Portal de Periódicos da CAPES

Caspase-10 is recruited to and activated at the native TRAIL and CD95 death-inducing signalling complexes in a FADD-dependent manner but can not functionally substitute caspase-8

2002; Springer Nature; Volume: 21; Issue: 17 Linguagem: Inglês

10.1093/emboj/cdf441

1460-2075

Martin R. Sprick, Eva Rieser, Heiko Stahl, Anne Große-Wilde, Markus A. Weigand, Henning Walczak,

NF-κB Signaling Pathways

Article2 September 2002free access Caspase-10 is recruited to and activated at the native TRAIL and CD95 death-inducing signalling complexes in a FADD-dependent manner but can not functionally substitute caspase-8 Martin R. Sprick Martin R. Sprick Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany Search for more papers by this author Eva Rieser Eva Rieser Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany Search for more papers by this author Heiko Stahl Heiko Stahl Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany Search for more papers by this author Anne Grosse-Wilde Anne Grosse-Wilde Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany Search for more papers by this author Markus A. Weigand Markus A. Weigand Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany University of Heidelberg, Clinic of Anaesthesiology, D-69120 Heidelberg, Germany Search for more papers by this author Henning Walczak Corresponding Author Henning Walczak Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany Search for more papers by this author Martin R. Sprick Martin R. Sprick Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany Search for more papers by this author Eva Rieser Eva Rieser Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany Search for more papers by this author Heiko Stahl Heiko Stahl Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany Search for more papers by this author Anne Grosse-Wilde Anne Grosse-Wilde Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany Search for more papers by this author Markus A. Weigand Markus A. Weigand Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany University of Heidelberg, Clinic of Anaesthesiology, D-69120 Heidelberg, Germany Search for more papers by this author Henning Walczak Corresponding Author Henning Walczak Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany Search for more papers by this author Author Information Martin R. Sprick1, Eva Rieser1, Heiko Stahl1, Anne Grosse-Wilde1, Markus A. Weigand1,2 and Henning Walczak 1 1Division for Apoptosis Regulation, Tumour Immunology Program, DKFZ, Heidelberg, Germany 2University of Heidelberg, Clinic of Anaesthesiology, D-69120 Heidelberg, Germany *Corresponding author. E-mail: [email protected] The EMBO Journal (2002)21:4520-4530https://doi.org/10.1093/emboj/cdf441 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The involvement of the death adaptor protein FADD and the apoptosis-initiating caspase-8 in CD95 and TRAIL death signalling has recently been demonstrated by the analysis of the native death-inducing signalling complex (DISC) that forms upon ligand-induced receptor cross-linking. However, the role of caspase-10, the other death-effector-domain-containing caspase besides caspase-8, in death receptor signalling has been controversial. Here we show that caspase-10 is recruited not only to the native TRAIL DISC but also to the native CD95 DISC, and that FADD is necessary for its recruitment to and activation at these two protein complexes. With respect to the function of caspase-10, we show that it is not required for apoptosis induction. In addition, caspase-10 can not substitute for caspase-8, as the defect in apoptosis induction observed in caspase-8-deficient cells could not be rescued by overexpression of caspase-10. Finally, we demonstrate that caspase-10 is cleaved during CD95-induced apoptosis of activated T cells. These results show that caspase-10 activation occurs in primary cells, but that its function differs from that of caspase-8. Introduction Apoptosis is an essential process during the development of the immune system and for the maintenance of T- and B-cell homeostasis. Besides tumour necrosis factor (TNF) itself, two other apoptosis-inducing members of the TNF family, CD95 (APO-1/Fas) ligand (CD95L) and TNF-related apoptosis-inducing ligand (TRAIL/APO-2L), have been shown to be involved in various immunological processes. The involvement of the CD95 receptor–ligand system in inflammation, activation-induced death of peripheral T cells, immune privilege, tumour evasion from the immune system, autoimmunity and AIDS is well established (Nagata, 1997; Krammer, 2000). The functional analysis of the TRAIL receptor–ligand system has been complicated by the fact that a total of five different receptors for this cytokine have been identified (Griffith and Lynch, 1998; Locksley et al., 2001). Recently, TRAIL has been shown to be functional in various processes of the immune system and tumour surveillance (Walczak and Krammer, 2000). Of particular interest was the recent finding that TRAIL was necessary for natural killer cell-mediated suppression of liver metastasis in a mouse tumour model (Smyth et al., 2001; Takeda et al., 2001). The early biochemical events that result in apoptosis induction by TRAIL and CD95L have been studied by analysis of the so-called death-inducing signalling complex (DISC) (Kischkel et al., 1995; Walczak and Sprick, 2001). Cross-linking of CD95 or the two apoptosis-inducing TRAIL receptors results in the recruitment of FADD and caspase-8 (Boldin et al., 1995; Chinnaiyan et al., 1995) to the respective DISC (Boldin et al., 1996; Muzio et al., 1996; Bodmer et al., 2000; Kischkel et al., 2000; Sprick et al., 2000). In a homotypic interaction, the death domain (DD) of FADD binds to the DD of CD95. The death effector domain (DED) of FADD in turn interacts with the DED of pro-caspase-8 and thereby recruits this pro-enzyme to the CD95 DISC (Medema et al., 1997). Pro-caspase-8 is proteolytically cleaved and thereby activated at the DISC. Activated caspase-8 then initiates the apoptosis-executing caspase cascade (Peter et al., 1999). Studies using cells from mice deficient in either FADD (Yeh et al., 1998) or caspase-8 (Varfolomeev et al., 1998) showed that these two proteins are essential for CD95L-induced apoptosis. However, Holler et al. (2000) and Matsumura et al. (2000) reported that human Jurkat cells deficient in caspase-8 underwent caspase-independent necrotic cell death after stimulation with highly active CD95L. The molecular mechanisms responsible for this type of cell death remain largely unclear, although the molecule RIP has been proposed to be essential for CD95L-induced necrosis (Holler et al., 2000). Until recently, the molecular explanation for a rare immunological disorder in children called autoimmune lymphoproliferative syndrome (ALPS) had been restricted to the CD95 receptor–ligand system (Straus et al., 1999; Fischer et al., 2000). However, two patients with severe ALPS have recently been identified who did not carry mutations in either the CD95 or the CD95L gene. Comprehensive analysis of apoptosis-associated genes in these patients revealed mutations in the caspase-10 gene. These mutations were shown to result in a disease termed ALPS II (Wang et al., 1999). In this study, it was proposed that TRAIL resistance of mature DC and activated peripheral T cells from ALPS II patients was due to the mutated non-functional caspase-10 and causative for the disease. Later, one of the mutations identified in this study was found to be a common polymorphism in the Danish population (Gronbaek et al., 2000), although to date, besides the ALPS II patient identified by Wang et al. (1999), no other individual homozygous for this caspase-10 mutation has been described. This finding raised the question whether this mutation in caspase-10 alone is causative for the disease. Caspase-10 (Mch4, FLICE-2) is the second known DED-containing caspase besides caspase-8, and was identified by homology cloning (Fernandes-Alnemri et al., 1996; Vincenz and Dixit, 1997). The FLICE-2 and Mch4 isoforms represent different splice variants derived from the same gene. Later, two additional isoforms were identified (Ng et al., 1999). Mch4 is now known as caspase-10a, FLICE-2 as caspase-10b, and the two recently identified variants as caspase-10c and caspase- 10d. While the caspase-10a, -10b and -10d isoforms represent proteins that