A Decade of Caspases

A decade of caspases

Alexei Degterev1, Michael Boyce1 and Junying Yuan*,1

1Department of Cell Biology, Harvard Medical School, 240 Longwood Ave., Boston, MA 02115, USA

Caspases are a family of cysteine proteases that play mentation, phosphatidylserine exposure and, finally, important roles in regulating apoptosis. A decade of fragmentation into membrane-enclosed apoptotic research has generated a wealth of information on the bodies sequestered by macrophages or other engulfing signal transduction pathways mediated by caspases, the cells (Wyllie et al., 1980). distinct functions of individual caspases and the mechan- Apoptosis is critical for the maintenance of normal isms by which caspases mediate apoptosis and a variety of homeostasis, on par with other important processes like physiological and pathological processes. proliferation and differentiation. The dysregulation of Oncogene (2003) 22, 8543–8567. doi:10.1038/sj.onc.1207107 such a powerful process can lead to disastrous consequences, and is involved in many diseases such as Keywords: caspase; apoptosis; protease cancer, autoimmunity and neurodegeneration. The central components of PCD machine in C. elegans are a troika of ced genes: ced-3, ced-4 and ced-9, where ced-9 protein is an inhibitor of apoptosis, Overview of apoptosis ced-4 protein is a proapoptotic adaptor molecule and ced-3 protein is a cysteine protease responsible for the The intriguing observation that cell death is a part of execution of the apoptotic program (Hengartner et al., normal development was first made more than 150 years 1992; Yuan and Horvitz, 1992; Yuan et al., 1993). ago (Vogt, 1842), but interest in this area remained Homologues of these genes have been clearly shown to marginal until the end of the 20th century. Lockshin and regulate apoptosis in higher eukaryotes, albeit utilizing Williams (1965) proposed the term ‘programmed cell much more evolved and complex mechanisms. The death’ (PCD) to describe an observation that some cells mammalian homologues of ced-9, the Bcl-2 family were destined to die during tadpole and insect meta- members, are critical regulators of the mitochondrial morphosis, as if driven by a cell-intrinsic program. In step in apoptosis (Gross et al., 1999). Apaf-1, a 1972, Kerr, Wyllie and Currie introduced the term mammalian ced-4 homologue (Zou et al., 1997), is a ‘apoptosis’ to describe cell death associated with a set of regulator of the postmitochondrial, cytochrome-c-de- common morphological features. In the late 1970s, the pendent step (Budihardjo et al., 1999). Finally, a large studies of C. elegans cell lineage by Horvitz and Sulston family of cysteine proteases, termed caspases, is led to the discovery that 131 of 1090 C. elegans somatic responsible for the execution of apoptosis in a multitude cells invariantly die during normal development, paving of higher eucaryotes (Cryns and Yuan, 1998). the way for genetic characterization of the critical This year is the 10th anniversary of the discovery of components of apoptotic molecular machinery. The the critical role of caspases in apoptosis (Miura et al., genetic analyses of C. elegans PCD, carried out mostly 1993; Yuan et al., 1993). In this review, we will in the laboratory by H. Robert Horvitz at MIT, led to summarize the current knowledge of this protease family the elucidation of genes that control this cellular suicide in apoptosis and other physiological and pathological mechanism (Ellis and Horvitz, 1986; Hengartner et al., processes. 1992). The molecular identities of two genes that are key to C. elegans cell death, ced-3 and ced-9, and their homology to vertebrate counterparts, caspase family General features and classification of caspases and Bcl-2 family (Yuan et al., 1993; Hengartner and Horvitz, 1994), thrust apoptosis from relative obscurity The history of caspases began with the discovery that into the mainstream of biological research. the C. elegans ced-3 gene (Yuan et al., 1993) encodes a Apoptosis is a genetically encoded, ubiquitous path- homologue of the independently identified human way enabling cells to undergo highly regulated death in interleukin-1b processing enzyme (ICE) (Cerretti et al., response to prodeath signaling. Apoptosis is character- 1992; Thornberry et al., 1992). The subsequent demon- ized by a number of unique distinguishing features, stration that overexpression of ICE (renamed caspase-1) including cytoplasmic shrinkage, membrane blebbing, in mammalian cells is sufficient to induce apoptosis nuclear fragmentation, intranucleosomal DNA frag- suggested that caspases may play a role analogous to that of ced-3 in mammalian PCD (Miura et al., 1993). *Correspondence: Junying Yuan; E-mail: [email protected] The demonstration that the expression of a caspase Decade of caspases A Degterev et al 8544 inhibitor protein in sensory neurons prevented neuronal cell death induced by trophic factor deprivation provided the first evidence for a functional role of caspases in neuronal cell death (Gagliardini et al., 1994). Finally, a member of the caspase family, caspase-3, was shown to be a critical mediator of the endogenous apoptotic pathway in mammalian cells (Nicholson et al., 1995). The identification of caspases in C. elegans, Drosophila and all studied vertebrate species demonstrated the evolutionary conservation of apoptotic machinery. Currently, 11 human caspases have been identified: caspase-1–10 and caspase-14 (Alnemri et al., 1996; Pistritto et al., 2002). The protein initially named caspase-13 was later found to represent a bovine homologue of caspase-4 (Koenig et al., 2001), and caspase-11 and -12 are murine enzymes that are most likely the homologues of human caspase-4 and -5(see below). There is a clear evolutionary tendency to increase the number of caspases over phylogenetic time, from four in C. elegans to seven in Drosophila and 11 in mice and humans (Lamkanfi et al., 2002). All caspases share a number of distinct features. These include the catalytic triad residues, consisting of the active site Cys285, which is a part of the conserved QACXG pentapeptide sequence, His237 and the backbone carbonyl of residue 177 (caspase-1 numbering) (Stennicke and Salvesen, 1999). A striking feature of the caspase family is its specificity for substrate cleavage Figure 1 Classification of caspases. (a) Classification of the after an Asp residue, which is unique among caspase family based on reported functions. (b) General structure mammalian proteases, except for the serine protease of caspases and classification based on the prodomain length. (c) Caspase substrate specificities. Data are based on Thornberry et al. granzyme B (see below). All caspases are synthesized as (1997). Preferred amino acids in P4-P1 positions are shown. Based inactive zymogens containing a prodomain followed by on the size of the S4 subsite and P4 residue, caspases can be divided p20 large and p17 small subunits (Figure 1b). The into three subfamilies. (d) Initial processing site sequences in activation of the zymogen precursor is mediated by a caspases. Note that they fit the preferred sequence for the initiator series of cleavage events, first separating the large and subfamily small subunits, followed by the removal of the prodomain (Figure 2c) (Ramage et al., 1995; Yamin et al., 1996). Multiple rubrics have been used to classify caspases typically processed and activated by upstream caspases. but, interestingly, all methods result in similar group- The downstream caspases form an ‘executioner’ class ings, suggesting tight structure-function correlations (caspase-3, -6, -7) and are characterized by the presence within the caspase family. The first classification method of a short prodomain. is based on caspase function. During apoptosis, uniform Protease families can be also classified according to execution machinery is triggered by a wide variety of their substrate preferences. Thornberry and co-workers extra- and intracellular signals. Therefore, some cas- developed a combinatorial method to assess caspase pases have evolved to link distinct upstream signaling substrate specificity in vitro using peptide-aminomethyl- pathways with the downstream execution steps. These coumarin (AMC) substrates (Rano et al., 1997). Using upstream family members are termed ‘initiator’ caspases this approach, the preferred tetrapeptide substrates for (Figure 1a). Initiator caspases possess long prodomains caspase-1–11 were determined (Thornberry et al., 1997; containing one of two characteristic protein–protein Kang et al., 2000) (Figure 1c). Caspases separate into interaction motifs: the death effector domain (DED) three classes based on the identity of the residue in the (caspase-8 and -10) and the caspase activation and P4 position of the substrate, which is a primary recruitment domain (CARD) (caspase-1, -2, -4, -5, -9, determinant of specificity. The first group, which -11 and -12) (Figure 1b), providing the basis for includes caspase-1, -4 and -5, prefers bulky, hydro- interaction with upstream adaptor molecules. Among phobic side chains in the P4 substrate position, with the the initiator caspases, caspase-1, -11 and -5form a optimal sequence of WEHD (class I). The second class, subclass of caspases that controls both apoptosis and consisting of caspase-6, -8, -9 and -11, has a preference certain inflammatory responses. On the other hand, for (L/V)EXD (class II). Finally, the optimal sequence caspases that perform the downstream execution steps for the third class, consisting of caspase-2, -3 and -7 is of apoptosis by cleaving multiple cellular substrates are DEXD, where X is V, T or H (class III) (Figure 1c).

Oncogene Decade of caspases A Degterev et al 8545 Such analysis on caspase-10 and -12 has not been caspase-3 (see below) (Kang et al., 2000) in addition to published. its role in inflammation. Second, these substrate Several observations can be made based on this preferences fit well with the sites found in the known substrate specificity classification. First, with the excep- preferred in vivo substrates of each functional class tion of caspase-2 and -6, this classification is also (Figure 1a), such as IL-1b and IL-18 (dual role initiator consistent with the above-mentioned functional subdivi- class), laminin A, caspase-3 and -7 (apoptotic initiator sions of the family (Figure 1a). The substrate prefer- class), and PARP and DNA-PK (effector class) (Thorn- ences of caspase-11 are most similar to the initiator berry et al., 1997). Most strikingly, the initial, most group II, consistent with the observation that this important (see below) processing sites between the large caspase can initiate apoptosis through direct cleavage of and small subunits in all initiator caspases conform to their own substrate preferences (Figure 1d), consistent with the ability of the initiator caspases to activate themselves, whereas processing sites in effector caspases fit the preferences of the class II caspases, consistent with being downstream activation targets of initiator caspases (see below). Interestingly, caspase-2 exhibits features of both an initiator caspase as well as an executioner caspase: the presence of a CARD domain in caspase-2 suggests its initiator role; however, both the substrate preference and the cleavage site between the large and small subunits of caspase-2 are similar to that of the executioner caspases, that is, DXXD. Such characteristics of caspase-2 suggest that it may act as a proximal responder to auto-activate in apoptosis signaling under certain conditions (Lassus et al., 2002; Paroni et al., 2002; Robertson et al., 2002) as well as function as an executioner caspase by cleaving cellular targets, rather than activating downstream caspases. The substrate specificity studies provide interesting data that may help explain the differences in the physiological roles of individual caspases, but as these results were derived using a short synthetic substrate in vitro, they should be interpreted with caution. Caspases may require cofactors in vivo, such as the apoptosome complex in the case of caspase-9, which may control the efficiency of catalysis as well as substrate specificity (Rodriguez and Lazebnik, 1999). Therefore, results obtained in vitro using purified recombinant proteins may not accurately reflect the enzymatic potential of the ‘native’ cellular protein. It should also be kept in mind that the in vitro results primarily show sites preferen- tially recognized by a particular caspase. They do not take into account intrinsic differences in caspases

Figure 2 Caspase structure and mechanism of action. (A) Ribbon representation of the caspase-3 structure showing positions of the active center forming loops (L1–L4, L20) (reproduced from Shi, 2002) based on the crystal structure of the complex of caspase-3 with Ac-DVAD-fmk (Mittl et al., 1997). The inhibitor is shown in purple. (B) The active site conformations of the caspases with known structures (1, 3, 7, 8 and 9). Loops L1 and L3 are highly conserved, whereas L2 and L4 are primarily responsible for the differences in substrate speciﬁcity. Reproduced from Shi (2002). (C) Generalized distribution of caspase catalytic center loops (L1– L4) on small and large subunits. Loops are shown in blue. Position of the activating cleavage processing are shown by arrows. Numbers represent the order of the activation cleavages. The active site Cys is shown by a red asterisk. Processing occurs in L2. The resulting large subunit portion of the L2 loop of one monomer and small subunit portion of the L2 loop of another monomer (L20) are involved in loop bundle formation (a, b). (D) Interactions in the active sites of caspases based on the crystal structures of active caspase-1, -3, -7 and -8 with DEVD-CHO (reproduced from Wei et al., 2000)

Oncogene Decade of caspases A Degterev et al 8546 activity or local concentration, which may result in some acid accepted in this position (Thornberry et al., 1997). caspase cleaving its ‘nonoptimal’ substrate at a higher At P3, Glu is preferred by all caspases (Thornberry et al., rate than another, less active caspase, which nevertheless 1997) due to its interaction with the surface-exposed prefers that site more strongly. Therefore, the in vitro Arg341 and, in the case of caspase-8, also with Arg177 data may not accurately point to the preferred caspase (Blanchard et al., 1999) (Figure 2d). This interaction is family member cleaving the particular substrate. There- predicted to contribute to the proper positioning of the fore, substrate recognition may primarily serve to substrate (Wei et al., 2000). Additional residues (Ser239 provide the optimal kinetics for regulating sequential and Arg341) coordinate the amino and carbonyl groups events in caspase cascades during apoptosis. of the amide bonds formed by P1 and P3 (Figure 2d). Whereas the S1 subsite interaction distinguishes the caspase family from other proteases (Barrett and Caspase structure and enzymatic mechanism Rawlings, 2001), the selectivity of different caspases is generated by the S4 subsite, formed by the L4 loop of The structures of active caspase-1, -3, -7, -8 and -9 have the small subunit (Shi, 2002; Thornberry et al., 1997). A been solved, providing valuable insights into the basis of comparison of the L4 loops from the available caspase caspase specificity and catalytic mechanisms (Walker structures shows that caspase-3 and -7 have the largest et al., 1994; Wilson et al., 1994; Rotonda et al., 1996; loops, followed by caspase-8 and -9 and then caspase-1 Mittl et al., 1997; Blanchard et al., 1999; Watt et al., (Figures 2b). In addition, the crystal structures of the 1999; Wei et al., 2000; Renatus et al., 2001). The active complexes of caspase-1, -3, -7 and -8 with Ac-DEVD- caspase is a homodimer, with each monomer consisting CHO show that caspase-1 forms the lowest number of of a large and a small subunit (Figure 2a). Six hydrogen bonds with the P4 Asp residue and possesses antiparallel b strands of each monomer form a the smallest (Val versus Trp for caspase-3, -7 and -8) continuous 12-stranded b sheet in the active enzyme residue at the bottom of the S4 pocket (Wei et al., 2000) with a separate active center formed by each monomer (Figures 2d). All of these factors contribute to caspase-1 symmetrically located on the opposite sides of the having the widest S4-binding site, which favors large enzyme (Figure 2a). This topology is different for hydrophobic residues. In contrast, caspase-3 and -7 caspase-9, in which case only one active center is formed possess the narrowest pocket, which, along with an (Renatus et al., 2001). extensive hydrogen bonding pattern (Figure 2d), creates Caspase substrate specificity is predominantly deter- a preference for the Asp residue in the P4 position mined by the four amino-acid residues N-terminal to the (Rano et al., 1997; Wei et al., 2000). The S4 pockets of scissile bond (P4-P1) (Margolin et al., 1997). The caspase-8 and -9 occupy intermediate positions and caspase cleavage mechanism is based on the catalytic show preference for small hydrophobic Val or Leu triad composed of Cys285, His237 and the carbonyl of (Rano et al., 1997). However, caspase-8 also efficiently residue 177, which is somewhat unusual, since the side recognizes the class II substrate DEVD (Wei et al., chain of the residue 177 does not appear to play a role 2000). A comparison of the available data also shows (Stennicke and Salvesen, 1999). In addition, an oxya- that the major structural differences among the caspases nion hole consisting of Gly238 and Cys285also are localized to the active sites, with the rest of the contributes to the catalysis. backbone being quite similar (Wei et al., 2000). The entire active center of caspases, consisting of the The use of extended caspase substrates demonstrated S4-S3-S2-S1 specificity subsites binding P4-P3-P2-P1 a surprising selectivity of the family for residues in the residues of the substrate, respectively, is formed by P10 position (Stennicke et al., 2000). Whereas small flexible loops (L1–L4) (Figures 2a–c) (Shi, 2002). Loop (Gly, Ser and Ala) and large aromatic (Phe and Tyr) L1 and a portion of L2, which contains the catalytic Cys amino acids are well tolerated, charged amino acids are residue, are a part of the large subunit, whereas L3 and completely prohibited. L4 come from the small subunit (Figure 2a, c). The In addition to the above-mentioned structural deter- activation-mediated cleavage of caspases occurs in the minants governing caspase substrate specificity, the loop L2, liberating the C-terminus of the large and the executioner caspase-3, -6 and -7 also possess a nine- N-terminus of the small subunits (Figure 2c). The size amino-acid insertion following residue 381, which is and topology of the L1 and L3 loops are highly absent in caspase-1 and -8 (Stennicke et al., 2000). This conserved among various caspases, whereas the L2 loop is postulated to block sterically the entry of and L4 loops display a greater degree of variation inappropriate substrates into the active sites of the (Figures 2b). The S1 subsite in all caspases is formed by executioner caspases, as exemplified by the cowpox the L1, L2 and L3 loops and is very narrow and deep. caspase-inhibitory protein CrmA (see below) (Renatus The restrictive geometry of the S1-binding pocket limits et al., 2000). P1 residue to Asp, providing the molecular basis for the distinctive specificity of the caspase family. The side- chain of the Asp substrate residue is coordinated by three invariant residues (Arg179, Gln283 and Arg341, Cellular pathways controlling caspase activation caspase-1 numbering) (Figure 2d). Since the P2 side chain is facing the outside of the active center, there is a As active caspases can cause rapid cell death, there is a greater degree of variation in the identity of the amino need for stringent mechanisms to control caspase

