TRAIL/Apo2L activates c-Jun NH2-terminal kinase (JNK) via caspase-dependent and caspase-independent pathways.

In this study we show that TRAIL (tumor necrosis factor-related apoptosis-inducing ligand), also called Apo2L, activates the c-Jun N-terminal kinase (JNK). Interestingly, TRAIL-induced JNK activation occurs in a cell type-specific manner. In HeLa cells, TRAIL-induced JNK activation can be completely blocked with the cysteine protease inhibitor zVAD-fmk, whereas the same inhibitor has no, or even a stimulatory, effect on JNK activation in Kym-1 cells. Hence, TRAIL can engage at least two independent pathways leading to JNK activation, one that is cysteine protease-dependent and one that is cysteine protease-independent. To investigate whether the cysteine protease-dependent signaling of TRAIL leading to JNK activation is related to the apoptotic pathway engaged by this ligand, we investigated HeLa cells stably overexpressing a dominant negative mutant of FADD (Fas-associating protein with death domain) (GFP(green fluorescent protein)DeltaFADD). In these cells, TRAIL-induced cell death and activation of the apoptosis executioner caspase-8 (FLICE/MACH) and caspase-3 (YAMA, CPP-32, Apopain), that belong to caspase subfamily of cysteine proteases, were abrogated, whereas JNK activation remained unaffected and was still sensitive toward z-VAD-fmk. Similar data were found in HeLa cells overexpressing Apo1/Fas and GFPDeltaFADD upon stimulation with agonistic antibodies. These data suggest that cross-linking of the TRAIL receptors and Apo1/Fas, respectively, engages a FADD-dependent pathway leading to the activation of apoptotic caspases and, in parallel, a FADD-independent pathway leading to the stimulation of one or more cysteine proteases capable to activate JNK but not sufficient for the induction of cell death.

Members of the tumor necrosis factor (TNF) 1 receptor superfamily of proteins are critically involved in inflammatory, immune regulatory, and pathophysiological reactions, including the induction of apoptosis (1,2). The receptors of this superfamily are characterized by two to six extracellular copies of a canonical motif of cysteine-rich pseudorepeats, each comprising six conserved cysteines in a stretch of about 40 amino acids (1,2). The ligands of these receptors belong to a complementary family of structurally related molecules, the TNF ligand family. Most of these ligands are primarily expressed as biologically active type II membrane proteins, from which soluble forms are produced by proteolytical cleavage or alternative splicing (1,2). A subgroup of the TNF receptor superfamily, comprising TNF-R1 (3), Apo1/Fas (4), Wsl/DR3/Apo-3/TRAMP/LARD (5)(6)(7)(8)(9), CAR1 (10), DR4/TRAIL-R1 (11), and DR5/TRAIL-R2/ TRICK2/KILLER (12)(13)(14)(15)(16)(17)(18)(19), can be defined by their capability to induce cell death in various cell lines. These receptors share a common intracellular 80 amino acid domain, called the death domain, that is indispensable for initiation of the intracellular signaling cascade leading to cell death (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18). The death domain motif is also found in the cytoplasmic adaptor proteins TNF-R1-associated death domain protein (TRADD; Ref. 20 26). Most of these molecules are involved in the mediation of cell death by the beforementioned death domain-containing receptors. All of these adaptor molecules are of at least bipartite structure. Aside from the death domain mediating binding to other death domain-containing proteins, these adaptor molecules have additional domains enabling the interaction with, e.g. the prodomains of apoptotic