contain both the large and small catalytic subunits, the caspase-10c variant represents a truncated form yielding a catalytically inactive molecule. While previous studies on the function of caspase-10 relied on overexpression of the different isoforms or catalytically inactive variants, no detailed study under native conditions with the endogenously expressed proteins has been conducted so far. Protein overexpression experiments are prone to artefacts, yielding a cautionary note to the results obtained in these systems. Furthermore, the expression and distribution of the different caspase-10 isoforms have been unclear, as only mRNA levels have been investigated. We therefore set out to study (i) the endogenous expression levels of caspase-10, (ii) whether it forms part of the TRAIL and the CD95 DISCs under native conditions, and (iii) whether caspase-8 and caspase-10 are functionally redundant or might fulfil different functions. Results Caspase-10 is expressed in three isoforms The role of caspase-10 in CD95- and TRAIL-mediated apoptosis has been controversial. At least four splice variants of caspase-10 have been reported (Fernandes-Alnemri et al., 1996; Vincenz and Dixit, 1997; Ng et al., 1999). Three of the reported caspase-10 isoforms contain the large and the small subunits of caspase-10 and yield potentially active enzymes. The fourth variant, caspase- 10c, codes for a truncated form that lacks the catalytic subunits and, thus, represents a catalytically inactive form of caspase-10 reminiscent of the CAP3 form of caspase-8 (Kischkel et al., 1995; Muzio et al., 1996). To investigate the potential role of caspase-10 during TRAIL- and CD95L-mediated apoptosis, we first set out to identify a caspase-10-specific antibody. Amongst the panel of commercially available antibodies that we tested, only one specifically recognized bands at 55 and 59 kDa in Chinese hamster ovary (CHO) cells transfected with expression plasmids encoding recombinant caspase-10a and -10d, respectively (Figure 1A). The epitope recognized by this antibody is localized in the DED-containing prodomain as the antibody also detects the recombinantly expressed His-tagged prodomain (Figure 1A). All splice variants of caspase-10 reported thus far contain this prodomain. Therefore, this antibody should recognize all known isoforms of caspase-10, while all other antibodies we tested did not react with either the recombinant prodomain or with overexpressed caspase-10a or -10d in CHO cells (data not shown). Figure 1.Three isoforms of caspase-10 are expressed in BJAB and Jurkat cell lines. (A) Specificity of the caspase-10 antibody was determined by blotting decreasing amounts of bacterially expressed His-tagged caspase-10 prodomain or lysates from CHO cells transfected with expression plasmids for human caspase-10d, -10a or a control plasmid. The asterisk denotes a co-purified protein resulting from inefficient termination of translation. (B) Three endogenous isoforms of caspase-10 can be detected in lysates from BJAB and Jurkat cells. Proteins (10 μg) from the indicated cell lines were resolved by SDS–PAGE, blotted and detected with the caspase-10 antibody. Lysates from BJAB cells were resolved on 2D gels spanning pH 5–8, blotted and subsequently detected with caspase-10 antibody. The pI and molecular weight (MW) of the large caspase-10 isoforms correspond to caspase-10d and caspase-10a, and the small protein to the pI and MW of caspase-10c. (C) Overview of the calculated MW and pI of the reported caspase-10 and caspase-8 isoforms. (D) As a control for the pI determination, the 2D-gel blot was reprobed with an antibody to caspase-8, showing the caspase-8 isoforms at the expected positions. Download figure Download PowerPoint Having identified a specific antibody for caspase-10, we investigated which isoforms of caspase-10 are expressed in the human Burkitt's lymphoma cell line BJAB and the human T-cell line Jurkat. Two large isoforms of caspase-10 (p55 and p59) were detected in the lysates derived from both cell lines (Figure 1B), reminiscent of the two large isoforms of caspase-8. At ∼28 kDa, an additional isoform of caspase-10 was detected. To determine to which of the published isoforms the