¹Department of Haematology and Oncology, Royal Children's Hospital, Parkville, Australia

²Murdoch Children's Research Institute, Royal Children's Hospital, Parkville, Australia

³Department of Paediatrics, University of Melbourne, Parkville, Australia

⁴Department of Surgery, Royal Melbourne Hospital, University of Melbourne, Parkville, Australia

Correspondence to: C Hawkins, Department of Haematology and Oncology, Royal Children's Hospital, Victoria, Australia 3052. E-mail: hawkinsc@cryptic.rch.unimelb.edu.au

FasL and TNF-related apoptosis-inducing ligand (TRAIL) belong to a subgroup of the TNF superfamily which induce apoptosis by binding to their death domain containing receptors. In the present study we have utilized a panel of seven cell lines derived from human malignant gliomas to characterize molecular pathways through which FasL and TRAIL induce apoptosis in sensitive glioma cells and the mechanisms of resistance in cell lines which survive the death stimuli. Our findings indicate that FADD and Caspase-8 are essential for FasL and TRAIL mediated apoptosis in glioma cells. One sensitive cell line (D270) can be protected from FasL and TRAIL induced death by anti-apoptotic Bcl-2 family members while another (D645) cannot, implying that these lines may represent glioma examples of type II and type I cells respectively. For the first time we demonstrate resistance to FasL but not to TRAIL within the one glioma cell line. Furthermore, we report distinct mechanisms of resistance within different glioma lines, including downregulation of Caspase-8 in U373MG. Cycloheximide sensitized four of the resistant cell lines suggesting the presence of labile inhibitors. None of the known apoptosis inhibitors examined accounted for the observed resistance, suggesting novel inhibitors may exist in glioma cells. Oncogene (2001) 20, 5789-5798.

apoptosis; glioma; resistance; TRAIL; Fas; Bcl-2

CHX, cycloheximide; DD, death domain; DED, death effector domain; DISC, death-inducing signaling complex; PBS, phosphate buffered saline; TNF, tumor necrosis factor; TRAIL, TNF-related apoptosis-inducing ligand

Introduction

Malignant gliomas are the most frequent primary brain tumors in adults and the prognosis of patients remains extremely poor (Hosli et al., 1998). Apoptosis is a highly evolutionarily conserved process fundamental to normal development and homeostasis. Defects in apoptosis regulation are implicated in a variety of human diseases, including cancer (Reed, 1999). It is likely alterations in apoptotic pathways play a role in the tumorigenesis of glioma and resistance to current therapies which induce apoptosis, such as radio- and chemotherapy. This study focuses on the death receptor mediated apoptotic pathways in glioma cells, induced by FasL (CD95L/Apo1L) and Tumor Necrosis Factor-related apoptosis-inducing ligand (TRAIL/Apo2L).

FasL and TRAIL belong to a subgroup of the Tumor Necrosis Factor (TNF) superfamily with the unique ability to induce apoptosis. They do so via their cognate death receptors Fas and DR4 (TRAIL-R1)/DR5 (TRAIL-R2/Trick2) respectively, which are characterized by a cytoplasmic death domain (reviewed by Reed, 2000). FasL and TRAIL can also bind to antagonist 'decoy' receptors. DcR3/TR6/M68 is a secreted soluble Fas decoy receptor which competes with Fas for binding to FasL (Bai et al., 2000; Pitti et al., 1998; Yu et al., 1999). Alternative splicing of the Fas gene can also give rise to mRNAs encoding membrane bound proteins which lack the death domain, such as the Fas Decoy Receptor (FDR) (Jenkins et al., 2000). TRAIL can bind to the decoy receptors DcR1 (TRAIL-R3/Trid/Lit) and DcR2 (TRAIL-R4/TRUNDD) which lack the intracellular death domain and are therefore incapable of inducing the apoptotic pathway. TRAIL decoy receptors are expressed almost exclusively in normal cells (reviewed by Reed 2000).

Studies in lymphoid cells suggest FasL and TRAIL engage similar apoptotic pathways in these cells. Aggregation of the death receptors by multimerized ligand induces the formation of a death-inducing signaling complex (DISC) (Kischkel et al., 1995; Sprick et al., 2000), involving the receptors' death domains (DDs), the adaptor molecule FADD (Mort-1) and Caspase-8 (FLICE, Mch5, MACH), which interact via their death effector domains (DEDs).

The subsequent activation of Caspase-8 at the Fas DISC has been shown to propagate alternate apoptotic pathways (Scaffidi et al., 1998). In type I cells, active Caspase-8 directly cleaves downstream effector caspases, such as Caspase-3, which in turn cleave an array of substrates, leading to the systematic destruction of the cell (Wolf and Green, 1999). However in type II cells, DISC formation is less efficient and mitochondrial amplification of the death receptor induced signal is required for apoptosis (Scaffidi et al., 1999b). Active Caspase-8 cleaves the BH3-only Bcl-2 family member Bid and truncated Bid (tBID) translocates to the mitochondria (Li et al., 1998). This results in the release of cytochrome-c, which binds to Apaf-1 and recruits Caspase-9 to form the 'apoptosome' (Luo et al., 1998; Zou et al., 1999). This complex is involved in the 'intrinsic' apoptotic pathway, which is triggered by serum withdrawal (Zhang et al., 2000b) and DNA damage (Coultas and Strasser, 2000). The Bcl-2 protein family regulates apoptotic signals involving the mitochondria (reviewed by Gross et al., 1999).

Seven glioma cell lines were used in this study and initially characterized for their sensitivity to FasL, TRAIL and cisplatin. The death receptor apoptotic signal transduction pathways used by sensitive cells and the mechanisms employed by resistant cells to elude apoptosis were investigated.

