Interleukin-1 Protects Transformed Keratinocytes from Tumor Necrosis Factor-related Apoptosis-inducing Ligand- and CD95-induced Apoptosis but Not from Ultraviolet Radiation-induced Apoptosis*

Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL), a new member of the tumor necrosis factor (TNF) family, induces apoptosis primarily of transformed cells. Interleukin-1 was previously found to protect the keratinocyte cell line KB from TRAIL-induced apoptosis, thus we studied whether interleukin-1 also protects from other apoptotic stimuli (ultraviolet radiation (UV), CD95-ligand). Interleukin-1 rescued KB cells from TRAIL- and CD95-induced apoptosis, which was critically dependent on nuclear factor κB, because cells transfected with a super-repressor form of the nuclear factor κB inhibitor IκB were less protected. In contrast, UV-mediated apoptosis was not only not prevented by interleukin-1 but even enhanced. This opposite effect of interleukin-1 was also observed for the expression of the inhibitor of apoptosis proteins (IAP). Whereas TRAIL- and CD95-mediated suppression of IAP expression was partially reversed by interleukin-1, UV-mediated down-regulation of IAPs was not reversed but even further enhanced. Increased apoptosis induced by interleukin-1 plus UV was accompanied by excessive TNFα release, implying that enhanced cytotoxicity is due to the additive effect of these two apoptotic stimuli. Accordingly, enhanced apoptosis was reduced by blocking the TNF receptor-1. The opposite effects of interleukin-1 indicate that different mechanisms are involved in UV-induced apoptosis compared with CD95- and TRAIL-mediated apoptosis. Furthermore, the data suggest that whether a signal acts in an antiapoptotic way or not does not only depend on the signal itself but also on the stimulus causing apoptosis.

During apoptosis, a complex death program becomes initiated that ultimately leads to the fragmentation of the cell. The death program can be either initiated by the cell itself when the time has come to die (programmed cell death) or by certain external stimuli activating death receptors on the cell surface (for review, see Ref. 1). Thus, apoptosis does not only play an important role in the development and maintenance of tissue homeostasis but also represents an effective mechanism by which harmful cells can be eliminated. Induction of apoptosis allows the organism to get rid of infected cells and also of tumor cells. Accordingly, resistance to apoptosis was identified as an important event in tumorigenesis. Apoptosis of tumor cells can be initiated by triggering cell death receptors, leading to activation of the intracellular apoptotic machinery (2). Chemotherapeutic drugs used in cancer treatment may exert their therapeutic effects by activating these pathways (3)(4)(5). On the other hand, it is known that defects in the apoptotic pathways or activation of antiapoptotic machineries can confer resistance to chemotherapy (for review, see Refs. 6 and 7). In addition, tumor cells can escape apoptotic elimination by down-regulation of apoptosis-related molecules on the cell surface (8,9). Consequently, control of the balance between pro-and antiapoptotic processes within the cell has been recognized as an important target for therapeutic intervention. Thus, elucidation of the molecular mechanisms regulating these processes is of primary interest.
Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL), 1 also called APO-2 ligand, is a recently identified molecule belonging to the tumor necrosis factor (TNF) family that was characterized by its ability to induce apoptosis (10,11). Among the TNF family members, TRAIL displays highest homology to CD95-ligand (CD95L), which induces apoptosis by triggering the surface receptor CD95 (Fas/APO-1) (12). In contrast to CD95L, TRAIL was found to induce apoptosis in numerous transformed cell lines but to kill normal cells less effectively (10,11). Because of the unique ability to induce apoptosis preferentially in cancer but not in normal cells, TRAIL may be highly efficient in specifically eradicating tumor cells in vivo (13). Thus, TRAIL has prospects of becoming an effective anticancer drug of the future. Recently, we observed that preincubation with the pro-inflammatory cytokine interleukin-1 (IL-1) renders transformed keratinocytes resistant to the apoptotic effect of TRAIL (14). The protective effect of IL-1 against TRAIL-induced apoptosis seems to be mediated via activation of the transcription factor nuclear factor B (NFB). The observation that transformed cells become resistant to TRAIL upon exposure to IL-1 was the first demonstration of a pathway that allows tumor cells to escape the killing effect of TRAIL. Because IL-1 is secreted by a variety of tumor cells (15) and is also released by inflammatory cells participating in the tumor-host immune response (16), tumors under these conditions could become resistant to TRAIL in vivo.