individual bands correspond, we performed two-dimensional (2D) gel electrophoretic analysis (Figure 1B). The calculated molecular weights and isoelectric points (pI) for the different isoforms of caspase-10 are shown in Figure 1C. As a control, the same blot was calibrated by detection of caspase-8, showing that the 2D-determined pI of the p55 and p53 caspase-8 isoforms exactly matched their predicted pI of 4.9 and 5.0, respectively (Figure 1D). Determination of the pI of the two expressed large isoforms of caspase-10 revealed that they represent caspase-10d (p59) and caspase-10a (p55). This analysis further excluded the possibility that one of them could represent caspase-10b (calculated pI 7.3). As the pI of the 28 kDa band matched that of caspase-10c, this protein spot most likely corresponded to this isoform of caspase-10. Caspase-10 is cleaved early during CD95- and TRAIL-induced apoptosis Caspase-8 has been shown to be the apoptosis-initiating caspase in the CD95 and the TRAIL DISCs and, thus far, caspase-8 has been thought to be the earliest caspase to become cleaved during death-receptor-mediated apoptosis. Knowing that caspase-8 is cleaved during TRAIL- and CD95L-induced apoptosis, we next analysed whether caspase-10 might also be processed during death-receptor-mediated apoptosis and whether this cleavage occurs prior to, concomitantly with, or following caspase-8 cleavage (Figure 2). During TRAIL- and CD95L-induced apoptosis, the two large isoforms of caspase-10 were cleaved, resulting in two intermediate cleavage products of caspase-10a and -10d, p43 and p47, respectively (Figure 2). Further cleavage of these two intermediates resulted in the appearance of a 25 kDa band representing the prodomain of caspase-10. Also, the 28 kDa caspase- 10c was cleaved during TRAIL- and CD95-induced apoptosis. Thus, caspase-10 is present in both BJAB and Jurkat cells in three different isoforms, and all of these isoforms are cleaved during TRAIL- and CD95-mediated apoptosis. Depending on cell type and stimulus, this cleavage was apparent between 15 and 30 min after onset of the incubation with TRAIL or CD95L, concomitantly with the appearance of caspase-8 cleavage in the cytoplasm. Figure 2.Caspase-10 is processed in TRAIL- and CD95L-sensitive cell lines after stimulation with the respective ligands. Cells were stimulated for the indicated time periods with TRAIL (1 μg/ml) or CD95L (500 ng/ml). Lysates were prepared, and analysed by western blotting with antibodies to caspase-8 and caspase-10. As a control, the blot containing lysates from cells stimulated with CD95L was reprobed with an antibody to ERK. Cleavage of all caspase-10 isoforms can be observed as early as 15–30 min for stimulation with CD95L (upper panel) or TRAIL (lower panel), respectively. Cleavage of caspase-10 leads to the appearance of two intermediate products, p47 and p43, and a smaller band corresponding to the prodomain at 25 kDa. The cleavage pattern is similar to that observed for caspase-8, which is included as a control. The lower amount of ERK control observed after 8 h of stimulation is the result of almost complete cell death observed at this time point. Download figure Download PowerPoint Caspase-10 is recruited to the CD95 and the TRAIL DISCS in BJAB and Jurkat cells Like caspase-8, caspase-10 contains the tandem DEDs necessary for homotypic interaction with the adapter protein FADD, making recruitment of caspase-10 to death receptor-bound FADD-containing complexes possible. Caspase-8 is recruited to the CD95, the TRAIL-R1 and the TRAIL-R2 signalling complexes after ligand-induced stimulation of the respective receptors in a FADD-dependent fashion. As results concerning protein–protein interactions that rely solely on transient protein expression are prone to artefacts, we analysed the receptor complexes under native conditions. We compared the DISCs that form after stimulation with CD95L and TRAIL in both Jurkat and BJAB cells, and tested whether caspase-10 is recruited to these complexes (Figure 3). To control for efficiency of DISC formation, we analysed recruitment of caspase-8 and FADD, two already known components of the TRAIL and the CD95 DISCs (Figure 3A). Whereas none of