Since its discovery, caspase-8 has been placed at the apex of the proteolytic cascade triggered by death receptor (DR) cross-linking. Because of its capacity to interact with the cytoplasmic portion of DR, it has been suggested that caspase-8 acts independently of other caspases in the initiation of Fas and other DR signaling. In this study, we demonstrate that in Jurkat cells, caspase-3 cleavage is an early step during Fas-induced apoptosis. We show that caspase-3 processing into its p20 occurs rapidly after Fas cross-linking, in the absence of mitochondrial depolarization and caspase-9 activation. Moreover, caspase-3 is present in lipid rafts of untreated Jurkat cells and peripheral T lymphocytes. Caspase-3, caspase-8, and Fas-associated death domain are further recruited to lipid rafts of Jurkat cells following anti-Fas treatment. Fas immunoprecipitation reveals that caspase-3 is a component of the death-inducing signaling complex, suggesting that this cysteine protease is in close proximity to caspase-8. Furthermore, transduction of Jurkat cells with a caspase-3 dominant-negative form inhibits caspase-8 processing and results in inhibition of apoptosis, suggesting that caspase-3 activity is required for caspase-8 activation. Overall, these findings support a model whereby caspase-3 is a component of the death-inducing signaling complex located in lipid rafts, and as such, is involved in the amplification of caspase-8 activity by the mitochondrion.

Apoptosis or programmed cell death plays a fundamental role in the development and homeostasis of the immune system. Alteration of the apoptotic machinery in peripheral T lymphocytes leads to abnormal T cell homeostasis and the development of autoimmune diseases in human and mice (1). Death receptors (DR)3 are TNFR family members such as TNFRI, Fas (CD95/APO-1), DR3, DR4, and DR5 that harbor a death domain on their intracellular portion (2, 3) and are potent inducers of apoptosis. Ligation of these receptors with their respective ligands or agonistic Abs leads to apoptosis through the activation of proteases of the caspase family (4, 5). These cysteine proteases trigger the apoptotic response by cleaving many substrates after aspartic acid residues. To date, 14 members have been identified within the caspase family (6, 7, 8). Cells from mice deficient in some caspases or expressing viral caspase inhibitors, such as the cytokine response modifier A or p35, are resistant to apoptosis mediated by DR, confirming the requirements of these cysteine proteases in DR-mediated apoptosis (9, 10, 11, 12, 13).

Fas ligation leads to caspase activation via the Fas-associated death domain (FADD) protein, an adaptor molecule that recruits caspase-8 to Fas because FADD interacts with Fas through its death domain and with caspase-8 through a death effector domain (DED). This results in the formation of a death-inducing signaling complex (DISC) (14, 15, 16, 17). Recently, it has been reported that Fas engagement on murine thymocytes induces the clustering of this receptor as well as the recruitment of FADD and caspase-8 in the lipid-rich plasma membrane compartments called lipid rafts (18). These plasma membrane structures are highly ordered microdomains containing sphingolipids, cholesterol, transmembrane proteins, and lipid-anchored proteins (19). They have long been proposed as the host of signal initiation of several immune receptors, including the TCR, the B cell receptor, and FcεRI (20, 21, 22, 23). NF-κB activation by TNFRI and Fas-mediated apoptosis in thymocytes are dependent on relocalization of these DR to lipid rafts (24).

Two pathways observed in different cell types have been proposed for subsequent steps of Fas signaling. In type I cells, the highly efficient recruitment of caspase-8 to the DISC allows this caspase to act directly on downstream caspases such as caspase-3, which is responsible for the cleavage of death substrates and execution of apoptosis (25). In type II cells, recruitment of caspase-8 to the DISC is weak, but sufficient to trigger mitochondrial events through processing of the proapoptotic Bcl-2 family member, Bid (25, 26). These events culminate in the release of cytochrome c in the cytosol, the formation of the apoptosome, a multimeric catalytic complex consisting of cytochrome c, Apaf-1, and caspase-9 that serves to amplify the weak initiating signal resulting from the altered DISC (25) and the release of second mitochondria-derived activator of caspases (Smac)/direct inhibitor of apoptosis-binding protein with a low isoelectric point, which antagonizes XIAP-mediated caspase-3 inhibition (27). Regardless of the cell type, caspase-8 is still considered to act independently of other caspases in the initiation phase of Fas-induced apoptosis. The poor recruitment of caspase-8 to the DISC in type II cells prompted us to dissect events that occur before Fas-mediated mitochondrial alteration and to seek the contribution of another major caspase, caspase-3, to Fas signaling. Our results reveal that this caspase is a component of the DISC along with caspase-8, that caspase-3 colocalizes in lipid rafts with caspase-8, and that caspase-3 activity is required for complete caspase-8 activation following Fas cross-linking.

The Jurkat cell line (clone E6-1) was purchased from the American Type Culture Collection (Manassas, VA). The CH11 and M3 anti-Fas Abs were obtained from Beckman Coulter (Fullerton, CA) and Immunex (Seattle, WA), respectively. Anti-FADD, anti-caspase-8, and anti-caspase-9 Abs were obtained from BD Transduction Laboratories (Mississauga, Ontario, Canada) and BD PharMingen (San Diego, CA). The anti-caspase-3 (28), anti-caspase-8, anti-LCK, anti-FLIP, and anti-DNA fragmentation factor (DFF) (29) Abs were generated in rabbits in our laboratory, using GST fusions with whole proteins. The anti-poly(ADP-ribose) polymerase (PARP) Ab was purchased from Biomol (Plymouth Meeting, PA), and the anti-Bcl-2 and anti-CD45 Abs from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-Bap-31 Ab was a gift from G. Shore (Department of Biochemistry, McGill University). HRP-conjugated goat anti-rabbit IgG and goat anti-mouse IgG used in Western blots were obtained from Jackson ImmunoResearch Laboratories (West Grove, PA), and N-octylglucoside from Sigma-Aldrich (St. Louis, MO).