Therefore, we were interested in expanding these observations by studying whether the antiapoptotic effect of IL-1 is restricted to TRAIL or whether other apoptotic stimuli are inhibited as well. Here, we show that IL-1 protects transformed keratinocytes from TRAIL-and CD95-mediated apoptosis. In contrast, apoptosis induced by ultraviolet (UV) irradiation was not only not prevented by IL-1 but was even further enhanced. This opposite effect of IL-1 was also observed when studying the expression of the antiapoptotic proteins c-IAP1 and c-IAP2. Thus, these data suggest that antiapoptotic activity of a stimulus does not only depend on its nature but also on the stimulus causing apoptosis.

EXPERIMENTAL PROCEDURES
Cells-The epitheloid carcinoma cell line KB (American Type Culture Collection, Manassas, VA) and the spontaneously transformed human keratinocyte cell line HaCaT (kindly provided by N. Fusenig, Deutsches Krebsforschungszentrum, Heidelberg, Germany) (17) were cultured in RPMI containing 10% fetal calf serum and 1% glutamine at 37°C with 5% CO 2 in a humidified atmosphere. Irradiation of cells with UV light was performed using a bank of 4 FS20 bulbs (Westinghouse Electric Corp., Pittsburgh, PA) which emit most of their energy within the UVB range (290 -320 nm) with an emission peak at 313 nm as described (18). Subconfluent cells were exposed through PBS to a dose of 300 J/m 2 , unless otherwise stated.
Reagents-Recombinant human TRAIL protein was provided from Immunex Corp., Seattle, WA. This is a leucine zipper form of TRAIL that requires no further cross-linking for induction of maximal apoptotic acitivity (19). Recombinant CD95L and an agonistic antibody against CD95 (CD95-Ab) were obtained from Alexis and Immunotech, respectively. Antibodies directed against caspase-3 and poly(ADP-ribose) polymerase (PARP) were obtained from Dianova, Hamburg and Roche Molecular Biochemicals, Mannheim, Germany, respectively. Antibodies directed against c-IAP1, c-IAP2, and the fluorescein isothiocyanate-conjugated secondary antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Recombinant human IL-1␤ was obtained from Roche Molecular Biochemicals. TNF␣ was measured by use of an ultrasensitive TNF␣ enzyme-linked immunosorbent assay (Diaclone, Besancon, France). Plasmids allowing overexpression of a mutated IB variant were kindly provided by K. Schulze-Osthoff, Mü nster, Germany (20).
Detection of Cell Death-For the detection of DNA fragmentation, a cell death detection enzyme-linked immunosorbent assay (Roche Molecular Biochemicals) was used. The enrichment of mono-and oligonucleosomes released into the cytoplasm is calculated using the formula: absorbance of sample cells/absorbance of control cells. Enrichment factor was used as a parameter of apoptosis and shown on the y axis as mean Ϯ S.D. of triplicates.
Quantitation of apoptosis by annexin V binding was performed using a commercially available kit (Bender Corp., Vienna, Austria). Briefly, cells were washed and resuspended in annexin V binding buffer. Fluorescein isothiocyanate-conjugated annexin V was added, and the samples were analyzed by flow cytometry (Epics XL, Coulter, Miami, FL).
Western Blot Analysis-Cells were harvested and lysed in RIPA buffer (10 mM Tris, pH 8, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM phenylmethylsulfonyl fluoride, 4 g/ml aprotinin, 1 mM sodium orthovanadate) for 15 min on ice. After centrifugation, supernatants were collected, and the protein content measured using a Bio-Rad Protein Assay kit (Bio-Rad). Protein samples were subjected to SDS-polyacrylamide gel electrophoresis, blotted onto nitrocellulose membranes, and incubated with antibodies of interest. To monitor equal loading, membranes were reprobed with an antibody directed against ␣-tubulin (Pharmingen, San Diego, CA). Signals were detected with an ECL kit (Amersham Pharmacia Biotech, Buckinghamshire, UK).