caspases (26 -28) or members of the TNF receptor-associated factor (TRAF) family (29,30), molecules involved in activation of NF-B and JNK (31,32). In particular, it has been shown that the lack of FADD in FADDϪ/Ϫ mice (33,34) or overexpression of a FADD molecule lacking the amino-terminal death effector domain, which mediates association with the prodomain of caspase-8, interferes with TNF-, Apo1/Fas-, and Wsl/ DR3-mediated apoptosis (6,29,35). This suggests that activation of the apoptotic caspase cascade by these death domaincontaining receptors requires recruitment of the adaptor molecule FADD into the respective receptor signaling complexes. As shown in detail for Apo1/Fas, FADD in turn recruits caspase-8 to the death-inducing receptor signaling complex leading to proteolytic activation of this proximal caspase and to initiation of the apoptotic program (27,28,36,37). Interestingly, all death domain-containing receptors are in principle able to activate the transcription factor NF-B (5, 7, 8, 13, 14, 38 -40), which is frequently associated with protection from apoptosis (41)(42)(43)(44)(45). In the case of TNF-R1, the activation of NF-B occurs via recruitment of RIP and TRAF2 into the receptor signaling complex through association with TRADD and subsequent activation of the TRAF-associated kinase, NF-B-inducing kinase. Subsequently, NF-B-inducing kinase phosphorylates the IB kinases ␣ and ␤ (IKK␣ and IKK␤), leading to activation of NF-B (46,47). In contrast, the molecular mechanisms involved in activation of NF-B by other death domain-containing receptors remain to be elucidated. The most recently identified death domain-containing receptors DR4/TRAIL-R1 and DR5/TRAIL-R2/TRICK/KILLER are receptors for the cytotoxic ligand TRAIL/Apo2L (48,49). Remarkably, TRAIL/Apo2L can also interact with two additional nonapoptotic members of the TNF receptor superfamily, the glycophospholipid-anchored cell surface protein DcRI/TRID/ TRAIL-R3 (12,15,18,50) and DcRII/TRAIL-R4 (51-52), a receptor containing a truncated death domain that is still able to activate NF-B (52). Overexpression of these nonapoptotic TRAIL receptors protects mammalian cells from TRAIL/ Apo2L-induced cell death (12, 15, 18, 50 -52), thereby defining a novel cell death control mechanism within the TNF receptor superfamily. In this report, we demonstrate the capability of TRAIL/Apo2L to induce JNK by a cysteine protease-dependent and by a cysteine protease-independent pathway. Moreover, using a dominant negative mutant of FADD as well as the cysteine protease inhibitor z-VAD-fmk, we are able to show that, in contrast to induction of apoptosis, the cysteine proteasedependent mode of JNK activation is FADD-independent. Hence, TRAIL/Apo2L has the capability to engage at least three different pathways: two independent pathways with distinct cysteine protease requirements leading to cell death or JNK activation and, in addition, a cysteine protease-independent pathway also linked to JNK activation.

EXPERIMENTAL PROCEDURES
Reagents and Cell Lines-The HeLa cells stably transfected with the green fluorescent protein (GFP) and a GFP-tagged dominant negative mutant of FADD, respectively, have been described elsewhere (53). For analysis of GFP a FACStarϩ (Becton and Dickinsion, San Jose, Ca) has been used. Purified recombinant FLAG-tagged human TRAIL was used as complex with anti-FLAG M2 antibody (Kodak International Biotechnologies) as described (53). HeLa and Kym-1 cells were grown in Clicks RPMI 1640 medium supplemented with 5% (HeLa) and 10% (Kym-1) heat-inactivated fetal calf serum, 2 mM L-glutamine, 50 units/ml penicillin, and 50 g/ml streptomycin.
