Granzyme B-Induced Apoptosis Requires Both Direct Caspase Activation and Relief of Caspase Inhibition

Transcription

1 Immunity, Vol. 18, , March, 2003, Copyright 2003 by Cell Press Granzyme B-Induced Apoptosis Requires Both Direct Caspase Activation and Relief of Caspase Inhibition Ing Swie Goping, 1 Michele Barry, 1,3 Peter Liston, 2 Tracy Sawchuk, 1 Gabriela Constantinescu, 1 Karolina M. Michalak, 1 Irene Shostak, 1 Darren L. Roberts, 1 Allison M. Hunter, 2 Robert Korneluk, 2 and R. Chris Bleackley 1, * 1 Department of Biochemistry it induces a death program that results in the destruction of the pathogenic cell. The enzymatic activity of grb is the key to its ability to induce cell death. It is a member of the serine proteinase family with an unusual specificity to cleave substrates at aspartic acid residues (Odake et al., 1991). The first University of Alberta substrate to be identified for grb was the cysteine prote- Edmonton, Alberta T6G 2H7 ase, caspase 3 (Darmon et al., 1995; Quan et al., 1996). 2 Children s Hospital of Eastern Ontario This caspase plays a central role in the effector stage Research Institute of apoptosis in a wide variety of systems. The activation 401 Smyth Road, Room 308 of caspase 3 by grb is a two-step process. Cleavage Ottawa K1H 8L1 occurs at aspartate 175, resulting in the production of Canada fragments p20 and p10. The larger subunit is then processed to p19/17 by the self-catalytic activity of the caspase (Martin et al., 1996), resulting in a fully active Summary dimer composed of p17 and p10. Once activated, mature caspase 3 acts on a number of substrates that are responsible for the biochemical and morphological Cytotoxic lymphocytes employ Granzyme B as a pofeatures of apoptosis. tent initiator of apoptosis to cleave and activate ef- In numerous apoptotic systems, mitochondria also fector caspases. Unexpectedly, cells transfected with play a pivotal role. Recent evidence has suggested that Bcl-2 were resistant to granzyme B-induced killing, these organelles are also involved in the grb pathway suggesting that a mitochondrial pathway was critical. (Heibein et al., 1999; MacDonald et al., 1999). Both mito- Utilizing cells expressing a dominant-negative caschondrial depolarization and cytochrome c leakage ocpase 9, the current study demonstrated that caspase cur in CTL-mediated lysis. Most importantly, these proactivation via the apoptosome was not required. Inceed in a caspase-independent fashion, suggesting that deed, cleavage of caspase 3 to p20 still occurred in other grb substrates are involved. Recently, it has been Bcl-2-transfectants but processing to p17 was blocked. shown that grb can cleave and activate the proapoptotic This blockade was recapitulated by the Inhibitormember of the Bcl-2 family, Bid (Barry et al., 2000). of-apoptosis-protein XIAP and relieved by Smac/ Granzyme-truncated bid (gtbid) has a different amino DIABLO. Thus granzyme B mediates direct cleavage of terminus than the caspase 8 cleaved form (tbid), but caspase 3 and also activates mitochondrial disruption, both translocate to the mitochondria in association with resulting in the release of proapoptotic proteins that Bax to bring about cytochrome c release (Heibein et al., suppress caspase inhibition. Engagement of both 2000; Sutton et al., 2000). The rate of Bid cleavage by pathways is critical for granzyme-induced killing. grb is faster than grb-mediated caspase 3 activation (Pinkoski et al., 2001). This has led to the hypothesis Introduction that the physiologically relevant route to active caspase 3 proceeded primarily through Bid, mitochondrial cytochrome Cytotoxic T lymphocytes (CTL) and Natural Killer (NK) c release, and caspase 9 activation through cells utilize two pathways to activate target cell death; formation of the apoptosome (Jiang and Wang, 2000). Fas and perforin/granzyme. Apoptosis mediated by Fas Caspase activity is regulated not only by control of is required for lymphocyte homeostasis, whereas granuinhibition zymogen activation as stated above, but also via direct lar killing is important for the control of viral infections, of active caspases by Inhibitor of Apoptosis tumor surveillance, and transplant rejection (Barry and Proteins (IAPs). IAPs were originally described in bacu- Bleackley, 2002; Russell and Ley, 2002). Granzymes A lovirus (Crook et al., 1993), although multiple mammalian and B are the major effector molecules of granularal., family members have now been identified (Duckett et mediated killing and initiate distinct pathways to target 1996; Liston et al., 1996; Rothe et al., 1995; Uren et cell apoptosis (Beresford et al., 1999; Shresta et al., al., 1996). The best-characterized IAP to date is XIAP. 1999). Granzyme B (grb) is responsible for rapid inducforms This protein can directly bind and inhibit the processed tion of cell death (Heusel et al., 1994; Shresta et al., of caspases 3, 7, and 9 (Chai et al., 2001; Huang 1995). CTLs and NK cells store granzymes in cytothus et al., 2001; Riedl et al., 2001; Srinivasula et al., 2001), plasmic granules that are vectorally exocytosed toward preventing cleavage of downstream substrates. a target such as a virus infected or tumor cell. GrB is Furthermore, the inhibitory activity of IAPs is modulated then taken up by receptor-mediated endocytosis into by other proteins that are released from mitochondria the cytoplasm of the target (Motyka et al., 2000) where along with cytochrome c, including Smac/DIABLO (Ver- hagen et al., 2000) and Omi/HtrA2 (Hegde et al., 2002; *Correspondence: chris.bleackley@ualberta.ca Martins et al., 2002; Suzuki et al., 2001; Verhagen et 3 Present address: Department of Medical Microbiology and Immunology, al., 2002). Thus, when these proteins are liberated from University of Alberta, Edmonton, Alberta T6G 2S2, Canada. mitochondria, they bind tightly to IAPs and release the

2 Immunity 356 suppression on caspase activity (Chai et al., 2000; Liu et al., 2000; Wu et al., 2000). In this report, we present evidence for a novel twostep regulatory mechanism for grb activation of caspase 3. The pathway depends on the action of grb on both the effector caspase and Bid and involves the release of proteins from mitochondria other than cytochrome c. Results Cells Overexpressing Bcl-2 Are Protected from Granzyme B-Induced Cell Death In order to activate a target cell s apoptotic program, the serine protease grb cleaves many cytosolic substrates, among which are the death-inducing caspases (Barry and Bleackley, 2002; Trapani, 2001). In addition, cells undergoing grb-mediated apoptosis display mitochondrial disruption such as cytochrome c release and loss of membrane potential (Heibein et al., 1999; MacDonald et al., 1999). To understand the functional relevance of the mitochondrial events, we investigated whether the antiapoptotic protein Bcl-2 influenced grb-induced cell death. Jurkat clones overexpressing Bcl-2 (J-Bcl-2) were generated and tested for sensitivity to activation of apoptosis. As expected, these clones were protected from various apoptosis-inducing regimes such as death receptor ligation (Fas/CD95/Apo-1), or activation of the intrinsic pathway with agents such as staurosporine (data not shown). Bcl-2-transfected cells were exposed to grb in the presence of replication-deficient adenovirus (AD). AD functioned, like the granular pore-forming protein perforin, to