Wild-type and Bcl-2-expressing Jurkat T cells (JIB2) were maintained at 37°C and 5% CO2 in RPMI 1640 (Life Technologies, Burlington, Ontario, Canada) supplemented with 10% FCS. Bcl-2-inducible expression in Jurkat cells was obtained, as follows. The Bcl-2-expressing system was produced by transferring an EcoRI/BglII fragment of the human Bcl-2 cDNA from the pIC-Bcl-2 plasmid, kindly provided by J. Adams (Walter and Eliza Hall Institute, Melbourne, Australia), and subcloned into the EcoRI and BamHI sites of the neo-CMVt vector, a kind gift from J. Hiscott (Lady Davis Institute, Montreal, Quebec, Canada). The CMVt-Bcl-2 construct was transfected by electroporation into a Jurkat clone carrying the CMVt-rtTA plasmid. The clones carrying both vectors were selected for puromycin and neomycin resistance and subcloned by limiting dilution. The JIB2 clone was selected for its low basal Bcl-2 expression, and high resistance to anti-Fas-induced apoptosis following Bcl-2 expression in the presence of 1 μg/ml of the tetracycline analog, doxycyclin (DOX; Sigma-Aldrich).

Human caspase-3 cDNA cloned into BamHI-EcoRI sites of the pBSK plasmid has served as a template to introduce a point mutation into caspase-3 catalytic site QA(C163A)RG to generate an inactive dominant-negative variant of caspase-3 (C3DN). This was achieved by a two-round PCR using two sets of overlapping primers. PCR products were then digested using KpnI and EcoRI restriction sites, then inserted in frame 3′ to the His6 stretch/TAT linker of the pTAT vector, a generous gift from S. Dowdy (Howard Hughes Medical Institute, Washington University, St. Louis, MO). After verification of the nucleic acid sequences, TAT-C3DN fusion protein production was conducted in BL21 pLys S strain (Novagen, Madison, WI). Highly expressing clones were selected and induced to produce fusion proteins using isopropyl β-d-thiogalactoside (1 mM). Protein was extracted from bacteria by sonication in a denaturing buffer composed of 20 mM of HEPES, 150 mM of NaCl, and 8 M of urea (pH 8). Resulting extracts were cleared by centrifugation for 15 min at 6000 rpm. TAT fusions were purified from cleared extracts by affinity chromatography using nickel NTA (Invitrogen, Burlington, Ontario, Canada) and eluted by imidazole (Sigma-Aldrich). The protein was further subjected to gel filtration chromatography on Sephadex G-25, PD-10 columns (Pharmacia/Apbiotech, Baie d’Urfé, Quebec, Canada) to remove urea and imidazole, then stored at 4°C until needed for functional assays.

Apoptosis was induced with anti-Fas Abs added for the indicated times. Cell aliquots (105 cells) from each condition were resuspended and incubated for 10 min in binding buffer (10 mM of HEPES, 150 mM of NaCl, 5 mM of KCl, 1 mM of MgCl2, and 1.8 mM of CaCl2/pH 7.4) containing 1 μg/ml propidium iodide (Sigma-Aldrich) and 1 μg/ml FITC-conjugated annexin V (AnV; BioLynx, San Antonio, TX). Apoptosis was monitored by FACS using the FL-1 and FL-2 detectors on a FACScan flow cytometer (BD Biosciences, Mississauga, Ontario, Canada). Assessment of the mitochondrial transmembrane potential was performed by incubation of cell suspensions for 30 min at 37°C in PBS containing 12 nM of dihexyloxacarbocyanine iodide (DIOC6) (Molecular Probes, Eugene, OR). The staining was monitored by FACS using the FL-1 detector following extensive washing in PBS buffer. At least 104 cells were acquired for each sample.

Cell lysates (20–50 μg protein/lane), immune complexes (for experiments on DISC collection), or lipid raft fractions were resolved by SDS-PAGE on 12–14% gels and transferred to nitrocellulose HyBond-C membranes (Amersham Pharmacia Biotech, Baie d’Urfé, Quebec, Canada). Membranes were blocked with 5% skimmed milk in PBS/0.05% Tween 20 (PBST) for 2 h at room temperature, then incubated with the appropriate Ab in the same solution overnight at 4°C. After three washes in PBST, blots were incubated for 1 h at room temperature with HRP-conjugated goat anti-rabbit IgG or goat anti-mouse IgG. Blots were then washed three times with PBST, revealed by incubation with ECL-Plus Western blotting detection kit (Amersham Pharmacia Biotech), and then developed on Kodak films.

Lipid rafts were prepared by sucrose gradient ultracentrifugation of cell lysates. Resting Jurkat cells (108 cells) or cells stimulated with 1 μg/ml anti-Fas Ab (CH11) for 7 min were washed in ice-cold PBS and lysed in 0.5 ml of cold buffer (1% Triton X-100, 20 mM of MES, and 150 mM of NaCl, pH 6.5, containing protease inhibitors (Roche, Basel, Switzerland)). The same procedure was used with 3 × 108 untreated or anti-CD3-stimulated PBMC. The lysates were cleared by centrifugation at 14,000 rpm for 30 s at 4°C, and the supernatants were subjected to sucrose gradient fractionation using ultracentrifugation (100,000 × g, 4°C, 17 h). Eleven to 12 fractions of 1 ml were collected, and N-octylglucoside (60 μg/ml) was added to dissolve lipids. A total of 10 μl of each fraction was subjected to dot-blot analysis using HRP-conjugated cholera toxin (Sigma-Aldrich) to detect GM1, a positive marker of rafts. A total of 15–20 μl from each fraction was resolved by SDS-PAGE and subjected to Western blot analysis for the screening of raft-associated proteins using the appropriate Abs.

Jurkat cells (20 × 106) were left untreated or stimulated with 1 μg/ml anti-Fas Ab (CH-11) for the indicated time points. The stimulation was then stopped with ice-cold PBS and cells were lysed, on ice for 30 min, in PBS containing 0.1% Nonidet P-40 detergent and protease inhibitors. Lysates from untreated cells were incubated with 1 μg/ml anti-Fas (CH11) on ice for 30 min. All the lysates were then centrifuged at 3500 rpm for 10 min at 4°C. The supernatants were incubated for 1 h at 4°C with agarose beads coupled to goat anti-mouse IgM (Sigma-Aldrich) to immunoprecipitate CH-11 immune complexes. The beads were washed extensively with lysis buffer and subsequently recovered by centrifugation, resuspended in nonreducing 2× loading buffer, and boiled for 10 min. The composition of Fas complexes in different DISC elements was then analyzed by Western blot using the appropriate Abs.