Transfection-Cells (1.5 ϫ 10 7 ) were washed once with PBS and resuspended in 600 l of PBS, 1.25% Me 2 SO. Cells were electroporated with 20 g of each plasmid DNA according to the method described by Melkonyan et al. (21). Transfection efficiency of cells cotransfected with a plasmid encoding ␤-galactosidase (pcDNA6/VS-His/lacZ; Invitrogen, San Diego, CA) was determined 24 h later by staining with X-gal (100 g/ml) in 5 mM potassium ferricyanide, 5 mM potassium ferrocyanide, and 1 mM MgCl 2 in PBS.
Staining of Intracellular Proteins-Aliquots of cells (2 ϫ 10 5 ) were harvested 16 h after stimulation, washed once with PBS, and fixed with 0.8% paraformaldehyde for 5 min on ice. After washing, cells were treated with 0.3% saponine for 5 min on ice. Following centrifugation cells were incubated with the antibodies directed against c-IAP1 or c-IAP2 in 0.3% saponine at 4°C overnight. Purified goat IgG was used as an isotype control. None of the stimuli used (TRAIL, CD95-Ab, UV radiation, IL-1) changed the isotype controls. Cells were washed with PBS and incubated with the respective fluorescein isothiocyanate-conjugated secondary antibody in 0.3% saponine for 30 min. Cells were washed, resuspended in 0.03% saponine/PBS, and subsequently analyzed in a flow cytometer.

IL-1 Protects Transformed Keratinocytes from TRAIL-and CD95-induced Apoptosis but Not from UV-induced Apoptosis-
Because we had recently observed that IL-1 protects cells from TRAIL-induced apoptosis (14) we were interested in studying whether this effect is specific for TRAIL or whether IL-1 also protects cells from other apoptotic stimuli. Therefore, KB cells were exposed to TRAIL or to an agonistic CD95-Ab, which induces apoptosis via activation of CD95 (12). Both TRAIL and CD95-Ab induced apoptotic cell death of KB cells, as determined by a cell death detection enzyme-linked immunosorbent assay (Fig. 1). When KB cells were preincubated with IL-1 for 15 min, cells were almost completely protected from the apo-FIG. 1. IL-1 protects from TRAILand CD95-induced apoptosis but not from UV-induced apoptosis. KB cells were exposed to TRAIL (32 ng/ml), CD95-Ab (1 g/ml), or UV light (300 J/m 2 ) in the absence or presence of recombinant human IL-1␤ (10 ng/ml). Control cells (Co) were left untreated. Apoptosis was examined 16 h later by determining nucleosomal DNA fragmentation using an apoptosis determination kit. The rate of apoptosis is reflected by the enrichment of nucleosomes in the cytoplasm shown by the values on the y axis. Data presented show the representative results of one of three independently performed experiments. ptotic effect of both TRAIL and the CD95-Ab. Similar data were obtained with HaCaT cells or when annexin V staining was used as a read-out system for apoptosis (data not shown). Thus, these data indicate that IL-1 rescues cells from TRAIL-mediated as well as from CD95-mediated apoptosis. Similar observations were obtained when instead of the CD95-Ab recombinant CD95L was used as the apoptotic stimulus (data not shown).
We next investigated whether IL-1 also protects KB cells from UV-mediated apoptosis. Therefore, KB cells were exposed to 300 J/m 2 in the absence or presence of IL-1 and apoptosis determined 16 h later. In contrast to CD95-and TRAIL-mediated apoptosis, prestimulation of KB cells with IL-1 did not protect cells from UV-induced apoptosis and even caused pronounced enhancement of cell death (Fig. 1). To exclude that the failure of IL-1 to rescue cells from UV-induced apoptosis is simply due to the fact that the dose of 300 J/m 2 represents a too severe insult to the cells, lower UV doses were tested. As shown in Fig. 2, UV irradiation induced apoptosis in a dose-dependent manner, and in all cases IL-1 enhanced the apoptotic response.
