The Apoptosome Activates Caspase-9 by Dimerization

Molecular Cell 22, 269–275, April 21, 2006 ª2006 Elsevier Inc. DOI 10.1016/j.molcel.2006.03.009 The Apoptosome Activates Caspase-9 by Dimerization Cristina Pop,1 John Timmer,1 Sabina Sperandio,1 and Guy S. Salvesen1,* 1 Program in Apoptosis and Cell Death Research The Burnham Institute for Medical Research 10901 North Torrey Pines Road La Jolla, California 92037 Summary The apical protease of the human intrinsic apoptotic pathway, caspase-9, is activated in a polymeric activation platform known as the apoptosome. The mechanism has been debated, and two contrasting hypotheses have been suggested. One of these postulates an allosteric activation of monomeric caspase-9; the other postulates a dimer-driven assembly at the surface of the apoptosome—the ‘‘induced proximity’’ model. We show that both Hofmeister salts and a reconstituted mini-apoptosome activate caspase-9 by a second-order process, compatible with a conserved dimer-driven process. Significantly, replacement of the recruitment domain of the apical caspase of the extrinsic apoptotic pathway, caspase-8, by that of caspase-9 allows activation of this hybrid caspase by the apoptosome. Consequently, apical caspases can be activated simply by directing their zymogens to the apoptosome, ruling out the requirement for allosteric activation and supporting an induced proximity dimerization model for apical caspase activation in vivo. Short Article (Boatright et al., 2003; Renatus et al., 2001) and, unlike the executioner caspases, do not require proteolysis in the linker region to become active (Boatright et al., 2003; Donepudi et al., 2003; Renatus et al., 2001; Rodriguez and Lazebnik, 1999; Shiozaki et al., 2003; Stennicke et al., 1999). Instead, caspase-9 is activated by small-scale rearrangements of surface loops that define the substrate cleft and catalytic residues (Renatus et al., 2001), and the same may be true for caspase-8. In vitro both caspases-8 and -9 can be activated by dimerization, and in the case of caspase-9, this is orchestrated when the dimer interface provides surfaces compatible with organization of the active site. On the grounds that dimerization activates caspase-9 in vitro, we suggested a conserved mechanism that utilizes the shared properties of the DISC and apoptosome activation platforms to drive dimerization of apical caspase zymogens in vivo (Boatright et al., 2003). However, there is a school of thought to suggest that caspase-9 is not activated by dimerization in the apoptosome but simply by ordering of a monomer into an active conformation (Rodriguez and Lazebnik, 1999; Shi, 2004; Shiozaki et al., 2002) (see Figure 1A). Here we test this hypothesis by determining the mechanism of induced activation of recombinant caspase-9 and by creating hybrid caspase constructs (Figure 1B) to probe the function of the apoptosome and to distinguish the allosteric (monomeric) from the induced proximity (dimeric) activation mechanism of caspase-9. Results Introduction In multicellular animals, the ordered dismantling of unwanted cells, also known as apoptosis, is dependent on a group of proteases called caspases (reviewed in Fuentes-Prior and Salvesen, 2004). The apoptotic caspases belong to two converging pathways where the apical (initiator) caspases-8 and -9 directly activate the common executioner (effector) caspases-3 and -7 (reviewed in Boatright and Salvesen, 2003; Riedl and Shi, 2004). In common with other proteolytic pathways, the caspases are secured in a latent conformation by structural constraints and therefore reside in cells as zymogens. Substantial evidence suggests that these constraints in the executioner caspase zymogens are relieved through specific limited proteolysis by direct action of the apical caspases. However, since there are no proteases to activate the apical caspases, mechanisms other than proteolysis must exist. The mechanism of activation of these apical caspases, a fundamental principle in cell death regulation, depends critically on cofactor recruitment platforms—the death-inducing signaling complex (DISC) for caspase-8 (reviewed in Debatin and Krammer, 2004) and the apoptosome for caspase-9 (reviewed in Jiang and Wang, 2004). At their cytosolic concentration in human cells (<50 nM), both procaspases-8 and -9 are monomeric *Correspondence: [email protected] Activation of Caspase-9 Is a Second-Order Process Previously we showed that recombinant caspases-8 and -9 are activated by high concentrations of certain salts that tend to organize molecular structures—the so-called ‘‘kosmotrope’’ or ‘‘Hofmeister’’ effect—interpreted in terms of a monomer-dimer transition that is necessary and sufficient to activate these apical caspases (Boatright et al., 2003). We confirm that crosslinking of caspase-9 in the presence of Na-citrate favors dimer formation and that salt driven caspase-9 activation is limited to kosmotropic salts (Figure S1 available with this article online). A chromatographically pure monomeric form of caspase-9 was used for all analyses described in this study. We determined whether the activation