Oncogene Decade of caspases A Degterev et al 8547 activation. These include the synthesis of caspases as inactive zymogens, highly evolved upstream regulatory pathways controlling the activation and availability of endogenous inhibitors that keep the active enzymes in check. The apoptotic signaling pathways that lead to caspase zymogen processing can be subdivided into two major categories: cell surface sensor-mediated and intracellular sensor-mediated. The former pathway is activated in response to extracellular signals, indicating that the cell’s existence is no longer needed for the well-being of the organism. Cell surface sensor-directed apoptotic signals are initiated by ligands binding to cell surface death-mediating receptors and are exemplified by the signaling of the death receptor family. The death receptor family, including CD95/Fas/ Apo1, TNFR1, TNFR2, DR3/Wsl-1/Tramp, DR4/ TRAIL-R1, DR5/TRAIL-R2/TRICK2/Killer and DR6 (see review on this topic in this issue) is characterized by the presence of multiple cysteine-rich repeats in the extracellular portion and the protein– protein interaction module known as the death domain (DD) in the cytoplasmic tails. Ligand-induced receptor multimerization results in the formation of the death inducing signaling complex (DISC) containing multiple adaptor molecules, including FADD (Fas associated DD), TRADD (TNF receptor associated DD), DAXX, RIP (receptor interacting protein kinase), RAIDD (RIP Figure 3 Pathways of caspase activation. Schematic representa- associated protein with a DD) and FLIP (FLICE-like tion of the cell surface receptor-mediated (receptor-mediated) and the intracellular receptor-mediated (internally initiated by damage, inhibitory protein) (see review on this topic in this issue stress, etc.) pathways of caspase activation. Domains mediating of Oncogene). FADD, which is recruited to the DISC critical interactions (DED, DD, CARD) are shown. Death receptor through its C-terminal DD, interacts through its N- signaling may involve direct caspase-8-mediated caspase-3 activa- terminal DED with the DED of caspase-8 (Muzio et al., tion (type I cells) or a Bid-cleavage-dependent mitochondrial 1996; Figure 3). The recruitment and oligomerization of amplification step (type II cells). Seven-spoked architecture of the apoptosome is based on Acehan et al. (2002) caspase-8 in the DISC results in its autocatalytic activation (see below) and is critical for the initiation of cell death (Juo et al., 1998; Varfolomeev et al., 1998). Caspase-10 and -2 may also be recruited to and activated at the DISC, but their role in death receptor signaling depends on a mitochondrial amplification step signaling remains controversial (Kischkel et al., 2001; (Figure 3), which may be necessary because of an Wang et al., 2001; Sprick et al., 2002). insufficient amount of active caspase-8 or downstream How does the activation of caspase-8 lead to cell caspases. In the type II pathway, caspase-8 cleaves the death? The most direct pathway, present in the so-called cytosolic BH3-only proapoptotic Bcl-2 family member type I cells, involves the processing and activation of the Bid (Li et al., 1998; Luo et al., 1998). The processing of executioner caspase-3 and -7 by caspase-8 (Scaffidi et al., Bid results in its translocation to mitochondria, where it 1998) (Figure 3). These latter caspases execute apoptosis causes cytochrome c release through oligomerization of by cleaving various cellular substrates (see below), the proapoptotic Bcl-2 family members Bax and Bak including other caspases, leading to the branched (Korsmeyer et al., 2000) (Figure 3). Released cyto- cascades of caspase activation (Hirata et al., 1998). chrome c induces the formation of the apoptosome The existence of such caspase cascades, which serve to complex in the cytosol (Figure 3). The apoptosome, amplify apoptotic signaling, may be necessary because which was initially defined by the works of Wang and of the inability of the initiator caspases to complete the co-workers (Liu et al., 1996; Li et al., 1997c; Zou et al., execution of apoptosis on their own. In addition, such 1997), was shown to be a heptamer comprised of seven pathways may enable cells to maintain a tighter control Apaf-1 adaptor molecules, each bound to one molecule over the cellular suicide mechanism at multiple points. of cytochrome c and a dimer of the initiator caspase-9 As noted, the above mechanism for receptor-induced (Cain et al., 2000; Acehan et al., 2002). Caspase-9, which apoptosis is present only in type I cells, which are is activated through an apoptosome-induced conforma- capable of producing a sufficient amount of activated tional change (see below), also processes the executioner caspase-8 to activate enough downstream caspases to caspase-3 and -7 to initiate the execution of apoptosis in complete apoptosis (Scaffidi et al., 1998). In contrast, in a manner similar to that of caspase-8 (Li et al., 1997c; the so-called type II cells, death receptor apoptotic Slee et al., 1999) (Figure 3).

Oncogene Decade of caspases A Degterev et al 8548 The other major category of apoptotic signaling, the injection, most likely by preventing the lymphocyte loss intracellular sensor-mediated pathway, is activated by due to apoptosis (Hotchkiss et al., 2000). stimuli such as DNA damage and cytotoxic drugs, In addition to the above-mentioned caspase-depen- which act inside the cell, and is very similar to the type II dent pathways, multiple noncaspase proteases were also death receptor mechanism in its requirement for signal shown to process and activate caspases directly. For amplification through mitochondrial damage. Cells example, granzyme B, a unique serine protease with possess multiple means of targeting mitochondria, but substrate specificity for aspartate residues, is critical for most of these signals integrate at the level of the Bax/ the killing of virally infected cells by cytotoxic lympho- Bak gateway (see review on Bcl-2 family in this issue of cytes (CTL) and directly activates caspase-3 in target Oncogene) (Figure 3). cells (Darmon et al., 1995). The Ca2 þ -activated protease An important issue with respect to apoptosis signaling m-calpain can process caspase-12 following ER stress pathways is the timing of the ‘point-of-no-return’. This (Nakagawa and Yuan, 2000), whereas the lysosomal event is defined as the step in the signaling process after protease cathepsin B was shown to process caspase-1 which termination of the apoptosis-inducing stimuli and -11 in vitro (Schotte et al., 1998; Vancompernolle does not prevent the execution of apoptosis. Elegant et al., 1998). In summary, cells possess diverse arrays of studies with nerve growth factor (NGF) deprivation- caspase activation cascades, which control the activation mediated apoptosis of sympathetic neurons suggested of specific caspase groups in response to a variety of that the ‘point-of-no-return’ occurs downstream of cytotoxic and inflammatory stimuli. cytochrome c release and coincides with executioner caspase activation, whereas the inhibition of caspase activity delays the commitment to death until mitochon- Functions of individual caspases drial depolarization (Deshmukh et al., 2000). The latter mechanism was further addressed by the analyses of the Our current knowledge suggests that the expansion of effect of cytochrome c readdition on the Fas-mediated the caspase family in higher organisms may reflect a loss of mitochondrial function (Mootha et al., 2001) and need to increase the specialization of cellular signaling in single cell analyses of mitochondrial respiratory func- apoptosis. Here, we will summarize our current knowl- tion following apoptotic induction by UV/cytotoxic edge of the specific roles of individual caspases, with drugs (Waterhouse et al., 2001). These studies suggested special emphasis on the lessons learned from mouse that, apart from caspase activation, cytochrome c knockout studies. release causes slow acquisition of the irreversible loss of mitochondrial function and respiration, which may Caspase-1, -11 and -5 lead to death. Therefore, the mitochondrial apoptotic pathway appears to result in a bipartite ‘point-of-no- Caspase-1 and -11-deficient mice provided clear insights return’ event, consisting of fast caspase activation and into their specific and nonredundant functions. The slow, caspase-independent death through mitochondrial deletion of caspase-1 showed no ill effects on animal dysfunction. development, and embryonic fibroblasts and thymocytes Since caspase-1, -5and -11 form a separate subfamily isolated from these mice displayed full sensitivity to with a dual role in regulating both cytokine processing various apoptotic stimuli (Li et al., 1995; Smith et al., and inflammatory signaling as well as apoptosis, they are 1997) (Table 1). Caspase-1/ mice, however, showed involved in unique activation networks. Mouse genetic specific defects in processing and secretion of the knockout studies have clearly shown that caspase-11 proinflammatory cytokines IL-1b and IL-18 and were mediates the activation of caspase-1, which in turn resistant to LPS-induced septic shock (Kuida et al., processes cytokines IL-1b and IL-18 (Li et al., 1995; 1995; Li et al., 1995; Ghayur et al., 1997), which is Ghayur et al., 1997; Wang et al., 1998). In the case of consistent with caspase-1’s original identification as an human cells, lipopolysaccharide (LPS) induces the IL-1b processing enzyme (Cerretti et al., 1992; Thorn- expression of caspase-5, like that of caspase-11 in murine berry et al., 1992). For as yet unknown reasons, IL-1a cells (Lin et al., 2000), and caspase-5was recently shown secretion, but not processing, is also prevented in the to be involved in the activation of human caspase-1 as a absence of caspase-1 (Kuida et al., 1995; Li et al., 1995). part of an ‘inflammasome’ complex, which also includes Interestingly, even though caspase-1 is most likely a the adaptor molecules NALP1 and Pycard (Martinon downstream target of caspase-11 (see below) (Wang et al., 2002). Based on these similarities, human et al., 1998), the two published lines of caspase-1 caspase-5is most likely a functional human homologue knockout mice cannot be induced to express caspase- of murine caspase-11. As for the induction of apoptosis, 11 (Kang et al., 2000). Thus, although caspase-1 most caspase-11 has been demonstrated to process caspase-3 likely is directly involved in the processing of IL-1b in directly during ischemia and septic shock in addition to vivo, the null phenotype of this caspase is still far from regulating caspase-1 activation (Kang et al., 2000, 2002). clear. The activation of caspase-3 by caspase-11 is directly The expression of an active site mutant caspase-1 in a responsible for the apoptosis of lymphocytes during transgenic mouse inhibited IL-1b secretion, similar to septic shock and, consistent with this conclusion, a the knockout results, and neurons from these mice were novel nonpeptide inhibitor of caspase-3 was found to partially resistant to trophic factor deprivation-induced decrease mortality and bacteremia in vivo following LPS cell death in vitro and ischemic brain damage in vivo

Oncogene Decade of caspases A Degterev et al 8549 Table 1 Phenotypes of the caspase knockout mice Caspase Embryonic lethality Developmental defects Apoptotic defects

1 No No Defects in processing/secretion of IL-1b, IL-1a, IL-18. Resistance to septic shock 11 No No Defects in lymphocyte apoptosis/caspase-3 activation during septic shock, inﬂammatory cytokine/caspase-1 processing. Resistant to septic shock, ischemic brain injury 2 No Reduced number of facial motor- Defects in apoptosis induction by chemotherapeutic neurons in neonatal stages, exces- drugs (only in oocytes), granzyme B (lymphocytes). sive oocyte generation Compensatory activation of other caspases was reported 3 Yes in C57/B6-129X1/SvJ mixed Neuronal hyperplasia Defects in DNA/nuclear fragmentation during apopto- background and 129X1/SvJ pure sis. Compensatory activation of other caspases was background. Most animals die reported during embryonic or early neonatal period. No in pure C57/B6 background 8 Yes, die at embryonic day 8.5Impaired formation of cardiac Defects in activation of death receptor-mediated muscles and abdominal hemor- apoptotic pathways rhage 9 Yes, most animals die during em- Neuronal hyperplasia Defects in apoptosis induction by stress and genotoxic bryonic or early neonatal period drugs in activation of caspase-3 and -6. Compensatory activation of other caspases was reported 12 No No Defects in ER stress apoptosis and Ab toxicity (cortical neurons)

See text for detailed descriptions

(Friedlander et al., 1997b; Hara et al., 1997). Interest- RAIDD-mediated interaction with Fas in overexpres- ingly, these mice also displayed a delayed onset of motor sion systems (Ahmad et al., 1997; Duan and Dixit, 1997) dysfunction when crossed with a transgenic mouse suggest its role as an initiator caspase. On the other expressing mutant SOD, a mouse model of amyotrophic hand, it may be activated by downstream executioner lateral sclerosis (ALS) (Friedlander et al., 1997a). The caspases such as caspase-3 (Li et al., 1997a) and its targets of this caspase-1 mutant, however, remain substrate preferences also belong to the effector class unclear as caspase-1 can interact and potentially (Figure 2c). One interesting feature of caspase-2 is its interfere with other caspases, such as caspase-11 (Wang intracellular localization to nuclei and the Golgi et al., 1998). Therefore, these promising data should be (Colussi et al., 1998; Mancini et al., 2000). Its nuclear considered with caution. import is mediated by two nuclear localization signals Murine caspase-11 displays a high degree of homol- (NLSs) present in the prodomain, but the significance of ogy with caspase-1. Caspase-11/ mice are devel- the nuclear pool of caspase-2 is currently unclear, since opmentally normal, but are resistant to endotoxic shock overexpression of the cytosolic form of caspase-2 with a and lack caspase-1 activation and IL-1b secretion, mutated NLS is fully capable of inducing apoptosis although caspase-11 does not directly process pro-IL- (Colussi et al., 1998; Baliga et al., 2002). In the Golgi, 1b (Wang et al., 1996) (Table 1). These data clearly caspase-2 may contribute to the fragmentation of this demonstrate that caspase-11 plays a critical role in the organelle associated with apoptosis through the cleavage activation of caspase-1. Caspase-11 and its human structural protein golgin-160 (Mancini et al., 2000). orthologue caspase-5have been shown to form a Therefore, caspase-2 may contribute to organelle- complex with caspase-1, which may be a critical step specific apoptotic events. However, these results were in mediating the activation of caspase-1 (Wang et al., obtained using overexpression systems and therefore, 1998; Martinon et al., 2002). However, there is no in the precise contribution of these caspase-2-mediated vitro or in vivo evidence that caspase-11 or -5activates events to the grand scheme of apoptosis remains to be caspase-1 through direct proteolytic cleavage. In con- confirmed in future. trast, caspase-11 activates caspase-3 by direct proteoly- The use of antisense constructs to inhibit the tic cleavage in mediating apoptosis following septic expression of caspase-2 demonstrated its prominent role shock and brain ischemia (Kang et al., 2000, 2002). in several in vitro cell death model systems, including neurotrophic factor deprivation of sympathetic neurons, Caspase-2 Ab-induced apoptosis in hippocampal neurons and Fas- mediated death of lymphocytes (Haviv et al., 1998; Caspase-2 may be the most tantalizing member of the Stefanis et al., 1998; Troy et al., 2000). However, family (Wang et al., 1994). Its long, CARD-containing caspase-2/ mice are viable and caspase-2/ cells prodomain, its early activation in response to a variety show normal induction of apoptosis in response to of apoptotic stimuli (Harvey et al., 1997) and its multiple stimuli (Bergeron et al., 1998) (Table 1). The

Oncogene Decade of caspases A Degterev et al 8550 only observed apoptotic defects in these mice are Yuan, 2000; Rao et al., 2001), whereas caspase-12 was excessive germ cells in the ovaries of female mutant proposed to mediate cytochrome c-independent activa- mice, the resistance of oocytes to chemotherapeutic tion of caspase-9 (Morishima et al., 2002; Rao et al., drug-induced apoptosis and the reduced death of 2002). lymphocytes treated with granzyme B (Bergeron et al., A human caspase in the caspase-1 chromosomal 1998). The lack of global neuron-specific apoptotic cluster at 11q23 with significant homology to murine defects in caspase-2-deficient mice may be explained by caspase-12 has been proposed to be the human compensation mechanisms, as suggested by a recent orthologue of caspase-12. Interestingly, although study by Troy et al. (2001), which showed that the mRNAs of this human caspase are expressed, the caspase-9 pathway may take over the functions of coding region has acquired apparent mutations in at caspase-2 in caspase-2/ sympathetic neurons. least certain human populations (Fischer et al., 2002). One novel approach, which may allow analysis of the However, since there is no functional data confirming protein cellular function without the limitations of this conclusion, it remains highly hypothetical. Other possible genetic compensation triggered in transgenic possibilities, including acquisition of the caspase-12 approaches, is the use of short interfering RNA function by another human caspase(s), and/or less a molecules (RNAi) to downregulate endogenous protein pronounced role of caspases in ER stress of human cells expression levels (Hunter, 1999; Sharp, 1999). This also remain a strong possibility. method indeed has recently been very successfully utilized for caspase-2, demonstrating the critical role Caspase-14 of caspase-2 in genotoxic stress-mediated cell death in cultured cells (Lassus et al., 2002). Therefore, unlike Caspase-14 appears to lack any clear designated role in transgenic mice, RNAi may help elucidate the true role cell death. It is present exclusively in simple and complex of caspase-2 in various apoptotic systems. epithelia (Ahmad et al., 1998; Hu et al., 1998; Van de Interestingly, caspase-2 knockout mice also showed Craen et al., 1998; Pistritto et al., 2002), distinguishing it fewer facial motor neurons at birth, which became from the other, relatively ubiquitously expressed cas- normal by day 7 (Bergeron et al., 1998). This defect may pases. Several studies suggested that caspase-14 is reflect in vivo antiapoptotic function of the short splice transcriptionally induced during terminal differentiation variants of caspase-2, which were previously shown to of keratinocytes (Eckhart et al., 2000a, b; Lippens et al., inhibit apoptosis in vitro (Wang et al., 1994). 2000), but another recent report suggested that high cell Several recent studies suggested that, rather than density and detachment from the extracellular matrix activating downstream caspase networks, as other are the primary causes for the induction, rather than the initiator caspases, caspase-2 may act primarily on differentiation per se (Pistritto et al., 2002). In any case, noncaspase substrates to mediate cell death. In that all the studies agree that caspase-14 function does not respect, caspase-2 was found to cause the release of appear to be related to apoptosis. apoptogenic factors from mitochondria directly, a process which may be mediated in part by caspase-2 Caspase-8 cleavage of Bid (Guo et al., 2002; Lassus et al., 2002; Robertson et al., 2002). Overall, these findings suggest The properties of caspase-8 are very well characterized that caspase-2 has many properties distinct from other and its role as an apical caspase in death receptor caspases, but we still have very little knowledge signaling has been well established through biochemical, regarding its true role in cell death in vivo. molecular and cell biological studies. For example, caspase-8-deficient Jurkat cells are completely resistant Caspase-12 to Fas and TRAIL-R2 and partially resistant to TNFa- induced apoptosis (Juo et al., 1998; Bodmer et al., 2000; Murine caspase-12 localizes specifically to the ER, Sprick et al., 2000). MEFs from the caspase-8/ which is unique among caspases (Nakagawa et al., embryos were resistant to Fas, TNF and DR3 receptor- 2000). Caspase-12-deficient mice are developmentally induced apoptosis, but remained sensitive to apoptosis normal (Nakagawa et al., 2000) (Table 1); however, through other pathways, such as UV irradiation and caspase-12 deletion renders cells specifically resistant to dexamethasone (Varfolomeev et al., 1998). Interestingly, ER stress-induced cell death in vitro and in vivo, caspase-8 inactivation in Jurkat cells may not comple- consistent with its subcellular localization (Nakagawa tely prevent death receptor-induced death, but rather et al., 2000). These data suggest that caspase-12 is a shift the balance to nonapoptotic PCD pathways specific sensor of ER perturbation. Ab cytotoxicity is (Kawahara et al., 1998; Holler et al., 2000; Matsumura also attenuated in caspase-12/ cortical neurons, et al., 2000). providing evidence for a role of ER stress in Alzheimer’s Caspase-8/ mice die at day 11.5of the embryonic disease (AD) (Nakagawa et al., 2000). development, displaying impaired formation of cardiac Although functional data clearly place caspase-12 in muscles and prominent abdominal hemorrhage due to the center of ER stress-mediated death, the pathways hyperemia (Varfolomeev et al., 1998) (Table 1). In upstream and downstream of caspase-12 still remain to addition, caspase-8/ embryos possess decreased be characterized. Thus far, calpain and caspase-7 have numbers of myeloid progenitor cells. FADD/ mice been proposed to activate caspase-12 (Nakagawa and displayed a similar phenotype, suggesting that death