RNase Protection Assay and RT-PCR-Total RNA was isolated from HeLa and Kym-1 cells (10 ϫ 10 6 ) with the RNA INSTAPURE kit (Eurogentech, Seraing, Belgium) according to the manufacturer's recommendations. The presence of transcripts of caspase-8, FasL, Fas, FADD, DR3, TRAIL-R1, TRAIL-R2, TRAIL-R3, TRAIL TNF-R1, TRADD, and RIP as well as the internal controls L32 and glyceraldehyde-3-phosphate dehydrogenase were analyzed using the hApo-3c Multi-Probe template set (PharMingen, Hamburg, Germany). Probe synthesis, hybridization, and RNase treatment were performed with the RiboQuant Multi-Probe RNase Protection Assay System (PharMingen, Hamburg, Germany) according to the manufacturer's recommendations. Finally, samples were resolved by electrophoresis on denaturing polyacrylamide gels (5%) and analyzed by phosphorimaging. Firststrand cDNA synthesis was performed with the First Strand Synthesis Kit (Amersham Pharmacia Biotech, Freiburg, Germany) and oligo(dT) as a primer according to the manufacturer's protocol. Aliquots of 1 l of cDNA were used as template in a final volume of 50 l in a standard High Fidelity PCR (Boehringer Mannheim, Mannheim, Germany) reaction. Samples were overlaid with 50 l of mineral oil and amplified as followed  -CCA CAG TGC AGT CTT TCA AAC AAA CAC-3Ј; TRAIL-R4: F-BspHI, 5Ј-CAG CAC TCA TGA GTC GGA AGA AAT TCA TTT CTT  ACC TCA AA-3Ј; R-NotI, 5Ј-GTG CTG GCG GCC GCT TTC CTG AAG  AGA TTC TTT CAC AGG CA-3Ј. Please obey that the primers for TRAIL-R1, TRAIL-R2, and TRAIL-R4 contain restriction sites for NotI, BspHI, and NcoI for cloning purposes not relevant for this study. Control reactions without or with 1 g of genomic DNA as template was performed in each experiment.
Cell Death Assays-HeLa cells (1.5 ϫ 10 4 /well) were cultivated in 96-well microtiter plates overnight. Next day TRAIL-M2 complex was titrated and 2.5 g/ml cycloheximide were added. After 18 h culture supernatants were discarded, and the cells were washed once with phosphate-buffered saline followed by crystal violet staining (20% methanol, 0.5% crystal violet) for 15 min. The wells were washed with H 2 O and air-dried. The dye was resolved with methanol for 15 min, and optical density at 550 nm was determined with a R5000 enzyme-linked immunosorbent assay plate reader (Dynatech, Guernsey, Great Britain). Kym-1 cells (1 ϫ 10 4 /well) were cultivated in 96-well microtiter plates overnight. Next day TRAIL-M2 complex was titrated and after additional 18 h of culture metabolic activity was measured by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide method.
Immune Complex JNK Assay-Following stimulation, cells (2-3 ϫ 10 6 ) were lysed in 1 ml of kinase lysis buffer (200 mM Tris, pH 7.4, 5 mM MgCl 2 , 1% Triton X-100, 150 mM NaCl, 0.1 mM phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 2 mM Na 3 VO 4 , and 10 mM NaF) for 20 min on ice. Cell debris was removed by centrifugation at 10,000 ϫ g for 10 min at 4°C. JNK was immunoprecipitated with 0.5 g of a rabbit polyclonal anti-JNK antiserum (Santa Cruz Biotechnology) and 20 l of protein A-Sepharose beads for 2 h. Beads were washed three times with kinase lysis buffer and twice in assay buffer (20 mM MOPS, pH 7.2, 10 mM EGTA, 10 mM MgCl 2 , 0.1% Triton X-100, and 1 mM dithiothreitol). After the last wash, beads were left in a 1:1 suspension, and kinase reactions were carried out at room temperature for 20 min after addition of 0.5 g GST-Jun(1-79) and ATP (100 M ATP and 5 Ci of [␥-32 P]ATP. Reactions were stopped by adding 25 l of 6-fold concentrated Laemmli buffer and boiling for 5 min. Samples were resolved on SDS-polyacrylamide gel, transferred to nitrocellulose, and analyzed using a PhosphoImager.