Results

Glioma cell lines differ in their susceptibility to death stimuli

A panel of seven glioma cell lines was characterized for FasL and TRAIL sensitivity (refer to Table 1 and Discussion) using the sensitive lymphoid cell line Jurkat as a control. Two distinct methods were used to assess cell viability with comparable results. The glioma cell lines displayed differential sensitivity to each of the death stimuli. D645, D270 and U118MG were sensitive to FasL, with the remainder being FasL resistant (Figure 1a,b). D645 was also very sensitive to TRAIL induced apoptosis. U118MG, D270 and U87MG displayed partial sensitivity to TRAIL, while U251MG and U373MG were resistant (Figure 1d,e). D54 was completely resistant to FasL, but sensitive to TRAIL. Death receptor mediated apoptosis has been shown to converge downstream with the 'intrinsic' apoptotic pathway which can be induced by chemotherapy drugs. To examine sensitivity to such an apoptotic stimulus, cells were treated with the chemotherapeutic agent cisplatin (Figure 1g,h). D270 was particularly sensitive to cisplatin with a second line, U118MG displaying intermediate sensitivity. D645, the most sensitive cell line to FasL and TRAIL, was resistant to cisplatin. U373MG and U251MG were resistant to all three death stimuli.

Caspase-3 activity correlates with sensitivity

Apoptosis triggered by all three agents studied has been demonstrated in other cell types to involve activation of caspases. The tetrapeptide DEVD is cleaved by effector caspases such as Caspase-3. Cleavage of the fluorogenic substrate DEVD-AMC was used to quantify activated Caspase-3 (and similar caspases) in each of the cell lines following treatment with FasL and TRAIL. Figure 1c,f demonstrate that the amount of Caspase-3-like activity correlates with the susceptibility of the cells to apoptosis. Absolute DEVD-ase activity was less in cisplatin treated cells, but the relative levels of Caspase-3-like activity correlated with sensitivity (Figure 1i).

Receptor expression does not correlate with sensitivity or resistance

Flow cytometry and immunoblotting were used to survey the expression of death and decoy receptors in the glioma cells. Fas was expressed on the surface of all cells (Figure 2a) and in each line the protein was full length (Figure 2b). The FasL decoy receptor DcR3 was not detected in lysates (Figure 6a) or conditioned media (data not shown) of any of the cell lines.

Mutations of Fas in melanoma cells have been described (Shin et al., 1999). Given the differential sensitivity of D54 to FasL and TRAIL, the Fas coding region from D54 was sequenced. It was found to be wild type, thus ruling out the possibility of a point mutation or small insertion/deletion of the death receptor being responsible for the resistance of D54. D54 was atypical with respect to TRAIL receptor expression, bearing high levels of DR5 and DR4 and intermediate levels of DcR2 (Figure 2c,d). DcR2 was also expressed in U87MG cells, and all cell lines expressed moderate levels of DR5 on their surface. DcR1 was non-detectable by immunoblotting (data not shown).

Elucidation of pathway components

Immunoblotting was used to evaluate the levels of the candidate FasL and TRAIL signaling molecules FADD, Caspase-8 and Caspase-3. FADD and Caspase-8 have been implicated as components of the FasL and TRAIL DISC. Formation of the DISC results in activation of effector proteases such as Caspase-3. All cell lines express similar levels of Caspase-3 (Figure 3a), however U373MG expressed very low levels of FADD and Caspase-8, as discussed below. As the sensitive lines expressed FADD and Caspase-8 we investigated whether these molecules are required for FasL and TRAIL apoptotic signaling in sensitive cells by introducing inhibitors. Transfection of a dominant negative mutant of FADD (FADD DN) or CrmA, a viral inhibitor of Caspase-8, abolished the apoptotic signals of both FasL and TRAIL in D645 and D270 cells (Figure 4a,b), suggesting that FADD and Caspase-8 are crucial for FasL and TRAIL signaling in glioma cells. As expected, FADD DN and CrmA transfected D270 cells were killed by cisplatin treatment as efficiently as control transfected cells (Figure 4a).

Transfection of the anti-apoptotic proteins Bcl-2 and Bcl-xL into D270 cells abolished cisplatin induced apoptosis (Figure 4a), a stimulus known to utilize the mitochondrial 'intrinsic' pathway. Interestingly, transfection of Bcl-2 and Bcl-x_L into D270 cells completely inhibited FasL and TRAIL induced apoptosis, while D645 cells remained sensitive after similar transfections (Figure 4b). This suggests the Bcl-2 family of proteins can regulate signals emanating from death receptors in some glioma cells. Expression levels of Bcl-2 family members may therefore account for the resistance to these stimuli exhibited by other glioma lines, so endogenous levels of Bcl-2 family members were examined by immunoblotting. However, resistant lines did not express significantly higher levels of the anti-apoptotic family members Bcl-2 and Bcl-x_L (Figure 3b), arguing against this hypothesis. Both D270 and D54 cell lines express large amounts of the pro-apoptotic protein Bid. D270 cells express very low levels of Bcl-x_L and Bad while the TRAIL sensitive cell line D54 expressed very low levels of Bcl-2.

Different mechanisms of resistance

There are two general hypothetical mechanisms for resistance to apoptotic stimuli: either a cell has a defective or underexpressed apoptotic pathway component, or it contains an inhibitor which blocks that pathway. To distinguish between these possibilities, apoptosis assays were performed in the presence of the potent translation inhibitor cycloheximide (CHX) (Figure 5). HeLa cells were used as a control as they have previously been shown to become sensitive to FasL following CHX treatment (Somia et al., 1999). Inhibition of translation by CHX substantially increased the sensitivity of D270, U118MG and U87MG lines to both TRAIL and FasL induced apoptosis after 6 h of incubation. Interestingly, D54 was dramatically sensitized to FasL-induced death following exposure to CHX. This finding suggests that D54 may express a labile inhibitor which specifically blocks FasL-induced apoptosis. U251MG and U373MG, which are resistant to both FasL and TRAIL, remain resistant in the presence of CHX, suggesting that their resistance is due to either a defective death pathway(s), or the presence of an inhibitor(s) with a longer half-life. This is consistent with the above finding that U373MG cells express low levels of FADD and Caspase-8. Sequencing of Caspase-8 and FADD in these resistant cell lines revealed them to be wild-type. Inhibition of translation by CHX did not sensitize any of the lines to cisplatin.