Oncogene Decade of caspases A Degterev et al 8551 receptor signaling pathways play an important role in Caspase-9 heart muscle development and hematopoiesis (Yeh et al., 1998). FADD deficiency was also shown to block the Caspase-9 is an important intracellular amplifier of activation of T cells and the development of B cells caspase signaling downstream of mitochondria (Zhang et al., 1998; Kabra et al., 2001), but similar (Figure 3). Caspase-9 deficiency results in embryonic- neonatal lethality with the onset of abnormalities at studies have not been possible for caspase-8/ mice et al et al due to their early embryonic lethality (Varfolomeev embryonic day 12.5(Hakem ., 1998; Kuida ., et al., 1998). Interestingly, caspase-8 deficiency was 1998) (Table 1). Interestingly, a few surviving homo- et al recently discovered in a human kindred with an zygous mutants showed no obvious defects (Kuida ., immunodeficiency syndrome (Chun et al., 2002). In- 1998), which may be due to strain-specific genetic dividuals with homozygous caspase-8 mutations modifiers of the mutant phenotype (Leonard et al., not only manifest defective lymphocyte apoptosis, 2002). Caspase-9/ embryos display an enlarged and characteristic of autoimmune lymphoproliferative malformed cerebellum due to a deficiency in apoptosis disorder (ALPS) found in patients with defects in and caspase-3 activation. It remains to be determined, Fas signaling but also display additional defects in however, whether the cause of lethality in caspase-9/ the activation of T, B and natural killer cells, which mice is truly a result of defective apoptosis or reflects an leads to immunodeficiency. While this finding unknown, nonapoptotic function of this caspase. remains to be confirmed by additional studies with Caspase-9/ thymocytes, MEFs, embryonic stem conditional mouse mutants of caspase-8, the above- (ES) cells and splenocytes showed defective apoptosis in mentioned data suggest a possible dual role of caspase-8 response to a variety of stimuli (g and UV irradiation, in both negative and positive regulation of immune cytotoxic drugs, dexamethasone, etc.), with the notable response in vivo. exception of death receptor pathways, which were not In addition to its role as an apical caspase, caspase-8 affected. Apaf-1/ mice displayed a phenotype similar may be activated by downstream caspases, such as to that of caspase-9 (Yoshida et al., 1998), consistent caspase-6, in a positive feedback loop (Cowling and with the model of caspase-9 and Apaf-1 being in the Downward, 2002), potentially resulting in Bid-depen- same pathway (Figure 3). One notable exception is dent targeting of mitochondria. Several reports have the retarded elimination of the interdigital webs in et al also suggested that cytotoxic drugs can induce activa- Apaf-1/ mice (Chautan ., 1999), which is not tion of the death receptor signaling mechanism through seen in caspase- 9/ mice. the Fas ligand (FasL)-independent induction of Fas As noted above for the case of caspase-2, the clustering and caspase-8 activation (Micheau et al., interpretation of mouse caspase knockout data may be 1999; Chen and Lai, 2001; Kulms et al., 2002). Finally, complicated by compensatory expression of different caspase-8 was recently shown to substitute for caspase-3 caspase family members. Also in this vein, the injection of anti-Fas antibody was lethal to both caspase-3 / function in caspase-3/ mice following middle cere- bral artery occlusion, a mouse model of stroke (Le et al., and caspase-9/ mice, challenging the conclusion that 2002). All of these data suggest that caspase-8 possesses Fas-induced hepatotoxicty involves the mitochondrial a wide variety of cell type-, cellular context- and signal- pathway (Yin et al., 1999). However, the explanation for specific apoptotic functions in addition to the death these results may lie in the compensatory induction of receptor pathway. caspase-6 and -7 in caspase-3/ and caspase-9/ mice (Zheng et al., 2000). Disruption of Apaf-1/caspase-9 signaling by knock- Caspase-10 outs or other methods may also result in the activation Caspase-10 is highly homologous to caspase-8 and of the alternative apoptotic apoptosome-independent is also recruited and activated by death receptors mechanism. This principle was recently demonstrated (Kischkel et al., 2001; Wang et al., 2001; Sprick et al., for caspase-9/ and Apaf-1/ using the reconstitu- 2002). Missense mutations in human caspase-10 have tion of the hematopoietic compartment of the wild-type been found to be genetically associated with the animals with transgenic fetal liver stem cells (Marsden et al autoimmune disease, ALPS type II, which is character- ., 2002). In addition, nonapoptotic death mechan- ized by the abnormal homeostasis of lymphocytes isms were reported to be activated in Apaf-1/ ES and dendritic cells caused by defective FasL/Fas cells (Haraguchi et al., 2000) and in caspase-3/ and et al signaling (Wang et al., 1999). Since such mutant alleles caspase-9/ motor neurons (Oppenheim ., 2001). of caspase-10 encode proteins with decreased enzymatic et al activity (Wang ., 1999), caspase-10 must play an Caspase-3, -6 and -7 important role in death receptor signaling at least in some cell types in vivo, even though current in vitro data These three highly homologous caspases form the regarding its role are not conclusive (Kischkel et al., execution subfamily. Analyses of apoptosis progression 2001; Wang et al., 2001; Sprick et al., 2002). Caspase-10 in MCF-7 cells, which lack caspase-3, showed that it is may also function as a tumor suppressor, as several specifically required for the activation of the CAD/ studies have reported finding missense mutations in DFF40-mediated internucleosomal DNA degradation caspase-10 in various tumors (Park et al., 2002; Shin (Tang and Kidd, 1998). Caspase-3/ mice in mixed et al., 2002a, b). 129X1/SvJ and C57BL/6J background were smaller

Oncogene Decade of caspases A Degterev et al 8552 than wild-type littermates, and most died prenatally, displaying neuronal hyperplasia due to the lack of apoptosis (Kuida et al., 1996). In a pure C57BL/6J background, however, caspase-3/ mice are viable with reduced fertility, suggesting the presence of a dominant strain-specific genetic suppressor (Leonard et al., 2002). Analysis of cell death in multiple cell types isolated from caspase-3 knockout animals revealed the complete inhibition of certain apoptotic hallmarks, such as membrane blebbing, DNA degradation and nuclear fragmentation (Woo et al., 1998; Zheng et al., 1998). On the other hand, caspase-3/ MEFs were not protected from eventual death and displayed multiple apoptotic features, such as nuclear condensation and phosphatidylserine exposure. Similarly, depletion of caspase-3 in a cell-free apoptotic system inhibited most of the downstream events, including various substrate cleavages, DNA fragmentation, chromatin marginalization, etc., whereas elimination of either caspase-6 or -7 had no effect (Slee et al., 2001). Thus, caspase-3 is important for the execution of certain, but not all, specific downstream events in apoptosis. In contrast to caspase-3, caspase-6 Figure 4 Structural data pertinent to the caspase substrate and -7 currently lack significant specialized functions. specificity and activation mechanism. (a) Zymogenicity values of The above-mentioned data create a paradox, since some caspases (Salvesen and Dixit, 1999). Zymogenicity is defined as a ratio of the activities of the processed caspase to the its none of the executioner caspases seem to fully control zymogen. (b) Model of caspase-7 activation based on the crystal execution of all the aspects of apoptosis. One possible structures of three forms of the enzyme (reproduced from Shi, explanation is that executioner caspases may have 2002). Cleavage of L20 results in the upward shift of the N-terminus distinct roles in specific pathways, such as ER stress of the small subunit, resulting in the formation of the loop bundle for caspase-7 (Rao et al., 2001), but have redundant and stabilization of the caspase active center. (c) Dimerization- driven caspase-9 activation mechanism (reproduced from Renatus functions in the majority of the general apoptotic events et al., 2001). Dimer formation results in the inhibitory loop binding and thus, abolition of all three may be required to to the site on the neighboring subunit, resulting in the rearrange- observe complete inhibition of all apoptotic features. ment of the active site and caspase activation Alternatively, caspase-3 (Slee et al., 2001) may play a primary role, whereas abundant compensatory mechanisms create a novel network in case ‘primary’ caspase is Yamin et al., 1996). A mechanistic explanation for missing, allowing most of the downstream events to autoactivation was provided by the observation that occur. Multiple specific examples of such compensatory individual initiator caspases’ processing sites often mechanisms have already been reported, for example, conform to their own substrate specificities (Figure 1d) caspase-3 activity can be substituted for by caspase-8 (Thornberry et al., 1997). (Le et al., 2002); inactivation of caspase-9 may lead to The ability to autoactivate implied the presence of loss of caspase-3 and -6 activation, but not caspase-7 residual activity in caspase zymogens, a possibility that processing (Marsden et al., 2002); caspase-9 can replace was directly confirmed by active site affinity labeling of caspase-2 (Troy et al., 2001); deletion of caspase-3 and - caspase-1 zymogen (Yamin et al., 1996). However, 9 results in the compensatory activation of caspase-7 caspase zymogens display greatly varying degrees of and -6 (Zheng et al., 2000). intrinsic activity (Figure 4a). Initiator caspases such as caspase-8 and -9 are quite active as zymogens, whereas the executioner caspase-3 is much less active (Salvesen and Dixit, 1999). In this sense, only executioner caspases Molecular basis of caspase activation are ‘true’ zymogens. The differences in zymogen caspase activity underscore the ability of the initiator caspases to Except for a number of special cases, caspase processing serve as proximal responders to proapoptotic signals is mediated primarily by caspases themselves. This due to their high propensity towards autoactivation and became obvious from the early observations that the necessitate the transactivation of the executioner cas- simple overexpression of wild type, but not active site pases by activator ones (see below). The high intrinsic mutant caspase zymogens, in bacteria results in their residual activity of initiator caspase zymogens also raises spontaneous activation (Orth et al., 1996; Stennicke and an intriguing question as to how such activity is Salvesen, 1997). Similarly, cleavage between the large suppressed in living cells. and small subunits with a concomitant increase in One possible molecular explanation for the enhanced activity was observed in highly purified samples of zymogenicity of the executioner caspases has been caspase-1 (Thornberry et al., 1992; Ramage et al., 1995; provided for caspase-3 (Roy et al., 2001). Roy et al.

Oncogene Decade of caspases A Degterev et al 8553 described an Asp–Asp–Asp tripeptide inhibitory ‘safety The elegant activation mechanism described above, catch’ present in the L2 loop of the caspase-3. This however, just primes caspase-7 for substrate binding. sequence, which is only found in caspase-3 and -7, forms Comparison of the free active and inhibitor-bound an extensive network of salt bridges, stabilizing the caspase-7 structures, performed in the same study (Chai inactive conformation and keeping zymogen activity et al., 2001), provided a striking conclusion that the under control. This mechanism is sensitive to low pH, substrate-binding groove in free processed caspase-7 is which may account for the observed requirement for still not properly formed because the L20 loop remains in cytosolic acidification during apoptosome-mediated cas- the ‘closed’ conformation. The binding of substrate/ pase-3 activation (Matsuyama et al., 2000). In addition, inhibitor triggers a 1801 flipping of the L20 loop, direct interaction of the RGD peptide with the safety resulting in its upward movement towards L1–L4, catch sequence may provide a mechanistic explanation as completing the formation of the ‘active bundle’. This to why the RGD tripeptide can induce apoptosis of observation has led to the proposal that the binding of endothelial cells (Buckley et al., 1999; Roy et al., 2001). the substrate by caspases is a dynamic ‘induced-fit’ process (Shi, 2002). Therefore, caspase processing and Transactivation substrate binding represent a continuous sequence of conformational changes leading to the ‘loop bundle’ As described in the preceding section, in many cases cells formation and molecular ordering of the active center. utilize intrinsically different mechanisms to activate Two additional interesting questions were answered initiator and executioner caspases (Figure 3). The by caspase-7 structural studies. First, it has long been former is achieved through interaction with the up- proposed that caspase heterodimers may form through stream signaling proteins, whereas the latter results from an interdigitation mechanism, which postulates that the direct processing by activated initiator caspases. The small and large subunits forming the structural unit in mechanistic differences in the initiator and executioner the heterodimer come from different caspase monomers caspase activation have been confirmed by structural (Stennicke and Salvesen, 2000). However, the zymogen analyses. It should be noted, however, that even though structure showed that each unit is composed of the there are clear differences in the activation and behavior subunits originating from the same polypeptide chain. of the two caspase classes, they are not absolute and Second, the concept of the loop bundle, formed by the significant overlaps exist, as will be discussed below. sequences contributed by different subunits of the same The process of executioner caspase activation is monomer (L4 and L2) and the neighboring polypeptide termed ‘transactivation’ because it is mediated in trans as well (L20) underscores the critical importance of by a heterologous caspase. The X-ray structures of dimerization for enzyme activity. active and zymogenic forms of caspase-7 have been The proposed mechanism for executioner caspase reported, providing the opportunity to visualize directly activation clearly postulates that the removal of the N- the changes associated with the activation process (Shi, terminal propeptide is neither required nor beneficial for 2002). Similar to the active caspases, procaspase-7 activation. This aspect of the model has been confirmed zymogen is a homodimer (Chai et al., 2001; Pop et al., experimentally for multiple caspases, including caspase- 2001; Riedl et al., 2001a). The overall structure of the 3 (Chai et al., 2001; Pop et al., 2001). However, a recent enzyme undergoes very minor changes upon activation, study suggested that propeptide cleavage is critical for except for the catalytic loops L2, L3 and L4, which caspase-6 activation, which may suggest that additional, undergo dramatic conformational transitions as yet unknown structural determinants, are involved in (Figure 4b). Three of the four loops (L2, L3 and L4) the activation process (Cowling and Downward, 2002). are away from each other in procaspase-7, in a conformation prohibitive to substrate binding, provid- Allosteric regulation ing a molecular mechanism for its zymogenicity. The cleavage of Asp198 in the activation process frees the Caspase-9 is unique among all the caspases as proteo- L20 loop from the neighboring small subunits in the lytic cleavage is neither sufficient nor necessary for its caspase-7 homodimer and allows it to form a tight activation (Rodriguez and Lazebnik, 1999; Stennicke ‘active bundle’ with the L2 and L4 loops, which in turn et al., 1999; Renatus et al., 2001). In living cells, caspase- produces the 901 rotation of L2, making the catalytic 9 is present in the cytosol predominantly in the Cys accessible to solvent. This initial cleavage is the first, monomeric form with a small population of dimers most important activation event. Subsequent processing (Renatus et al., 2001). The active site of the monomer is steps (steps 2 and 3 in Figure 2c) may or may not be in the inactive conformation and its activation is driven needed for full activation depending on the caspase. by a dimerization-dependent conformational change, Overall, the activation process triggers the necessary which was characterized through X-ray structural conformation change in caspase-7 to permit substrate studies (Renatus et al., 2001). It is not clear whether a binding (Chai et al., 2001) (Figure 4b). While this small amount of caspase-9 dimer is actually present in dynamic process of proteolytic activation is still to be living cells or formed as a result of the isolation process. confirmed with other caspases, the high degree of According to the model of caspase-9 activation, the structural conservation in the caspase family suggests interaction surface of the neighboring monomer pro- that this model is likely to be true for other caspases as vides an acceptor site for the activation loop (Figure 4c), well. and this binding results in the rearrangement in the