facilitate the release of grb into the target cell (Froelich et al., 1996). These GrB/AD-treated Bcl-2-transfected cells were resistant to apoptosis (Figure 1). After incubation with grb/ad, 53% of the control Jurkat-neo (J-neo) cells displayed DNA fragmentation as assessed by TUNEL, whereas J-Bcl-2 cells displayed only 0.3% cell death (Figure 1A). Similar results were obtained with grb and perforin. Utilizing the tritium re- lease assay, which measures the production of small DNA fragments, J-neo cells treated for 2 hr with grb/ AD produced 72% tritium release, versus 0.1% tritium from J-Bcl-2 cells (Figure 1B). In addition, J-Bcl-2 cells treated with grb/ad had minimal membrane damage Figure 1. Bcl-2 Expression Protects Cells from GrB/AD-Mediated Cell Death DNA fragmentation as assessed by TUNEL and tritium release assays. (A) 10 6 Jurkat (J), Jurkat-neo (J-neo), or Jurkat-Bcl-2 (J-Bcl-2) cells were treated with 600 ng/ml of granzyme B and adenovirus at an moi of 100 for 2 hr at 37 C. Cells were fixed and then processed for the TUNEL reaction. Percent specific TUNEL positive cells was calculated by Fluorescence Activated Cell Sorting analysis as described in Experimental Procedures. Data are the mean SD of five independent experiments. (B) J, J-neo, and J-Bcl-2 cells were labeled with 3 H-thymidine overas assessed by 51 Cr-release (data not shown). Multiple night and then treated with grb/ad as above. Percent specific 3 H- independent Bcl-2 transfected clones were analyzed, thymidine release was calculated as described in Experimental Procedures. and all exhibited resistance to the induction of apoptosis by granzyme B. Over a longer time frame (24 hr), the J-Bcl-2-expressing cells did display DNA fragmentation, possibly mediated via a caspase-independent route, al- Granzyme B Can Bypass the Protective Effect though there was no apparent cytolysis as assessed by of Dominant-Negative Caspase 9 51 Cr release. Additionally, we assessed other cell targets The first mitochondrial apoptotic pathway to be fully that are not dependent on the mitochondrial pathway elucidated was the assembly of the apoptosome, a for Fas-mediated cell death, such as SKW6.4 Type I multisubunit complex comprised of mitochondrialcells (Scaffidi et al., 1998), and determined that Bcl-2 released cytochrome c, Apaf-1, caspase 9, and datp/ also protected these cells from grb-mediated cell death ATP (Li et al., 1997; Liu et al., 1996; Zou et al., 1997). (data not shown). These results all suggest that a Bcl- Given the Bcl-2 inhibition described above, it might be 2-sensitive pathway, likely mitochondria, was indeed expected that activation of the apoptosome was impornecessary for the full execution of grb-mediated cell death. tant to the grb pathway. In order to test this hypothesis, Jurkat clones overexpressing a dominant-negative ver-

3 Granzyme B-Induced Cleavage of Caspase Figure 2. DN-C9 Expression Protects Cells from Fas Oligomerization-Induced Apoptosis (A) Level of DN-C9 expression. Western blot analysis of J-neo or J-DN-C9 lysates. The epitope-tagged DN-caspase 9, which was of slower mobility than the endogenous caspase 9, was visualized by immunoblotting with anti-caspase 9 antibody. (B) Caspase 9 activity assay. J, J-neo, and J-DN-C9 clones were treated with ( ) or without ( ) anti-fas antibody (250 ng/ml) for 8 hr prior to lysis. Cell lysates were incubated with the caspase 4-, 5-, and 9-specific colorimetric substrate, Ac-LEHD-pNA, for 2 3 hr, and Ac-LEHD-ase activity was measured at 405 nm. Data are the mean SD of three independent experiments. sion of caspase 9 (C287A) (Li et al., 1997) were established. We characterized the Jurkat dominant-negative caspase 9 (J-DN-C9) cell lines with respect to protein expression and caspase 9 catalytic activity. In an analysis of total cell protein, the amount of dominant-negative protein greatly exceeded that of the barely detectable endogenous procaspase 9 protein (Figure 2A). When control cells were exposed to Fas antibody, the cell lysates displayed a considerable amount of caspase 9 catalytic activity as assessed by the cleavage of the peptide substrate Ac-LEHD-pNA (Figure 2B). In con- trast, lysates from the J-DN-C9 cells had a significantly reduced amount of LEHD-ase activity, indicating that the overexpression of the dominant-negative protein did indeed attenuate the enzymatic activity of the wild-type protein. The J-DN-C9 cells expressed wild-type levels of Fas, procaspase 8, FADD, Bax, and Bid and also showed resistance to UV-mediated apoptosis (data not shown), suggesting that these cells were able to overcome Fas-dependent killing due to the ablation of caspase 9 activity and not due to secondary mutations of the Fas pathway. The J-DN-C9 clones were then tested for sensitivity to apoptosis triggered by grb. Cells were exposed to grb/ad and assessed for DNA fragmentation and caspase 3 cleavage (Figure 3). The percent of TUNEL-positive grb-treated Jurkat cells (27%) was similar to that observed with grb-treated J-DN-C9 cells (26%), although the level was slightly reduced relative to J-neo cells (44%) (Figure 3A). Similarly, the production of small DNA fragments as assessed by the detection of 3 H-thymidine released from grb-treated cells was 77% by J-neo cells and 60% by J-DN-C9 cells (Figure 3B). Membrane damage of grb-treated cells was also measured by the release of 51 Cr from labeled cells, and again, J-DN-C9 and control cells displayed similar levels (data not shown). These results contrast with those seen with anti-fas antibody treatment. Anti-Fas was able to induce apoptosis of J-neo cells (59% TUNEL positive, 64% 3 H- thymidine release), whereas the J-DN-C9 cells were largely resistant (6% TUNEL positive, 0% 3 H-thymidine release). The caspase 3 cleavage profiles of the cells was assessed by Western blotting of cellular proteins from untreated, grb-, or anti-fas-treated cells. As can be seen in Figure 3C, treatment of J-neo cells with grb/ AD produced the mature caspase 3 cleavage products, p19/17 (Figure 3C, compare lanes 1 and 2). The same pattern was seen in grb-treated J-DN-C9 cells (Figure 3C, lane 8). In contrast, anti-fas treated J-DN-C9 cells showed no caspase 3 processing (Figure 3C, lane 9). Therefore, although the presence of a dominant-nega- tive caspase 9 molecule in cells inhibited death via the Fas pathway, those same cells were not protected from grb-mediated apoptosis. Caspase 3 Processing Is Blocked at p20 in Bcl-2 Overexpressing Cells Since grb has been shown to directly cleave caspase 3 (Atkinson et al., 1998; Darmon et al., 1995), inhibition of apoptosis by Bcl-2 was unexpected. A somewhat surprising observation was made when we analyzed the caspase 3 cleavage profile of grb/ad-treated J-Bcl-2 cells. Even though these cells did not die (see Figure 1), procaspase 3 was partially processed to the p20 large subunit. As seen in Figure 4A, treatment of J-neo cells with grb/ad resulted in the appearance of the smaller fragments of processed caspase 3 (p20 and p19/17) (Figure 4, lanes 2 and 3). Cells that overexpressed Bcl-2, however, produced only the p20 frag- ment, albeit in lower amounts, and showed no further processing of caspase 3 to p19/17 (Figure 4, lanes 5 and 6). Similar results were obtained with grb and perforin. The level of caspase 3 enzymatic activity in lysates from treated cells was measured using Ac-DEVD-pNA.