A Bcl-2-inducible expression system in Jurkat cells under the control of DOX was generated to determine whether the cleavage of caspase-3 occurs before or following mitochondrial events during Fas-induced apoptosis. Bcl-2-negative (−DOX) or Bcl-2-expressing Jurkat cells (+DOX) were subjected to anti-Fas treatment and apoptosis levels, mitochondrial transmembrane potential loss, and caspase activation were analyzed using AnV assay, DIOC6 staining, and Western blotting, respectively. Four hours after Fas cross-linking, up to 70% of Bcl-2-negative cells underwent apoptosis, while only 13% of DOX-treated cells displayed signs of apoptosis (Fig. 1,A). In Bcl-2-expressing cells, anti-Fas treatment did not trigger a decrease in DiOC6 incorporation (Fig. 1,B), confirming the inhibitory effect of Bcl-2 on the mitochondrial response to Fas ligation. In the presence of Bcl-2, procaspase-9 and PARP degradation in response to Fas cross-linking were totally inhibited and caspase-8 cleavage was almost abrogated (Fig. 1,C). However, in three independent experiments, generation of the caspase-3 p20 cleavage product still occurred following treatment with anti-Fas (Fig. 1 C). Therefore, a first cleavage step between the large and small subunits of caspase-3 occurs in the absence of mitochondrial events, indicating that caspase-3 processing into p20 is triggered either independently or upstream of the mitochondrial cytochrome c release. In contrast, removal of the prodomain of caspase-3, complete processing of caspase-8, and progression to cell death are fully dependent on mitochondrial events.

FIGURE 1.

Fas-induced processing of caspase-3 into p20 cleavage product occurs despite blockade of mitochondrial events. A, Bcl-2-transfected cells were stimulated with 1 μg/ml DOX for 36 h to stimulate Bcl-2 expression. Induced (+DOX) and noninduced cells (−DOX) were then subjected to anti-Fas treatment for 3 h with 250 ng/ml CH-11 anti-Fas Ab. Apoptosis in each condition was monitored by FACS analysis after AnV staining. B, Cells from the same experiment also served to monitor Fas-mediated mitochondrial depolarization by FACS analysis using the mitochondrial dye DiOC6. C, Protein extracts from each time point in the same experiment were immunoblotted to monitor Bcl-2 expression and the cleavage status of caspase-3, caspase-8, caspase-9, and PARP. Results shown are representative of three independent experiments.

FIGURE 1.

Fas-induced processing of caspase-3 into p20 cleavage product occurs despite blockade of mitochondrial events. A, Bcl-2-transfected cells were stimulated with 1 μg/ml DOX for 36 h to stimulate Bcl-2 expression. Induced (+DOX) and noninduced cells (−DOX) were then subjected to anti-Fas treatment for 3 h with 250 ng/ml CH-11 anti-Fas Ab. Apoptosis in each condition was monitored by FACS analysis after AnV staining. B, Cells from the same experiment also served to monitor Fas-mediated mitochondrial depolarization by FACS analysis using the mitochondrial dye DiOC6. C, Protein extracts from each time point in the same experiment were immunoblotted to monitor Bcl-2 expression and the cleavage status of caspase-3, caspase-8, caspase-9, and PARP. Results shown are representative of three independent experiments.

Close modal

Cleavage of caspase-3 in the presence of Bcl-2 suggested that it occurred upstream of the mitochondria and possibly at the plasma membrane. Therefore, we explored the possibility of association of caspases-3, caspase-8, and other apoptotic components with the plasma membrane, particularly lipid rafts. Lipid rafts were purified from lysates of untreated Jurkat cells by ultracentrifugation on a sucrose gradient (19, 30). Fig. 2,A shows the distribution of GM-1, a well-established raft-associated molecule (31), in fractions obtained from Jurkat cell extracts. Sucrose fractions 4–7 were positive for GM-1, and hence contained proteins from lipid rafts. Surprisingly, caspase-3 and -7 were reproducibly (n = 3) detected in the same fractions, indicating that a proportion of the caspase-3 and -7 intracellular pools is constitutively located at the membrane level in lipid rafts, even in the absence of Fas engagement (Fig. 2,B). Caspase-8, FLIP-long isoform (FLIPL), and FADD were also found in these raft-rich fractions, before Fas cross-linking. Localization of these molecules in lipid-rich compartment was specific because caspase-9 was not found in GM-1-positive fractions (Fig. 2,B). Furthermore, Lamin B, a protein exclusively found in the nucleus, did not colocalize with caspase-3, -7, and -8, confirming that localization of these caspases in lipid raft fractions did not result from a contamination by nonraft fractions (fractions 10 and 11) during the process of extraction. Lipid rafts were extracted from peripheral blood T cells to determine whether the localization of caspase-3 in lipid rafts could be observed in a more physiological system. Caspase-3 was found as a constituent of lipid rafts along with the well-known raft-associated molecule p56lck in untreated, nonapoptotic peripheral lymphocytes (Fig. 2, C and D). Caspase-3 localization in these signaling platforms persisted during at least 16 h of stimulation with an anti-CD3 Ab. These data demonstrate that effector caspases are localized in lipid rafts in the absence of Fas cross-linking, and that caspase-3 colocalizes with p56lck in lipid rafts of normal peripheral T lymphocytes.

FIGURE 2.

Caspase-3 and other apoptotic molecules are associated with lipid rafts in resting Jurkat cells and peripheral blood T lymphocytes. A, Untreated Jurkat cells were lysed in 1% Triton X-100 buffer and subjected to lipid raft isolation by sucrose density-gradient ultracentrifugation, as described in Materials and Methods. The distribution of the lipid raft marker, the ganglioside GM-1 in different sucrose gradient fractions was controlled by slot-blot analysis using an HRP-conjugated cholera toxin. B, The same sucrose fractions analyzed in A were subjected to Western blot analysis to verify the distribution of caspase-3, caspase-7, caspase-8, FLIPL, caspase-9, FADD, and Lamin B between raft and nonraft plasma membrane compartments. Results are representative of three independent experiments. C, Human PBMC were left untreated or stimulated with 1 μg/ml OKT3 for 8 or 16 h. Apoptosis levels (percentage of annexin-positive cells) and expression of CD69 and Fas were assessed by flow cytometry. D, Lipid raft proteins (fractions 4–6) from resting or activated lymphocytes were extracted and subjected to Western blot analysis for the presence of caspase-3 and p56lck. Results are from one representative experiment among two.

FIGURE 2.