Oncogene Decade of caspases A Degterev et al 8554 substrate-binding pocket and reorientation of the similar to the dimerization-driven caspase-9 activation catalytic Cys to form a functional active site (Micheau et al., 2002). The L2 loop of caspase-8 may be (Figure 4c) (Renatus et al., 2001). Interestingly, struc- long and flexible enough to be subsequently auto- tural studies showed that the caspase-9 dimer possesses cleaved, resulting in the generation of the partially only one active site, but the significance of this finding is processed, but fully active heterodimer of FLIPL and currently unknown (Renatus et al., 2001). p43/41 (prodomain plus large subunit) and p12 subunits The apoptosome complex, consisting of Apaf-1, of caspase-8. Thus, caspase-8 activation may include cytochrome c and caspase-9, is proposed to serve as an features of both allosteric and processing-based mechan- allosteric regulator of caspase-9, promoting its homo- isms. In contrast to that of caspase-8, the L2 loop of dimerization and activation (Renatus et al., 2001; caspase-9 may be extended enough to eliminate the need Shiozaki et al., 2002). Thus, the apoptosome complex for proteolytic activation completely (Renatus et al., can be considered as a holoenzyme, in which caspase-9 is 2001). 1000-fold more active than its monomeric form (Rodriguez and Lazebnik, 1999). The allosteric mechanism may also be utilized by other caspases. For example, it may contribute to the A complex network of caspase cross-activation activation of caspase-1 and -5in the ‘inflammasome’ complex involving the Apaf-1-like adaptor NALP1 As described above, the initiator–executioner dichotomy (Martinon et al., 2002). of the caspase family fits well with their established activation mechanisms. However, multiple studies Induced proximity mechanism clearly demonstrate that this subdivision primarily applies to the initial stages of the apoptotic signaling, Caspase-8 processing may represent yet another me- with the complex network of caspase cross-activation chanism of caspase activation with features of both occurring in the downstream steps of apoptosis. proteolytic and allosteric regulation. The caspase-8 Recombinant caspase-1, -2, -3, -6, -7 and -8 can zymogen possesses significant activity, but proteolytic process various in vitro translated caspase zymogens cleavage is still clearly required for its full activation (Van de Craen et al., 1999). For example, caspase pairs (Muzio et al., 1998). Since caspase-8 is the most 8/3, 6/7 and 6/3 were found to form reciprocal feedback upstream caspase in the death receptor signaling path- loops. Caspase-8 was shown to process all caspases way and its activation is brought about as a result of the efficiently. Caspase-1 and -11 were found to be able to adaptor-mediated multimerization, an ‘induced proxi- activate executioner caspases, but not vice versa. mity’ mechanism for its processing has been suggested Caspase-2 was shown to be activated by a host of (Muzio et al., 1998). In this model, the increased local caspases, but it failed to process any tested caspases. concentration of the caspase-8 zymogen induced by Overall, these findings underscore the complexity of the FADD-mediated trimerization results in the intermole- caspase transactivation patterns, with executioner cas- cular processing of caspase-8. Consistent with this pases capable of processing the normally upstream model, FK1012-induced dimerization of caspase-8 fused activator molecules in vitro. to the FKBP12 protein indeed resulted in enzyme An intricate network of caspase transactivation has activation (Muzio et al., 1998). also been demonstrated in a cell-free apoptosis system The induced proximity model may need to be updated (Slee et al., 1999). The addition of cytochrome c to based on the recent studies of FLIPL-mediated caspase- postnuclear Jurkat (JK) cell lysates resulted in the 8 activation (Chang et al., 2002; Micheau et al., 2002). initiation of the caspase-9/caspase-3 axis, with caspase-3 FLIP is a catalytically inactive, close homologue of subsequently processing caspase-2 and -6 and the latter caspase-8 and plays an important role in the regulation activating apical caspase-8 and -10. In addition to of caspase-8 activation and inhibition of cell death in caspase-3, caspase-9 also processed caspase-7, while vivo (Yeh et al., 2000). Two splice variants of FLIPs caspase-3 formed a feedback loop to cleave caspase-9. have been described: short and long (Irmler et al., 1997). The primary function of the latter event is to remove the The former is the homologue of the caspase-8 prodo- binding site for XIAP (X-linked inhibitor of apoptosis) main only, whereas the long form also contains a from the L20 loop of the caspase-9, making it refractive pseudocatalytic domain lacking the active site Cys. to inhibition (Srinivasula et al., 2001). FLIPs can directly block caspase-8 recruitment into the The above observations made in vitro provide just a DISC (Kirchhoff et al., 2000). In contrast, FLIPL brief insight into the potential complexity of the putative cannot prevent the recruitment of caspase-8 into the downstream caspase cross-activation networks induced DISC, and its presence in the DISC even stimulates the inside cells. Although upstream activation may initially first cleavage step of caspase-8 zymogen, but blocks involve particular caspases, it is likely that a vast array further processing (Krueger et al., 2001), resulting in of additional family members is eventually activated as local activation of caspase-8 within the DISC complex well. Such transactivation may blur the distinctions (Chang et al., 2002; Micheau et al., 2002). Based on the between the ‘initiator’ and ‘executioner’ caspases and results of the molecular modeling of caspase-8/ FLIPL may serve (a) to expand the repertoire of accessible complex, it has been proposed that FLIPL can induce a caspase substrates, (b) to stimulate diverse downstream reordering of the caspase-8 catalytic center in a manner signaling processes (e.g. cytokine processing versus

Oncogene Decade of caspases A Degterev et al 8555 apoptosis), (c) to activate specific apoptotic execution for a review, see Kim et al., 2002). Therefore, the subroutines, (d) to generate sustained signaling through outcome of NO generation may depend on the level of positive feedback loops, (e) to provide multiple redun- its production and cellular context. dant pathways to ensure execution of the apoptotic Similarly, excessive generation of reactive oxygen program in the event of the inhibition or malfunction of species (ROS) may lead to the inhibition of caspases, a specific subprogram and (f) to accelerate execution of even though ROS possesses high intrinsic toxicity. In apoptosis through massive caspase activation. many cases, inhibition of caspases by such stimuli may not result in the alleviation of death, but rather in a switch from apoptosis to necrosis, which is defined as a Additional regulatory mechanisms for caspases nonspecific, caspase-independent death caused by the overwhelming stress (Nicotera et al., 1999). Most caspases are constitutively expressed in multiple cell types. The activation mechanisms described above Phosphorylation represent by far the most important ways to regulate Phosphorylation is the most widely used cellular caspase activity. However, additional regulatory me- regulatory mechanism, but the only available example chanisms have also been uncovered in recent years. of phosphorylation directly regulating caspase activity is the inactivation of human caspase-9 by Akt (Cardone Gene expression et al., 1998). Interestingly, this mechanism may have The requirement for new protein synthesis in certain been developed late in evolution, since murine caspase-9 apoptotic paradigms provided the first hint of an apparently lacks the cognate phosphorylation site endogenous suicide mechanism (Lockshin, 1969). Even present in human caspase-9 (Fujita et al., 1999). though subsequent studies suggested that apoptosis, in Phosphorylation, however, has been shown to play an most cases, does not require new protein synthesis, important role in the regulation of the other steps in regulation of caspase transcription does play a role in apoptosis signaling (see other reviews in this issue of certain circumstances. A dramatic demonstration of this Oncogene). One possible explanation for the lack of notion came from the analysis of murine caspase-11. direct phosphorylation-dependent caspase modulation The level of this caspase in healthy resting cells is very is that phosphorylation is best suited for the rapid and low, but treatment with LPS or other pathological reversible modulation of signaling networks and there- stimuli such as ischemia in vivo leads to a dramatic fore may not be appropriate for regulating caspases, transcriptional upregulation of caspase-11 expression whose activation often initiates an irreversible process. (Wang et al., 1996; Kang et al., 2000). The putative human caspase-11 homologue, caspase-5, is also tran- Ubiquitination and degradation scriptionally upregulated by LPS (Lin et al., 2000), but it remains to be determined if levels of the protein also Increasing attention has been paid in recent years to the increase. Caspase-14 is another example of transcrip- ubiquitination of caspases by IAPs (see review of IAP tionally regulated caspase (see above) (Eckhart et al., family in this issue of Oncogene). 2000a). A role of the oncogenic E2F transcription factor in regulating caspase expression was recently proposed by Caspase inhibitors Nahle et al. (2002). It was found that E2F deregulation (caused by E2A oncogene overexpression or loss of the Caspases are the only proteases that are not compart- tumor suppressor Rb) resulted in the accumulation of mentalized in living cells. Their potent proapoptotic zymogenic forms of multiple caspases, which appears to activity mandates that they must be kept under tight potentiate p53-mediated apoptotic signals. These results control in order for healthy cells to survive. In this may explain the increased sensitivity of oncogene- section, we discuss viral caspase inhibitors that target expressing cells to multiple apoptotic stimuli, which is the cellular apoptosis machinery that is activated upon in many cases exploited by cancer therapy. viral infection, as well as the synthetic inhibitors that have been widely used to examine the roles of caspases Nitrosylation and oxidation in regulating cellular processes.

The catalytic cysteine of caspases is very active and p35 susceptible to modifications. It has been demonstrated that nitric oxide (NO) donors can directly inhibit Apoptosis is the most potent host defense mechanism to caspase activity through active site nitrosylation (Li eliminate virally infected cells. It is not surprising, et al., 1997b; Mohr et al., 1997) and this mechanism was therefore, that many viruses have developed counter- implicated in the NO-mediated inhibition of apoptosis measures to suppress apoptosis. The baculoviral p35 or inflammatory cytokine production. On the other protein was first discovered through its ability to inhibit hand, multiple reports also suggested that NO can cause apoptosis in virally infected insect cells (Clem et al., nitrosative stress, DNA and mitochondrial damage and, 1991) and later shown to inhibit multiple active caspases as a result, induce apoptosis and caspase activation (see directly by acting as a pseudosubstrate (Bump et al.,

Oncogene Decade of caspases A Degterev et al 8556 1995; Xue and Horvitz, 1995). p35 is first cleaved by caspases (Ray et al., 1992). CrmA displays a preference caspases, recognizing the DQMD sequence in the for caspase-1 and -8 in vitro (Table 2). Consistent with reactive site loop (RSL) of p35, followed by the this observation, CrmA prevented the processing of IL- formation of the covalent adduct between the P1 1b by caspase-1 and inhibited TNFa- and Fas-mediated Asp87 residue of p35and active site Cys of caspases apoptosis (Enari et al., 1995; Los et al., 1995; Miura (Bump et al., 1995). et al., 1995). On the other hand, despite a high second- The molecular mechanism of p35activity was order rate constant and a low nanomolar Ki against established through X-ray crystallographic analysis of caspase-9 in vitro, caspase-9 most likely is not a the p35/caspase-8 complex (Xu et al., 2001). The initial physiological target of CrmA, based on the inability of cleavage by caspases results in significant conforma- CrmA to block caspase-9-mediated mitochondria-de- tional changes in p35, including the formation of an pendent cell death pathways (Datta et al., 1997a; Orth internal b sheet (Xu et al., 2001), surface exposure of the and Dixit, 1997). Cys2 residue of p35, normally buried in the hydrophobic Structurally, CrmA belongs to the serpin class of core (Fisher et al., 1999) and protein stabilization (Riedl inhibitors, but unlike other members of this class, which et al., 2001b). The released Cys2 residue inserts into the target serine proteases, CrmA inhibits cysteine procaspase-8 active site and likely forms a hydrogen bond teases. What makes this inhibitor truly unique is its with the catalytic His residue. This interaction comple- ability to also inhibit the serine protease granzyme B, tely blocks the solvent accessibility of the His residue which shows similar substrate specificity to caspases and also induces its rotation into a position unfavorable (Quan et al., 1995), making CrmA a dual Ser/Cys for catalysis, disrupting the catalytic triad and resulting protease inhibitor. in the stabilization of the covalent adduct formed p35and CrmA are both pseudosubstrate inhibitors with the catalytic Cys287 (Xu et al., 2001). Consistent and their interaction with caspases involves caspase with this mechanism, Cys2 mutations convert p35 substrate recognition subsites S1–S4 and corresponding into an efficient caspase substrate, abrogating P1–P4 inhibitor residues (Ekert et al., 1999b; Xu et al., covalent adduct formation (Riedl et al., 2001b; Xu 2001), with additional extended interaction surfaces et al., 2001). In the case of executioner caspases, p35also surrounding S1–S4/P1–P4 sites proposed to further appears to interact with the surface loop 381, which is increase the affinity of binding (Xu et al., 2001). The missing in initiator caspases, explaining the preference caspase recognition motifs present in the RSL loops of of p35for the executioner subgroup (Eddins et al., p35and CrmA are DQMD and LVAD, respectively, 2002). and were shown to play a significant role in their specificity towards particular caspase family members CrmA (Ekert et al., 1999b). However, additional parameters, such as the geometry of the active centers of individual CrmA (a cytokine response modifier gene) protein of caspases, may also contribute to their relative affinities cowpox virus was the first identified inhibitor of for CrmA and p35(Renatus et al., 2000).

Table 2 Inhibitory constants of various classes of caspase inhibitors Caspase CrmA XIAP c-IAP1 c-IAP2 p35 zVAD-fmk Ac-DEVD-CHO Ac-YVAD-CHO

1 4105 2 107 8 103 2.5 109 15–18 109 7.6 1010 2 4105 2.4 106 1.71 106 3 7 1010 1.08 107 3.5 108 4105 2 102 7 105 4.3 108 0.23–2.2 109 4 1.3 107 1.32 107 3.62 107 5 5.3 109 2.05 107 1.63 107 6 4105 4105 4.4 102 1 105 9.8 108 3.1 108 7 o10 2 1010 4.2 108 2.9 108 4105 1 105 3.9 108 1.6 109 8 4105 9 104 6 104 2.5 109 9.2 1010 3.52 107 9 1 109 1 105 3.9 109 6 108 9.7 107 10 4105 o102 2 103 1.2 108 4.08 107

Inhibitory constants of different caspase inhibitors against various caspases. Results are compiled from Ekert et al. (1999a), Salvesen and Duckett 1 1 (2002), Stennicke et al. (2002). In case of p35and CrmA, second-order rate constants are shown ( Kobs/I in M s ), for other inhibitors Ki values are indicated (Ki,app in M). For further details refer to Ekert et al. (1999a), Salvesen and Duckett (2002), Stennicke et al. (2002) and references therein. In addition to the inhibitors described in detail in the text, inhibitory constants of three cellular IAP factors (XIAP (X-linked IAP), c-IAP1 and c-IAP2) are also shown for the reference (see review of IAP family in this issue of Oncogene for further details)

Oncogene Decade of caspases A Degterev et al 8557 IAPs neonatal hypoxic-ischemic brain injury (Han et al., 2002). This new and rapidly developing area may Inhibitor of apoptosis protein (IAP) was also lead to the clinical utilization of the extensive informa- originally identiﬁed in baculovirus (Crook et al., 1993). tion generated in the caspase ﬁeld over the past Homologues of IAP proteins have subsequently been 10 years. discovered in all eucaryotes from yeast to humans, making them very important physiological regulators of cell death (see review in this issue of Oncogene). Caspase targets

Peptide-based caspase inhibitors It is important to establish several basic concepts when The development of peptide-based caspase inhibitors examining the role of caspases in the execution of cell (Thornberry et al., 1992) provided valuable tools for death. Firstly, caspases do not degrade other proteins. studies of caspase family members and apoptosis in Due to their strict substrate recognition specificity, general. The design of these inhibitors is based on the caspases cleave targets at one or a few highly selective tetrapeptide caspase recognition motif and therefore the sites. Secondly, even though caspases do cleave some selectivity of the inhibitors parallels the caspase sub- structural proteins required for the maintenance of cell strate specificities described in the previous sections structure, they also frequently target signaling proteins, (Figure 1c, Table 2) (Garcia-Calvo et al., 1998). such as kinases. The cleavage of such signaling However, there are also differences in the mechanism molecules results in the activation or suppression of of action of p35and CrmA. For example, whereas specific downstream pathways, which in turn execute caspase cleavage is an integral part of inhibition by p35 cell death. Therefore, caspases serve as signaling (see above), in case of CrmA, caspase cleavage has been mediators that orchestrate a complex web of down- shown to be inactivating (Komiyama et al., 1994). The stream execution pathways. In that sense, caspases can introduction of an aldehyde group at the C-terminus of act in a manner analogous to kinases, with the notable the tetrapeptide results in the generation of reversible exception that protein modifications introduced by inhibitors (Graybill et al., 1994), whereas chloromethyl caspases are irreversible, underscoring the terminal ketone (Estrov et al., 1995), diazomethyl ketone nature of apoptosis. (Thornberry et al., 1992) and acyloxymethyl ketone Consistent with the substrate specificity considerations (Dolle et al., 1994; Thornberry et al., 1994) at this presented above, caspases most likely cleave only a very position create inhibitors that irreversibly inactivate the small subset of cellular proteins. Close to 100 substrates enzymes. The latter group comprises the most selective have been identified thus far (see reviews for detailed and potent caspase inhibitors, with the second-order listings Earnshaw et al., 1999; Li and Yuan, 1999; Utz rate constants reaching 2 106/M/s (Dolle et al., 1995). and Anderson, 2000). Here, we provide only a brief However, all these inhibitors possess several disadvan- overview of the current state of the field. Based on the tages, such as toxicity of the leaving group, low half-life analyses of their cellular function, the caspase targets can in cells, poor cell permeability, modest selectivity be subdivided into six major categories: (1) proteins towards individual caspases (or even caspases versus directly involved in the regulation of apoptosis, (2) other proteases) and a lack of oral bioavailability, which proteins mediating/regulating apoptotic signal transduc- limits their use to mostly cell-based studies. Since these tion (e.g. protein kinases), (3) structural and essential- synthetic peptide inhibitors are relatively poor in their function proteins, (4) proteins required for cellular repair, ability to discriminate among different caspase family (5) proteins regulating the cell cycle and (6) proteins members, these reagents are best suited to demonstrate involved in human pathologies (Figure 5). the involvement of caspases and apoptosis in general in It should be noted that in many cases, there is very a particular process, rather than pinpointing a role for a limited information regarding the cleavage of specific specific caspase. For further detailed reviews, see substrates, making it difficult to determine whether it Livingston, (1997), Ekert et al. (1999a), Rudel (1999), represents a universally important event in apoptosis or Talanian et al. (2000). is restricted to a particular cell type or stimulus. This consideration is especially relevant for substrate classes Nonpeptide caspase inhibitors 2, 5and 6, whereas proteolysis of many apoptotic, structural and repair proteins, such as caspases, ICAD/ Small molecule nonpeptide inhibitors of caspases are DFF45, keratins, lamins and poly(ADP-ribose) poly- currently in development and may overcome some of the merase, is considered general characteristic feature of limitations of the peptide-based inhibitors. The efficacy apoptotic cell death. In addition, very frequently the role of nonpeptide inhibitors has been demonstrated in of putative caspase substrates in regulation/execution of various animal models of pathology. Such inhibitors apoptosis is investigated via the overexpression of were shown to block apoptosis in neutrophils, chon- mutated, caspase-uncleavable versions, which may not drocytes and primary dopaminergic neurons efficiently accurately represent the endogenous functions of these in vitro (Lee et al., 2000; Bilsland et al., 2002) as well as proteins. Therefore, at least some conclusions regarding in vivo in animal models of septic shock (Grobmyer et al., the involvement of specific caspase substrates in down- 1999; Hotchkiss et al., 2000), liver injury (Natori et al., stream apoptotic events remain to be confirmed in the 1999; Deaciuc et al., 2001; Hoglen et al., 2001) and future.