4 Immunity 358 Figure 3. Granzyme B Can Mediate Cell Death in J-DN-C9 Cells DNA fragmentation as assessed by TUNEL assay. (A) J, J-neo, or J-DN-C9 cells were left untreated (open bars) or treated with grb/ad (gray bars) or anti-fas antibody (black bars). The percent positive TUNEL cells is graphed as the mean SD of three independent experiments. (B) J, J-neo, or J-DN-C9 cells were treated with grb/ad (gray bars) or anti-fas antibody (black bars). Percent specific 3 H-thymidine release is graphed as the mean SD of three independent experiments. (C) Western blot analysis of J, J-neo, and J-DN-C9 cells that were untreated ( ) or treated with 600 ng/ml grb/ad (grb) or 250 ng/ml anti- Fas (Fas). The membranes were blotted with anti-caspase 3 antibody. The full-length procaspase 3 and processed p19 and p17 fragments are indicated with arrows. The experiment was performed on three separate occasions and produced similar results. The effect of XIAP on grb-induced death was therefore evaluated. Initially, this was tested in an in vitro grb- dependent caspase 3 cleavage assay. HeLa cell lysates (S100) were treated with increasing amounts of purified grb in the presence of either exogenously added GST control protein, or GST-XIAP, and then the endogenous caspase 3 was visualized by Western blot (Figure 5A). GrB addition resulted in the cleavage of caspase 3 to p20 and p19 (Figure 5A, lanes 1 3). Caspase 3 was not further processed to p17, suggesting the requirement of other factors not present in the in vitro system. Never- theless, caspase 3 was enzymatically active as assessed by activity assay. When GST-XIAP was added to the reaction, however, processing of caspase 3 was arrested at the p20 fragment (Figure 5A, lane 9). Addition of GST as a control did not inhibit caspase 3 cleavage (Figure 5A, compare lanes 3 and 6). These results clearly Caspase activity assays revealed a significantly reduced amount of DEVD-ase activity in J-Bcl-2 cells (Figure 4B). Presence of the p20 fragment of caspase 3 in grbtreated J-Bcl-2 cells correlated with a lack of enzymatic activity. Therefore, cells treated with grb/ad were susceptible to initial procaspase 3 processing, but high levels of Bcl-2 prevented subsequent maturation and activation of the caspase. XIAP Overexpression Can Protect Cells from grb-mediated Cell Death The observation that grb-treated Bcl-2 overexpressing cells were unable to completely process caspase 3 sug- gested that inhibitor molecules might be responsible for the arrest of caspase 3 maturation. Such IAPs have been implicated in the regulation of caspases 3, 7, and 9 (Holcik and Korneluk, 2001; Salvesen and Duckett, 2002).

5 Granzyme B-Induced Cleavage of Caspase vented the formation of the p19 fragment of caspase 3 (Figure 5C, compare caspase 3 pattern of XIAP-expressing cells in lane 4 with LacZ-expressing control cells in lane 6). This caspase 3 cleavage pattern was similar to that of grb-treated J-Bcl-2 cells. Furthermore, these XIAP-expressing cells were protected from grb-mediated cell death as determined by TUNEL. Figure 5D is a representative histogram of TUNEL results depicting the inhibition of DNA fragmentation in cells expressing XIAP (1% TUNEL positive), compared to cells expressing LacZ (63% TUNEL positive). Similar results were seen in 3 H-thymidine release assays (data not shown). Therefore, overexpression of XIAP arrests the apoptotic program initiated by grb, and this occurs via a block in caspase 3 activation. Smac/DIABLO Can Relieve XIAP Suppression of Procaspase 3 Proteolytic Cleavage Recent studies have identified a number of proteins, in addition to cytochrome c, that are released from the mitochondria of apoptotic cells. In particular, Smac/DIA- BLO (Du et al., 2000; Verhagen et al., 2000) and Omi/ HtrA2 (Hegde et al., 2002; Martins et al., 2002; Suzuki et al., 2001; Verhagen et al., 2002) have been shown to counter the inhibitory effects of XIAP. Given the mitochondrial involvement in the granzyme pathway, it is possible these molecules are involved in this system. First, the release of Smac/DIABLO in grb-treated cells was monitored. Untreated cells or grb/ad-treated cells were lysed with digitonin, and the cytosolic fractions which were depleted of mitochondria were analyzed. Western blot analysis showed the presence of Smac/ DIABLO in the cytosol from grb-treated cells, but little cytosolic Smac/DIABLO from non-treated cells (Figure 6A). This Smac/DIABLO release from mitochondria correlated with full activation of grb-cleaved caspase 3. In order to test whether Smac/DIABLO could indeed counter the negative effects of XIAP with respect to Figure 4. Bcl-2 Prevents grb-mediated Activation of Caspase 3 grb-mediated apoptosis, we again utilized the in vitro (A) Western blot. J-neo cells (lanes 1 3) or J-Bcl-2 cells (lanes 4 6) system. Peptides generated from the amino-terminal were mock treated (lanes 1 and 4) or treated with grb/ad for 1 hr amino acids of mature Smac/DIABLO have been shown (lanes 2 and 5) or 2 hr (lanes 3 and 6). Cell lysates were separated to promote XIAP-suppressed caspase activation (Srinion SDS-PAGE, and the membranes were blotted with anti-caspase vasula et al., 2000). Therefore, we added Smac/DIABLO 3 antibody. The full-length procaspase 3 and processed p20, p19, and p17 fragments are indicated with arrows. peptide to the granzyme-dependent caspase activation (B) Caspase 3 activity assay. J, J-neo, and J-Bcl-2 cells were treated assay described earlier. In the presence of GST control with ( ) or without ( ) 600 ng/ml grb/ad for 3 hr prior to lysis. Cell protein, purified grb was capable of initiating maturation lysates were incubated with the caspase 3- and 7-specific colori- of caspase 3 in S100 lysates (Figure 6B, compare lanes metric substrate, DEVD-pNA, for 2 hr, and DEVD-ase activity was 1 and 4). When GST-XIAP was incubated with grb and measured at 405 nm. Data are the mean SD of three independent S100, caspase 3 cleavage was blocked at p20 (Figure experiments. 6B, lane 6). The addition of Smac/DIABLO peptide, however, relieved the XIAP-mediated inhibition of caspase demonstrated that XIAP could indeed block grb-dependent caspase 3 maturation but does not, in itself, directly p19 fragment was again produced. Therefore, in vitro, 3 processing (Figure 6B, lane 7). In these reactions, the inhibit grb activity. the XIAP-mediated block in full caspase 3 maturation In order to determine in vivo whether the presence of could be overcome by the amino terminus of mature XIAP would affect the apoptotic ability of grb, we utilized Smac/DIABLO. an adenoviral expression system to overexpress XIAP This led to the testable hypothesis that the presence in HeLa cells. Infection with Adeno-XIAP resulted in a of cytosolic Smac/DIABLO in live cells could overcome significant increase in the amount of XIAP protein as the Bcl-2 block of grb-mediated death. Using an in vivo assessed by Western blotting (Figure 5B, compare lanes system, we transfected HeLa cells with plasmids driving 2 and 3). When these cells were then treated with grb/ the expression of Bcl-2 and either GFP alone, or a GFP- AD and assayed by Western blot for caspase 3 cleavage IETD-Smac fusion construct (Srinivasula et al., 2000). products, the presence of excess XIAP in the cells pre- The mature N terminus of Smac/DIABLO would remain