Caspase-3 and other apoptotic molecules are associated with lipid rafts in resting Jurkat cells and peripheral blood T lymphocytes. A, Untreated Jurkat cells were lysed in 1% Triton X-100 buffer and subjected to lipid raft isolation by sucrose density-gradient ultracentrifugation, as described in Materials and Methods. The distribution of the lipid raft marker, the ganglioside GM-1 in different sucrose gradient fractions was controlled by slot-blot analysis using an HRP-conjugated cholera toxin. B, The same sucrose fractions analyzed in A were subjected to Western blot analysis to verify the distribution of caspase-3, caspase-7, caspase-8, FLIPL, caspase-9, FADD, and Lamin B between raft and nonraft plasma membrane compartments. Results are representative of three independent experiments. C, Human PBMC were left untreated or stimulated with 1 μg/ml OKT3 for 8 or 16 h. Apoptosis levels (percentage of annexin-positive cells) and expression of CD69 and Fas were assessed by flow cytometry. D, Lipid raft proteins (fractions 4–6) from resting or activated lymphocytes were extracted and subjected to Western blot analysis for the presence of caspase-3 and p56lck. Results are from one representative experiment among two.

Close modal

Upon receptor engagement, redistribution of signaling molecules between raft and nonraft compartments is indicative of the involvement of these molecules in early steps of receptor’s signal transduction. Hence, we examined the impact of Fas cross-linking on caspase-3 distribution in lipid rafts of Jurkat cells. Results shown in Fig. 3,B consistently showed in three independent experiments that caspase-3 was localized in lipid rafts of untreated cells and that 7 min following Fas ligation, increased levels of caspase-3 were associated to these domains. Fas cross-linking also resulted in a strong increase in caspase-8 and FADD levels in the fractions corresponding to lipid rafts (Fig. 3,B). On the contrary, CD45 was not recruited to these fractions either before or after Fas cross-linking (Fig. 3 A). Therefore, Fas-induced recruitment of caspase-3 to lipid rafts, which are well-established signaling platforms, suggests that this caspase plays a role in Fas signaling at the plasma membrane level.

FIGURE 3.

Fas ligation induces an increase in caspase-3 levels in lipid rafts. A, Lipid raft were extracted from Jurkat cells that were left untreated (−) or stimulated with anti-Fas Ab (+), and raft fraction was analyzed by Western blot for the presence of a raft marker, p56Lck, and the absence of a receptor excluded from rafts, the CD45. B, To analyze the effect of Fas ligation on redistribution of death-signaling molecules, fractions from the same experiment were analyzed for the presence of FADD, caspase-8, and caspase-3 by Western blotting.

FIGURE 3.

Fas ligation induces an increase in caspase-3 levels in lipid rafts. A, Lipid raft were extracted from Jurkat cells that were left untreated (−) or stimulated with anti-Fas Ab (+), and raft fraction was analyzed by Western blot for the presence of a raft marker, p56Lck, and the absence of a receptor excluded from rafts, the CD45. B, To analyze the effect of Fas ligation on redistribution of death-signaling molecules, fractions from the same experiment were analyzed for the presence of FADD, caspase-8, and caspase-3 by Western blotting.

Close modal

Because caspase-3 was found in lipid rafts where the DISC is formed, the presence of this protease in the DISC following Fas aggregation was investigated. Jurkat cells were subjected to a time course treatment with anti-Fas Ab and lysed in a mild lysis buffer, and cell surface cross-linked Fas was further isolated by immunoprecipitation. Interestingly, in two independent experiments, caspase-3 was detected in immune complexes of untreated cells (time 0 in Fig. 4), supporting the constitutive localization of this caspase in proximity to Fas and its signaling elements (Fig. 4). Caspase-3 levels in the DISC gradually increased upon Fas cross-linking with a peak at 20-min postligation. Constitutive association and further recruitment to the DISC were specific to caspase-3 because a protein endowed with a cytoplasmic distribution such as NF-κB (p65) was not detected in the DISC at any time point during the kinetic (Fig. 4). The increase in FADD levels, but not NF-κB, confirms that only proteins from the DISC were immunoprecipitated. Furthermore, the absence of FADD in the immunoprecipitated DISC before Fas cross-linking (time 0) excludes the possibility of a contamination by cytoplasmic proteins at this time point. Together with data from Figs. 2 and 3, these results demonstrate that caspase-3 is associated to Fas DISC at the plasma membrane.

FIGURE 4.

Analysis of the Fas-signaling complex (DISC) in the Jurkat cell line. Untreated cells (time 0) were first lysed and then incubated with anti-Fas Ab before Fas immunoprecipitation, as described in Materials and Methods. Stimulated cells were incubated with CH11 for the indicated time points at 37°C before cell lysis, and cell surface Fas was immunoprecipitated using anti-mouse IgG-coupled agarose beads. Immune complexes were resolved by SDS-PAGE and subjected to Western blot analysis for the presence of FADD, caspase-8, caspase-3, and NF-κB (p65) using respective Abs. An aliquot of the total lysate (TL) was loaded as a positive control. These data are from one of two independent experiments giving the same results.

FIGURE 4.

Analysis of the Fas-signaling complex (DISC) in the Jurkat cell line. Untreated cells (time 0) were first lysed and then incubated with anti-Fas Ab before Fas immunoprecipitation, as described in Materials and Methods. Stimulated cells were incubated with CH11 for the indicated time points at 37°C before cell lysis, and cell surface Fas was immunoprecipitated using anti-mouse IgG-coupled agarose beads. Immune complexes were resolved by SDS-PAGE and subjected to Western blot analysis for the presence of FADD, caspase-8, caspase-3, and NF-κB (p65) using respective Abs. An aliquot of the total lysate (TL) was loaded as a positive control. These data are from one of two independent experiments giving the same results.

Close modal

The localization of caspase-3 in lipid rafts and its association with elements of the DISC suggest that this caspase is mobilized in the early steps of Fas signaling. To verify this hypothesis, we analyzed the activation status of caspase-3 and -8 in response to Fas ligation on Jurkat T cells. The 20-kDa cleavage product (p20) of caspase-3 was detected in lysates from Jurkat cells 60 min following Fas cross-linking, even before phosphatidylserine expression (Fig. 5, A and B). At this time point, two caspase-3 substrates, DFF and PARP, were already partially cleaved. Caspase-8 p45, p43, and p18 cleavage products appeared only 120 min after anti-Fas treatment. To further confirm this early processing of caspase-3, Jurkat cells were treated with anti-Fas Ab for 2 h, and AnV (viable) or AnV+ (apoptotic) cells were sorted by flow cytometry. Results shown in Fig. 5,C clearly demonstrate that caspase-3 processing into p20 occurs in Fas-treated viable cells (AnV), whereas caspase-8 cleavage was only observed in AnV+ apoptotic cells. These results demonstrate that the processing of caspase-3 into its p20 form precedes caspase-8 detectable cleavage during Fas-induced apoptosis, which corroborates the previous observation that caspase-3 (but not caspase-8) cleavage still occurs in the presence of Bcl-2 in anti-Fas-treated Jurkat cells (Fig. 1).