TRAIL/Apo2L
Activates JNK-It has been shown that JNK activity is induced upon cross-linking of non-death domaincontaining receptors of the TNF receptor superfamily, e.g. CD40 (54), as well as by stimulation of the death domaincontaining receptors TNF-R1 and Apo1/Fas (55)(56)(57)(58). Accordingly, we were interested to know whether TRAIL/Apo2L is also able to activate this pathway. For this purpose we analyzed HeLa and Kym-1 cells, as both cell lines are susceptible toward the cytotoxic action of TRAIL and should therefore express at least one of the two death domain-containing receptors of TRAIL. To get first insights in the expression status of the various TRAIL receptors in these cells, we analyzed the transcription level by RNase protection assays and RT-PCR. Using a template set including TRAIL-R1, TRAIL-R2, and TRAIL-R3-specific probes we found significant level of TRAIL-R1 and TRAIL-R2 mRNA but no detectable amounts of TRAIL-R3 mRNA (Fig. 1A). Nevertheless, with RT-PCR transcripts of all four TRAIL-R were detectable within a reasonable number of cycles in RNA from both cell lines (Fig. 1B). For analysis of JNK activation HeLa and Kym-1 cells were cultured for 0 -5 h in the presence of an agonistic TRAIL-FLAG-M2 complex or TNF, and JNK activity was determined by an immunocomplex kinase assay using an amino-terminal (residues 1-79) c-Jun-GST fusion protein as a substrate ( Fig. 2A). A low basal level of JNK activity was found in unstimulated cells, which was significantly increased by treatment with TRAIL-FLAG-M2 complex as well as TNF. However, whereas TNF induced the activation of JNK immediately and transiently for 10 -30 min, TRAIL-FLAG-M2 complex-induced activation of JNK could be discerned only after 2 h and reached a plateau after around 3-4 h of stimulation ( Fig. 2A). Next, Apo1/Fas-mediated JNK activation was investigated in HeLa cells. Because HeLa cells express only low levels of Apo1/Fas (data not shown), we analyzed a HeLa transfectant stably overexpressing this receptor. In these cells, cross-linking of Apo1/Fas using the agonistic antibody anti-Apo1 resulted in a significant activation of JNK with kinetics resembling that of TRAIL/Apo2Linduced JNK activation ( Fig. 2A). Interestingly, TRAIL-FLAG-M2 complex induced apoptosis in Kym-1 but not in HeLa cells without further treatment. Nevertheless, HeLa cells became sensitive for TRAIL/Apo2L-induced cell death after addition of cycloheximide (Fig. 2B). In contrast, TRAIL-mediated JNK activation occurs in the absence of cycloheximide or other metabolic inhibitors in both cell lines. In the presence of cycloheximide TRAIL-mediated JNK activation remained unaffected, whereby the basal level of JNK activity was somewhat increased (data not shown). This suggests that, at least in HeLa cells, activation of JNK by TRAIL/Apo2L is not sufficient for induction of cell death. As expected, in both cell lines TRAIL/Apo2L-induced cell death could be inhibited by the addition of the caspase inhibitor z-VAD-fmk (Fig. 2B).
TRAIL/Apo2L-induced Activation of JNK Can Occur via a Caspase-dependent and via a Caspase-independent Pathway-Induction of apoptosis by death domain-containing receptors critically depends on the proteolytic activation of members of the caspase family, as indicated by the ability of caspase inhibitors, e.g. the cowpox viral protein cytokine response modifier A (CrmA) or the inhibitory peptide z-VAD-fmk, to prevent cell death. In addition, it has been demonstrated that a dominant negative mutant of the FADD-associated caspase-8 also counteracts the apoptotic signaling of the death domain receptors (13,14,16,29,35). Moreover, for Apo1/Fas, the involvement of FIG. 1. Detection of TRAIL receptor transcripts. A, RNase protection assay analysis of steady state levels of TRAIL receptors in HeLa and Kym-1 cells. Whole RNAs were isolated and 10 g of each RNA was analyzed with the hApo-3c Multi-Probe template set to detect caspase-8, FasL, Fas, FADD, DR3, TRAIL-R1, TRAIL-R2, TRAIL-R3, TRAIL TNF-R1, TRADD, and RIP as well as the internal controls L32 and glyceraldehyde-3-phosphate dehydrogenase. Samples were resolved by electrophoresis on denaturing polyacrylamide gels (5%) and analyzed by phosphorimaging. B, detection of TRAIL receptor mRNA by RT-PCR. 1 l of oligo(dT)-primed cDNA was amplified with the indicated number of cycles and analyzed by agarose gel electrophoresis. As a control for cross-contaminations a reaction without template was performed. In addition, control reactions with 1 g of genomic DNA indicate that residual genomic DNA does not interfere with the detection of specific TRAIL receptor transcripts.