6 Immunity 360 Figure 5. XIAP Overexpression Protects Cells from grb-mediated Cell Death (A) In vitro caspase 3 activation assay. 30 g of HeLa S100 was incubated with 20 g/ml purified GST (lanes 4 6) or 20 g/ml purified GST- XIAP (lanes 7 9) in the presence of 0.3 g/ml (lanes 2, 5, and 8) or 3 g/ml (lanes 3, 6, and 9) granzyme B at 37 C for 30 min. Reaction products were separated on SDS-PAGE, and the membranes were blotted with anti-caspase 3 antibody. The full-length procaspase 3 and processed p20 and p19 fragments are indicated with arrows. The experiment was performed on four separate occasions and produced similar results. (B) XIAP expression levels. HeLa cells were infected with adeno-xiap (lane 2) or adeno-lacz (lane 3) at a moi of 1, for 48 hr. Cell lysate products were separated on SDS-PAGE, and the membranes were blotted with anti-xiap antibody. The XIAP product is indicated with an arrow. (C) Cells were infected with adeno-xiap (lanes 3 4) or adeno-lacz (lanes 5 6) as described in (B). After a 48 hr incubation, the cells were mock treated (lanes 1, 3, and 5) or treated with 600 ng/ml GrB/AD (lanes 2, 4, and 6) as described in Experimental Procedures. Reaction products were separated on SDS-PAGE, and the membranes were blotted with anti-caspase 3 antibody. The full-length procaspase 3 and processed p20, p19, and p17 fragments are indicated with arrows. The experiment was performed on four separate occasions and produced similar results. (D) TUNEL assay. Cells were infected with adeno-xiap or adeno-lacz as described in (B). After a 48 hr incubation, the cells were mock treated ( ) or treated with 600 ng/ml grb/ad ( ) as described in Experimental Procedures. The cells were then fixed and processed for TUNEL. The experiment was performed on three separate occasions and produced similar results. transfected HeLa cells, the presence of Bcl-2 protected cells from grb-mediated cell death; however, this pro- tection was reversed in the presence of cytosolic Smac/ DIABLO. Similar results were obtained with grb and perforin. blocked by GFP unless the IETD recognition site was cleaved by granzyme B or by activated caspase 8 (Barry et al., 2000). These cells were then treated with grb/ AD, and the morphology of the GFP-positive cells was analyzed by microscopy. As seen in Figure 6C, in the

7 Granzyme B-Induced Cleavage of Caspase Discussion In the model for death induced by CTL, GrB is transferred from the effector to the target where its ability to proteolytically activate caspase 3 provides a potent apoptotic stimulus. However, the ability of Bcl-2 to inhibit the grb pathway (Figure 1) was a confounding finding but has been seen in other reports (MacDonald et al., 1999; Pinkoski et al., 2001; Sutton et al., 1997). The realization that grb also induced a mitochondrial death program (Heibein et al., 1999; MacDonald et al., 1999) that was inhibited by Bcl-2 provided a plausible explanation for the dilemma. GrB cleaves the proapoptotic protein Bid to form gtbid (Barry et al., 2000). The production of gtbid caused the translocation of the proapoptotic protein Bax to mitochondria (Heibein et al., 2000) and induced release of cytochrome c (Alimonti et al., 2001; Heibein et al., 2000; Thomas et al., 2001) and activation of caspase 9 and then 3 (Li et al., 1997). Bak has also been implicated by the demonstration that Bak-deficient cells are resistant to grb-mediated cytochrome c release (Wang et al., 2001). Bcl-2 did not inhibit Bid cleavage (Heibein et al., 2000; Sutton et al., 2000) but aborted the translocation step, thus thwarting cytochrome c release (Heibein et al., 2000). Bcl-2 expression would result in no translocation of cytochrome c and consequently attenuation of apoptosome formation and caspase 9 activation. Thus no caspase amplification cycle could occur and apoptosis would be blocked. This model predicted that caspase 9 activity was required for the granzyme pathway, but this is controversial. Studies utilizing caspase 9 / cells pointed to the importance of caspase 9 (Alimonti et al., 2001). However, the observations that caspase 9 was cleaved relatively late during grb-induced apoptosis (Sutton et al., 2000) and that inhibition of apoptosome formation did not affect the ability of grb to activate caspases in vitro (Bruey et al., 2000), suggested that the apoptosome was not absolutely required. Here, we used a dominantnegative caspase 9 construct transfected into Jurkat to address this issue. We showed, by analysis of both caspase 3 processing and DNA fragmentation, that grb killing was unaffected by the presence of the DN-C9, whereas the Fas pathway was subverted. In all probability though, assembly of a functional apoptosome does Figure 6. Smac/DIABLO Can Relieve XIAP or Bcl-2 Inhibition of grbdid slightly inhibit death (see Figure 3). accelerate the pathway, as the presence of the DN-C9 Mediated Caspase 3 Cleavage (A) Jurkat cells were mock treated (lane 1) or treated with grb/ad An important clue came from analysis of the caspase (lane 2) for 4 hr prior to lysis in buffer A sucrose digitonin. The 3 activation profile in control and Bcl-2 transfected cells cytosolic fractions devoid of mitochondria were separated on SDS- (Figure 4). Whereas empty vector transfected Jurkat PAGE, and the membranes were blotted with anti-smac/diablo cells revealed a normal pattern of caspase fragments antibody. Location of Smac/DIABLO is indicated by an arrow. with a majority of p19, the Bcl-2-expressing cells (B) In vitro caspase 3 cleavage assay. HeLa S100 lysates were treated with 3 g/ml grb (lanes 2 7) in the presence of 20 g/ml showed a build-up of p20. It appeared that the initial GST or 20 g/ml GST-XIAP with (lanes 3, 5, and 7) or without (lanes cleavage event by grb was normal but the self-cata- 1, 2, 4, and 6) 1 mm N7-Smac/DIABLO peptide. The products were lytic activity was blocked. This requirement for caspase separated on SDS-PAGE, and the membranes were blotted with 3 activity to produce further cleavage of the p20 fraganti-caspase 3 antibody. The full-length procaspase 3 and p20 and p19 fragments are indicated with arrows. The experiment was performed on two separate occasions and produced similar results. (C) HeLa cells were transfected on coverslips with 0.5 g GFP- treated with or without grb/ad for 2 hr. Cells were fixed with 4% expressing or GFP-IETD-Smac-expressing plasmid with ( ) or with- paraformaldehyde in PBS and analyzed by microscopy. Data are out ( ) 0.5 g Bcl-2-expressing plasmid. After 24 hr, the cells were the mean SD of four independent experiments.