the proteins analysed was bound to the non-stimulated receptors, we observed a stimulation-dependent recruitment of FADD and caspase-8 to CD95 and the TRAIL death receptors, as described before (Bodmer et al., 2000; Kischkel et al., 2000; Sprick et al., 2000). For caspase-10 the analysis revealed ligand-induced recruitment of all three expressed caspase-10 isoforms to the native TRAIL DISC and the native CD95 DISC in both cell types (Figure 3B). As with caspase-8, the products of the first activation step of caspase-10a and -10d (p43 and p47) were also associated with the TRAIL and the CD95 DISCs. These two forms represent the prodomain with the p17 subunit of caspase-10a and -10d, respectively, after separation of the p12 subunit by proteolytic cleavage. In addition, the prodomain is also apparent in both DISCs (Figure 3B, lower panel). Figure 3.Caspase-10 is recruited to the native TRAIL and CD95 DISCs. Precipitation of the DISC complexes, which form upon stimulation with TRAIL and CD95L, was performed in BJAB and Jurkat cells. The resulting precipitated protein complexes were separated by SDS–PAGE, and analysed by western blotting. (A) To control for DISC formation, the western blots were probed for the known DISC components caspase-8 and FADD, showing recruitment of these proteins to the TRAIL and the CD95 DISC in BJAB and Jurkat cells. (B) The DISC complexes were also analysed for the presence of caspase-10, showing that caspase-10 is recruited to the TRAIL DISC and the CD95 DISC in both cell lines tested. All three caspase-10 isoforms associate with both complexes. In addition, the intermediate products p47 and p43 and the fragment corresponding to the prodomain p25 are also associated with the CD95 and the TRAIL DISCs. The light chain of the M2 anti-FLAG antibody migrates at the same MW as the caspase-10 isoform. Correlation of heavy and light chain levels reveals caspase-10c association with cross-linked receptors. (C) Caspase-10 associates with homomeric TRAIL-R1 and TRAIL-R2 complexes. Differential TRAIL DISC analysis was carried out by blocking either TRAIL-R1 or TRAIL-R2 with mAbs (10 μg/ml for 20 min at 25°C) against the respective receptors before DISC analysis was performed. The resulting complexes were separated by SDS–PAGE and analysed by western blotting. While precipitation of the unstimulated receptors shows no association of the proteins analysed (lane 1), under conditions where both receptors are stimulated (lane 2), caspase-10 associates with both TRAIL-R2 complexes (lane 3) and TRAIL-R1 complexes (lane 4). Blockage of both receptors completely abolishes DISC formation (lane 5). Download figure Download PowerPoint Caspase-10 is recruited to the TRAIL-R1 and the TRAIL-R2 DISCs So far, no biochemical differences in recruitment of adaptor and effector proteins between TRAIL-R1 and TRAIL-R2 have been identified. To investigate whether caspase-10 is recruited selectively to one of the apoptosis-inducing TRAIL receptors, we performed differential DISC analysis for TRAIL-R1 and TRAIL-R2 on BJAB cells (Figure 3C). Either one of the two receptors was blocked by incubation with a monoclonal antibody (mAb) specific for TRAIL-R1 or TRAIL-R2 before stimulation of the other receptor by the ligand. Efficiency of inhibition of the receptors was demonstrated by complete inhibition of FADD and caspase-8 recruitment upon simultaneous TRAIL-R1 and TRAIL-R2 blockage. The analysis of the resulting complexes shows that both caspase-10a and -10d were recruited to both receptors independently of the other receptor. FADD is necessary for the recruitment of caspase-10 to the CD95, the TRAIL-R1 and the TRAIL-R2 DISCs FADD has been shown to be required for the recruitment and subsequent activation of caspase-8 to the TRAIL-R2 and the CD95 DISCs. To test whether FADD is also required for recruitment of caspase-10 to the CD95, TRAIL-R1 and TRAIL-R2 DISCs, we investigated the respective protein complexes in cells that lack FADD but express all three receptors (Figure 4). As the original FADD-deficient Jurkat cells only expressed TRAIL-R2 (Bodmer et al., 2000; Sprick et al., 2000), we established a stable subclone of this cell line transfected with a TRAIL-R1 expression