Analysis of expression of candidate inhibitors

Figure 6a,b show expression levels of previously identified FasL or TRAIL inhibitors in the glioma cell lines. Neither splice variant of the Caspase-8 dominant negative inhibitor cFLIP (long or short) was detected in the lysates of any of the glioma cell lines, thus being excluded as a possible mechanism of resistance. Similarly, no expression of the Fas specific candidate inhibitors DcR3 or LFG mRNA was detected. The expression levels of the caspase inhibitor XIAP and PEA-15 mRNA did not differ significantly between the different cell lines.

Sensitizing resistant cell lines

Rendering resistant cell lines sensitive to apoptosis is a means to confirm the proposed mechanism of resistance. The results described above suggest the glioma cell line U373MG is resistant to both TRAIL and FasL induced apoptosis due to low levels of the crucial signaling components FADD and/or Caspase-8. Figure 7 demonstrates that enforced expression of Caspase-8 into U373MG sensitized the cells equally to FasL and TRAIL mediated apoptosis. Similar enforced expression of FADD had no effect (data not shown).

Discussion

Previous studies in glioma cell lines have shown that differential sensitivities exist between the tumor cell lines to both TRAIL and FasL induced apoptosis (Rieger et al., 1998) and the data presented here confirm these findings. In addition, we demonstrate that glioma cell lines potentially utilize differential apoptotic pathways and exhibit a number of mechanisms of resistance to death receptor induced apoptosis, summarized in Table 1.

Apoptotic pathway components

Although the glioma cells vary in their susceptibility, FasL and TRAIL appear to utilize similar apoptotic pathway components. The present study agrees with previous findings that both Caspase-8 and FADD are essential for Fas (Juo et al., 1998, 1999) and DR5 mediated apoptosis (Sprick et al., 2000), and Caspase-3-like caspases are important effector caspases for both death stimuli (Wrone-Smith et al., 2001).

The Bcl-2 family of proteins plays a more controversial role in regulating death receptor mediated apoptosis. Reports of the inhibitory capacity of anti-apoptotic Bcl-2 family members on FasL and TRAIL killing range from no effect (Huang et al., 1999; Kim et al., 2001), to substantial effect (Hinz et al., 2000; Scaffidi et al., 1998). These contrasting results have been attributed variously to differences in the cells used and the experimental agent used to aggregate the death receptors.

The sensitivity of D270 cells to cisplatin, FasL and to a lesser extent TRAIL, may reflect the endogenous protein expression levels of the Bcl-2 family members, expressing high levels of the pro-apoptotic Bid and low levels of the anti-apoptotic Bcl-x_L. Enforced expression of the anti-apoptotic proteins Bcl-2 and Bcl-x_L into D270 cells rendered them resistant to apoptosis, suggesting death receptor mediated apoptosis is routed via the mitochondria in these cells, as in type II cells. D645 cells remained sensitive to FasL and TRAIL in the presence of Bcl-2 and Bcl-x_L, which may indicate that they are type I cells. The fact that glioma cell lines may be either type I or type II cells and thus differentially regulated is currently being investigated further. This observation that the glioma cell lines may employ alternate apoptotic pathways complicates the determination of resistance mechanisms. However, both cell types utilize some common pathway components: receptors, DISC and downstream effector caspases. Therefore mutations or inhibitors of any of these signaling components could render the cells resistant to apoptosis.

Mechanisms of resistance

The most proximal step to suppress a death receptor pathway is inhibition of ligand binding. This could be achieved by lack of or mutations in death receptors or the presence of decoy receptors. Prior reports using glioma cell lines have examined TRAIL receptor mRNA levels by RT-PCR (Rieger et al., 1998) and Northern blot (Wu et al., 2000). However, decoy receptor localization studies in melanoma cells suggest that mRNA levels of the receptors do not correlate with surface protein expression (Zhang et al., 2000a). Therefore in the present study, we determined total protein levels of the receptors and surface expression levels.

Our findings suggest there is no correlation between TRAIL receptor surface expression and susceptibility to TRAIL induced apoptosis in glioma cell lines. This is in accordance with some previous observations in various tumor types (Keane et al., 1999; Tschopp et al., 1998; Wen et al., 2000) but not with others (Zhang et al., 1999). All cell lines expressed the highest affinity death receptor DR5 (Truneh et al., 2000) on their surface, with D54 having the highest levels of expression. D54 is also the only cell line with detectable levels of DR4. While this cell line is susceptible to TRAIL, D645 is the most sensitive and has similar DR5 surface expression levels to the remaining cell lines. In the present study, the levels of decoy receptor expression were very low in all cell lines and would therefore not account for TRAIL resistance.

Expression of Fas on the surface of glioma cells has been reviewed elsewhere (Roth and Weller, 1999). In the present study, no correlation was found between Fas surface expression and susceptibility to FasL induced apoptosis. DcR3 is a secreted soluble protein which competes for binding to Fas with an equal affinity to FasL, inhibiting apoptosis (Yu et al., 1999). DcR3 gene amplification has been implicated in human lung and colon carcinoma (Pitti et al., 1998) and its protein is over-expressed in gastro-intestinal tract tumors (Bai et al., 2000). Using an antibody which readily detects DcR3 protein in 293T cells transfected with a DcR3 expression plasmid, it was not possible to detect DcR3 in the supernatants (data not shown) or lysates of any of the glioma lines (Figure 6a). Surprisingly, detection of DcR3 protein in concentrated supernatants of U87MG and U373MG cells has recently been published (Roth et al., 2001). It is therefore possible, assuming that the antibody in those experiments did not detect residual bovine DcR3 from the culture medium, that DcR3 is expressed in these lines at levels below our detection threshold.