FIG. 2. TRAIL/Apo-2L induces the JNK pathway and apoptosis.
A, time course of JNK activation by members of the TNF receptor superfamily. Cell lysates were prepared from HeLa, HeLa-Fas, and Kym-1 cells, stimulated for the indicated times with TNF (5 ng/ml), TRAIL-FLAG-M2 complex (200 ng/ml), and anti-Apo1 antibody (50 ng/ml), respectively. JNK activity was measured by immunocomplex kinase assay with GSTc-Jun(1-79) as substrate. B, TRAIL/ Apo-2L induces apoptosis. HeLa and Kym-1 cells were plated in triplicates overnight at 37°C in 96-well microtiter plates (HeLa, 1.5 ϫ 10 4 cells/well; Kym-1, 1 ϫ 10 4 cells/well). Next day the cells were treated for additional 16 h with various concentrations of TRAIL-FLAG-M2 complex in the presence or absence of 20 M z-VAD-fmk. In addition, HeLa cells were treated with 2.5 g/ml cycloheximide. Viable HeLa cells were quantified by staining with crystal violet, and viable Kym-1 cells were quantitated using the 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide assay. caspases in the activation of JNK and the p38 mitogen-activated protein kinase has been clearly demonstrated (56,57,59). To unravel a possible regulatory role of caspases in activation of JNK by other members of the TNF receptor superfamily, we stimulated HeLa and Kym-1 cells with TRAIL-FLAG-M2 complex, anti-Apo1, or TNF in the absence or presence of z-VAD-fmk. Interestingly, aside from its inhibitory activity on the induction of apoptosis in both cell lines (Fig. 2B), z-VAD-fmk was able to completely block the activation of JNK in HeLa cells induced by TRAIL/Apo2L and anti-Apo1, but not TNF (Fig. 3, A and C). However, in Kym-1 cells, no inhibitory effect of z-VAD-fmk on TRAIL/Apo2L-mediated JNK activation was observed (Fig. 3, B and C). These data clearly indicate that: (i) in HeLa cells, TRAIL/Apo2L and FasL on the one side and TNF on the other side utilize different pathways for activation of JNK. (ii) TRAIL/Apo2L-induced JNK activation occurs in a cell type-specific manner via two distinct pathways, of which one is cysteine protease-dependent and one cysteine proteaseindependent. The dose dependence of the inhibitory effect of z-VAD-fmk on JNK activation and induction of cell death by TRAIL is nearly identical (data not shown). As the latter is mediated by the caspase subgroup of the cysteine protease family, this argues also for an involvement of caspases in TRAIL-mediated JNK activation in HeLa cells.