plasmid. This clone, which we named Jurkat FADDdef-TR1/2, expressed TRAIL-R1 and TRAIL-R2 on its surface (Figure 4A). The protein complexes that formed upon stimulation with CD95L and TRAIL are shown in Figure 4B. While the CD95 and the TRAIL DISCs were efficiently formed in the parental Jurkat cells by incubation of these cells with CD95L and TRAIL, respectively, the absence of FADD resulted in failure of the cross-linked receptors to recruit caspase-8 and caspase-10. Thus, FADD is not only necessary for recruitment and activation of caspase-8, but also for recruitment and activation of caspase-10 at the TRAIL and the CD95 DISCs. Figure 4.FADD is the adapter protein for caspase-10 recruitment to TRAIL-R1, TRAIL-R2 and CD95. (A) FACS stain of the FADD- deficient cell line FADDdef-TR1/2. A surface stain for TRAIL-R1 (thin line) and TRAIL-R2 (thick line) shows expression of both TRAIL-R1 and TRAIL-R2. Dotted line, IgG control. (B) DISC analysis was performed with TRAIL and CD95L on wild-type and FADDdef-TR1/2 Jurkat cells. Western blot analysis of the precipitated protein complexes shows that caspase-8, caspase-10 and FADD are efficiently recruited to the TRAIL DISC and the CD95 DISC in the parental wild-type (wt) cells. In the absence of the adapter protein FADD in the cell line Jurkat FADDdef-TR1/2 (FADD def), caspase-10 as well as caspase-8 are not recruited to either the TRAIL DISC or to the CD95 DISC. Download figure Download PowerPoint Caspase-10 is not necessary for death-receptor-mediated apoptosis and caspase-8 activation at the DISC As both caspase-8 and caspase-10 are present in the DISC complexes, it could be possible that both are needed for apoptosis to be initiated at the CD95 death receptors and the TRAIL receptors. We noted that several cell lines, including the Burkitt's lymphoma cell line BL60 and the CD95-transfected BL60 subclone K50, do not express detectable levels of caspase-10. Although K50 cells are devoid of detectable levels of caspase-10 (Figure 5A, upper panel), stimulation with CD95L or TRAIL resulted in normal apoptosis induction (Figure 5B). In addition, caspase-8 is recruited to both the CD95 and the TRAIL DISCs in this cell line, and is activated as indicated by the presence of the cleavage products p43 and p41 in the DISC (Figure 5C). Thus, caspase-10 is dispensable for activation of caspase-8 at the DISC and concomitant apoptosis induction. Figure 5.Caspase-10 is not required for TRAIL- and CD95L-mediated cell death and activation of caspase-8. (A) Western blotting of lysates of different cell lines shows that some cell lines do not express detectable levels of caspase-10. (B) Cells from the cell line K50 were incubated in CD95L or TRAIL at the concentrations indicated. Cell death was measured after 16 h by pI exclusion and FSC/SSC analysis. Both ligands induced cell death in the absence of detectable levels of caspase-10. (C) Analysis of the TRAIL and CD95 DISCs in K50 cells shows that caspase-8 is recruited and activated at both DISC complexes in the absence of caspase-10. The DISCs were precipitated and analysed as in Figure 4B. Download figure Download PowerPoint Caspase-10 can not functionally substitute caspase-8, although it is recruited to the DISC in the absence of caspase-8 It has been shown previously by us and others that Jurkat cells deficient in caspase-8 are defective in apoptosis induction after stimulation with CD95L or TRAIL. However, analysis of expression of caspase-10 in this cell line revealed that the levels of caspase-10 are substantially reduced compared with the parental cell line (Figure 6A). To test whether caspase-10 can induce apoptosis in the absence of caspase-8, we established stable clones that expressed caspase-10d in the absence of caspase-8. To expand our analysis of the TRAIL DISC to TRAIL-R1, we first generated a caspase-8-deficient cell line stably transfected with a TRAIL-R1 expression plasmid, as the Jurkat caspase-8def cell lines normally express only TRAIL-R2. A resulting clone was transfected with an expression plasmid for caspase-10. Several independent caspase-10-expressing clones were chosen for analysis. The amounts of caspase-10 expressed in these clones ranged from