A further downstream level at which apoptosis can be inhibited is that of the apoptotic signaling components. The U373MG cell line appears to be resistant to death receptor mediated apoptosis due to lack of crucial signaling components, and expression of Caspase-8 sensitized this cell line to FasL and TRAIL mediated apoptosis, while overexpression of FADD did not. Many recent studies have reported methylation of the Caspase-8 gene as a mechanism for decreased levels of protein expression in neuroblastomas, rendering cells resistant to apoptosis (Hopkins-Donaldson et al., 2000; Teitz et al., 2000). In U373MG cells however, demethylation with 5-aza-2'-deoxycytidine did not increase Caspase-8 expression (data not shown), implying that methylation of the Caspase-8 gene is not responsible for its low expression levels. U251MG was similarly resistant to FasL and TRAIL. The lack of sensitization by CHX would imply that its resistance is due to a mutated component or the expression of an inhibitor with a long half-life.

The presence of anti-apoptotic intracellular proteins which interfere with signaling is another potential mechanism for tumor cell resistance. Similarly to previous reports in glioma cell lines (Hao et al., 2001; Rieger et al., 1998; Wu et al., 2000), inhibition of protein translation by CHX dramatically sensitized four of the glioma cell lines to TRAIL and FasL induced death. This suggests the presence of a labile inhibitor in these lines and expression levels of a number of known inhibitors were examined. cFLIP is a dominant negative form of Caspase-8 which lacks residues important for substrate catalysis rendering it devoid of proteolytic activity. Levels of cFLIP_L have been correlated with resistance to FasL in lymphoid (Perlman et al., 1999) and melanoma cells (Irmler et al., 1997) and to TRAIL induced apoptosis in melanoma cells (Griffith et al., 1998) and various tumor types (Kim et al., 2000; Leverkus et al., 2000). However, none of the glioma cell lines in the present study expressed cFLIP.

PEA-15/PED is a DED containing phosphoprotein expressed in brain which protects astrocytes from TNF- induced apoptosis (Kitsberg et al., 1999), through interruption of FADD-Caspase-8 binding (Condorelli et al., 1999). Twofold higher levels of PEA-15 have been implicated recently in the resistance of some glioma lines, including U373MG (Hao et al., 2001). While our attempts to obtain an antibody to evaluate PEA-15 protein levels have been unsuccessful, RNA analysis suggests equal PEA-15 transcript levels in all cell lines. This, coupled with the inability of overexpression of PEA-15 to confer resistance on the sensitive lines (data not shown), suggests that PEA-15 does not account for the resistance of the glioma cell lines used in the present study. The U373MG cells used in this study differ from those used by Hao et al. (2001) (it appears that our cells may express lower levels of FADD and Caspase-8). It is therefore possible that the apparent discrepancies are due to the analysis of different sublines of U373MG. The ATCC catalog warns that mislabeling of U251MG cells as U373MG cells may have occurred in the past. Our U251MG and U373MG lines are morphologically distinct, but this confusion may partially explain the differences observed between this study and that of Hao et al. (2001)

The fact that D54 is remarkably sensitized to FasL while being sensitive to TRAIL in the absence of CHX suggests the presence of a labile inhibitor in D54 that is specific to the Fas apoptotic pathway. LFG is an intracellular protein which inhibits Fas but not TNF- or TRAIL mediated apoptosis and could therefore account for this differential sensitivity (Somia et al., 1999), however mRNA for LFG was not detected in any of the glioma cell lines. Further described mechanisms of resistance to Fas mediated apoptosis include Fas-associated phosphotase-1 (FAP-1) overexpression in tumors (Myc et al., 1999; Sato et al., 1995; Yanagisawa et al., 1997) and Toso, a membrane bound inhibitor of the Fas pathway confined to lymphoid cells (Hitoshi et al., 1998).

Members of the IAP family of apoptosis inhibitors have been shown to directly inhibit effector Caspases-3, 7 and 9 and to inhibit various cell death stimuli. The inhibition of Caspase-3 by XIAP (K_i = 0.7 nM) (Deveraux et al., 1997) probably reflects physiological inhibitory potential. All the cell lines expressed similar levels of XIAP suggesting this is not responsible for the differential sensitivities.

This study characterized the differential sensitivity of glioma cell lines to FasL and TRAIL, demonstrating that sensitive glioma cells can exhibit characteristics of type I or type II cells. It is important to elucidate the varying methods glioma cells have adopted to elude apoptosis induced by both conventional treatments and death receptor ligands, as this may have important implications for future multi-modal treatments. A number of hypotheses for resistance based on mechanisms of resistance previously described in other cell types were examined. We have identified that glioma cells exhibit varying mechanisms of resistance to the death receptor apoptotic pathway, including downregulation of Caspase-8. None of the previously reported apoptosis inhibitors or resistance mechanisms examined here account for the resistance of the glioma cells to FasL or TRAIL mediated apoptosis, suggesting the resistance mechanisms may be novel or distinct in glioma cells.

Materials and methods

Cell lines and plasmids

The 'U' series of glioma cell lines was established at the Wallenbery laboratory in Uppsala, Sweden, by Ponten and Westermark and was obtained from ATCC. The 'D' series was obtained from the Duke University Medical Center, USA (Bigner et al., 1981). U87MG cells were maintained in Zinc Option Media (Life Technologies) with 10% fetal bovine serum (FBS). Other cell lines were maintained in DMEM with 10% FBS.