TRAIL/Apo2L-and Apo1/Fas-mediated Cell Death Is Prevented by Overexpression of GFP⌬FADD, but JNK Activation
Still Occurs in a Caspase-dependent Manner-From the data described so far it was unclear whether TRAIL-R-and Apo1/ Fas-mediated JNK activation in HeLa cells occurs via the apoptotic caspase cascade or via a distinct pathway with other cysteine protease requirements. As mentioned above, several recent reports have demonstrated that overexpression of a dominant negative mutant of FADD, which lacks the aminoterminal death effector domain and is thus unable to recruit caspase-8 to receptor signaling complexes, efficiently abrogates apoptotic signaling induced by death domain-containing receptors (13, 14, 16, 29, 35, 53). We therefore analyzed JNK activation in HeLa cells stably transfected with a GFP-tagged However, as shown in Fig. 4C, in these cells the TRAIL-FLAG-M2 complex still induced JNK activation in a cysteine protease-dependent manner. As GFP⌬FADD prevented activation of the apoptotic caspases, caspase-3 and caspase-8 (Fig.  4D), these data strongly argue for a TRAIL-initiated pathway that leads to activation of FADD-independent caspases, which in turn initiate JNK activation. Moreover, as shown in Fig. 5 for HeLa cells stably transfected with Apo1/Fas and GFP⌬FADD (Fig. 5A), the concept of different caspase(s) requirements of JNK activation and apoptosis also holds true for Apo1/Fas (Fig. 5, B and C). D, GFP⌬FADD prevents activation of caspase-3 and caspase-8. 2 ϫ 10 6 HeLa, HeLa-GFP, and HeLa-GFP⌬FADD cells were treated with TRAIL-FLAG-M2 complex (200 ng/ml) for 5 h in the presence or absence of cycloheximide (1 g/ml). Protein lysates and immunoblotting of proteins were performed as described under "Experimental Procedures." Blots were probed with a polyclonal IgG fraction directed against caspase-3 or a monoclonal caspase-8-specific antibody.
as well as studies with transgenic mice (62,63), have suggested that TRAF molecules are indispensable for JNK activation by these receptors. Interestingly, the TNF receptor superfamily member Apo1/Fas, representing the prototype of a death-inducing receptor, does not associate with TRAFs, but is nevertheless able to induce JNK activation (56 -58, 64 -66). Both receptors, TNF-R1, which uses TRAF2 for signaling, and Apo1/Fas, which signals TRAF independently, belong to a death-inducing subgroup of the TNF receptor superfamily that is defined by a common intracellular domain, called the death domain. It was therefore of interest to define whether TRAIL/Apo2L, which binds to two additional members of this receptor subgroup, is also able to activate JNK and, if yes, whether TRAIL/Apo2Linduced JNK activation resembles the TNF-R1 or the Apo1/Fas pathway. HeLa and Kym-1 cells are highly susceptible toward the cytotoxic action of a TRAIL-FLAG-M2 complex (Fig. 2B) and express all four TRAIL receptors, in particular the two recently cloned death domain-containing receptors for TRAIL/ Apo2L (Fig. 1). In addition, cross-linking of the TRAIL receptor(s) in these cell lines resulted in a stimulation of JNK ( Fig.  2A). TRAIL-induced JNK activation could be detected 1-2 h after stimulation, reaching a plateau after 3-4 h. Hence, the time course of JNK activation was clearly different from the rapid and transient JNK stimulation by TNF ( Fig. 2A), but resembled the sustained JNK activation found after stimulation of Apo1/Fas ( Fig. 2A). It has been reported that both the overexpression of the serpin family protease inhibitor CrmA and treatment with the cysteine protease inhibitor z-VAD-fmk can interfere with Apo1/Fas-mediated p38 and JNK activation (56,57,59,66). We therefore analyzed the effect of z-VAD-fmk on TRAIL/Apo2L-induced activation of JNK. TRAIL/Apo2Land anti-Apo1-, but not TNF-induced, JNK activation was blocked by treatment with z-VAD-fmk (Fig. 3, A and C) in HeLa cells. This suggests a caspase-dependent pathway of TRAILand anti-Apo1-induced JNK activation. In contrast, in Kym-1 cells, treatment with z-VAD-fmk had a rather stimulatory effect on JNK activation (Fig. 3C). Although at present we cannot rule out that the apparent stimulatory effect of caspase inhibitors on TRAIL-induced JNK-activation in Kym-1 cells is just a consequence of protection from the apoptotic effect of TRAIL, it is evident that JNK activation in Kym-1 cells does not required caspases or other cysteine proteases.