Oncogene Decade of caspases A Degterev et al 8558 factors (Cheng et al., 1997; Clem et al., 1998). However, it is not currently clear whether this cleavage represents a ubiquitous positive feedback mechanism or a cell type or stimulus-specific event. Caspase inhibitors, such as IAPs (see review in this issue of Oncogene) and FLIPL (Scaffidi et al., 1999), are also cleaved by caspases, but the functional significance of these observations is currently unclear. Finally, although proapoptotic Bcl-2 family members may induce cytochrome c release upstream of (and, therefore, independent of) caspase activity (Kluck et al., 1997; Yang et al., 1997), additional mitochondrial damage, as detected by the loss of inner membrane potential, is often caspase-dependent. However, the specific targets in this case are unknown (Bossy-Wetzel et al., 1998; Li et al., 1998).

Protein kinases A large number of protein kinases are cleaved by caspases. Some are involved in the inhibition of apoptosis, while others may potentiate apoptosis or mediate speciﬁc downstream execution events. The most notable antiapoptotic kinase target of caspases is Akt, which is cleaved following matrix detachment in epithelial cells (Bachelder et al., 1999, 2001). Akt mediates cell survival through multiple Figure 5 Caspase substrates. The list of the most relevant caspase pathways, including the phosphorylation of proapopto- substrates (see text for further details). Speciﬁc information regarding the role of each substrate and precise processing site tic Bcl-2 family member Bad (Datta et al., 1997c; del can be found elsewhere (Earnshaw et al., 1999; Utz and Anderson, Peso et al., 1997) and transcription factor Forkhead 2000) (Brunet et al., 1999). RIP kinase, the cell survival signaling component of the DISC complex, is also Apoptotic proteins cleaved by caspase-8, causing its inactivation and the suppression of antiapoptotic RIP-mediated NF-kB Caspases are the most obvious examples of apoptotic induction (Lin et al., 1999; Martinon et al., 2000). factors targeted by caspases (see above). In addition, FAK kinase transduces cell survival signals from the multiple additional upstream and downstream media- extracellular matrix through a Raf-1 pathway involving tors of apoptosis are also cleaved by caspases. They can MAP kinases Erk-1 and Erk-2 and phosphatidylinositol be further subdivided into pro- and antiapoptotic 3-kinase, which is an activator of Akt (see for a review categories based on their functions. Schlaepfer et al., 1999). FAK is cleaved and inactivated In the proapoptotic category, the Bcl-2 family during apoptosis induced by multiple stimuli (e.g. c-myc member Bid is cleaved by caspase-8 following death overexpression, death receptor signaling, staurosporine receptor signaling, which results in its activation and and camptothecin), which contributes both to the loss of mitochondrial translocation (Li et al., 1998; Luo et al., survival signals and matrix detachment (Crouch et al., 1998). Recently, caspase-2 was also shown to cleave Bid 1996; Wen et al., 1997; Gervais et al., 1998). Multiple (Guo et al., 2002). Another prominent target of caspases additional components of this pathway, such as Raf-1, is DNA fragmentation factor 45kDa subunit (DFF45/ Akt, Cbl and Cbl-b, were also reported to be inactivated ICAD) (Liu et al., 1997; Enari et al., 1998). Caspase by caspase cleavage following Fas engagement (Wid- cleavage removes the N-terminal CIDE-N interaction mann et al., 1998b). domain, abrogating DFF45/ICAD-mediated suppres- The activation of transcription factor NF-kBisan sion of the catalytic DFF40/CAD subunit (Inohara important prosurvival mechanism (for reviews see de et al., 1999; Lugovskoy et al., 1999), which plays a Martin et al., 1999; Foo and Nolan, 1999). The p65/ critical role in the internucleosomal DNA degradation RelA subunit of NF-kB (Levkau et al., 1999) and characteristic of apoptosis (Sakahira et al., 1999). upstream activating kinase IKKb (Tang et al., 2001) In addition to proapoptotic factors, caspases also may be inhibited through caspase-dependent cleavage, cleave and, in most cases, inactivate antiapoptotic providing a way of tipping the balance towards cell proteins. For example, the antiapoptotic Bcl-2 family death. members, Bcl-2 and Bcl-xL, are cleaved between their A number of proapoptotic kinases were reported to N-terminal BH4 and BH3 domains following Sinbis be cleaved and activated by caspases. Fas signaling virus infection or IL-3 deprivation, resulting in the results in caspase-3-mediated cleavage of PAK2 (p21- conversion of these proteins into the proapoptotic activated kinase 2), separating the N-terminal regulatory

Oncogene Decade of caspases A Degterev et al 8559 and C-terminal catalytic domains and generating the by caspase-6 appears to be one of the very few reported constitutively active form of PAK2, which contributes examples of the requirement for a nonredundant activity to the formation of apoptotic bodies (Rudel and of a specific executioner caspase (Ruchaud et al., 2002). Bokoch, 1997). MEKK1 is another kinase activated The proteolytic disruption of multiple additional nucle- during genotoxic stress and Fas-induced apoptosis ar proteins, such as mitotic apparatus proteins NuMA, (Deak et al., 1998; Widmann et al., 1998a). Apoptotic topoisomerases I and II, RNA polymerase I and MEKK1 signaling involves JNK activation (Deak et al., upstream binding factor (UBF), may also contribute 1998). Mst1 kinase is both phosphorylated and cleaved to nuclear collapse (Casiano et al., 1996; Greidinger during apoptosis induced by Fas and staurosporine et al., 1996). (Graves et al., 1998; Lee et al., 1998). The activation of Finally, the basic mRNA splicing machinery is also this kinase also enhances apoptosis through the activa- affected during apoptosis through processing of several tion of JNK (Watabe et al., 1999) and/or the nuclear splicing factors by caspases. These include the 70 kDa translocation of the cleaved kinase, specifically con- U1-specific factor (Casciola-Rosen et al., 1994) and the tributing to chromatin condensation (Ura et al., 2001). heteronuclear riboproteins C1 and C2 (Waterhouse Mst1-related, Mst2, SLK and SPAK kinases, are also et al., 1996). cleaved and activated in response to multiple stimuli et al et al et al (Graves ., 1998; Lee ., 1998; Johnston ., Proteins required for cellular repair 2000; Sabourin et al., 2000). ROCK1 kinase was recently reported to be cleaved and activated in response Failure to repair or restore cellular functions in response to TNFa and this event was shown to be both necessary to cellular distress is one of the main triggers of and sufficient for apoptotic membrane blebbing (Cole- apoptosis and therefore induction of apoptosis in many man et al., 2001). Finally, a number of protein kinase C cases may follow activation of the cellular repair isoforms and related proteins are also proteolytically networks. Both apoptosis and cellular maintenance/ activated during apoptosis (Emoto et al., 1995; Ghayur repair pathways are resource-demanding processes and et al., 1996; Cryns et al., 1997; Datta et al., 1997b; Shao may compete for valuable cellular assets, such as ATP. et al., 1997; Endo et al., 2000). Overall, these results In addition, apoptotic execution itself causes cellular paint a picture of a complex network of downstream damage, which may trigger further activation of the kinase pathways controlling various aspects of apopto- repair pathways and create a feedback loop unfavorable sis execution activated or inactivated by caspases, for apoptosis. Finally, cellular maintenance pathways reinforcing the critical role played by caspases in may also interfere with apoptosis through activation of apoptotic signaling. downstream events that are not compatible with apoptosis. For instance, repair responses typically Structural and essential function proteins trigger cell cycle arrest, whereas apoptosis is associated with the upregulation and/or activation of many cell In the demolition phase of apoptosis, caspases cycle related proteins (see below). make a few strategic cuts in a number of critical The shutdown of cellular repair may accelerate the locations in order to dismantle the cellular architecture, execution of apoptosis. DNA-dependent protein kinase a process that is critical for the completion of apo- (DNA-PK), which plays a critical role in the induction ptosis. of DNA repair in response to double-strand breaks For example, cleavage of gelsolin leads to its (Lees-Miller, 1996), is one of the upstream caspase activation and, subsequently, gelsolin-mediated disrup- targets during apoptosis (Casciola-Rosen et al., 1995; tion of the actin filament network, which may specifi- Song et al., 1996). Human Rad51 and ATM proteins, cally contribute to cells rounding up and detaching from involved in recombination DNA repair, are also cleaved the matrix during apoptosis (Kothakota et al., 1997). A (Flygare et al., 1998; Hotti et al., 2000; Tong et al., number of intermediate filament proteins (keratins 18 2000). Interestingly, the same DNA repair proteins that and 19, vimentin) are also cleaved by caspases (Caulin are inactivated by caspases also appear to mediate the et al., 1997; Ku et al., 1997; Prasad et al., 1998). induction of apoptosis. For example, DNA-PK Interestingly, keratin 18 cleavage has been recently used cooperates with p53 in the induction of cell death in to demonstrate a role of scaffolding protein DEDD in response to DNA damage (Wang et al., 2000). These directing activated caspase-3 towards specific substrates observations suggest that many DNA repair enzymes (Lee et al., 2002), which may represent yet another possess multiple functions including repair, cell cycle mechanism regulating caspase function. Caspase clea- arrest and induction of apoptosis, and the extent of vage of the adherent junction proteins b-catenin and cellular damage may determine which function takes plakoglobin/g-catenin leads to the disruption of cell–cell over (Bernstein et al., 2002). interactions, which may in turn attenuate their prosur- The significance of cellular energy balance for vival signaling (Brancolini et al., 1998; Schmeiser et al., apoptosis execution is best illustrated by the studies of 1998). another DNA repair enzyme: poly(ADP-ribose) poly- Nuclear fragmentation critically depends on the merase (PARP). PARP is activated by single- and disruption of the nuclear filamentous network through double-strand DNA breaks and catalyses the attach- the cleavage of lamins A, B1 and C by caspases ment of ADP-ribose polymers to multiple nuclear (Lazebnik et al., 1995; Rao et al., 1996). Lamin cleavage factors, facilitating repair (for a review see Smulson

Oncogene Decade of caspases A Degterev et al 8560 et al., 2000). When activated, this process consumes caspase-8 and -10 as well as other proteins (Sanchez large amounts of NAD þ , thereby indirectly depleting et al., 1999; U et al., 2001). the cellular ATP store. Since the loss of cellular ATP HD is just one of the eight polyglutamine expansion- was demonstrated to abrogate apoptosis, resulting in the induced neurodegenerative disorders (Ross et al., 1999). induction of necrosis (Leist et al., 1997; Nicotera et al., Other disease loci-encoded factors, atropin-1 (DRPLA 1998), this may explain the observed caspase-mediated disease), ataxin-3 (SCA-3) and androgen receptor cleavage and inactivation of PARP during apoptosis (Kennedy’s disease), have also been found to be caspase (Kaufmann et al., 1993). targets and mutations of the putative caspase target sites in atropin-1 and androgen receptor were shown to diminish their cytotoxicity (Ellerby et al., 1999a, b). Cell cycle proteins The oligomerization of Ab 1–42 and Ab 1–40 peptides, which arise from the b and g-secretase cleavage A number of cell cycle inhibitory factors are cleaved of the transmembrane portion of the amyloid precursor during apoptosis. Those include Cdc27, a component of protein (APP), is critical to the progressive dysfunction the ubiquitin ligase complex degrading mitotic cyclins, and loss of cholinergic neurons in AD (Selkoe, 2000, Wee1, a kinase providing inhibitory phosphorylation of 2001). Caspase-3 has been found to cleave the cytosolic Cdks, and two Cdk inhibitors, p21CIP1 and p27KIP1 tail of APP at the residue 720, which is C-terminal to the (Levkau et al., 1998; Zhou et al., 1998). These g-secretase site, stimulating the subsequent g-secretase degradation events result in the accumulation of active cleavage step approximately fivefold (Gervais et al., Cdc2 and Cdk2, but surprisingly, do not result in 1999). The released cytosolic portion of APP has also mitotic entry (Zhou et al., 1998). In addition, death been suggested to play a separate role in the disease receptor signaling was found to cause caspase-depen- pathophysiology (Nishimura et al., 2002). In addition, dent inactivation of another proliferation inhibitor, Rb oligomeric Ab was shown to mediate caspase-dependent (An and Dou, 1996; Janicke et al., 1996). Overall, apoptosis in isolated neuronal cultures (LaFerla et al., induction of apoptosis appears to coincide with the 1995; Troy et al., 2000). However, the role of caspases in activation of multiple cell cycle-associated factors APP processing and neuronal degeneration caused by through cleavage and inactivation of their inhibitors. Ab peptide remains to be verified in vivo. Cleavage of cell cycle regulators may promote apoptosis signaling by multiple means as illustrated by recent studies. In one example, Cdc2 was shown to Nonapoptotic functions of caspases phosphorylate the proapoptotic Bcl-2 family member Bad at a site, which inhibits its interaction with Akt and In addition to their roles in apoptosis, multiple caspases thereby activates its proapoptotic activity (Konishi et al., possess additional functions not related to the cell death, 2002). In another example, loss of Rb was found to which may or may not involve other components of the promote accumulation of procaspases (Nahle et al., cell death machinery. 2002). The cleavage of cell cycle regulators may have Selected examples of such functions have already been evolved as a specific tumor suppressor mechanism aimed described in preceding sections. For example, caspase-1 at ensuring induction of cell death in response to and -11 play a role in the regulation of inflammatory improper cell cycle signaling by oncogenes. response. The components of the apoptotic death receptor pathway have been found to both down- Disease-related proteins regulate the immune response, by promoting apoptosis, as well as to stimulate the immune response, by Since the improper regulation of apoptosis contributes promoting lymphocyte proliferation and activation, as to a variety of diseases, caspases represent promising evidenced by the requirement for FADD and caspase-8 therapeutic targets (Nicholson, 2000; Talanian et al., in lymphocyte activation (Kabra et al., 2001; Chun et al., 2000). In addition to their role in apoptosis, caspases 2002). may also be involved in the direct cleavage of disease- Caspases may also be involved in terminal differentia- associated factors. Two of the most striking examples tion. A recent study showed that caspase-3-mediated are pertinent to neurodegenerative disorders: Hunting- activation of MST1 kinase is critical for proper muscle ton’s disease (HD) and Azheimer’s disease (AD). differentiation, and this process is defective in caspase- HD is caused by the genetic amplification of CAG 3/ mice (Fernando et al., 2002). Since MST1 repeats, which encode polyglutamine (polyQ), in the N- activation is also a part of the apoptotic response, it is terminus of the huntingtin protein (Ross et al., 1999). unclear how these two pathways are separated. In Caspases, including caspase-2 and -3, may contribute to addition, myeloid, monocyte and erythrocyte differen- the generation of expanded polyQ fragment of hunting- tiation may also require caspase activity (Pandey et al., tin through cleavage in the cluster of DXXD sites 2000; Zermati et al., 2001; Sordet et al., 2002). (Goldberg et al., 1996; Wellington et al., 2002). Keratinocyte differentiation may involve caspase-14 Processed expanded polyQ forms intranuclear aggre- (see above). Overall, these results indicate that caspases gates that are proposed to play a key role in the are involved in a variety of processes other than progressive neuronal dysfunction and neuronal loss apoptosis and this list will likely grow in the coming characteristic of HD by recruiting and activating years.

Oncogene Decade of caspases A Degterev et al 8561 Caspases in pathophysiology of disease chemoresistance (Kim et al., 2001; Liu et al., 2002; Shin et al., 2002b; Yang, 2002), whereas inactivation of Apaf- Apoptosis is a major mechanism aimed at insuring 1 has been observed in malignant melanomas (Soengas proper development and organism homeostasis. Due to et al., 2001). its role in the elimination of virally infected and While the activation or lack of caspases have been damaged cells, apoptosis also plays a central role in detected in many pathological conditions, it remains to the prevention of diseases. The enormous power of this be tested in the majority of the cases mentioned above process, however, makes the dysregulation of apoptosis whether caspases play an indispensable role in these deleterious for the organism and can lead to a host diseases. Due to the potential redundancy in apoptosis of pathologies (Thompson, 1995). Abnormal activation pathways and, in many cases, persistence of the cellular of apoptosis is implicated in tissue degeneration dysfunction, we urge caution when considering caspases during ischemia, AIDS, neurodegeneration and as therapeutic targets. septic shock. The induction of apoptosis in these diseases may significantly contribute to their pathogenesis. Caspase activation has been demonstrated in many Conclusions diseases characterized by abnormal cell death. For example, cleaved caspases have been detected in AD In the past decade, the number of publications on (Chan et al., 1999; LeBlanc et al., 1999; Stadelmann caspases has snowballed from zero to more than 10 000. et al., 1999), ALS (Pasinelli et al., 1998), Parkinson’s The critical roles of caspases in regulating signaling and disease (Hartmann et al., 2000; Tatton, 2000), HD execution of apoptosis are well accepted. However, (Sanchez et al., 1999) and ischemic brain (Namura et al., many important questions still remain unanswered. 1998; Matsushita et al., 2000). Inappropriate caspase- Firstly, it is very likely that we have uncovered only a mediated fodrin cleavage has been implicated in the small fraction of the biological functions of caspases. pathophysiology of Sjogren’s syndrome, an autoim- While the role of caspases in regulating apoptosis is mune disorder associated with excessive tissue destruc- unequivocal, nonapoptotic functions have been demon- tion (Hayashi et al., 2000; Inoue et al., 2001). Similarly, strated only for caspase-1, -3, -5, -8, -11 and -14 thus far. caspase activation was shown to correlate with spinal/ It is possible that all caspases will be found to regulate head trauma (Clark et al., 1999; Springer et al., 1999) functions other than apoptosis. Secondly, there is still and epilepsy (Henshall et al., 2000). Finally, the caspase- much to be learned about the mechanism of caspase 1 subfamily of caspases specifically contributes to activation initiated by different signaling processes. The inflammatory disorders associated with defects in DISC and apoptosome complexes provide the proto- cytokine processing, such as rheumatoid arthritis (RA) typic examples of such mechanisms, but the activation and inflammatory bowel disease (Crohn’s disease) mechanisms for other initiator caspases remain to be (Randle et al., 2001; Siegmund, 2002). explored. Thirdly, although the functional roles of The lack of proper caspase activity can also lead to caspases in regulating a variety of pathological processes pathologies. For example, the inactivation of caspase-10 have been demonstrated using cell culture and animal may give rise to ALPS II (see above). Mutations of models, the functional roles of caspases in human caspase-8 and -10 have been found in multiple human diseases remain to be examined directly by specific small tumors (see above and Liu et al., 2002; Shin et al., molecule inhibitors. We are optimistic that a decade 2002b). In addition, IAPs and FLIP are frequently from now, we will have caspase inhibitors as therapies overexpressed in cancers and may contribute to tumor for human diseases.