Caspase-3 is a member of the family of interleukin-1␤ converting enzyme proteases that is involved in CD95-, TRAIL-, and UV-mediated cell killing (22)(23)(24)(25). For activation, caspase-3 must be cleaved from its 32-kDa proform into its 17-kDa active form (26). Thus, caspase-3 cleavage can serve as an additional read-out system to evaluate apoptosis. Therefore, KB cells were exposed to TRAIL, CD95-Ab, or to UV light either in the absence or presence of IL-1. Cell lysates were prepared 16 h later for Western blot analysis using an antibody against caspase-3. Because this antibody is directed against the caspase-3 proform, it cannot recognize the processed 17-kDa form, and a loss of the immunoreactive band in samples in which caspase-3 is activated can be observed. Significant reduction of the caspase-3 proform was observed in protein extracts of KB cells that were treated either with TRAIL, CD95-Ab, or UV light (Fig. 3). Preincubation of cells with IL-1 almost completely prevented TRAIL-and CD95-induced caspase-3 activation. In contrast, IL-1 enhanced UV-induced caspase-3 activation, which is shown by the complete disappearance of the immunoreactive band.
Caspase-3 cleaves the death substrate PARP (27). Accordingly, PARP was cleaved from its intact 116-kDa form into the inactive 85-kDa fragment in samples of TRAIL-, CD95-Ab-, and UV-exposed KB cells (Fig. 3). IL-1 pretreatment significantly reduced TRAIL-and CD95-Ab-induced PARP cleavage, whereas UV-mediated cleavage of PARP was again enhanced in the presence of IL-1. Taken together, these data clearly demonstrate that IL-1 rescues cells from CD95-and TRAILinduced apoptosis, whereas UV-induced apoptosis is even enhanced by IL-1.
The Protective Effect of IL-1 Is Critically Dependent on NFB-We recently proposed that IL-1 might rescue cells from TRAIL-induced apoptosis via activation of NFB, because IL-1 activated NFB in KB cells. Furthermore, inhibition of NFB activation by the proteasome inhibitor MG132 antagonized the protective effect of IL-1 (14). Activation of NFB is associated with degradation of the inhibitory protein IB by the proteasome pathway. Upon triggering of the IL-1 receptor, IB becomes phosphorylated at the serine residues 32 and 36, which acts as a signal for ubiquination and subsequent degradation of IB by the 26 S proteasome (28,29). Thus, IB degradation and consequently NFB activation can be blocked by proteasome inhibitors such as MG132 or lactacystin (29). However, the approach using proteasome inhibitors provides only indirect evidence that the protective effect of IL-1 is due to activation of NFB, because one cannot exclude that the inhibitors like MG132 may affect other pathways as well. Thus, we addressed whether IL-1 protects KB cells from TRAIL-and CD95-mediated apoptosis via activation of NFB by overexpressing a super-repressor form of IB. In this mutant form, two point mutations (Ser-32 3 Ala, Ser-36 3 Ala) prevent phosphorylation and subsequent proteasomal degradation of IB (20). As a  3. IL-1 differentially affects cleavage of caspase-3 and PARP. KB cells were exposed to TRAIL (32 ng/ml) (lanes 2 and 3), CD95-Ab (1 g/ml) (lanes 4 and 5), or UV radiation (300 J/m 2 ) (lanes 6 and 7) in the absence (lanes 1, 2, 4, and 6) or presence (lane 3, 5, and 7) of 10 ng/ml IL-1␤. Control cells were left untreated (lane 1). Proteins were extracted 16 h after treatment, and Western blot analysis was performed using antibodies directed against caspase-3, PARP, or ␣-tubulin.
consequence, NFB release, nuclear translocation, and functional DNA binding is prevented. Whereas KB cells transfected with the empty cytomegalovirus vector were still rescued by IL-1, the protective effect of IL-1 on TRAIL-and CD95-induced apoptosis was significantly reduced in cells transfected with the IB super-repressor (Fig. 4). These results support the concept that IL-1-mediated protection from TRAIL-and CD95induced apoptosis is dependent on NFB activation.
