashkenazi A and Weller M. . 1998 FEBS Letts 427: 124–128.
Roth W, Fontana A, Trepel M, Dichgans J, Reed JC and Weller M. . 1997 Cancer Immunol. Immunother. 44: 55–63.
Schulz JB, Bremen D, Reed JC, Lommatzsch J, Takayama S, Wüller U, Löschmann P-A, Klockgether T and Weller M. . 1997 J. Neurochem. 69: 2075–2086.
Tamura M, Gu J, Matsumoto K, Aota S-i, Parsons R and Yamada K. . 1998 Science 280: 1614–1617.
Teng D-F, Hu R, Lin H, Davis T, Iliev D, Frye C, Swedlund B, Hansen K, Vinson V, Gumpper K, Ellis L, El-Nagger A, Frazier M, Jasser S, Langford L, Lee J, Mills G, Pershouse M, Pollack R, Tornos C, Troncoso P, Yung W, Fujii G, Berson A, Bookstein R, Bolen J, Tavtigian S and Steck P. . 1997 Cancer Res. 57: 5221–5225.
Wang S, Puc J, Li J, Bruce J, Cairns P, Sidransky D and Parsons R. . 1997 Cancer Res. 57: 4183–4186.
Weller M, Frei K, Groscurth P, Krammer PH and Yonekawa Y. . 1994 J. Clin. Invest. 94: 954–964.
Weller M, Rieger J, Grimmel C, Van Meir E, De Tribolet N, Krajewski S, Reed J, Von Deimling A and Dichgans J. . 1998 Int. J. Cancer 79: 1–5.
Weller M, Trepel M, Grimmel C, Schabet M, Bremen D, Krajewski S and Reed JC. . 1997 Neurol. Res. 19: 459–470.
Supported by the German Research Foundation (We 1502/5-2) and the IKFZ Tübingen
About this article
Molecular Cell (2019)
Scientific Reports (2019)
Molecular Therapy - Methods & Clinical Development (2018)
Journal of Molecular Biology (2017)