FIGURE 5.

Caspase-3 cleavage producing the p20 species occurs early during Fas-induced apoptosis. A, Cells were treated with 250 ng/ml CH11 for indicated times, and apoptosis was monitored by FACS analysis using AnV/propidium iodide double staining. AnV cells were counted as viable, whereas AnV+ cells were counted as apoptotic. B, Cell extracts from each time point in the same experiment were subjected to Western blot analysis using anti-caspase-3, anti-caspase-3, -caspase-8, -PARP, and -DFF Abs (nonspecific bands recognized by the anti-caspase-8 antiserum are indicated by an asterisk). C, Two hours following anti-Fas treatment, Jurkat cells were stained with AnV, and positive and negative cells were purified using a cell sorter. Protein extracts from each population were analyzed by Western blot using anti-caspase-3 or -caspase-8 Abs.

FIGURE 5.

Caspase-3 cleavage producing the p20 species occurs early during Fas-induced apoptosis. A, Cells were treated with 250 ng/ml CH11 for indicated times, and apoptosis was monitored by FACS analysis using AnV/propidium iodide double staining. AnV cells were counted as viable, whereas AnV+ cells were counted as apoptotic. B, Cell extracts from each time point in the same experiment were subjected to Western blot analysis using anti-caspase-3, anti-caspase-3, -caspase-8, -PARP, and -DFF Abs (nonspecific bands recognized by the anti-caspase-8 antiserum are indicated by an asterisk). C, Two hours following anti-Fas treatment, Jurkat cells were stained with AnV, and positive and negative cells were purified using a cell sorter. Protein extracts from each population were analyzed by Western blot using anti-caspase-3 or -caspase-8 Abs.

Close modal

The strong inhibition of caspase-8, but not caspase-3 processing by Bcl-2 prompted us to analyze the effect of caspase-3 inhibition on Fas-induced caspase-8 activation. We took advantage of the widely used protein transduction system using the HIV-1 TAT peptide that has the ability to enter treated cells by passive diffusion through plasma membrane (32). A chimeric protein consisting of the TAT peptide and a catalytically inactive variant of caspase-3 (TAT-cas3c/s) was used to specifically block caspase-3 activity and study the effect of this inhibition on caspase-8 activation in response to Fas ligation. Fig. 6,A shows that 100% of Jurkat cells treated with a FITC-conjugated chimera became FITC positive, demonstrating the high efficiency of the transduction. Pretreatment of Jurkat cells with TAT-cas3c/s resulted in a 65% inhibition of apoptosis (Fig. 6,B), whereas Jurkat cells treated with an irrelevant fusion protein TAT-p16 remained sensitive to Fas-mediated killing (data not shown). Western blot analysis of caspase-8-carried on samples from the same experiment clearly shows that Fas-mediated caspase-8 processing was inhibited by TAT-cas3c/s (Fig. 6 C). Together with the inhibition of caspase-8 (but not caspase-3) cleavage by Bcl-2, these results demonstrate that caspase-3 activity is essential for complete caspase-8 activation during Fas-induced cell death.

FIGURE 6.

Caspase-8 processing in response to Fas-induced apoptosis is inhibited by a dominant-negative form of caspase-3. A, TAT-C3c/s fusion protein was labeled with FITC and incubated with Jurkat cells at a concentration of 0.7 μM for 1 h. The transduction capacity of FITC-conjugated TAT-C3c/s was monitored by FACS analysis. B, Jurkat cells treated or not with 0.7 μM of TAT-C3c/s fusion protein were either cultured alone or stimulated for the indicated times with anti-Fas Ab. Apoptosis in each condition was monitored by flow cytometry using AnV staining. C, Protein extracts from the same experiment described in A were immunoblotted with an anti-caspase-8 Ab.

FIGURE 6.

Caspase-8 processing in response to Fas-induced apoptosis is inhibited by a dominant-negative form of caspase-3. A, TAT-C3c/s fusion protein was labeled with FITC and incubated with Jurkat cells at a concentration of 0.7 μM for 1 h. The transduction capacity of FITC-conjugated TAT-C3c/s was monitored by FACS analysis. B, Jurkat cells treated or not with 0.7 μM of TAT-C3c/s fusion protein were either cultured alone or stimulated for the indicated times with anti-Fas Ab. Apoptosis in each condition was monitored by flow cytometry using AnV staining. C, Protein extracts from the same experiment described in A were immunoblotted with an anti-caspase-8 Ab.

Close modal

Some members of the TNFR family, including nerve growth factor/p75, CD40, TNFR, and Fas, are associated with lipid rafts (33, 34, 35). Experiments performed using murine thymocytes have shown that caspase-8 and FADD are also recruited to lipid rafts following Fas engagement and that disruption of these structures by cholesterol depletion abolishes Fas-triggered DISC formation and cell death (18). Furthermore, in a different study, a small amount of caspase-8 and FADD was detected in the DISC before Fas engagement in murine peripheral T lymphocytes (36). Our data show that significant amounts of caspase-3, caspase-7, caspase-8, and FADD are associated with rafts before Fas cross-linking in Jurkat cells. Coprecipitation of caspase-3 with Fas also demonstrated that it is a DISC component. These findings are consistent with the presence of preassembled Fas trimers at the cell surface, in the absence of Fas ligand, through domains called preligand assembly domains (37, 38). The colocalization of caspase-3 and -8 with lipid rafts and the fact that caspase-8 processing was blocked by Bcl-2, whereas caspase-3 cleavage was still partially processed, suggested a role of caspase-3 in the amplification of caspase-8 activation during Fas signaling. Indeed, using a caspase-3 dominant-negative form (Fig. 6), we have shown that caspase-3 activity is necessary for complete caspase-8 processing, confirming the active participation of caspase-3 in the amplification of Fas-mediated caspase-8 activation. Therefore, this is the first evidence showing the recruitment of an effector caspase to the signaling complex of a death receptor.