The plasmid CMV-lacZ has been previously described (Hawkins et al., 1996). A mammalian expression plasmid with an amino terminal FLAG epitope tag (F-pIRES) was constructed by annealing oligonucleotides 1 and 2 and ligating into NotI + EcoRI digested pIRESneo (Clontech). FpIRES-Casp8 was generated by excising Caspase-8 from Casp8-pCUP1-(LEU2) (Hawkins et al., 1999) and inserting into FpIRES. Oligonucleotides 3 and 4 were annealed and ligated into FpIRES to give FpIRES-LZ. The extracellular region of human TRAIL was amplified using oligonucleotides 5 and 6 and cloned into F-pIRES-LZ to give FpIRES-LZ-TRAIL, the sequence of which was verified by automated DNA sequencing. The CrmA, Bcl-2 and FADD dominant negative expression plasmids for apoptosis inhibition studies have been previously published (Newton et al., 1998; Strasser et al., 1995). pEF-FLAG-Bcl-x_L was generated by PCR cloning and verified by sequencing. The coding region of Bcl-x_L was amplified using oligonucleotides 7 and 8 and the product digested with BglII and XbaI prior to insertion into BamHI-XbaI cut pEF-FLAG-FLIP_L puro (Scaffidi et al., 1999a).

Oligonucleotides used in the construction of the plasmids described above were:
1: GGCCACCATGGACTACAAGGACGACGATGACAAGGCGGCC
GCAGCTAGCG;
2: AATTCGCTAGCTGCGGCCGCCTTGTCATCGTCGTCCTTGT
AGTCCATGGT;
3: GGCCGCCATGAAACAGATCGAAGACA AAATAGAG GAGATCCTTAGCAA GATCTACCATATAGAAAACGAGATAGCTCGTATCAAAAAG
CTTATTGGTGAAG;
4: TTCTTCACCAATAAGCTTTTTGATACGAGCTATCTCGTTT
TCTATATGGTAGATCTTGCTAAGG ATCTCCTCTATTTT GTCTTCGATCTGTTTCATGGC;
5: GGAATTCGTGAGAGAAAGAGGTCC;
6: GCGGATCCTTAGCCAACTAAAAAGGCCCCG;
7: GCAGATCTATGTCTCAGAGCAACCGGG;
8: GCTCTAGACTATTTCCGACTGAAGAGTG.

Antibodies

The following antibodies were used: anti-TRAIL-R1, R2, R3, R4, Bcl-x_L, Bad and XIAP (R&D Systems); anti-Fas DX2, FADD A662 and Caspase-3 (Pharmingen); anti-DcR3 TR6 (Prosci); anti-Caspase-8 (Alexis); anti-Bid (Cell Signaling Technology); anti-cFLIP NF6 (a kind gift from Marcus Peter); anti-Bcl-2 (Oncogene); anti-FLAG M2 and anti-mouse Ig-HRP (Sigma); anti-goat IgG HRP, anti-rabbit HRP and anti-goat IgG FITC (Pierce); anti-mouse Ig-FITC (Silenus). Antibody specificity was verified by Western blot analysis of transfected 293T cell lysates or by flow cytometric analysis of transfected 293Ts (data not shown).

Generation of aggregated FasL and TRAIL

Supernatant from the Neuro2A-CD95L cell line (Schneider et al., 1998) was the source of active aggregated FasL, and was routinely used at a one in 10 dilution. A one in 20 dilution was used for inhibitor transfection experiments. Supernatant from an empty vector control stable Neuro2A line was used for a negative control.

Inclusion of both the FLAG and leucine zipper tags increased TRAIL cytotoxicity following anti-FLAG antibody treatment relative to either tag alone (data not shown). FpIRES-LZ-TRAIL was transiently transfected into 293T cells then 4 days later the media was removed and centrifuged at 13 k.r.p.m. to remove remaining cells. The supernatant (verified to contain F-LZ-TRAIL by Western blot) was diluted one in three and pre-incubated with anti-FLAG antibody 1 g/ml immediately prior to use in apoptosis and caspase activity assays.

Measurement of apoptosis

Cells were plated in triplicate and allowed to adhere overnight prior to experiments.

MTT assay: 5 ´ 10³ cells well^-1 were plated in a 96-well plate. Following treatment, MTT stock solution (Cell Titer 96^Ò Aqueous, Promega) was added to the medium and incubated with cells at 37°C for 1 h to allow cell-mediated reduction of MTT. Absorbance was measured at 495 nm.

Propidium iodide uptake: 3 ´ 10⁴ cells well^-1 were plated in a 24-well plate. Following treatment, cells were harvested and resuspended in buffer (1% BSA, 0.05% Na₃Az in PBS) containing 25 g/ml propidium iodide (Sigma) and analysed by flow cytometry (Nicoletti et al., 1991).

Annexin-V binding: 3 ´ 10⁴ cells well^-1 were plated in a 24-well plate. Following treatment, cells were harvested and resuspended in buffer (10 mM HEPES, 140 mM NaCl, 5 mM CaCl₂, pH 7.4) containing 1 : 100 Annexin-V-Alexa 568 (Roche) and analyzed by flow cytometry. Cycloheximide (CHX) (10 g/ml) was added simultaneously with the death stimuli or with control media.

Determination of Caspase activity

Cells were treated with FasL and TRAIL for 6 h and cisplatin 10 g/ml for 12 h. The cells were washed, lysed in buffer A (Miller et al., 1997) and the protein concentration measured. Lysate (30 l) containing 25 g of protein was mixed with 70 l of the fluorogenic caspase substrate DEVD-AMC (Calbiochem) in buffer B (Miller et al., 1997), which is cleaved by Type II caspases such as Caspase-3. The DEVD-ase activity in the extracts was measured fluorometrically at 508 nm for a 2 h period.

Western blots

Cells were lysed in triple lysis buffer (Sambrook et al., 1989) then sonicated to shear DNA. Protein quantitation was performed using Protein Assay Dye Reagent Concentrate (Biorad) and 50 g of protein was loaded on 12% SDS-PAGE gels. Equal loading was confirmed by coomassie staining (data not shown). The proteins were transferred to Hybond-P membrane (Amersham), blocked in 1% Blocking Reagent (Roche) in PBS and probed with the indicated antibodies in 1% Blocking Reagent in PBS with 0.05% Tween 20. HRP conjugated antibodies were detected by chemiluminescence using Super Signal^Ò West Dura Extended Duration Substrate (Pierce).