Our data, showing a cell type-specific mode of JNK activation by TRAIL, are in agreement with results described by different groups for Apo1/Fas. While Cahill et al.  Fig. 3. B, HeLa-Fas-GFP⌬FADD cells are resistant toward anti-Apo1-induced cell death. 1.5 ϫ 10 4 cells/well were cultured overnight in a 96-well microtiter plate. The cells were then incubated for additional 16 h with 100 ng/ml anti-Apo1 in the presence of 2.5 g/ml cycloheximide. Viable cells were quantified after staining with crystal violet. C, HeLa-Fas and HeLa-Fas-GFP⌬FADD cells were treated with the indicated concentrations of anti-Apo1 monoclonal antibody alone or in combination with z-VAD-fmk (20 M) for 4 h. JNK activity was measured in an immunocomplex kinase assay using GST-c-Jun(1-79) as a substrate. transient transfection assays in 293 cells. In the case of TNF, the induction of cell death via the caspase cascade and the TRAF2-dependent activation of JNK by TNF-R1 bifurcate at the level of the receptor-associated protein TRADD (43). Hence, it is apparent that JNK activation is dispensable for TNFinduced cell death. However, several reports point to a critical role of JNK activation for induction of apoptosis by other stimuli (e.g. Refs. [67][68][69]. The question therefore arose whether JNK activation has an essential function for TRAIL/Apo2Linduced cell death in the cellular system studied here. Two recent studies have described the generation of FADDϪ/Ϫ mice (33)(34). In one of these studies it has been shown in embryonic fibroblasts that the lack of FADD interferes with TNF-R1, Apo1/Fas, and DR3, but not with TRAIL-R1-induced apoptosis (33). This argues for a FADD-independent mechanism of TRAIL-R1-induced apoptosis; however, it is still unclear whether FADD is involved in TRAIL-R2-induced apoptosis as suggested by some studies based on transient transfection assays. We have shown recently that HeLa cells expressing a dominant negative GFP⌬FADD fusion protein are completely resistant toward induction of apoptosis by TNF, agonistic Apo1/ Fas-specific antibodies, and TRAIL (53). This indicates that FADD itself or a FADD-related factor is involved in TRAILinduced apoptosis in these cells. More important for this study, TRAIL-mediated processing of the apoptotic caspase-8 and caspase-3 (Fig. 4D) is blocked by overexpression of GFP⌬FADD. As JNK activation by TRAIL/Apo2L is not impaired in the GFP⌬FADD-expressing cells, but is still z-VADfmk-sensitive (Fig. 4C), the cysteine protease(s) involved in this process must be distinct from caspase-3 and caspase-8. Moreover, in HeLa cells, TRAIL-induced processing of caspase-3 and caspase-8, and therefore induction of cell death, only occur in the presence of cycloheximide, whereas JNK activation takes place without further treatment. Together, these data suggest (i) that a FADD-dependent as well as a FADD-independent pathway of cysteine protease activation exist and (ii) that both pathways activate different subsets of cysteine protease, one not sufficient to induce apoptosis but essential for JNK activation and one dispensable for JNK activation but involved in apoptosis (Fig. 6). Transcripts of all known TRAIL receptors containing an intracellular domain (TRAIL-R1/DR4, TRAIL-R2/DR5, TRAIL-R4) have been easily detected in HeLa and Kym-1 cells using RNase protection assays and/or RT-PCR (Fig. 1). It is therefore rather unlikely that the cysteine protease-dependent and cysteine protease-independent modes of JNK activation in HeLa and Kym-1 cells reflect the utilization of distinct TRAIL receptors. Moreover, the similarities to the Apo1L/Apo1 system rather point to distinct cellular pathways involved in JNK activation, because triggering of Apo1/Fas, the only known receptor for Apo1L, also results in cell type-specific, caspase-dependent and caspase-independent activation of JNK. At least in HeLa cells, TRAIL-R-and Apo1/Fas-mediated JNK activation is independent of FADD. How this pathway is molecularly linked to the respective receptors remains to be elucidated. Two possible candidates are the recently cloned molecules RAIDD and DAXX, proteins that are involved in TNF-R1 and Apo1/Fas signaling. However, overexpression of dominant negative forms of these molecules in HeLa cells exerted no effect on apoptosis or JNK activation initiated by TRAIL/Apo2L (data not shown).