References

Acehan D, Jiang X, Morgan DG, Heuser JE, Wang X and Bachelder RE, Wendt MA, Fujita N, Tsuruo T and Mercurio Akey CW. (2002). Mol. Cell, 9, 423–432. AM. (2001). J. Biol. Chem., 276, 34702–34707. Ahmad M, Srinivasula SM, Hegde R, Mukattash R, Baliga BC, Colussi PA, Read SH, Dias MM, Jans DA and Fernandes-Alnemri T and Alnemri ES. (1998). Cancer Kumar S. (2002). J. Biol. Chem., 10, 10. Res., 58, 5201–5205. Barrett AJ and Rawlings ND. (2001). Biol. Chem., 382, 727– Ahmad M, Srinivasula SM, Wang L, Talanian RV, Litwack 733. G, Fernandes-Alnemri T and Alnemri ES. (1997). Cancer Bergeron L, Perez GI, Macdonald G, Shi L, Sun Y, Jurisicova Res., 57, 615–619. A, Varmuza S, Latham KE, Flaws JA, Salter JC, Hara H, Alnemri ES, Livingston DJ, Nicholson DW, Salvesen G, Moskowitz MA, Li E, Greenberg A, Tilly JL and Yuan J. Thornberry NA, Wong WW and Yuan J. (1996). Cell, 87, (1998). Genes Dev., 12, 1304–1314. 171. Bernstein C, Bernstein H, Payne CM and Garewal H. (2002). An B and Dou QP. (1996). Cancer Res., 56, 438–442. Mutat. Res., 511, 145–178. Bachelder RE, Ribick MJ, Marchetti A, Falcioni R, Soddu S, Bilsland J, Roy S, Xanthoudakis S, Nicholson DW, Han Y, Davis KR and Mercurio AM. (1999). J. Cell. Biol., 147, Grimm E, Hefti F and Harper SJ. (2002). J. Neurosci., 22, 1063–1072. 2637–2649.

Oncogene Decade of caspases A Degterev et al 8562 Blanchard H, Kodandapani L, Mittl PR, Marco SD, Krebs Coleman ML, Sahai EA, Yeo M, Bosch M, Dewar A and JF, Wu JC, Tomaselli KJ and Grutter MG. (1999). Struct. Olson MF. (2001). Nat. Cell Biol., 3, 339–345. Fold Des., 7, 1125–1133. Colussi PA, Harvey NL and Kumar S. (1998). J. Biol. Chem., Bodmer JL, Holler N, Reynard S, Vinciguerra P, Schneider P, 273, 24535–24542. Juo P, Blenis J and Tschopp J. (2000). Nat. Cell Biol., 2, 241– Cowling V and Downward J. (2002). Cell Death Differ., 9, 243. 1046–1056. Bossy-Wetzel E, Newmeyer DD and Green DR. (1998). Crook NE, Clem RJ and Miller LK. (1993). J. Virol., 67, EMBO J., 17, 37–49. 2168–2174. Brancolini C, Sgorbissa A and Schneider C. (1998). Cell Death Crouch DH, Fincham VJ and Frame MC. (1996). Oncogene, Differ., 5, 1042–1050. 12, 2689–2696. Brunet A, Bonni A, Zigmond MJ, Lin MZ, Juo P, Hu LS, Cryns V and Yuan J. (1998). Genes Dev., 12, 1551–1570. Anderson MJ, Arden KC, Blenis J and Greenberg ME. Cryns VL, Byun Y, Rana A, Mellor H, Lustig KD, Ghanem (1999). Cell, 96, 857–868. L, Parker PJ, Kirschner MW and Yuan J. (1997). J. Biol. Buckley CD, Pilling D, Henriquez NV, Parsonage G, Threlfall Chem., 272, 29449–29453. K, Scheel-Toellner D, Simmons DL, Akbar AN, Lord JM Darmon AJ, Nicholson DW and Bleackley RC. (1995). and Salmon M. (1999). Nature, 397, 534–539. Nature, 377, 446–448. Budihardjo I, Oliver H, Lutter M, Luo X and Wang X. (1999). Datta R, Kojima H, Banach D, Bump NJ, Talanian RV, Annu. Rev. Cell Dev. Biol., 15, 269–290. Alnemri ES, Weichselbaum RR, Wong WW and Kufe DW. Bump NJ, Hackett M, Hugunin M, Seshagiri S, Brady K, (1997a). J. Biol. Chem., 272, 1965–1969. Chen P, Ferenz C, Franklin S, Ghayur T and Li P, Licari J, Datta R, Kojima H, Yoshida K and Kufe D. (1997b). J. Biol. Mankovich L, Shi L, Greenberg AH, Miller LK and Wong Chem., 272, 20317–20320. Science WW. (1995). , 269, 1885–1888. Datta SR, Dudek H, Tao X, Masters S, Fu H, Gotoh Y and Cain K, Bratton SB, Langlais C, Walker G, Brown DG, Sun Greenberg ME. (1997c). Cell, 91, 231–241. XM and Cohen GM. (2000). J. Biol. Chem., 275, 6067–6070. de Martin R, Schmid JA and Hofer-Warbinek R. (1999). Cardone MH, Roy N, Stennicke HR, Salvesen GS, Franke Mutat. Res., 437, 231–243. TF, Stanbridge E, Frisch S and Reed JC. (1998). Science, Deaciuc IV, D’Souza NB, de Villiers WJ, Burikhanov R, 282, 1318–1321. Sarphie TG, Hill DB and McClain CJ. (2001). Alcohol Clin. Casciola-Rosen LA, Anhalt GJ and Rosen A. (1995). J. Exp. Exp. Res., 25, 935–943. Med., 182, 1625–1634. Deak JC, Cross JV, Lewis M, Qian Y, Parrott LA, Distelhorst Casciola-Rosen LA, Miller DK, Anhalt GJ and Rosen A. CW and Templeton DJ. (1998). Proc. Natl. Acad. Sci. USA, (1994). J. Biol. Chem., 269, 30757–30760. 95, 5595–5600. Casiano CA, Martin SJ, Green DR and Tan EM. (1996). del Peso L, Gonzalez-Garcia M, Page C, Herrera R and Nunez J. Exp. Med., 184, 765–770. G. (1997). Science, 278, 687–689. Caulin C, Salvesen GS and Oshima RG. (1997). J. Cell. Biol., Deshmukh M, Kuida K and Johnson Jr EM. (2000). J. Cell. 138, 1379–1394. Biol., 150, 131–143. Cerretti DP, Kozlosky CJ, Mosley B, Nelson N, Van Ness K, Dolle RE, Hoyer D, Prasad CV, Schmidt SJ, Helaszek CT, Greenstreet TA, March CJ, Kronheim SR, Druck T, Miller RE and Ator MA. (1994). J. Med. Chem., 37, 563– Cannizzaro LA, Huebner K and Black RA. (1992). Science, 564. 256, 97–100. Dolle RE, Singh J, Whipple D, Osifo IK, Speier G, Graybill Chai J, Wu Q, Shiozaki E, Srinivasula SM, Alnemri ES and TL, Gregory JS, Harris AL, Helaszek CT, Miller RE and Shi Y. (2001). Cell, 107, 399–407. Ator MA. (1995). J. Med. Chem., 38, 220–222. Chan SL, Griffin WS and Mattson MP. (1999). J. Neurosci. Duan H and Dixit VM. (1997). Nature, 385, 86–89. Res., 57, 315–323. Earnshaw WC, Martins LM and Kaufmann SH. (1999). Annu. Chang DW, Xing Z, Pan Y, Algeciras-Schimnich A, Barnhart Rev. Biochem., BC, Yaish-Ohad S, Peter ME and Yang X. (2002). EMBO 68, 383–424. J., 21, 3704–3714. Eckhart L, Ban J, Fischer H and Tschachler E. (2000a). Chautan M, Chazal G, Cecconi F, Gruss P and Golstein P. Biochem. Biophys. Res. Commun., 277, 655–659. (1999). Curr. Biol., 9, 967–970. Eckhart L, Declercq W, Ban J, Rendl M, Lengauer B, Mayer Chen Y and Lai MZ. (2001). J. Biol. Chem., 276, 8350–8357. C, Lippens S, Vandenabeele P and Tschachler E. (2000b). J. Invest. Dermatol., 115, 1148–1151. Cheng EH, Kirsch DG, Clem RJ, Ravi R, Kastan MB, Bedi A, Ueno K and Hardwick JM. (1997). Science, 278, Eddins MJ, Lemongello D, Friesen PD and Fisher AJ. (2002). 1966–1968. Acta Crystallogr. D: Biol. Crystallogr., 58, 299–302. Chun HJ, Zheng L, Ahmad M, Wang J, Speirs CK, Siegel Ekert PG, Silke J and Vaux DL. (1999a). Cell Death Differ., 6, RM, Dale JK, Puck J, Davis J, Hall CG, Skoda-Smith S, 1081–1086. Atkinson TP, Straus SE and Lenardo MJ. (2002). Nature, Ekert PG, Silke J and Vaux DL. (1999b). EMBO J., 18, 330– 419, 395–399. 338. Clark RS, Kochanek PM, Chen M, Watkins SC, Marion DW, Ellerby LM, Andrusiak RL, Wellington CL, Hackam AS, Chen J, Hamilton RL, Loeffert JE and Graham SH. (1999). Propp SS, Wood JD, Sharp AH, Margolis RL, Ross CA, FASEB J., 13, 813–821. Salvesen GS, Hayden MR and Bredesen DE. (1999a). Clem RJ, Cheng EH, Karp CL, Kirsch DG, Ueno K, J. Biol. Chem., 274, 8730–8736. Takahashi A, Kastan MB, Griffin DE, Earnshaw WC, Ellerby LM, Hackam AS, Propp SS, Ellerby HM, Rabizadeh Veliuona MA and Hardwick JM. (1998). Proc. Natl. Acad. S, Cashman NR, Trifiro MA, Pinsky L, Wellington CL, Sci. USA, 95, 554–559. Salvesen GS, Hayden MR and Bredesen DE. (1999b). Clem RJ, Fechheimer M and Miller LK. (1991). Science, 254, J. Neurochem., 72, 185–195. 1388–1390. Ellis HM and Horvitz HR. (1986). Cell, 44, 817–829.

Oncogene Decade of caspases A Degterev et al 8563 Emoto Y, Manome Y, Meinhardt G, Kisaki H, Kharbanda S, Hakem R, Hakem A, Duncan GS, Henderson JT, Woo M, Robertson M, Ghayur T, Wong WW, Kamen R and Soengas MS, Elia A, de la Pompa JL, Kagi D, Khoo W, Weichselbaum R. (1995). EMBO J., 14, 6148–6156. Potter J, Yoshida R, Kaufman SA, Lowe SW, Penninger JM Enari M, Hug H and Nagata S. (1995). Nature, 375, 78–81. and Mak TW. (1998). Cell, 94, 339–352. Enari M, Sakahira H, Yokoyama H, Okawa K, Iwamatsu A Han BH, Xu D, Choi J, Han Y, Xanthoudakis S, Roy S, Tam and Nagata S. (1998). Nature, 391, 43–50. J, Vaillancourt J, Colucci J, Siman R, Giroux A, Robertson Endo K, Oki E, Biedermann V, Kojima H, Yoshida K, GS, Zamboni R, Nicholson DW and Holtzman DM. (2002). Johannes FJ, Kufe D and Datta R. (2000). J. Biol. Chem., J. Biol. Chem., 277, 30128–30136. 275, 18476–18481. Hara H, Fink K, Endres M, Friedlander RM, Gagliardini V, Estrov Z, Black RA, Sleath PR, Harris D, Van Q, LaPushin R, Yuan J and Moskowitz MA. (1997). J. Cereb. Blood Flow Estey EH and Talpaz M. (1995). Blood, 86, 4594–4602. Metab., 17, 370–375. Fernando P, Kelly JF, Balazsi K, Slack RS and Megeney LA. Haraguchi M, Torii S, Matsuzawa S, Xie Z, Kitada S, (2002). Proc. Natl. Acad. Sci. USA, 99, 11025–11030. Krajewski S, Yoshida H, Mak TW and Reed JC. (2000). Fischer H, Koenig U, Eckhart L and Tschachler E. (2002). J. Exp. Med., 191, 1709–1720. Biochem. Biophys. Res. Commun., 293, 722–726. Hartmann A, Hunot S, Michel PP, Muriel MP, Vyas S, Fisher AJ, Cruz W, Zoog SJ, Schneider CL and Friesen PD. Faucheux BA, Mouatt-Prigent A, Turmel H, Srinivasan A, (1999). EMBO J., 18, 2031–2039. Ruberg M, Evan GI, Agid Y and Hirsch EC. (2000). Proc. Flygare J, Armstrong RC, Wennborg A, Orsan S and Hellgren Natl. Acad. Sci. USA, 97, 2875–2880. D. (1998). FEBS Lett., 427, 247–251. Harvey NL, Butt AJ and Kumar S. (1997). J. Biol. Chem., 272, Foo SY and Nolan GP. (1999). Trends Genet., 15, 229–235. 13134–13139. Friedlander RM, Brown RH, Gagliardini V, Wang J and Haviv R, Lindenboim L, Yuan J and Stein R. (1998). Yuan J. (1997a). Nature, 388, 31. J. Neurosci. Res., 52, 491–497. Friedlander RM, Gagliardini V, Hara H, Fink KB, Li W, Hayashi Y, Yayagi K and Haneji N. (2000). Arch. Immunol. MacDonald G, Fishman MC, Greenberg AH, Moskowitz Ther. Exp. (Warsz), 48, 399–403. MA and Yuan J. (1997b). J. Exp. Med., 185, 933–940. Hengartner MO, Ellis RE and Horvitz HR. (1992). Nature, Fujita E, Jinbo A, Matuzaki H, Konishi H, Kikkawa U and 356, 494–499. Momoi T. (1999). Biochem. Biophys. Res. Commun., 264, Hengartner MO and Horvitz HR. (1994). Cell, 76, 550–555. 665–676. Gagliardini V, Fernandez PA, Lee RK, Drexler HC, Rotello Henshall DC, Clark RS, Adelson PD, Chen M, Watkins SC RJ, Fishman MC and Yuan J. (1994). Science, 263, 826–828. and Simon RP. (2000). Neurology, 55, 250–257. Garcia-Calvo M, Peterson EP, Leiting B, Ruel R, Nicholson Hirata H, Takahashi A, Kobayashi S, Yonehara S, Sawai H, DW and Thornberry NA. (1998). J. Biol. Chem., 273, 32608– Okazaki T, Yamamoto K and Sasada M. (1998). J. Exp. 32613. Med., 187, 587–600. Gervais FG, Thornberry NA, Ruffolo SC, Nicholson DW and Hoglen NC, Hirakawa BP, Fisher CD, Weeks S, Srinivasan A, Roy S. (1998). J. Biol. Chem., 273, 17102–17108. Wong AM, Valentino KL, Tomaselli KJ, Bai X, Karanews- Gervais FG, Xu D, Robertson GS, Vaillancourt JP, Zhu Y, ky DS and Contreras PC. (2001). J. Pharmacol. Exp. Ther., Huang J, LeBlanc A, Smith D, Rigby M, Shearman MS, 297, 811–818. Clarke EE, Zheng H, Van Der Ploeg LH, Ruffolo SC, Holler N, Zaru R, Micheau O, Thome M, Attinger A, Valitutti Thornberry NA, Xanthoudakis S, Zamboni RJ, Roy S and S, Bodmer JL, Schneider P, Seed B and Tschopp J. (2000). Nicholson DW. (1999). Cell, 97, 395–406. Nat. Immunol., 1, 489–495. Ghayur T, Banerjee S, Hugunin M, Butler D, Herzog L, Hotchkiss RS, Chang KC, Swanson PE, Tinsley KW, Hui JJ, Carter A, Quintal L, Sekut L, Talanian R, Paskind M, Klender P, Xanthoudakis S, Roy S, Black C, Grimm E, Wong W, Kamen R, Tracey D and Allen H. (1997). Nature, Aspiotis R, Han Y, Nicholson DW and Karl IE. (2000). Nat. 386, 619–623. Immunol., 1, 496–501. Ghayur T, Hugunin M, Talanian RV, Ratnofsky S, Quinlan Hotti A, Jarvinen K, Siivola P and Holtta E. (2000). Oncogene, C, Emoto Y, Pandey P, Datta R, Huang Y, Kharbanda S, 19, 2354–2362. Allen H, Kamen R, Wong W and Kufe D. (1996). J. Exp. Hu S, Snipas SJ, Vincenz C, Salvesen G and Dixit VM. (1998). Med., 184, 2399–2404. J. Biol. Chem., 273, 29648–29653. Goldberg YP, Nicholson DW, Rasper DM, Kalchman MA, Hunter CP. (1999). Curr. Biol., 9, R440–R442. Koide HB, Graham RK, Bromm M, Kazemi-Esfarjani P, Inohara N, Koseki T, Chen S, Benedict MA and Nunez G. Thornberry NA, Vaillancourt JP and Hayden MR. (1996). (1999). J. Biol. Chem., 274, 270–274. Nat. Genet., 13, 442–449. Inoue H, Tsubota K, Ono M, Kizu Y, Mizuno F, Takada K, Graves JD, Gotoh Y, Draves KE, Ambrose D, Han DK, Yamada K, Yanagi K, Hayashi Y and Saito I. (2001). Wright M, Chernoff J, Clark EA and Krebs EG. (1998). J. Immunol., 166, 5801–5809. EMBO J., 17, 2224–2234. Irmler M, Thome M, Hahne M, Schneider P, Hofmann K, Graybill TL, Dolle RE, Helaszek CT, Miller RE and Ator Steiner V, Bodmer JL, Schroter M, Burns K, Mattmann C, MA. (1994). Int. J. Pept. Protein Res., 44, 173–182. Rimoldi D, French LE and Tschopp J. (1997). Nature, 388, Greidinger EL, Miller DK, Yamin TT, Casciola-Rosen L and 190–195. Rosen A. (1996). FEBS Lett., 390, 299–303. Janicke RU, Walker PA, Lin XY and Porter AG. (1996). Grobmyer SR, Armstrong RC, Nicholson SC, Gabay C, EMBO J., 15, 6969–6978. Arend WP, Potter SH, Melchior M, Fritz LC and Nathan Johnston AM, Naselli G, Gonez LJ, Martin RM, Harrison LC CF. (1999). Mol. Med., 5, 585–594. and DeAizpurua HJ. (2000). Oncogene, 19, 4290–4297. Gross A, McDonnell JM and Korsmeyer SJ. (1999). Genes Juo P, Kuo CJ, Yuan J and Blenis J. (1998). Curr. Biol., 8, Dev., 13, 1899–1911. 1001–1008. Guo Y, Srinivasula SM, Druilhe A, Fernandes-Alnemri T and Kabra NH, Kang C, Hsing LC, Zhang J and Winoto A. Alnemri ES. (2002). J. Biol. Chem., 277, 13430–13437. (2001). Proc. Natl. Acad. Sci. USA, 98, 6307–6312.