Surface protein expression

Surface expression of TRAIL receptors and Fas was determined by flow cytometric analysis by measuring the binding of anti-TRAIL receptor and anti-Fas antibodies. One ´ 10⁵ cells were incubated in 100 l with 1 g/ml of primary antibody (TRAIL receptor, Fas or control) in buffer (1% BSA, 0.05% Na₃Az in PBS) for 30 min at room temperature. The cells were washed twice with buffer, before incubation with secondary antibody for 30 min at room temperature. After further washing, the cells were resuspended in 200 l of buffer and analysed by flow cytometry. The mean fluorescence ratio (MFR) of cells labeled with non-immune control antibody and cells labeled with anti-receptor antibody was calculated. The difference between the percentage of receptor-labeled and isotype control-labeled cells above the point of intersection of the two profiles defines the percentage of cells positive for receptor expression.

Northern blots

Total RNA was isolated from the glioma cell lines (Chomczynski and Sacchi, 1987). Positive control RNA was isolated from 293T cells transiently transfected with F-pIRES LFG or PEA-15. Ten micrograms of RNA was resolved on a 1% agarose-formaldehyde gel and transferred to Hybond-N membrane (Amersham) overnight. ³²P-labeled (Gigaprime, Geneworks) cDNA was hybridized to the membrane overnight at 68°C with Express Hyb (Clontech). The membrane was washed twice with 2 ´ SSC, 0.1% SDS at 50°C and bands visualized by autoradiography. Equal loading was confirmed by the 28S ribosomal band intensity which has been shown to be a more consistent method than GAPDH probing (de Leeuw et al., 1989).

Transfection studies

One ´ 10⁵ cells well^-1 were seeded in 24-well plates and allowed to adhere overnight. Cells were transfected using Lipofectamine 2000 (Life Technologies) according to manufacturer's instructions. Triplicate coded wells were transfected with 0.1 g of the reporter plasmid CMV-lacZ (Hawkins et al., 1996) and 0.9 g of expression plasmid bearing the gene of interest or no insert. The following day, the media was replaced with media containing supernatants from empty vector, FasL or F-LZ-TRAIL transfectants (+1 g/ml M2 antibody for TRAIL) for 6 h, or cisplatin for 24 h. The death stimuli were used at a concentration required to kill approximately 60% of cells at the above timepoints (FasL 1 : 20, F-LZ-TRAIL 1 : 3 and cisplatin 3 g/ml). The cells were fixed (2% formamide, 2% gluteraldehyde in PBS) for 5 min and stained overnight (0.1% X-gal, 5 mM potassium ferricyanide, 5 mM potassium ferrocyanide, 2 mM MgCl₂ in PBS). Blue cells were scored for apoptosis using morphological criteria as previously described (Miura et al., 1993). To examine the effect of expression of Caspase-8 on FasL and TRAIL sensitivity of U373MG cells, the media was replaced 6 h following transfection and 24 h later cells were fixed, stained and scored as above.

Acknowledgements

The authors acknowledge the advice and generous donations of reagents from the following people: Drs David Vaux, David Huang, Andreas Strasser, Paul Ekert, Marcus Peter, Andriano Fontano, Darell Bigner, Mark Schmitt and also Janna Stickland for her help with the preparation of Figures. This study was funded by a James S McDonnell Foundation Program Grant and a National Health and Medical Research Council Project Grant.

Bai C, Connolly B, Metzker ML, Hilliard CA, Liu X, Sandig V, Soderman A, Galloway SM, Liu Q, Austin CP, Caskey CT. (2000). Proc. Natl. Acad. Sci. USA 97, 1230-1235. MEDLINE

Bigner D, Bigner S, Ponten J, Westermark B, Mahaley M, Ruoslahti E, Herschman H, Eng L, Wikstrand C. (1981). J. Neuropathol. Exp. Neurol. 40, 201-229. MEDLINE

Chomczynski P, Sacchi N. (1987). Anal. Biochem. 162, 156-159. Article MEDLINE

Condorelli G, Vigliotta G, Cafieri A, Trencia A, Andolo P, Oriente F, Miele C, Caruso M, Formisano P, Beguinot F. (1999). Oncogene 18, 4409-4415. MEDLINE

Coultas L, Strasser A. (2000). Apoptosis 5, 491-507. Article MEDLINE

de Leeuw WJ, Slagboom PE, Vijg J. (1989). Nucleic Acids Res. 17, 10137-10138. MEDLINE

Deveraux Q, Takahashi R, Salvesen G, Reed J. (1997). Nature 388, 300-304. Article MEDLINE

Griffith TS, Chin WA, Jackson GC, Lynch DH, Kubin MZ. (1998). J. Immunol. 161, 2833-2840. MEDLINE

Gross A, McDonnell J, Korsmeyer S. (1999). Genes Dev. 13, 1899-1911. MEDLINE

Hao C, Beguinot F, Condorelli G, Trencia A, Van Meir EG, Yong VW, Parney IF, Roa WH, Petruk KC. (2001). Cancer Res. 61, 1162-1170. MEDLINE

Hawkins CJ, Uren AG, Hacker G, Medcalf RL, Vaux DL. (1996). Proc. Natl. Acad. Sci. USA 93, 13786-13790. Article MEDLINE

Hawkins CJ, Wang SL, Hay BA. (1999). Proc. Natl. Acad. Sci. USA 96, 2885-2890. Article MEDLINE

Hinz S, Trauzold A, Boenicke L, Sandberg C, Beckmann S, Bayer E, Walczak H, Kalthoff H, Ungefroren H. (2000). Oncogene 19, 5477-5486. MEDLINE

Hitoshi Y, Lorens J, Kitada SI, Fisher J, LaBarge M, Ring HZ, Francke U, Reed JC, Kinoshita S, Nolan GP. (1998). Immunity 8, 461-471. MEDLINE