Differential Regulation of Antiapoptotic Proteins by IL-1-Because IL-1 protected KB cells from CD95-and TRAIL-induced apoptosis but enhanced UV-mediated cell death, we were interested in elucidating the mechanism underlying this heterogenous effect. It has recently been reported that NFB may exert its antiapoptotic effect via induction of several proteins including the inhibitor of apoptosis proteins (IAP) c-IAP1 and c-IAP2 (30). Therefore we investigated how IL-1 affects c-IAP1 and c-IAP2 expression in KB cells exposed to the different apoptotic stimuli. Intracellular protein expression evaluated by fluorescence-activated cell sorter analysis revealed that KB cells express c-IAP1 and c-IAP2 constitutively (Fig. 5). Exposure of KB cells to either TRAIL, CD95-Ab, or UV light caused a down-regulation of both c-IAP proteins. IL-1 alone only marginally enhanced constitutive c-IAP1 and c-IAP2 expression (data not shown), but reversed CD95-and TRAIL-induced down-regulation of c-IAP1 and c-IAP2 partially (Fig. 5). The opposite effect was exerted by IL-1 when cells were exposed to UV light. In this case, IL-1 did not reverse reduced c-IAP1 and c-IAP2 protein levels, but even slightly further enhanced the down-regulation. Taken together, IL-1 appears to affect c-IAP levels differentially depending on the stimulus that induces apoptosis. The further reduction of c-IAP expression in UVexposed cells by IL-1 might be an explanation why IL-1 enhances UV-induced apoptosis.
Enhancement of UV-induced Apoptosis by IL-1 Is Due to Endogenous Release of TNF␣-TNF␣ can generate two types of signals, one that induces apoptosis (31) and one that activates NFB (32). The overall result in a specific cell type under specific conditions appears to be dependent on the balance of the two signals (32). When KB cells were treated with TNF␣ in addition to TRAIL, an enhancement of apoptosis was observed (data not shown), indicating that under these conditions the death signal caused by TNF␣ overrules activation of NFB. Consequently, both stimuli induce apoptosis in an additive way. Because TNF␣ can be released by KB cells (18)  ptosis might be due to autocrine release of TNF␣. To address this issue, KB and HaCaT cells, respectively, were exposed to TRAIL, CD95-Ab, or UV radiation in the absence or presence of IL-1. KB and HaCaT cells did not constitutively release TNF␣. IL-1, TRAIL, or CD95-Ab alone or in any combination induced TNF␣ secretion only marginally. whereas UV irradiation induced moderate release of TNF␣. In contrast, when cells were stimulated with UV light plus IL-1, dramatically enhanced TNF␣ release was observed (Table I). To substantiate that TNF␣ released under these conditions is involved in the enhanced induction of apoptosis, an antibody neutralizing the TNF receptor type 1 which mediates apoptosis (34) was used. Preincubation with the antibody caused a complete loss of the enhancing effect of IL-1 on UV-induced apoptosis (Fig. 6). This indicates that increased apoptosis of UV-exposed KB and HaCaT cells in the presence of IL-1 is due to the autocrine release of TNF␣, which then enhances apoptosis. DISCUSSION Recently, IL-1 was shown to protect tumor cells from the apoptotic effect of TRAIL (14). Because IL-1 can be released by a variety of tumor cells (15) and is also released by inflammatory cells participating in the tumor-host immune response (16), cancer cells under these conditions may escape the possible therapeutic effect of TRAIL. In this study, we demonstrate that IL-1 does not only protect transformed cells from TRAILbut also from CD95-mediated apoptosis. This effect of IL-1 is clearly due to activation of NFB, because inhibition of NFB by overexpression of a dominant negative mutant of IB prevented protection by IL-1. Protection from CD95-induced apoptosis by IL-1 might also affect the activity of those chemotherapeutic drugs that exert their cytotoxic effect via the CD95/ CD95L system (3)(4)(5). This assumption is also supported by the observation that activation of NFB protects cells against apoptosis induced by the cytostatic drug daunorubicin (35).
Because NFB rescues cells from death induced by TNF␣ (35)(36)(37)(38), TRAIL (13), ionizing radiation (35), chemotherapeutic drugs (35), and, as demonstrated in this study, from CD95mediated cell death, NFB appears to protect from apoptosis in general. Thus, it was quite surprising to observe that IL-1 did not protect but even enhanced apoptosis induced by UV radiation.
Recently it was reported that CD95 is critically involved in UV-mediated apoptosis (24,39,40). UV radiation can directly activate the death receptor by inducing functional aggregation of CD95 (24,39). Inhibition of CD95 clustering following UV exposure reduces but does not completely block apoptosis (24), implying that besides the CD95 pathway other mechanisms must be involved as well. This assumption is supported by the present findings: if UV-induced apoptosis was to be exclusively mediated by CD95, both CD95-and UV-induced apoptosis would be inhibitable by the same interventions. In this case, IL-1 should have protected KB and HaCaT cells not only from CD95-but also from UV-induced apoptosis.