The molecular basis for such recruitment is not yet clear because the short prodomain of caspase-3 does not carry the homotypic DED motif found in caspase-8. However, other groups have reported the simultaneous association of caspase-3 and -8 with FLIP (39), suggesting that this protein might recruit caspase-3 to the DISC. Supporting this hypothesis, we have observed the presence of FLIPL in lipid rafts of untreated Jurkat cells (Fig. 2). FLIPL is a catalytically inactive caspase-8-like molecule that was initially proposed in several studies to inhibit Fas-mediated apoptosis by interfering with caspase-8 activation following their concomitant recruitment to the DISC (40, 41). However, it has recently been shown that once recruited to the DISC, FLIPL is an activator of caspase-8 and can promote apoptosis (42, 43). Therefore, FLIPL could exert its proapoptotic activity through the recruitment of caspase-3 in the DISC. How a caspase lacking a DED or a caspase recruitment domain homotypic motif such as caspase-3 becomes cleaved into p20 following Fas ligation remains unclear. This activation could be mediated by autoprocessing as a consequence of caspase-3 oligomerization in the rafts, in an induced proximity manner as originally proposed for the activation of caspase-8 (44).

Previous reports have suggested that during Fas-mediated apoptosis in type II cells, caspase-3 activation is triggered downstream of mitochondrial events (25, 26, 45). This suggested that caspase-3 was only required for the execution of death signals triggered by caspase-8 at the site of initiation. In this study, we show that inhibiting the intrinsic pathway by Bcl-2 overexpression leads to a drastic inhibition of Fas-mediated caspase-8 and -9 activation, but does not prevent caspase-3 from being processed to its p20 form. Our results confirm a previous study also showing the generation of the p20 cleavage product of caspase-3 in Bcl-2 stable transfectants (46). These results imply that processing of caspase-3 into p20 and p12 is independent of the mitochondrial potential disruption and caspase-9 activity, whereas removal of caspase-3 prodomain (which results in the appearance of the p17), complete caspase-8 activation, and execution of apoptosis depend on these events.

It was previously shown that in the presence of Bcl-2, the p20 form of caspase-3 remained associated to XIAP, because Bcl-2 inhibits the release of Smac from the mitochondria into the cytosol (46). We have demonstrated in this study that caspase-3 activity was required for complete caspase-8 processing following Fas engagement. This result suggests that before Smac release into the cytosol, which is inhibited by Bcl-2, caspase-3 is associated to XIAP and cannot remove its prodomain or trigger the complete cleavage of caspase-8. Interestingly, in Jurkat cytosolic extracts, caspase-8 activation induced by addition of cytochrome c was completely abrogated when caspase-3 was depleted (47). This observation corroborates our results obtained in intact Jurkat cells and confirms the requirement for caspase-3 activity in caspase-8 activation.

The results presented in this work are consistent with a model in which critical components of Fas signaling are preassembled in the plasma membrane, notably in the lipid-rich compartment, as summarized in Fig. 7. Upon Fas cross-linking, the association of FADD, caspase-8, and caspase-3 in the DISC increases and leads to caspase-3 processing into its 20- and 12-kDa subunits. The partially processed caspase-3 is sequestered by XIAP until released by mitochondrial proteins such as Smac/direct inhibitor of apoptosis-binding protein with a low isoelectric point. Caspase-3 then removes its prodomain, and its activity results in the complete cleavage of caspase-8 and amplification of Smac and cytochrome c release into the cytosol. This model suggests that caspase-3 is not only an effector caspase, but is also a central molecule in the amplification of Fas signaling in T lymphocytes.

FIGURE 7.

The preassembly model. In contrast to the type I/II models, the preassembly model proposes the preassociation of caspase-3, caspase-8, and other signaling molecules with the plasma membrane. Upon Fas ligation, higher levels of these molecules are recruited and stabilized in lipid-rich compartments, leading to the formation of a signaling core, the DISC, already containing effector caspases. However, caspase-3 autoprocessing is inhibited by XIAP, and the release of Smac from the mitochondria allows caspase-3 to dissociate from XIAP, remove its own prodomain, and trigger caspase-8 processing. This results in the execution of apoptosis through the cleavage of caspase substrates. Therefore, caspase-3 contributes to the amplification of caspase-8 activation.

FIGURE 7.

The preassembly model. In contrast to the type I/II models, the preassembly model proposes the preassociation of caspase-3, caspase-8, and other signaling molecules with the plasma membrane. Upon Fas ligation, higher levels of these molecules are recruited and stabilized in lipid-rich compartments, leading to the formation of a signaling core, the DISC, already containing effector caspases. However, caspase-3 autoprocessing is inhibited by XIAP, and the release of Smac from the mitochondria allows caspase-3 to dissociate from XIAP, remove its own prodomain, and trigger caspase-8 processing. This results in the execution of apoptosis through the cleavage of caspase substrates. Therefore, caspase-3 contributes to the amplification of caspase-8 activation.

Close modal

We thank Dr. S. F. Dowdy (Howard Hughes Medical Institute), Dr. D. Lynch (Immunex), Dr. G. C. Shore (McGill University), Dr. J. M. Adams (Walter and Eliza Hall Institute), Dr. S. Roy (Merck-Frosst, Kirkland, Quebec, Canada), and Dr. J. Hiscott (Lady Davis Institute) for providing valuable reagents and useful comments.

1

This work was supported by the Medical Research Council of Canada, Grant MOP38105 (to R.-P.S.). R.-P.S. holds a Canada Research Chair in Human Immunology, and is a senior scientist of the Canadian Institutes for Health Research.

3

Abbreviations used in this paper: DR, death receptor; AnV, annexin V; DED, death effector domain; DFF, DNA fragmentation factor; DISC, death-inducing signaling complex; DOX, doxycyclin; FADD, Fas-associated death domain; FLIPL, FLIP-long isoform; PARP, poly(ADP-ribose) polymerase; XIAP, X-linked inhibitor of apoptosis protein; Smac, second mitochondria-derived activator of caspases.