Hopkins-Donaldson S, Bodmer JL, Bourloud KB, Brognara CB, Tschopp J, Gross N. (2000). Cancer Res. 60, 4315-4319. MEDLINE

Hosli P, Sappino A, de Tribolet N, Dietrich P. (1998). Ann. Oncol. 9, 589-600. MEDLINE

Huang D, Hahne M, Schroeter M, Frei K, Fontana A, Villunger A, Newton K, Tschopp J, Strasser A. (1999). Proc. Natl. Acad. Sci. USA 96, 14871-14876. Article MEDLINE

Irmler M, Thome M, Hahne M, Schneider P, Hofmann B, Steiner V, Bodmer JL, Schroter M, Burns K, Mattmann C, Rimoldi D, French LE, Tschopp J. (1997). Nature 388, 190-195. Article MEDLINE

Jenkins M, Keir M, McCune JM. (2000). J. Biol. Chem. 275, 7988-7993. Article MEDLINE

Juo P, Kuo CJ, Yuan J, Blenis J. (1998). Curr. Biol. 8, 1001-1008. MEDLINE

Juo P, Woo MS, Kuo CJ, Signorelli P, Biemann HP, Hannun YA, Blenis J. (1999). Cell Growth Differ. 10, 797-804. MEDLINE

Keane MM, Ettenberg SA, Nau MM, Russell EK, Lipkowitz S. (1999). Cancer Res. 59, 734-741. MEDLINE

Kim EJ, Suliman A, Lam A, Srivastava RK. (2001). Int. J. Oncol. 18, 187-194. MEDLINE

Kim K, Fisher MJ, Xu SQ, el-Deiry WS. (2000). Clin. Cancer Res. 6, 335-346. MEDLINE

Kischkel FC, Hellbardt S, Behrmann I, Germer M, Pawlita M, Krammer PH, Peter ME. (1995). EMBO J. 14, 5579-5588. MEDLINE

Kitsberg D, Formstecher E, Fauquet M, Kubes M, Cordier J, Canton B, Pan G, Rolli M, Glowinski J, Chneiweiss H. (1999). J. Neurosci. 19, 8244-8251. MEDLINE

Leverkus M, Neumann M, Mengling T, Rauch CT, Brocker EB, Krammer PH, Walczak H. (2000). Cancer Res. 60, 553-559. MEDLINE

Li H, Zhu H, Xu C, Yuan J. (1998). Cell 94, 491-501. MEDLINE

Luo X, Budihardjo I, Zou H, Slaughter C, Wang X. (1998). Cell 94, 481-490. MEDLINE

Miller TM, Moulder KL, Knudson CM, Creedon DJ, Deshmukh M, Korsmeyer SJ, Johnson EM. (1997). J. Cell Biol. 139, 205-217. MEDLINE

Miura M, Zhu H, Rotello R, Hartweig EA, Yuan J. (1993). Cell 75, 653-660. MEDLINE

Myc A, Arscott PL, Bretz JD, Thompson NW, Baker JR. (1999). Endocrinol. 140, 5431-5434.

Newton K, Harris AW, Bath ML, Smith KG, Strasser A. (1998). EMBO J. 17, 706-718. Article MEDLINE

Nicoletti I, Migliorati G, Pagliacci MC, Grignani F, Riccardi C. (1991). J. Immunol. Meth. 139, 271-279.

Perlman H, Pagliari LJ, Georganas C, Mano T, Walsh K, Pope RM. (1999). J. Exp. Med. 190, 1679-1688. MEDLINE

Pitti RM, Marsters SA, Lawrence DA, Roy M, Kischkel FC, Dowd P, Huang A, Donahue CJ, Sherwood SW, Baldwin DT, Godowski PJ, Wood WI, Gurney AL, Hillan KJ, Cohen RL, Goddard AD, Botstein D, Ashkenazi A. (1998). Nature 396, 699-703. Article MEDLINE

Reed J. (1999). J. Clin. Oncol. 17, 2941-2953. MEDLINE

Reed JC. (2000). Am. J. Pathol. 157, 1415-1430. MEDLINE

Rieger J, Naumann U, Glaser T, Ashkenazi A, Weller M. (1998). FEBS Lett. 427, 124-128. MEDLINE

Roth W, Isenmann S, Nakamura M, Platten M, Wick W, Kleihues P, Bahr M, Ohgaki H, Ashkenazi A, Weller M. (2001). Cancer Res. 61, 2759-2765. MEDLINE

Roth W, Weller M. (1999). Cell. Mol. Life Sci. 56, 481-506. MEDLINE

Sambrook J, Fritsch E, Maniatis T. (1989). Molecular Cloning: A Laboratory Manual 2nd edn. Cold Spring Harbor Laboratory Press,

Sato T, Irie S, Kitada S, Reed JC. (1995). Science 268, 411-415. MEDLINE

Scaffidi C, Fulda S, Srinivasen A, Friesen C, Li F, Tomaselli K, Debatin K-M, Krammer P, Peter M. (1998). EMBO J. 17, 1675-1687. Article MEDLINE

Scaffidi C, Schmitz I, Krammer P, Peter M. (1999a). J. Biol. Chem. 274, 1541-1548. Article MEDLINE

Scaffidi C, Schmitz I, Zha J, Korsmeyer SJ, Krammer PH, Peter ME. (1999b). J. Biol. Chem. 274, 22532-22538. Article MEDLINE

Schneider P, Holler N, Bodmer JL, Hahne M, Frei K, Fontana A, Tschopp J. (1998). J. Exp. Med. 187, 1205-1213. MEDLINE

Shin M, Park W, Kim S, Kim H, Kang S, Song K, Park J, Dong S, Pi J, Oh R, Lee J, Yoo N, Lee S. (1999). Am. J. Pathol. 154, 1785-1791. MEDLINE