8 Immunity 362 ment has been reported previously (Sharif-Askari et al., chondrial released Omi/HtrA2 can enhance apoptosis, 2001). Clearly, the caspase 9 amplification pathway is via its BIR3 binding domain, and also through its serine dampened in the Bcl-2-transfected cells. However, the protease activity (Hegde et al., 2002; Martins et al., 2002; critical processing step of caspase 3 from p20 to p19/ Suzuki et al., 2001; Verhagen et al., 2002). Our initial 17 was blocked, thus preventing progression of the apo- observations suggest that Omi, like Smac/DIABLO, can ptotic program. also overcome Bcl-2 block (data not shown). In addition, Candidate proteins that may cause such a caspase- the roles of Endonuclease G (Li et al., 2001) and AIF processing arrest could be members of the IAP family (Susin et al., 1999) have not yet been explored with of proteins (Holcik and Korneluk, 2001) that bind and respect to the grb death pathway. Although most of our inhibit the processed forms of caspase 3, 7, and 9 (De- discussion has focused on the mitochondrial involveveraux et al., 1997; Srinivasula et al., 2001). In particular, ment mediated by Bid, it is possible that other routes XIAP binding to caspase 3 occurs preferentially to the to mitochondrial protein release are possible. Notably in p20 subunit (Deng et al., 2002; Sun et al., 2002), thereby Bid-deficient mice, cytochrome c release is attenuated inhibiting its further maturation to p19/17 (Li et al., 2002; after granzyme B treatment but death still occurs Srinivasula et al., 2000). Therefore, we decided to test (Thomas et al., 2001), and this could be due to release whether XIAP had any effect on the ability of grb to of Smac/DIABLO, Omi, AIF, or Endonuclease G. generate the active components of caspase 3. In an in In total, our results suggest a two-stage model of vitro system of grb-dependent caspase 3 activation, caspase activation that involves a direct contribution of XIAP proved to be a potent inhibitor of caspase 3 pro- grb to create the first cleavage in caspase 3. However, cessing (Figure 5). The activation profile observed was there is a second, indirect role for the proteinase that identical to the earlier pattern observed with grb-treated involves cleavage of Bid and effects on the release of Bcl-2 transfected cells. This suggested that the Bcl-2 mitochondrial proteins. We propose that cytosolic transfected cells had suboptimal caspase activation be- Smac/DIABLO and Omi/HtrA2 contribute to the grb cause the presence of XIAP caused a build-up of p20 pathway by relieving the inhibition of IAPs. The coordior more importantly a lack of processing to p19. nated action of these molecules results in the rapid In order to establish the physiological significance of induction of an efficient apoptotic death program. IAPs these results, we tested whether overexpression of XIAP and Bcl-2 family members could act to disrupt this coorin live cells could attenuate grb-mediated killing. The dinated effort and thus limit lymphocyte-induced killing. results exactly mirrored what had been observed in the Such molecules have been shown to be overexpressed in vitro assays. Infection of cells with an adenovirus that in tumors. Perhaps by interfering with their action in encodes XIAP completely blocked apoptotic cell death vivo it will be possible to re-coordinate the response. as revealed by TUNEL. This could be explained by inhibi- Alternatively, influencing levels of expression of such tion of caspase 9 resulting in less caspase 3 activation. molecules may represent an interesting therapeutic tar- Undoubtedly this occurs, but the critical point was re- get for immunosuppression. vealed by analysis of the pattern of caspase 3 activation. The striking feature in the infected cells was the complete blockage of caspase 3 proteolysis at p20, indicating Experimental Procedures an effect of XIAP on caspase 3. Cell Lines and Reagents Other reports show that caspase 3-processing block The human T cell lymphoma line Jurkat and transfected J-Bcl-2 can be overcome by the protein Smac/DIABLO (Deng cells were maintained as described previously (Heibein et al., 2000). et al., 2002; Li et al., 2002; Sun et al., 2002). Smac/ The J-DN-C9 cells were derived by transfecting Jurkat cells with a DIABLO was first identified as a protein that was rethat plasmid encoding human procaspase 9, C287A. To isolate cells overexpressed the dominant-negative protein, we treated the leased from mitochondria after exposure of cells to UV neomycin-resistant bulk population 2 with anti-fas antibody (250 (Du et al., 2000; Verhagen et al., 2000). Cytosolic Smac/ ng/ml) and created clonal lines of the surviving cells. The levels of DIABLO interacts with IAPs and reduces IAP protection DN-C9 expression were assessed by Western blotting, using anti- (Du et al., 2000; Srinivasula et al., 2001; Verhagen et al., caspase 9 antibody. Three clones were analyzed and showed similar 2000). In fact, the N terminus alone of mature Smac/ results. The data presented are from clone 6. HeLa cells were trans- DIABLO displaces XIAP from caspase 9 to promote casfacturer s fected with Lipofectamine 2000 (Invitrogen) according to the manupase activation in vitro (Srinivasula et al., 2001). Our instructions. GrB was purified from the cytolytic granules of the NK line YT-Indy, as previously described (Caputo et al., 1999). results showed that the N terminus of Smac/DIABLO Human replication-deficient AD type 5d170-3 was purified from incould relieve XIAP suppression of caspase 3 (Figure 6). fected 293 cells, as