1
Lenardo, M., K. M. Chan, F. Hornung, H. McFarland, R. Siegel, J. Wang, L. Zheng.
1999
. Mature T lymphocyte apoptosis: immune regulation in a dynamic and unpredictable antigenic environment.
Annu. Rev. Immunol.
17
:
221
.
2
Ashkenazi, A., V. M. Dixit.
1998
. Death receptors: signaling and modulation.
Science
281
:
1305
.
3
Ashkenazi, A..
2002
. Targeting death and decoy receptors of the tumor-necrosis factor superfamily.
Nat. Rev. Cancer
2
:
420
.
4
Wallach, D., E. E. Varfolomeev, N. L. Malinin, Y. V. Goltsev, A. V. Kovalenko, M. P. Boldin.
1999
. Tumor necrosis factor receptor and Fas signaling mechanisms.
Annu. Rev. Immunol.
17
:
331
.
5
Strasser, A., L. O’Connor, V. M. Dixit.
2000
. Apoptosis signaling.
Annu. Rev. Biochem.
69
:
217
.
6
Kumar, S..
1999
. Mechanisms mediating caspase activation in cell death.
Cell Death Differ.
6
:
1060
.
7
Rathmell, J. C., C. B. Thompson.
1999
. The central effectors of cell death in the immune system.
Annu. Rev. Immunol.
17
:
781
.
8
Shi, Y..
2002
. Mechanisms of caspase activation and inhibition during apoptosis.
Mol. Cell
9
:
459
.
9
Woo, M., R. Hakem, M. S. Soengas, G. S. Duncan, A. Shahinian, D. Kagi, A. Hakem, M. McCurrach, W. Khoo, S. A. Kaufman, et al
1998
. Essential contribution of caspase 3/CPP32 to apoptosis and its associated nuclear changes.
Genes Dev.
12
:
806
.
10
Enari, M., H. Hug, S. Nagata.
1995
. Involvement of an ICE-like protease in Fas-mediated apoptosis.
Nature
375
:
78
.
11
Tewari, M., L. T. Quan, K. O’Rourke, S. Desnoyers, Z. Zeng, D. R. Beidler, G. G. Poirier, G. S. Salvesen, V. M. Dixit.
1995
. Yama/CPP32 β, a mammalian homolog of CED-3, is a CrmA-inhibitable protease that cleaves the death substrate poly(ADP-ribose) polymerase.
Cell
81
:
801
.
12
Varfolomeev, E. E., M. Schuchmann, V. Luria, N. Chiannilkulchai, J. S. Beckmann, I. L. Mett, D. Rebrikov, V. M. Brodianski, O. C. Kemper, O. Kollet, et al
1998
. Targeted disruption of the mouse caspase 8 gene ablates cell death induction by the TNF receptors, Fas/Apo1, and DR3 and is lethal prenatally.
Immunity
9
:
267
.
13
Hisahara, S., T. Araki, F. Sugiyama, K. Yagami, M. Suzuki, K. Abe, K. Yamamura, J. Miyazaki, T. Momoi, T. Saruta, et al
2000
. Targeted expression of baculovirus p35 caspase inhibitor in oligodendrocytes protects mice against autoimmune-mediated demyelination.
EMBO J.
19
:
341
.
14
Boldin, M. P., T. M. Goncharov, Y. V. Goltsev, D. Wallach.
1996
. Involvement of MACH, a novel MORT1/FADD-interacting protease, in Fas/APO-1- and TNF receptor-induced cell death.
Cell
85
:
803
.
15
Fernandes-Alnemri, T., R. C. Armstrong, J. Krebs, S. M. Srinivasula, L. Wang, F. Bullrich, L. C. Fritz, J. A. Trapani, K. J. Tomaselli, G. Litwack, E. S. Alnemri.
1996
. In vitro activation of CPP32 and Mch3 by Mch4, a novel human apoptotic cysteine protease containing two FADD-like domains.
Proc. Natl. Acad. Sci. USA
93
:
7464
.
16
Muzio, M., A. M. Chinnaiyan, F. C. Kischkel, K. O’Rourke, A. Shevchenko, J. Ni, C. Scaffidi, J. D. Bretz, M. Zhang, R. Gentz, et al
1996
. FLICE, a novel FADD-homologous ICE/CED-3-like protease, is recruited to the CD95 (Fas/APO-1) death-inducing signaling complex.
Cell
85
:
817
.
17
Kischkel, F. C., S. Hellbardt, I. Behrmann, M. Germer, M. Pawlita, P. H. Krammer, M. E. Peter.
1995
. Cytotoxicity-dependent APO-1 (Fas/CD95)-associated proteins form a death-inducing signaling complex (DISC) with the receptor.
EMBO J.
14
:
5579
.
18
Hueber, A. O., A. M. Bernard, Z. Herincs, A. Couzinet, H. T. He.
2002
. An essential role for membrane rafts in the initiation of Fas/CD95-triggered cell death in mouse thymocytes.
EMBO Rep.
3
:
190
.
19
Brown, D. A., E. London.
1998
. Functions of lipid rafts in biological membranes.
Annu. Rev. Cell. Dev. Biol.
14
:
111
.
20
Janes, P. W., S. C. Ley, A. I. Magee.
1999
. Aggregation of lipid rafts accompanies signaling via the T cell antigen receptor.
J. Cell Biol.
147
:
447
.
21
Drevot, P., C. Langlet, X. J. Guo, A. M. Bernard, O. Colard, J. P. Chauvin, R. Lasserre, H. T. He.
2002
. TCR signal initiation machinery is pre-assembled and activated in a subset of membrane rafts.
EMBO J.
21
:
1899
.
22
Guo, B., R. M. Kato, M. Garcia-Lloret, M. I. Wahl, D. J. Rawlings.
2000
. Engagement of the human pre-B cell receptor generates a lipid raft-dependent calcium signaling complex.
Immunity
13
:
243
.
23
Stauffer, T. P., T. Meyer.
1997
. Compartmentalized IgE receptor-mediated signal transduction in living cells.
J. Cell Biol.
139
:
1447
.
24
Legler, D. F., O. Micheau, M. A. Doucey, J. Tschopp, C. Bron.
2003
. Recruitment of TNF receptor 1 to lipid rafts is essential for TNFα-mediated NF-κB activation.
Immunity
18
:
655
.
25
Scaffidi, C., S. Fulda, A. Srinivasan, C. Friesen, F. Li, K. J. Tomaselli, K. M. Debatin, P. H. Krammer, M. E. Peter.
1998
. Two CD95 (APO-1/Fas) signaling pathways.
EMBO J.
17
:
1675
.
26