Oncogene Decade of caspases A Degterev et al 8564 Kang SJ, Wang S, Hara H, Peterson EP, Namura S, Amin- MS, Ryan MD, Haltiwanger RC, Liang PH, Janson CA, Hanjani S, Huang Z, Srinivasan A, Tomaselli KJ, Thorn- McDevitt PJ, Johanson K, Concha NO, Chan W, Abdel- berry NA, Moskowitz MA and Yuan J. (2000). J. Cell. Biol., Meguid SS, Badger AM, Lark MW, Nadeau DP, Suva LJ, 149, 613–622. Gowen M and Nuttall ME. (2000). J. Biol. Chem., 275, Kang SJ, Wang S, Kuida K and Yuan J. (2002). Cell Death 16007–16014. Differ., 9, 1115–1125. Lee JC, Schickling O, Stegh AH, Oshima RG, Dinsdale D, Kaufmann SH, Desnoyers S, Ottaviano Y, Davidson NE and Cohen GM and Peter ME. (2002). J. Cell. Biol., 158, 1051– Poirier GG. (1993). Cancer Res., 53, 3976–3985. 1066. Kawahara A, Ohsawa Y, Matsumura H, Uchiyama Y and Lee KK, Murakawa M, Nishida E, Tsubuki S, Kawashima S, Nagata S. (1998). J. Cell. Biol., 143, 1353–1360. Sakamaki K and Yonehara S. (1998). Oncogene, 16, 3029–3037. Kim PK, Mahidhara R and Seol DW. (2001). Drug Resist. Lees-Miller SP. (1996). Biochem. Cell Biol., 74, 503–512. Updat., 4, 293–296. Leist M, Single B, Castoldi AF, Kuhnle S and Nicotera P. Kim SJ, Ju JW, Oh CD, Yoon YM, Song WK, Kim JH, Yoo (1997). J. Exp. Med., 185, 1481–1486. YJ, Bang OS, Kang SS and Chun JS. (2002). J. Biol. Chem., Leonard JR, Klocke BJ, D’Sa C, Flavell RA and Roth KA. 277, 1332–1339. (2002). J. Neuropathol. Exp. Neurol., 61, 673–677. Kirchhoff S, Muller WW, Li-Weber M and Krammer PH. Levkau B, Koyama H, Raines EW, Clurman BE, Herren B, (2000). Eur. J. Immunol., 30, 2765–2774. Orth K, Roberts JM and Ross R. (1998). Mol. Cell, 1, 553–563. Kischkel FC, Lawrence DA, Tinel A, LeBlanc H, Virmani A, Levkau B, Scatena M, Giachelli CM, Ross R and Raines EW. Schow P, Gazdar A, Blenis J, Arnott D and Ashkenazi A. (1999). Nat. Cell Biol., 1, 227–233. (2001). J. Biol. Chem., 276, 46639–46646. Li H, Bergeron L, Cryns V, Pasternack MS, Zhu H, Shi L, Kluck RM, Bossy-Wetzel E, Green DR and Newmeyer DD. Greenberg A and Yuan J. (1997a). J. Biol. Chem., 272, (1997). Science, 275, 1132–1136. 21010–21017. Koenig U, Eckhart L and Tschachler E. (2001). Biochem. Li H and Yuan J. (1999). Curr. Opin. Cell Biol., 11, 261–266. Biophys. Res. Commun., 285, 1150–1154. Li H, Zhu H, Xu CJ and Yuan J. (1998). Cell, 94, 491–501. Komiyama T, Ray CA, Pickup DJ, Howard AD, Thornberry Li J, Billiar TR, Talanian RV and Kim YM. (1997b). Biochem. NA, Peterson EP and Salvesen G. (1994). J. Biol. Chem., Biophys. Res. Commun., 240, 419–424. 269, 19331–19337. Li P, Allen H, Banerjee S, Franklin S, Herzog L, Johnston C, Konishi Y, Lehtinen M, Donovan N and Bonni A. (2002). McDowell J, Paskind M, Rodman L, Salfeld J, Towne E, Mol. Cell, 9, 1005–1016. Tracey D, Wardwell S, Wei F-Y, Wong W, Kamen R and Korsmeyer SJ, Wei MC, Saito M, Weiler S, Oh KJ Seshadri T. (1995). Cell, 80, 401–411. and Schlesinger PH. (2000). Cell Death Differ., 7, Li P, Nijhawan D, Budihardjo I, Srinivasula SM, Ahmad M, 1166–1173. Alnemri ES and Wang X. (1997c). Cell, 91, 479–489. Kothakota S, Azuma T, Reinhard C, Klippel A, Tang J, Chu Lin XY, Choi MS and Porter AG. (2000). J. Biol. Chem., 275, K, McGarry TJ, Kirschner MW, Koths K, Kwiatkowski DJ 39920–39926. and Williams LT. (1997). Science, 278, 294–298. Lin Y, Devin A, Rodriguez Y and Liu ZG. (1999). Genes Dev., Krueger A, Schmitz I, Baumann S, Krammer PH and 13, 2514–2526. Kirchhoff S. (2001). J. Biol. Chem., 276, 20633–20640. Lippens S, Kockx M, Knaapen M, Mortier L, Polakowska R, Ku NO, Liao J and Omary MB. (1997). J. Biol. Chem., 272, Verheyen A, Garmyn M, Zwijsen A, Formstecher P, 33197–33203. Huylebroeck D, Vandenabeele P and Declercq W. (2000). Kuida K, Haydar TF, Kuan CY, Gu Y, Taya C, Karasuyama Cell Death Differ., 7, 1218–1224. H, Su MS, Rakic P and Flavell RA. (1998). Cell, 94, 325– Liu B, Peng D, Lu Y, Jin W and Fan Z. (2002). J. Biol. Chem., 337. 277, 30159–30164. Kuida K, Lippke JA, Ku G, Harding MW, Livingston DJ, Su Liu X, Kim CN, Yang J, Jemmerson R and Wang X. (1996). MS and Flavell RA. (1995). Science, 267, 2000–2003. Cell, 86, 147–157. Kuida K, Zheng TS, Na S, Kuan C, Yang D, Karasuyama H, Liu X, Zou H, Slaughter C and Wang X. (1997). Cell, 89, 175– Rakic P and Flavell RA. (1996). Nature, 384, 368–372. 184. Kulms D, Dussmann H, Poppelmann B, Stander S, Schwarz A Livingston DJ. (1997). J. Cell. Biochem., 64, 19–26. and Schwarz T. (2002). Cell Death Differ., 9, 598–608. Lockshin RA. (1969). J. Insect Physiol., 15, 1505–1516. LaFerla FM, Tinkle BT, Bieberich CJ, Haudenschild CC and Lockshin RA and Williams CM. (1965). J. Insect Physiol., 11, Jay G. (1995). Nat. Genet., 9, 21–30. 803–809. Lamkanﬁ M, Declercq W, Kalai M, Saelens X and Vandena- Los M, Van de Craen M, Penning LC, Schenk H, Westendorp beele P. (2002). Cell Death Differ., 9, 358–361. M, Baeuerle PA, Droge W, Krammer PH, Fiers W and Lassus P, Opitz-Araya X and Lazebnik Y. (2002). Science, Schulze-Osthoff K. (1995). Nature, 375, 81–83. 297, 1352–1354. Lugovskoy AA, Zhou P, Chou JJ, McCarty JS, Li P and Lazebnik YA, Takahashi A, Poirier GG, Kaufmann SH Wagner G. (1999). Cell, 99, 747–755. and Earnshaw WC. (1995). J. Cell Sci., 19 (Suppl.), Luo X, Budihardjo I, Zou H, Slaughter C and Wang X. 41–49. (1998). Cell, 94, 481–490. Le DA, Wu Y, Huang Z, Matsushita K, Plesnila N, Mancini M, Machamer CE, Roy S, Nicholson DW, Thorn- Augustinack JC, Hyman BT, Yuan J, Kuida K, Flavell berry NA, Casciola-Rosen LA and Rosen A. (2000). J. Cell. RA and Moskowitz MA. (2002). Proc. Natl. Acad. Sci. Biol., 149, 603–612. USA, 99, 15188–15193. Margolin N, Raybuck SA, Wilson KP, Chen W, Fox T, Gu Y LeBlanc A, Liu H, Goodyer C, Bergeron C and Hammond J. and Livingston DJ. (1997). J. Biol. Chem., 272, 7223–7228. (1999). J. Biol. Chem., 274, 23426–23436. Marsden VS, O’Connor L, O’Reilly LA, Silke J, Metcalf D, Lee D, Long SA, Adams JL, Chan G, Vaidya KS, Francis TA, Ekert PG, Huang DC, Cecconi F, Kuida K, Tomaselli KJ, Kikly K, Winkler JD, Sung CM, Debouck C, Richardson S, Roy S, Nicholson DW, Vaux DL, Bouillet P, Adams JM and Levy MA, DeWolf Jr WE, Keller PM, Tomaszek T, Head Strasser A. (2002). Nature, 419, 634–637.

Oncogene Decade of caspases A Degterev et al 8565 Martinon F, Burns K and Tschopp J. (2002). Mol. Cell, 10, Pandey P, Nakazawa A, Ito Y, Datta R, Kharbanda S and 417–426. Kufe D. (2000). Oncogene, 19, 3941–3947. Martinon F, Holler N, Richard C and Tschopp J. (2000). Park WS, Lee JH, Shin MS, Park JY, Kim HS, Kim YS, Lee FEBS Lett., 468, 134–136. SN, Xiao W, Park CH, Lee SH, Yoo NJ and Lee JY. (2002). Matsumura H, Shimizu Y, Ohsawa Y, Kawahara A, Oncogene, 21, 2919–2925. Uchiyama Y and Nagata S. (2000). J. Cell. Biol., 151, Paroni G, Henderson C, Schneider C and Brancolini C. (2002). 1247–1256. J. Biol. Chem., 277, 15147–15161. Matsushita K, Wu Y, Qiu J, Lang-Lazdunski L, Hirt L, Pasinelli P, Borchelt DR, Houseweart MK, Cleveland DW and Waeber C, Hyman BT, Yuan J and Moskowitz MA. (2000). Brown Jr RH. (1998). Proc. Natl. Acad. Sci. USA, 95, J. Neurosci., 20, 6879–6887. 15763–15768. Matsuyama S, Llopis J, Deveraux QL, Tsien RY and Reed JC. Pistritto G, Jost M, Srinivasula SM, Baffa R, Poyet JL, Kari (2000). Nat. Cell Biol., 2, 318–325. C, Lazebnik Y, Rodeck U and Alnemri ES. (2002). Cell Micheau O, Solary E, Hammann A and Dimanche-Boitrel Death Differ., 9, 995–1006. MT. (1999). J. Biol. Chem., 274, 7987–7992. Pop C, Chen YR, Smith B, Bose K, Bobay B, Tripathy A, Micheau O, Thome M, Schneider P, Holler N, Tschopp J, Franzen S and Clark AC. (2001). Biochemistry, 40, 14224– Nicholson D, Briand C and Grutter MG. (2002). J. Biol. 14235. Chem., 4, 4. Prasad SC, Thraves PJ, Kuettel MR, Srinivasarao GY, Mittl PR, Di Marco S, Krebs JF, Bai X, Karanewsky DS, Dritschilo A and Soldatenkov VA. (1998). Biochem. Priestle JP, Tomaselli KJ and Grutter MG. (1997). J. Biol. Biophys. Res. Commun., 249, 332–338. Chem., 272, 6539–6547. Quan LT, Caputo A, Bleackley RC, Pickup DJ and Salvesen Miura M, Friedlander RM and Yuan J. (1995). Proc. Natl. GS. (1995). J. Biol. Chem., 270, 10377–10379. Acad. Sci. USA, 92, 8318–8322. Ramage P, Cheneval D, Chvei M, Graff P, Hemmig R, Heng Miura M, Zhu H, Rotello R, Hartwieg EA and Yuan J. (1993). R, Kocher HP, Mackenzie A, Memmert K, Revesz L and Cell, 75, 653–660. Wishart W. (1995). J. Biol. Chem., 270, 9378–9383. Mohr S, Zech B, Lapetina EG and Brune B. (1997). Biochem. Randle JC, Harding MW, Ku G, Schonharting M and Kurrle Biophys. Res. Commun., 238, 387–391. R. (2001). Expert Opin. Invest. Drugs, 10, 1207–1209. Mootha VK, Wei MC, Buttle KF, Scorrano L, Panoutsako- Rano TA, Timkey T, Peterson EP, Rotonda J, Nicholson DW, poulou V, Mannella CA and Korsmeyer SJ. (2001). EMBO Becker JW, Chapman KT and Thornberry NA. (1997). J., 20, 661–671. Chem. Biol., 4, 149–155. Morishima N, Nakanishi K, Takenouchi H, Shibata T and Rao L, Perez D and White E. (1996). J. Cell. Biol., 135, 1441– Yasuhiko Y. (2002). J. Biol. Chem., 277, 34287–34294. 1455. Muzio M, Chinnaiyan AM, Kischkel FC, O’Rourke K, Shevchenko A, Ni J, Scafﬁdi C, Bretz JD, Zhang M, Gentz Rao RV, Castro-Obregon S, Frankowski H, Schuler M, Stoka R, Mann M, Krammer PH, Peter ME and Dixit VM. (1996). V, del Rio G, Bredesen DE and Ellerby HM. (2002). J. Biol. Chem., Cell, 85, 817–827. 277, 21836–21842. Muzio M, Stockwell BR, Stennicke HR, Salvesen GS and Rao RV, Hermel E, Castro-Obregon S, del Rio G, Ellerby Dixit VM. (1998). J. Biol. Chem., 273, 2926–2930. LM, Ellerby HM and Bredesen DE. (2001). J. Biol. Chem., Nahle Z, Polakoff J, Davuluri RV, McCurrach ME, Jacobson 276, 33869–33874. MD, Narita M, Zhang MQ, Lazebnik Y, Bar-Sagi D and Ray CA, Black RA, Kronheim SR, Greenstreet TA, Lowe SW. (2002). Nat. Cell Biol., 4, 859–864. Sleath PR, Salvesen GS and Pickup DJ. (1992). Cell, 69, Nakagawa T and Yuan J. (2000). J. Cell. Biol., 150, 887–894. 597–604. Nakagawa T, Zhu H, Morishima N, Li E, Xu J, Yankner BA Renatus M, Stennicke HR, Scott FL, Liddington RC and and Yuan J. (2000). Nature, 403, 98–103. Salvesen GS. (2001). Proc. Natl. Acad. Sci. USA, 98, 14250– Namura S, Zhu J, Fink K, Endres M, Srinivasan A, Tomaselli 14255. KJ, Yuan J and Moskowitz MA. (1998). J. Neurosci., 18, Renatus M, Zhou Q, Stennicke HR, Snipas SJ, Turk D, 3659–3668. Bankston LA, Liddington RC and Salvesen GS. (2000). Natori S, Selzner M, Valentino KL, Fritz LC, Srinivasan A, Struct. Fold Des., 8, 789–797. Clavien PA and Gores GJ. (1999). Transplantation, 68, 89– Riedl SJ, Fuentes-Prior P, Renatus M, Kairies N, Krapp S, 96. Huber R, Salvesen GS and Bode W. (2001a). Proc. Natl. Nicholson DW. (2000). Nature, 407, 810–816. Acad. Sci. USA, 98, 14790–14795. Nicholson DW, Ali A, Thornberry NA, Vaillancourt JP, Ding Riedl SJ, Renatus M, Snipas SJ and Salvesen GS. (2001b). CK, Gallant M, Gareau Y, Grifﬁn PR, Labelle M and Biochemistry, 40, 13274–13280. Lazebnik YA, Munday NA, Raju SM, Smulson ME, Yamin Robertson JD, Enoksson M, Suomela M, Zhivotovsky B and T-T, Yu VL and Miller DK. (1995). Nature, 376, 37–43. Orrenius S. (2002). J. Biol. Chem., 277, 29803–29809. Nicotera P, Leist M and Ferrando-May E. (1998). Toxicol. Rodriguez J and Lazebnik Y. (1999). Genes Dev., 13, 3179– Lett., 102–103, 139–142. 3184. Nicotera P, Leist M and Ferrando-May E. (1999). Biochem. Ross CA, Wood JD, Schilling G, Peters MF, Nucifora Jr FC, Soc. Symp., 66, 69–73. Cooper JK, Sharp AH, Margolis RL and Borchelt DR. Nishimura I, Uetsuki T, Kuwako K, Hara T, Kawakami T, (1999). Philos. Trans. R. Soc. Lond. B. Biol. Sci., 354, 1005– Aimoto S and Yoshikawa K. (2002). Cell Death Differ., 9, 1011. 199–208. Rotonda J, Nicholson DW, Fazil KM, Gallant M, Gareau Y, Oppenheim RW, Flavell RA, Vinsant S, Prevette D, Kuan CY Labelle M, Peterson EP, Rasper DM, Ruel R, Vaillancourt and Rakic P. (2001). J. Neurosci., 21, 4752–4760. JP, Thornberry NA and Becker JW. (1996). Nat. Struct. Orth K and Dixit VM. (1997). J. Biol. Chem., 272, 8841–8844. Biol., 3, 619–625. Orth K, O’Rourke K, Salvesen GS and Dixit VM. (1996). Roy S, Bayly CI, Gareau Y, Houtzager VM, Kargman S, J. Biol. Chem., 271, 20977–20980. Keen SL, Rowland K, Seiden IM, Thornberry NA and