described (Bett et al., 1994). Caspase 9 antibod- This suggested that in the in vitro system, caspase 9 ies were from Alexis. Caspase 3 antibodies were a kind gift from could promote caspase 3 maturation from p20 to p19. Dr. D. Nicholson. Smac/DIABLO antibodies were generously provided In vivo, however, grb stimulated caspase 3 maturation by Drs. M. MacFarlane and G. Cohen. Anti-human Fas IgM even in the presence of DN-C9, implying that in cells the antibody (clone CH11) was from Upstate Biotechnology. Immuproduction of mature caspase subunits is autocatalytic. noblotted proteins were detected using the Enhanced Chemillum- inescent System (Amersham). The caspase substrates Ac-DEVD- More importantly, however, we show in vivo that overexpna and Ac-LEHD-pNA were from Biomol Research Laboratories. pression of cytosolic Smac/DIABLO abolished the pro- Smac/DIABLO peptide (NH 2 -AVPIAQK-COOH) was from the Alberta tective effect of Bcl-2 on grb-mediated cell death. Peptide Institute. Therefore, mature Smac/DIABLO alone enabled grb to bypass the requirement for mitochondrial dysfunction, Cell Death Assays ultimately allowing cell death. GrB/AD-mediated cell death was carried out as described previously Other proteins that are released from mitochondria (Motyka et al., 2000). In brief, cell/ml were treated with 600 may also be involved in the grb death pathway. Mito- ng of grb and 100 p.f.u. of AD in serum-free media supplemented

9 Granzyme B-Induced Cleavage of Caspase with 0.1% BSA. The cells were incubated from 2 3 hr at 37 C, before the gift of the anti-smac/diablo antibody. This work was supported being harvested and assessed for apoptosis. by grants from the Canadian Institute of Health Research (CIHR) Cells killed via the Fas pathway were treated with anti-human Fas and the National Cancer Institute of Canada. R.C.B. is a CIHR Distinguished antibody at 250 ng/ml for 8 hr prior to being harvested and assessed Scientist, a Medical Scientist of the AHFMR, a Howard for apoptosis. Hughes International Scholar, and a Canada Research Chair. M.B. is an Alberta Heritage Foundation of Medical Research Scholar. Apoptosis Detection Systems Analysis of DNA fragmentation from Jurkat cells by TUNEL has been Received: September 4, 2002 previously described (Heibein et al., 1999). For HeLa cells, after grb/ Revised: December 20, 2002 AD treatment, the cells were again subjected to trypsin digestion and fixed overnight at 20 C in 95% methanol:5% acetic acid. Cells References were permeabilzed in PBS 0.1% TWEEN 20. TUNEL reactions (Roche Diagnostics) were done according to the manufacturer s Alimonti, J.B., Shi, L., Baijal, P.K., and Greenberg, A.H. (2001). Graninstructions. Percent specific TUNEL-positive cells was calculated zyme B induces BID-mediated cytochrome c release and mitochon- as follows: [(% positive labeled cells with grb and AD % positive drial permeability transition. J. Biol. Chem. 276, labeled cells without grb and AD)/(100 % positive labeled cells Atkinson, E.A., Barry, M., Darmon, A.J., Shostak, I., Turner, P.C., without grb and AD)] 100. Moyer, R.W., and Bleackley, R.C. (1998). Cytotoxic T lymphocyte- The 3 H-thymidine release assays have been previously described assisted suicide. Caspase 3 activation is primarily the result of the (Darmon et al., 1996). Percent specific 3 H-thymidine release was direct action of granzyme B. J. Biol. Chem. 273, calculated as follows: [(sample cpm spontaneous cpm)/(total Barry, M., and Bleackley, R.C. (2002). Cytotoxic T lymphocytes: all cpm spontaneous cpm)] 100. roads lead to death. Nat. Rev. Immunol. 2, Cell morphology of HeLa cells was assessed on a LSM510 laser Barry, M., Heibein, J.A., Pinkoski, M.J., Lee, S.F., Moyer, R.W., scanning confocal microscope mounted on a Zeiss Axiovert 100 M Green, D.R., and Bleackley, R.C. (2000). Granzyme B short-circuits microscope. GFP-positive cells were visualized by excitation at 488 the need for caspase 8 activity during granule-mediated cytotoxic nm, with the fluorescence signal collected using a nm band T-lymphocyte killing by directly cleaving Bid. Mol. Cell. Biol. 20, pass filter Caspase Activity Assay Beresford, P.J., Xia, Z., Greenberg, A.H., and Lieberman, J. (1999). After treatment of cells with grb/ad or anti-fas, cells were Granzyme A loading induces rapid cytolysis and a novel form of lysed in 200 l of Lysis Buffer (50 mm HEPES [ph 7.4], 10 mm MgCl 2, DNA damage independently of caspase activation. Immunity 10, and 0.25% NP-40) and processed as described (Hughes et al., 1997) Bett, A., Haddara, W., Prevec, L., and Graham, F. (1994). An efficient In Vitro Cleavage of Endogenous Procaspase and flexible system for construction of adenovirus vectors with in- For the in vitro caspase 3 cleavage assay, 10 l of the S100 con- sertions or deletions in early regions 1 and 3. Proc. Natl. Acad. Sci. taining 30 g of protein was incubated with either 0.2 g of GST or USA 91, GST-XIAP (Liston et al., 2001) in the presence of 30 ng of purified Bruey, J.M., Ducasse, C., Bonniaud, P., Ravagnan, L., Susin, S.A., granzyme B in a total volume of 15 l. Where indicated, 1 mm of Diaz-Latoud, C., Gurbuxani, S., Arrigo, A.P., Kroemer, G., Solary, E., Smac/DIABLO peptide was added. After a 30 min incubation at and Garrido, C. (2000). Hsp27 negatively regulates cell death by 30 C, the lysates were analyzed for the presence of procaspase 3 interacting with cytochrome c. Nat. Cell Biol. 2, cleavage products by Western blotting. Caputo, A., Parrish, J.C., James, M.N., Powers, J.C., and Bleackley, R.C. (1999). Electrostatic reversal of