relatively low levels to amounts greatly exceeding that of caspase-10 present in the parental cell line, in the parental Jurkat caspase-8def and also in primary T cells (data not shown). Expression of the TRAIL receptors and CD95 was comparable in the different cell lines, as determined by surface staining (data not shown). We next analysed TRAIL sensitivity of these cell lines in comparison to the parental clones. Surprisingly, apoptosis induction was not significantly higher in caspase-10-transfected versus non-transfected caspase-8-deficient cell lines (Figure 6B, upper panel). Figure 6.Caspase-10 can not functionally substitute loss of caspase-8. (A) Comparison of the expression levels of caspase-10 in different cell lines and clones stably transfected with caspase-10. Cell lysates from the cell lines indicated were resolved by SDS–PAGE and blotted. The blots were probed with antibodies against caspase-10 and caspase-8. (B) High expression levels of caspase-10 do not sensitize caspase-8-deficient cell lines to TRAIL- or CD95L-induced apoptosis. Different cell lines were incubated with either TRAIL (upper panel) or CD95L (lower panel) at the concentrations indicated. Cell death was assessed after 24 h by measuring the decrease in FSC/SSC and pI exclusion. (C) Caspase-10 is recruited to and activated at the CD95 DISC in the absence of caspase-8. Cells were stimulated with CD95L at 100 ng/ml for 15 min. After lysis, the DISC complexes were precipitated with an antibody against CD95 (anti-APO-1), resolved by SDS–PAGE and blotted. The blots were probed for caspase-10, showing recruitment of caspase-10 in the caspase-8-deficient cell line 1D2 and in the caspase-10-transfected caspase-8-deficient cell lines (clones 41, 23 and 30). Caspase-10 is recruited to the CD95 DISC in all cell lines even in the absence of caspase-8. Also, in all cases, caspase-10 is activated in these complexes, as indicated by the appearance of the p47/p43 fragments. Lysate denotes the lane containing control lysate from unstimulated clone 41. Download figure Download PowerPoint Stimulation with CD95L led to the induction of cell death in the caspase-8-deficient cell lines, although much higher amounts of CD95L were needed when comparing caspase-8-deficient cell lines with the caspase-8-expressing control cell line (Figure 6B, lower panel). Recently, cell death induction by CD95L in caspase-8-deficient Jurkat cells was described (Holler et al., 2000; Matsumura et al., 2000). However, as for stimulation with TRAIL, overexpression of caspase-10 did not result in a sensitization to CD95L-induced apoptosis when compared with the parental caspase-8-deficient cell line. The inability of caspase-10 to substitute for caspase-8 could be due to its inability to be activated at the DISC complexes in the absence of caspase-8. Yet, analysis of the DISC shows that caspase-10 is also recruited to and activated at the CD95 DISC in the absence of caspase-8 (Figure 6C). Together with the findings presented in Figure 6B, this result shows that although caspase-10 is recruited to and activated at the DISC, it can not functionally substitute for caspase-8 in apoptosis induction. Caspase-10 is expressed in primary human T cells and is cleaved upon CD95-induced apoptosis of pre-activated human T cells Mutations in caspase-10 have been shown to result in the human ALPS II that is characterized by a defect in T-cell and dendritic cell apoptosis (Wang et al., 1999). These results prompted us to test whether caspase-10 might play a role in apoptosis induction in primary cells. We therefore analysed whether caspase-10 is cleaved during CD95-induced apoptosis of pre-activated human T cells. This analysis showed that caspase-10a, -10d and -10c were present in lysates obtained from these cells, and that all caspase-10 isoforms were cleaved following CD95 stimulation (Figure 7). The observed cleavage pattern was essentially identical to the pattern observed in BJAB and Jurkat cells (Figure 2). The proteolytic activation of caspase-10 during apoptosis of activated T cells is indicative of a role for this caspase in the transmission of signals emanating from death receptors not only in transformed cell lines but also in