Li, H., H. Zhu, C. J. Xu, J. Yuan.
1998
. Cleavage of BID by caspase 8 mediates the mitochondrial damage in the Fas pathway of apoptosis.
Cell
94
:
491
.
27
Srinivasula, S. M., P. Datta, X. J. Fan, T. Fernandes-Alnemri, Z. Huang, E. S. Alnemri.
2000
. Molecular determinants of the caspase-promoting activity of Smac/DIABLO and its role in the death receptor pathway.
J. Biol. Chem.
275
:
36152
.
28
Alam, A., M. Y. Braun, F. Hartgers, S. Lesage, L. Cohen, P. Hugo, F. Denis, R. P. Sekaly.
1997
. Specific activation of the cysteine protease CPP32 during the negative selection of T cells in the thymus.
J. Exp. Med.
186
:
1503
.
29
Sharif-Askari, E., A. Alam, E. Rheaume, P. J. Beresford, C. Scotto, K. Sharma, D. Lee, W. E. DeWolf, M. E. Nuttall, J. Lieberman, R. P. Sekaly.
2001
. Direct cleavage of the human DNA fragmentation factor-45 by granzyme B induces caspase-activated DNase release and DNA fragmentation.
EMBO J.
20
:
3101
.
30
Brown, D. A., E. London.
2000
. Structure and function of sphingolipid- and cholesterol-rich membrane rafts.
J. Biol. Chem.
275
:
17221
.
31
Dietrich, C., L. A. Bagatolli, Z. N. Volovyk, N. L. Thompson, M. Levi, K. Jacobson, E. Gratton.
2001
. Lipid rafts reconstituted in model membranes.
Biophys. J.
80
:
1417
.
32
Nagahara, H., A. M. Vocero-Akbani, E. L. Snyder, A. Ho, D. G. Latham, N. A. Lissy, M. Becker-Hapak, S. A. Ezhevsky, S. F. Dowdy.
1998
. Transduction of full-length TAT fusion proteins into mammalian cells: TAT-p27Kip1 induces cell migration.
Nat. Med.
4
:
1449
.
33
Bilderback, T. R., R. J. Grigsby, R. T. Dobrowsky.
1997
. Association of p75(NTR) with caveolin and localization of neurotrophin-induced sphingomyelin hydrolysis to caveolae.
J. Biol. Chem.
272
:
10922
.
34
Hostager, B. S., I. M. Catlett, G. A. Bishop.
2000
. Recruitment of CD40 and tumor necrosis factor receptor-associated factors 2 and 3 to membrane microdomains during CD40 signaling.
J. Biol. Chem.
275
:
15392
.
35
Cottin, V., J. E. Doan, D. W. Riches.
2002
. Restricted localization of the TNF receptor CD120a to lipid rafts: a novel role for the death domain.
J. Immunol.
168
:
4095
.
36
Jones, R. G., A. R. Elford, M. J. Parsons, L. Wu, C. M. Krawczyk, W. C. Yeh, R. Hakem, R. Rottapel, J. R. Woodgett, P. S. Ohashi.
2002
. CD28-dependent activation of protein kinase B/Akt blocks Fas-mediated apoptosis by preventing death-inducing signaling complex assembly.
J. Exp. Med.
196
:
335
.
37
Chan, F. K., H. J. Chun, L. Zheng, R. M. Siegel, K. L. Bui, M. J. Lenardo.
2000
. A domain in TNF receptors that mediates ligand-independent receptor assembly and signaling.
Science
288
:
2351
.
38
Siegel, R. M., J. K. Frederiksen, D. A. Zacharias, F. K. Chan, M. Johnson, D. Lynch, R. Y. Tsien, M. J. Lenardo.
2000
. Fas preassociation required for apoptosis signaling and dominant inhibition by pathogenic mutations.
Science
288
:
2354
.
39
Shu, H. B., D. R. Halpin, D. V. Goeddel.
1997
. Casper is a FADD- and caspase-related inducer of apoptosis.
Immunity
6
:
751
.
40
Krueger, A., I. Schmitz, S. Baumann, P. H. Krammer, S. Kirchhoff.
2001
. Cellular FLICE-inhibitory protein splice variants inhibit different steps of caspase-8 activation at the CD95 death-inducing signaling complex.
J. Biol. Chem.
276
:
20633
.
41
Scaffidi, C., I. Schmitz, P. H. Krammer, M. E. Peter.
1999
. The role of c-FLIP in modulation of CD95-induced apoptosis.
J. Biol. Chem.
274
:
1541
.
42
Micheau, O., M. Thome, P. Schneider, N. Holler, J. Tschopp, D. W. Nicholson, C. Briand, M. G. Grutter.
2002
. The long form of FLIP is an activator of caspase-8 at the Fas death-inducing signaling complex.
J. Biol. Chem.
277
:
45162
.
43
Chang, D. W., Z. Xing, Y. Pan, A. Algeciras-Schimnich, B. C. Barnhart, S. Yaish-Ohad, M. E. Peter, X. Yang.
2002
. c-FLIP(L) is a dual function regulator for caspase-8 activation and CD95-mediated apoptosis.
EMBO J.
21
:
3704
.
44
Muzio, M., B. R. Stockwell, H. R. Stennicke, G. S. Salvesen, V. M. Dixit.
1998
. An induced proximity model for caspase-8 activation.
J. Biol. Chem.
273
:
2926
.
45
Luo, X., I. Budihardjo, H. Zou, C. Slaughter, X. Wang.
1998
. Bid, a Bcl2 interacting protein, mediates cytochrome c release from mitochondria in response to activation of cell surface death receptors.
Cell
94
:
481
.
46
Sun, X. M., S. B. Bratton, M. Butterworth, M. MacFarlane, G. M. Cohen.
2002
. Bcl-2 and Bcl-xL inhibit CD95-mediated apoptosis by preventing mitochondrial release of Smac/DIABLO and subsequent inactivation of X-linked inhibitor-of-apoptosis protein.
J. Biol. Chem.
277
:
11345
.
47
Slee, E. A., M. T. Harte, R. M. Kluck, B. B. Wolf, C. A. Casiano, D. D. Newmeyer, H. G. Wang, J. C. Reed, D. W. Nicholson, E. S. Alnemri, et al
1999
. Ordering the cytochrome c-initiated caspase cascade: hierarchical activation of caspases-2, -3, -6, -7, -8, and -10 in a caspase-9-dependent manner.
J. Cell Biol.
144
:
281
.