serine proteinase substrate Adenovirus-XIAP Infection of HeLa Cells specificity. Proteins 35, Infection of HeLa cells with adeno-lacz and adeno-xiap was done Chai, J., Du, C., Wu, J.W., Kyin, S., Wang, X., and Shi, Y. (2000). as described (Li et al., 1998; Xu et al., 1997). Subconfluent HeLa Structural and biochemical basis of apoptotic activation by Smac/ cells (approximately cells/well of a six-well dish) were in- DIABLO. Nature 406, fected with Adeno-XIAP or Adeno-LacZ control virus at an moi of Chai, J., Shiozaki, E., Srinivasula, S.M., Wu, Q., Dataa, P., Alnemri, 1, in 600 l of complete media for 90 min at 37 C. After 48 hr, the E.S., and Shi, Y. (2001). Structural basis of caspase-7 inhibition by cells were trypsinized, left to recover at 37 C for 1 hr, and then XIAP. Cell 104, processed for susceptibility to grb-mediated killing as described above. Crook, N., Clem, R., and Miller, L. (1993). An apoptosis-inhibiting baculovirus gene with a zinc finger-like motif. J. Virol. 67, Cell Fractionation Darmon, A.J., Nicholson, D.W., and Bleackley, R.C. (1995). Activa- The separation of cytosolic fractions from heavy membrane (mitogranzyme tion of the apoptotic protease CPP32 by cytotoxic T-cell-derived chondria containing) fractions that were analyzed for Smac/DIABLO B. Nature 377, release, were performed as follows. Jurkat cells that were untreated Darmon, A.J., Ley, T.J., Nicholson, D.W., and Bleackley, R.C. (1996). or treated with grb/ad were harvested, washed, and resuspended Cleavage of CPP32 by granzyme B represents a critical role for in Buffer A sucrose (20 mm HEPES [ph 7.5], 10 mm KCl, 1.5 mm granzyme B in the induction of target cell DNA fragmentation. J. MgCl 2, 1 mm EDTA, 1 mm EGTA, 250 mm sucrose) 0.019% digito- Biol. Chem. 271, nin and lysed on ice for 15 min, then fractionated according to Deng, Y., Lin, Y., and Wu, X. (2002). TRAIL-induced apoptosis re- Heibein et al. (2000). quires Bax-dependent mitochondrial release of Smac/DIABLO. HeLa S100 lysates (Yang et al., 1997) were prepared from Genes Dev. 16, cells lysed in 900 l of Buffer A sucrose with 15 strokes of a Dounce homogenizer fitted with a B pestle. The supernatant recovered from Deveraux, Q.L., Takahashi, R., Salvesen, G.S., and Reed, J.C. (1997). a 15 min 100,000 g spin was designated as S100. X-linked IAP is a direct inhibitor of cell-death proteases. Nature 388, Acknowledgments Du, C., Fang, M., Li, Y., Li, L., and Wang, X. (2000). Smac, a mitochondrial protein that promotes cytochrome c-dependent caspase activation We thank Dr. B. Motyka for helpful suggestions, K. Veugelers for by eliminating IAP inhibition. Cell 102, information on TUNEL with HeLa cells, Dr. X.J. Sun, C. Frantz, and Duckett, C.S., Nava, V.E., Gedrich, R.W., Clem, R.J., Van Dongen, G. Baron for advice with microscopy, Drs. E. Alnemri and S. Srinivasula J.L., Gilfillan, M.C., Shiels, H., Hardwick, J.M., and Thompson, C.B. for the dominant-negative caspase 9 and GFP-IETD-Smac plas- (1996). A conserved family of cellular genes related to the baculovi- mids, Drs. M. Peter and A. Strasser for SKW6.4neo and Bcl-2 cells, Dr. D. Hudig for perforin, and Drs. M. MacFarlane and G. Cohen for rus iap gene and encoding apoptosis inhibitors. EMBO J. 15,

10 Immunity 364 Froelich, C.J., Orth, K., Turbov, J., Seth, P., Gottlieb, R., Babior, B., C.A., O Brien, G.A., Fitzgerald, P., Tan, E.M., Bokoch, G.M., Shah, G.M., Bleackley, R.C., Dixit, V.M., and Hanna, W. (1996). New Greenberg, A.H., and Green, D.R. (1996). The cytotoxic cell protease paradigm for lymphocyte granule-mediated cytotoxicity. Target granzyme B initiates apoptosis in a cell-free system by proteolytic cells bind and internalize granzyme B, but an endosomolytic agent processing and activation of the ICE/CED-3 family protease, CPP32, is necessary for cytosolic delivery and subsequent apoptosis. J. via a novel two-step mechanism. EMBO J. 15, Biol. Chem. 271, Martins, L.M., Iaccarino, I., Tenev, T., Gschmeissner, S., Totty, N.F., Hegde, R., Srinivasula, S.M., Zhang, Z., Wassell, R., Mukattash, R., Lemoine, N.R., Savopoulos, J., Gray, C.W., Creasy, C.L., Dingwall, Cilenti, L., DuBois, G., Lazebnik, Y., Zervos, A.S., Fernandes- C., and Downward, J. (2002). The serine protease Omi/HtrA2 regu- Alnemri, T., and Alnemri, E.S. (2002). Identification of Omi/HtrA2 as lates apoptosis by binding XIAP through a reaper-like motif. J. Biol. a mitochondrial apoptotic serine protease that disrupts inhibitor of Chem. 277, apoptosis protein-caspase interaction. J. Biol. Chem. 277, Motyka, B., Korbutt, G., Pinkoski, M.J., Heibein, J.A., Caputo, A., Heibein, J.A., Barry, M., Motyka, B., and Bleackley, R.C. (1999). Hobman, M., Barry, M., Shostak, I., Sawchuk, T., Holmes, C.F., et al. Granzyme B-induced loss of mitochondrial inner membrane poten- (2000). Mannose 6-phosphate/insulin-like growth factor II receptor is tial (Delta Psi m) and cytochrome c release are caspase indepen- a death receptor for granzyme B during cytotoxic T cell-induced dent. J. Immunol. 