Somia N, Schmitt M, Vetter D, Van Antwerp D, Heinemann S, Verma I. (1999). Proc. Natl. Acad. Sci. USA 96, 12667-12672. MEDLINE

Sprick MR, Weigand MA, Rieser E, Rauc CT, Juo P, Blenis J, Krammer PH, Walczak H. (2000). Immunity 12, 599-609. MEDLINE

Strasser A, Harris AW, Huang D, Krammer PH, Cory S. (1995). EMBO J. 14, 6136-6147. MEDLINE

Teitz T, Wei T, Valentine MB, Vanin EF, Grenet J, Valentine VA, Behm FG, Look AT, Lahti JM, Kidd VJ. (2000). Nat. Med. 6, 529-535. Article MEDLINE

Truneh A, Sharma S, Silverman C, Khandekar S, Reddy MP, Deen KC, McLaughlin MM, Srinivasula SM, Livi GP, Marshall LA, Alnemri ES, Williams WV, Doyle ML. (2000). J. Biol. Chem. 275, 23319-23325. MEDLINE

Tschopp J, Irmler M, Thome M. (1998). Curr. Opin. Immunol. 10, 552-558. MEDLINE

Wen J, Ramadevi N, Nguyen D, Perkins C, Worthington E, Bhalla K. (2000). Blood 96, 3900-3906. MEDLINE

Wolf B, Green D. (1999). J. Biol. Chem. 274, 20049-20052. Article MEDLINE

Wrone-Smith T, Izban KF, Ergin M, Cosar EF, Hsi ED, Alkan S. (2001). Exp. Hematol. 29, 572-581. MEDLINE

Wu M, Das A, Tan Y, Zhu C, Cui T, Wong MC. (2000). J. Neurosci. Res. 61, 464-470. MEDLINE

Yanagisawa J, Takahashi M, Kanki H, Yano-Yanagisawa H, Tazunoki T, Sawa E, Nishitoba T, Kamishohara M, Kobayashi E, Kataoka S, Sato T. (1997). J. Biol. Chem. 272, 8539-8545. MEDLINE

Yu KY, Kwon B, Ni J, Zhai Y, Ebner R, Kwon BS. (1999). J. Biol. Chem. 274, 13733-13736. Article MEDLINE

Zhang X, Franco A, Myers K, Gray C, Nguyen T, Hersey P. (1999). Cancer Res. 59, 2747-2753. MEDLINE

Zhang XD, Franco AV, Nguyen T, Gray CP, Hersey P. (2000a). J. Immunol. 164, 3961-3970. MEDLINE

Zhang XJ, Yan J, Cuttle L, Endre Z, Gobe G. (2000b). Biochem. Biophys. Res. Commun. 277, 487-493.

Zou H, Li Y, Liu X, Wang X. (1999). J. Biol. Chem. 274, 11549-11556. Article MEDLINE

Figure 1 Analysis of FasL (a,b,c), F-LZ-TRAIL (d,e,f) and cisplatin (g,h,i) induced apoptosis in the glioma line panel. Glioma cell lines were treated with TRAIL, FasL or control medium for 24 h (a,b,d,e) or 6 h (c,f) and 10 g/ml cisplatin for 24 h (g,h) or 12 h (i). (a,d,g) Cell viability was assessed by colorimetric MTT assay. (b,e,h) Membrane integrity was measured by propidium iodide uptake. The data are expressed as percentage of viability relative to the cells exposed to the control media. Graphs represent the average and standard deviation of three individual experiments. (c,f,i) Caspase activity was measured fluorometrically by DEVD-AMC cleavage. The data are expressed as relative fluorescence units (RFU) per hour corrected for control

Figure 2 Expression of FasL (a,b) and TRAIL (c,d) receptors. (a,c) Surface expression of Fas and the four apoptosis-related TRAIL receptors on the glioma cell line panel was analysed using flow cytometry on unfixed cells. The mean fluorescence ratio (MFR) and percentage positive cells are indicated. (b,d) Using immunoblotting, the size of the proteins were assayed

Figure 3 Expression of candidate pathway components. The levels of the apoptotic pathway components were assayed by immunoblotting. (a) FADD, Caspase-8 and Caspase-3 protein levels. (b) Bcl-2 family members

Figure 4 Enforced expression of inhibitors implicates apoptosis pathway components. (a) D270 and (b) D645 cells were transfected with either empty vector or plasmids encoding CrmA, FADD DN, Bcl-2 or Bcl-x_L together with a lacZ reporter plasmid. Transfected cells were treated with control medium or media containing FasL, TRAIL or cisplatin 3 g/ml. Following X-gal staining, the percentages of transfected (blue) cells which were apoptotic were determined microscopically. Error bars represent standard deviations from three replicates from one representative experiment

Figure 5 Sensitization to apoptosis by cycloheximide. Cells were incubated with control medium, cycloheximide (CHX) 10 g/ml and FasL, TRAIL or cisplatin 10 g/ml as indicated. Apoptosis was assessed by flow cytometry using Alexa 568-conjugated annexin-V and the percentage of annexin-V positive cells is indicated. Control media and CHX treatment alone killed less than 10% of cells and the percentage values have been subtracted from the treatment group data

Figure 6 (a) XIAP, cFLIP and DcR3 protein expression levels. (b) LFG and PEA-15 mRNA were analysed and loading is indicated by 28S rRNA. 293T transfectant positive controls are shown

Figure 7 Enforced expression of Caspase-8 increases susceptibility of U373MG to FasL and TRAIL-induced apoptosis. U373MG cells were transfected with either FpIRESneo or FpIRES-Casp-8 together with a lacZ reporter plasmid. Transfected cells were treated with normal medium (white columns) or media containing FasL (grey columns) or TRAIL (black columns) and, following X-gal staining, the percentages of transfected (blue) cells which were apoptotic were determined microscopically. Error bars represent standard deviations from at least five replicates from one representative experiment

Table 1 Summary of glioma cell line characteristics and apoptotic pathway analysis