According to recent reports (30,33,41,42), NFB appears to exert its antiapoptotic effects via the induction of antiapoptotic proteins, including c-IAP1, c-IAP2, X-linked IAP, and IEX-1L. Therefore, we were interested in the effect IL-1 may exert on the expression of IAPs in cells exposed either to TRAIL, CD95-Ab, or UV light. Intracellular protein measurements revealed that all three apoptotic stimuli reduced levels of c-IAP1 and c-IAP2, although at different levels. c-IAP1 and c-IAP2 appear to exert their antiapoptotic activity by specifically binding to the terminal effector domains of caspase-3 and -7 (43). In contrast to the mode of action of these proteins, little is known about how c-IAPs are regulated, except that they are under control of NFB and that c-IAP2 can exert a positive feedback control on NFB via an IB targeting mechanism (44). This study for the first time demonstrates negative regulation of c-IAP expression by apoptotic stimuli. Because this was assessed only by determining the protein expression, we do not as yet know whether this inhibition is transcriptionally regulated. Although the mechanisms by which c-IAPs are down-regulated remain to be determined, this phenomenon might be of rele-FIG. 6. Enhancement of UV-induced apoptosis by IL-1 is inhibited by blocking the TNF receptor type 1. KB and HaCaT cells, respectively, were exposed to UV radiation (300 J/m 2 ) in the absence or presence of IL-1␤ (10 ng/ml). An antibody blocking the TNF receptor type 1 was added (500 ng/ml). Apoptosis was examined 16 h after treatment by determining nucleosomal DNA fragmentation using an apoptosis determination kit. The rate of apoptosis is reflected by the enrichment of nucleosomes in the cytoplasm shown by the values on the y axis. Data presented show the representative results of one of three independently performed experiments. a KB and HaCaT cells were left untreated (Ϫ) or exposed to TRAIL (32 ng/ml), CD95-Ab (1 g/ml), or UV radiation (300 J/m 2 ) in the absence or presence of IL-1␤ (10 ng/ml). vance for the execution of apoptosis. CD95-and TRAIL-mediated down-regulation of both c-IAP1 and c-IAP2 was partially antagonized by IL-1. Thus, induction of c-IAP1 and c-IAP2 might be the mechanism by which IL-1 rescues cells from CD95-and TRAIL-induced apoptosis. In contrast, UV-mediated down-regulation of c-IAPs was not restored by IL-1 but even further enhanced slightly. Although we do not yet know the mechanism by which IL-1 affects c-IAP expression in such a different way, the present data indicate that whether a signal acts in an antiapoptotic way or not does not only depend on the signal itself but also on the stimulus causing apoptosis.
The apoptotic effects of TNF␣ on most cells can only be observed when protein synthesis is blocked, suggesting that de novo protein synthesis protects cells by induction of antiapoptotic genes (35). Keratinocytes are able to release TNF␣ (18), and we therefore hypothesized that enhancement of UV-induced apoptosis by IL-1 may be mediated by the autocrine release of TNF␣. Because protective c-IAPs are down-regulated upon UV-exposure, under these conditions the apoptotic effect of TNF␣ could dominate the activation of NFB. Indeed, the enhancing effect of IL-1 on UV-induced apoptosis was completely reversed when the experiment was performed in the presence of an antibody, which blocks the TNF-receptor type 1. Accordingly, excessively enhanced levels of TNF␣ were detected when cells were exposed to UV radiation in the presence of IL-1. Neither TRAIL nor CD95-Ab synergized with IL-1 to induce TNF␣ release from KB or HaCaT cells. Thus, this observation is compatible with our hypothesis that down-regulation of c-IAPs preconditions the cells for the killing effect of TNF␣.
Taken together, the present study demonstrates that IL-1 can exert diverse effects on apoptosis. TRAIL-and CD95-mediated apoptosis is significantly reduced by IL-1 via activation of NFB, whereas UV-mediated apoptosis is remarkably enhanced via the autocrine release of TNF␣. Thus, IL-1 cannot be regarded as an antiapoptotic cytokine in general. Furthermore, these findings suggest that whether a stimulus affects apoptosis in a positive or negative way does not exclusively depend on the nature of the stimulus itself but also on the signal that induces apoptosis. As demonstrated here, completely opposite effects can be obtained in different apoptosis systems.