163, apoptosis. Cell 103, Heibein, J.A., Goping, I.S., Barry, M., Pinkoski, M.J., Shore, G.C., Odake, S., Kam, C.M., Narasimhan, L., Poe, M., Blake, J.T., Krahen- Green, D.R., and Bleackley, R.C. (2000). Granzyme B-mediated cyto- buhl, O., Tschopp, J., and Powers, J.C. (1991). Human and murine chrome c release is regulated by the Bcl-2 family members bid and cytotoxic T lymphocyte serine proteases: subsite mapping with pep- Bax. J. Exp. Med. 192, tide thioester substrates and inhibition of enzyme activity and cytolysis Heusel, J.W., Wesselschmidt, R.L., Shresta, S., Russell, J.H., and by isocoumarins. Biochemistry 30, Ley, T.J. (1994). Cytotoxic lymphocytes require granzyme B for the Pinkoski, M.J., Waterhouse, N.J., Heibein, J.A., Wolf, B.B., Kuwana, rapid induction of DNA fragmentation and apoptosis in allogeneic T., Goldstein, J.C., Newmeyer, D.D., Bleackley, R.C., and Green, D.R. target cells. Cell 76, (2001). Granzyme B-mediated apoptosis proceeds predominantly Holcik, M., and Korneluk, R.G. (2001). XIAP, the guardian angel. Nat. through a Bcl-2-inhibitable mitochondrial pathway. J. Biol. Chem. Rev. Mol. Cell Biol. 2, , Huang, Y., Park, Y.C., Rich, R.L., Segal, D., Myszka, D.G., and Wu, Quan, L.T., Tewari, M., O Rourke, K., Dixit, V., Snipas, S.J., Poirier, H. (2001). Structural basis of caspase inhibition by XIAP: differential G.G., Ray, C., Pickup, D.J., and Salvesen, G.S. (1996). Proteolytic roles of the linker versus the BIR domain. Cell 104, activation of the cell death protease Yama/CPP32 by granzyme B. Hughes, F.M., Jr., Bortner, C.D., Purdy, G.D., and Cidlowski, J.A. Proc. Natl. Acad. Sci. USA 93, (1997). Intracellular K suppresses the activation of apoptosis in Riedl, S.J., Renatus, M., Schwarzenbacher, R., Zhou, Q., Sun, C., lymphocytes. J. Biol. Chem. 272, Fesik, S.W., Liddington, R.C., and Salvesen, G.S. (2001). Structural Jiang, X., and Wang, X. (2000). Cytochrome c promotes caspase-9 basis for the inhibition of caspase-3 by XIAP. Cell 104, activation by inducing nucleotide binding to Apaf-1. J. Biol. Chem. Rothe, M., Pan, M.G., Henzel, W.J., Ayres, T.M., and Goeddel, D.V. 275, (1995). The TNFR2-TRAF signaling complex contains two novel pro- Li, J., Kim, J.M., Liston, P., Li, M., Miyazaki, T., Mackenzie, A.E., teins related to baculoviral inhibitor of apoptosis proteins. Cell 83, Korneluk, R.G., and Tsang, B.K. (1998). Expression of inhibitor of apoptosis proteins (IAPs) in rat granulosa cells during ovarian follicu- Russell, J.H., and Ley, T.J. (2002). Lymphocyte-mediated cytotoxiclar development and atresia. Endocrinology 139, ity. Annu. Rev. Immunol. 20, Li, L.Y., Luo, X., and Wang, X. (2001). Endonuclease G is an apoptotic Salvesen, G.S., and Duckett, C.S. (2002). IAP proteins: blocking the DNase when released from mitochondria. Nature 412, road to death s door. Nat. Rev. Mol. Cell Biol. 3, Li, P., Nijhawan, D., Budihardjo, I., Srinivasula, S.M., Ahmad, M., Scaffidi, C., Fulda, S., Srinivasan, A., Friesen, C., Li, F., Tomaselli, Alnemri, E.S., and Wang, X. (1997). Cytochrome c and datp-depen- K.J., Debatin, K.M., Krammer, P.H., and Peter, M.E. (1998). Two dent formation of Apaf-1/caspase-9 complex initiates an apoptotic CD95 (APO-1/Fas) signaling pathways. EMBO J. 17, protease cascade. Cell 91, Sharif-Askari, E., Alam, A., Rheaume, E., Beresford, P.J., Scotto, C., Li, S., Zhao, Y., He, X., Kim, T.-H., Kuharsky, D.K., Rabinowich, H., Sharma, K., Lee, D., DeWolf, W.E., Nuttall, M.E., Lieberman, J., and Chen, J., Du, C., and Yin, X.-M. (2002). Relief of extrinsic pathway Sekaly, R.P. (2001). Direct cleavage of the human DNA fragmentation inhibition by the Bid-dependent mitochondrial release of Smac in factor-45 by granzyme B induces caspase-activated DNase release Fas-mediated hepatocyte apoptosis. J. Biol. Chem. 277, and DNA fragmentation. EMBO J. 20, Shresta, S., MacIvor, D.M., Heusel, J.W., Russell, J.H., and Ley, T.J. Liston, P., Roy, N., Tamai, K., Lefebvre, C., Baird, S., Cherton-Horvat, (1995). Natural killer and lymphokine-activated killer cells require G., Farahani, R., McLean, M., Ikeda, J.E., MacKenzie, A., and Kor- granzyme B for the rapid induction of apoptosis in susceptible target neluk, R.G. (1996). Suppression of apoptosis in mammalian cells by cells. Proc. Natl. Acad. Sci. USA 92, NAIP and a related family of IAP genes. Nature 379, Shresta, S., Graubert, T.A., Thomas, D.A., Raptis, S.Z., and Ley, T.J. Liston, P., Fong, W.G., Kelly, N.L., Toji, S., Miyazaki, T., Conte, D., (1999). Granzyme A initiates an alternative pathway for granule- Tamai, K., Craig, C.G., McBurney, M.W., and Korneluk, R.G. (2001). mediated apoptosis. Immunity 10, Identification of XAF1 as an antagonist of XIAP anti-caspase activ- Srinivasula, S.M., Datta, P., Fan, X.J., Fernandes-Alnemri, T., Huang, ity. Nat. Cell Biol. 3, Z., and Alnemri, E.S. (2000). Molecular determinants of the caspase- Liu, X., Kim, C.N., Yang, J., Jemmerson, R., and Wang, X. (1996). promoting activity of Smac/DIABLO and its role in the death receptor Induction of apoptotic program in cell-free extracts: requirement for pathway. J. Biol. Chem. 275, datp and cytochrome c. Cell 86, Srinivasula, S.M., Hegde, R., Saleh, A., Datta, P., Shiozaki, E., Chai, Liu, Z., Sun, C., Olejniczak, E.T., Meadows, R.P., Betz, S.F., Oost, J., Lee, R.A., Robbins, P.D., Fernandes-Alnemri, T., Shi, Y., and T., Herrmann, J., Wu, J.C., and Fesik, S.W. (2000). Structural basis Alnemri, E.S. (2001). A conserved XIAP-interaction motif in for binding of Smac/DIABLO to the XIAP BIR3 domain. Nature 408, caspase-9 and Smac/DIABLO regulates caspase activity and apo ptosis. Nature 410, MacDonald, G., Shi, L., Vande Velde, C., Lieberman, J., and Sun, X.M., Bratton, S.B., Butterworth, M., MacFarlane, M., and Co- Greenberg, A.H. (1999). Mitochondria-dependent and -independent hen, G.M. (2002). Bcl-2 and Bcl-xL inhibit CD95-mediated apoptosis regulation of Granzyme B-induced apoptosis. J. Exp. Med. 189, by preventing mitochondrial release of Smac/DIABLO and subse quent inactivation of X-linked inhibitor-of-apoptosis protein. J. Biol. Martin, S.J., Amarante-Mendes, G.P., Shi, L., Chuang, T.H., Casiano, Chem. 277,