Oncogene Decade of caspases A Degterev et al 8566 Nicholson DW. (2001). Proc. Natl. Acad. Sci. USA, 98, Sprick MR, Rieser E, Stahl H, Grosse-Wilde A, Weigand MA 6132–6137. and Walczak H. (2002). EMBO J., 21, 4520–4530. Ruchaud S, Korfali N, Villa P, Kottke TJ, Dingwall C, Sprick MR, Weigand MA, Rieser E, Rauch CT, Juo P, Blenis Kaufmann SH and Earnshaw WC. (2002). EMBO J., 21, J, Krammer PH and Walczak H. (2000). Immunity, 12, 599– 1967–1977. 609. Rudel T. (1999). Herz, 24, 236–241. Springer JE, Azbill RD and Knapp PE. (1999). Nat. Med., 5, Rudel T and Bokoch GM. (1997). Science, 276, 1571–1574. 943–946. Sabourin LA, Tamai K, Seale P, Wagner J and Rudnicki MA. Srinivasula SM, Hegde R, Saleh A, Datta P, Shiozaki E, Chai (2000). Mol. Cell. Biol., 20, 684–696. J, Lee RA, Robbins PD, Fernandes-Alnemri T, Shi Y and Sakahira H, Enari M, Ohsawa Y, Uchiyama Y and Nagata S. Alnemri ES. (2001). Nature, 410, 112–116. (1999). Curr. Biol., 9, 543–546. Stadelmann C, Deckwerth TL, Srinivasan A, Bancher C, Salvesen GS and Dixit VM. (1999). Proc. Natl. Acad. Sci. Bruck W, Jellinger K and Lassmann H. (1999). Am. J. USA, 96, 10964–10967. Pathol., 155, 1459–1466. Sanchez I, Xu CJ, Juo P, Kakizaka A, Blenis J and Yuan J. Stefanis L, Troy CM, Qi H, Shelanski ML and Greene LA. (1999). Neuron, 22, 623–633. (1998). J. Neurosci., 18, 9204–9215. Scaffidi C, Fulda S, Srinivasan A, Friesen C, Li F, Tomaselli Stennicke HR, Deveraux QL, Humke EW, Reed JC, Dixit VM KJ, Debatin KM, Krammer PH and Peter ME. (1998). and Salvesen GS. (1999). J. Biol. Chem., 274, 8359–8362. EMBO J., 17, 1675–1687. Stennicke HR, Renatus M, Meldal M and Salvesen GS. Scaffidi C, Schmitz I, Krammer PH and Peter ME. (1999). (2000). Biochem. J., 350, 563–568. J. Biol. Chem., 274, 1541–1548. Stennicke HR and Salvesen GS. (1997). J. Biol. Chem., 272, Schlaepfer DD, Hauck CR and Sieg DJ. (1999). Prog. Biophys. 25719–25723. Mol. Biol., 71, 435–478. Stennicke HR and Salvesen GS. (1999). Cell Death Differ., 6, Schmeiser K, Hammond EM, Roberts S and Grand RJ. 1054–1059. (1998). FEBS Lett., 433, 51–57. Stennicke HR and Salvesen GS. (2000). Biochim. Biophys. Schotte P, Van Criekinge W, Van de Craen M, Van Loo G, Acta, 1477, 299–306. Desmedt M, Grooten J, Cornelissen M, De Ridder L, Talanian RV, Brady KD and Cryns VL. (2000). J. Med. Vandekerckhove J, Fiers W, Vandenabeele P and Beyaert R. Chem., 43, 3351–3371. (1998). Biochem. Biophys. Res. Commun., 251, 379–387. Tang D and Kidd VJ. (1998). J. Biol. Chem., 273, 28549– Selkoe DJ. (2000). Ann. NY Acad. Sci., 924, 17–25. 28552. Selkoe DJ. (2001). Physiol. Rev., 81, 741–766. Tang G, Yang J, Minemoto Y and Lin A. (2001). Mol. Cell, 8, Shao RG, Cao CX and Pommier Y. (1997). J. Biol. Chem., 1005–1016. 272, 31321–31325. Tatton NA. (2000). Exp. Neurol., 166, 29–43. Sharp PA. (1999). Genes Dev., 13, 139–141. Thompson CB. (1995). Science, 267, 1456–1462. Shi Y. (2002). Mol. Cell, 9, 459–470. Thornberry NA, Bull HG, Calaycay JR, Chapman KT, Shin MS, Kim HS, Kang CS, Park WS, Kim SY, Lee SN, Lee Howard AD, Kostura MJ, Miller DK, Molineaux SM, JH, Park JY, Jang JJ, Kim CW, Kim SH, Lee JY, Yoo NJ Weidner JR, Aunins J, Euiston KO, Ayala JM, Casano FJ, and Lee SH. (2002a). Blood, 99, 4094–4099. Chin J, Ding GJF, Egger LA, Gaffney EP, Limjuco G, Shin MS, Kim HS, Lee SH, Lee JW, Song YH, Kim YS, Park Palyha OC, Raju SM, Rolando AM, Salley JP, Yamin T-T, WS, Kim SY, Lee SN, Park JY, Lee JH, Xiao W, Jo KH, Lee TD, Shively JE, MacCross M, Mumford RA, Schmidt Wang YP, Lee KY, Park YG, Kim SH, Lee JY and Yoo NJ. JA and Tocci MJ. (1992). Nature, 356, 768–774. (2002b). Oncogene, 21, 4129–4136. Thornberry NA, Peterson EP, Zhao JJ, Howard AD, Griffin Shiozaki EN, Chai J and Shi Y. (2002). Proc. Natl. Acad. Sci. PR and Chapman KT. (1994). Biochemistry, 33, 3934–3940. USA, 99, 4197–4202. Thornberry NA, Rano TA, Peterson EP, Rasper DM, Timkey Siegmund B. (2002). Biochem. Pharmacol., 64, 1–8. T, Garcia-Calvo M, Houtzager VM, Nordstrom PA, Roy S, Slee EA, Adrain C and Martin SJ. (2001). J. Biol. Chem., 276, Vaillancourt JP, Chapman KT and Nicholson DW. (1997). 7320–7326. J. Biol. Chem., 272, 17907–17911. Slee EA, Harte MT, Kluck RM, Wolf BB, Casiano CA, Tong X, Liu B, Dong Y and Sun Z. (2000). Int. J. Radiat. Newmeyer DD, Wang HG, Reed JC, Nicholson DW, Biol., 76, 1387–1395. Alnemri ES, Green DR and Martin SJ. (1999). J. Cell. Troy CM, Rabacchi SA, Friedman WJ, Frappier TF, Brown Biol., 144, 281–292. K and Shelanski ML. (2000). J. Neurosci., 20, 1386–1392. Smith DJ, McGuire MJ, Tocci MJ and Thiele DL. (1997). Troy CM, Rabacchi SA, Hohl JB, Angelastro JM, Greene LA J. Immunol., 158, 163–170. and Shelanski ML. (2001). J. Neurosci., 21, 5007–5016. Smulson ME, Simbulan-Rosenthal CM, Boulares AH, Ya- U M, Miyashita T, Ohtsuka Y, Okamura-Oho Y, Shikama Y kovlev A, Stoica B, Iyer S, Luo R, Haddad B, Wang ZQ, and Yamada M. (2001). Cell Death Differ., 8, 377–386. Pang T, Jung M, Dritschilo A and Rosenthal DS. (2000). Ura S, Masuyama N, Graves JD and Gotoh Y. (2001). Proc. Adv. Enzyme Regul., 40, 183–215. Natl. Acad. Sci. USA, 98, 10148–10153. Soengas MS, Capodieci P, Polsky D, Mora J, Esteller M, Utz PJ and Anderson P. (2000). Cell Death Differ., 7, 589–602. Opitz-Araya X, McCombie R, Herman JG, Gerald WL, Van de Craen M, Declercq W, Van den brande I, Fiers W and Lazebnik YA, Cordon-Cardo C and Lowe SW. (2001). Vandenabeele P. (1999). Cell Death Differ., 6, 1117–1124. Nature, 409, 207–211. Van de Craen M, Van Loo G, Pype S, Van Criekinge W, Van Song Q, Lees-Miller SP, Kumar S, Zhang Z, Chan DW, Smith den brande I, Molemans F, Fiers W, Declercq W and GC, Jackson SP, Alnemri ES, Litwack G, Khanna KK and Vandenabeele P. (1998). Cell Death Differ., 5, 838–846. Lavin MF. (1996). EMBO J., 15, 3238–3246. Vancompernolle K, Van Herreweghe F, Pynaert G, Van de Sordet O, Rebe C, Plenchette S, Zermati Y, Hermine O, Craen M, De Vos K, Totty N, Sterling A, Fiers W, Vainchenker W, Garrido C, Solary E and Dubrez-Daloz L. Vandenabeele P and Grooten J. (1998). FEBS Lett., 438, (2002). Blood, 8, 8. 150–158.

Oncogene Decade of caspases A Degterev et al 8567 Varfolomeev EE, Schuchmann M, Luria V, Chiannilkulchai Widmann C, Gibson S and Johnson GL. (1998b). J. Biol. N, Beckmann JS, Mett IL, Rebrikov D, Brodianski VM, Chem., 273, 7141–7147. Kemper OC, Kollet O, Lapidot T, Soffer D, Sobe T, Wilson KP, Black JA, Thomson JA, Kim EE, Grifﬁth JP, Avraham KB, Goncharov T, Holtmann H, Lonai P and Navia MA, Murcko MA, Chambers SP, Aldape RA, Wallach D. (1998). Immunity, 9, 267–276. Raybuck SA and Livingston DJ. (1994). Nature, 370, 270– Vogt C. (1842). Sdothum: Jent and Gassman, 130. 275. Walker NP, Talanian RV, Brady KD, Dang LC, Bump NJ, Woo M, Hakem R, Soengas MS, Duncan GS, Shahinian A, Ferenz CR, Franklin S, Ghayur T, Hackett MC, Hammill Kagi D, Hakem A, McCurrach M, Khoo W, Kaufman SA, LD, Herzog L, Hugunin M, Houy W, Mankovich JA, Senaldi G, Howard T, Lowe SW and Mak TW. (1998). McGuiness L, Orlewicz E, Paskind M, Pratt CA, Reis P, Genes Dev., 12, 806–819. Summani A, Terranova M, Welch JP, Xiong L, Mo¨ller A, Wyllie AH, Kerr JF and Currie AR. (1980). Int. Rev. Cytol., Tracey, DE, Kamen R and Wong WW. (1994). Cell, 78, 68, 251–306. 343–352. Xu G, Cirilli M, Huang Y, Rich RL, Myszka DG and Wu H. Wang J, Chun HJ, Wong W, Spencer DM and Lenardo MJ. (2001). Nature, 410, 494–497. (2001). Proc. Natl. Acad. Sci. USA, 98, 13884–13888. Xue D and Horvitz HR. (1995). Nature, 377, 248–251. Wang J, Zheng L, Lobito A, Chan FK, Dale J, Sneller M, Yao Yamin TT, Ayala JM and Miller DK. (1996). J. Biol. Chem., X, Puck JM, Straus SE and Lenardo MJ. (1999). Cell, 98, 271, 13273–13282. 47–58. Yang J, Liu X, Bhalla K, Kim CN, Ibrado AM, Cai J, Peng Wang L, Miura M, Bergeron L, Zhu H and Yuan J. (1994). TI, Jones DP and Wang X. (1997). Science, 275, 1129–1132. Cell, 78, 739–750. Yang X. (2002). Cancer Biol. Ther., 1, 407–408. Wang S, Guo M, Ouyang H, Li X, Cordon-Cardo C, Yeh WC, Itie A, Elia AJ, Ng M, Shu HB, Wakeham A, Kurimasa A, Chen DJ, Fuks Z, Ling CC and Li GC. Mirtsos C, Suzuki N, Bonnard M, Goeddel DV and Mak (2000). Proc. Natl. Acad. Sci. USA, 97, 1584–1588. TW. (2000). Immunity, 12, 633–642. Wang S, Miura M, Jung Y, Zhu H, Gagliardini V, Shi L, Yeh WC, Pompa JL, McCurrach ME, Shu HB, Elia AJ, Greenberg AH and Yuan J. (1996). J. Biol. Chem., 271, Shahinian A, Ng M, Wakeham A, Khoo W, Mitchell K, El- 20580–20587. Deiry WS, Lowe SW, Goeddel DV and Mak TW. (1998). Wang S, Miura M, Jung YK, Zhu H, Li E and Yuan J. (1998). Science, 279, 1954–1958. Cell, 92, 501–509. Yin XM, Wang K, Gross A, Zhao Y, Zinkel S, Klocke B, Watabe M, Kakeya H and Osada H. (1999). Oncogene, 18, Roth KA, and Koismeyer SJ. (1999). Nature, 400, 5211–5220. 886–891. Waterhouse N, Kumar S, Song Q, Strike P, Sparrow L, Yoshida H, Kong YY, Yoshida R, Elia AJ, Hakem A, Dreyfuss G, Alnemri ES, Litwack G, Lavin M and Watters Hakem R, Penninger JM and Mak TW. (1998). Cell, 94, D. (1996). J. Biol. Chem., 271, 29335–29341. 739–750. Waterhouse NJ, Goldstein JC, von Ahsen O, Schuler M, Yuan J and Horvitz HR. (1992). Development, 116, 309–320. Newmeyer DD and Green DR. (2001). J. Cell. Biol., 153, Yuan J, Shaham S, Ledoux S, Ellis HM and Horvitz HR. 319–328. (1993). Cell, 75, 641–652. Watt W, Koeplinger KA, Mildner AM, Heinrikson RL, Zermati Y, Garrido C, Amsellem S, Fishelson S, Bouscary D, Tomasselli AG and Watenpaugh KD. (1999). Struct. Fold Valensi F, Varet B, Solary E and Hermine O. (2001). J. Exp. Des., 7, 1135–1143. Med., 193, 247–254. Wei Y, Fox T, Chambers SP, Sintchak J, Coll JT, Golec JM, Zhang J, Cado D, Chen A, Kabra NH and Winoto A. (1998). Swenson L, Wilson KP and Charifson PS. (2000). Chem. Nature, 392, 296–300. Biol., 7, 423–432. Zheng TS, Hunot S, Kuida K, Momoi T, Srinivasan A, Wellington CL, Ellerby LM, Gutekunst CA, Rogers D, Warby Nicholson DW, Lazebnik Y and Flavell RA. (2000). Nat. S, Graham RK, Loubser O, van Raamsdonk J, Singaraja R, Med., 6, 1241–1247. Yang YZ, Gafni J, Bredesen D, Hersch SM, Leavitt BR, Zheng TS, Schlosser SF, Dao T, Hingorani R, Crispe IN, Roy S, Nicholson DW and Hayden MR. (2002). Boyer JL and Flavell RA. (1998). Proc. Natl. Acad. Sci. J. Neurosci., 22, 7862–7872. USA, 95, 13618–13623. Wen LP, Fahrni JA, Troie S, Guan JL, Orth K and Rosen Zhou BB, Li H, Yuan J and Kirschner MW. (1998). Proc. GD. (1997). J. Biol. Chem., 272, 26056–26061. Natl. Acad. Sci. USA, 95, 6785–6790. Widmann C, Gerwins P, Johnson NL, Jarpe MB and Johnson Zou H, Henzel WJ, Liu X, Lutschg A and Wang X. (1997). GL. (1998a). Mol. Cell. Biol., 18, 2416–2429. Cell, 90, 405–413.

Oncogene