of caspase-9 in high concentrations of Na-citrate followed first-order or second-order kinetics because a first-order response (unimolecular) would describe allosteric activation of a monomer, whereas a second-order (bimolecular) activation would describe dimerization. We assayed under steady-state conditions, where activity at equilibrium was measured in 0.7 M Na-citrate as a function of enzyme concentration. The activity/concentration plot of caspase-3—a stable dimer – displays a linear dependence in the low nanomolar range (Figure 2A). In contrast, the activity of caspase-9 demonstrated a second-order component of activation, indicating a bimolecular process (Figure 2, Equation 1). However, the Molecular Cell 270 Figure 1. Proposed Activation Models and Constructs Used in The Study (A) The two contrasting caspase-9 activation models. The apoptosome is depicted by the seven Apaf-1 monomers that arrange in a ring to form this specific activation platform (Acehan et al., 2002). In the allosteric model, a single caspase-9 molecule is activated by rearrangements of the active site induced by interactions with apoptosome. In the induced proximity model the local concentration of caspase-9 overcomes a kinetic barrier to dimerization, and monomer/monomer contacts within the catalytic domain of the caspase-9 dimer cause the activating rearrangements. (B) The constructs used in the study. In caspase-9/8, the recruitment domains (DEDs) of caspase-8 are replaced by the CARD of caspase-9 to generate a hybrid protein capable of recruitment to the apoptosome. Mutation of the interchain Asp residues to Ala prevents processing to the two-chain form. Unless specifically stated, all experiments were performed with two-chain caspases-8, -9, or 9/8. best fit to the activation revealed a mixed order of 1.47 (60.02), indicating that a first-order process also takes place in the Na-citrate-induced activation. This may represent ordering of the relatively mobile active site loops of caspase-9, which accompanies dimer-induced activation. It may also be a consequence of approach to the Kd for dimerization in Na-citrate, which would tend to skew the order from 2 toward 1. We cannot distinguish between these two possibilities, but we can conclude that in 0.7 M Na-citrate, dimerization accounts for a substantial component of the activation of caspase-9. The estimated Kd for caspase-9 dissociation in presence of 0.7 M citrate and 50 mM substrate is 79 (610) nM (see Supplemental Experimental Procedures; Equation 3). To test whether this process occurs in a more natural setting, we examined the activation of caspase-9 by a constitutively active form of recombinant Apaf-1. This form lacks the C-terminal 603 residues that contain the regulatory WD40 repeats and is able to activate caspase-9 in a dATP-dependent manner (Riedl et al., 2005). In agreement with Riedl et al. (Riedl et al., 2005), we find that caspase-9 requires substantially more mini-Apaf-1 than full-length Apaf-1 (Zou et al., 1999) for optimal activation but forms the expected high-order oligomers in the presence of Mg2+ and dATP (Figure S2). Significantly, the enzyme produced by Na-citrate has an 8-fold higher kcat/KM value than the enzyme produced by the mini-apoptosome, and this is primarily due to a reduction in KM (Figure 3). More importantly, the concentration-dependent activity of caspase-9 in the miniapoptosome revealed an order of 1.8 (60.1), indicating mainly a bimolecular process (Figure 2B). To confirm that this second-order relationship was not an artifact of decreasing the caspase concentration, we kept the total caspase concentration constant by including an appropriate amount of a caspase-9 Cys/Ala catalytic mutant (Figure 2B). Cleavage of a Natural Caspase-9 Substrate Is Independent of Activation Inducer We were surprised to find that caspase-9 activated by Na-citrate had higher activity on Ac-LEHD-AFC than when activated by the mini-apoptosome. To determine whether this holds for natural substrates we compared the relative efficiencies of the two differentially activated forms of caspase-9 on the Cys/Ala catalytic mutant of procaspase-3. Figure 3 demonstrates that the apoptosome-activated caspase-9 and Na-citrate-activated caspase-9 have roughly the same activity on procaspase-3, in the range 1.0–2.1 3 103 M21s21. Thus, we conclude that activation of caspase-9 by the mini-apoptosome resembles the activation induced by Na-citrate in that the major components of each are bimolecular. The resulting apoptosome-activated enzymes have comparable activities on procaspase-3, but Na-citrate-activated caspase-9 gives an additional 8-fold enhancement on a synthetic substrate compared Caspase-9 Activation 271 Figure 2. Kinetics of Caspase-9 Activation (A) Concentration dependence of caspase-9 activity in 0.7 M Na-citrate (A). The fit shows a n w 1.5 demonstrating a mixed first and secondorder mechanism for caspase activation in Na-citrate. Caspase-3 activity in assay buffer is the control (>). (B) Concentration dependence of caspase-9 activity in the presence of Apaf-1 (4 mM) (d). As a control for varying caspase-9 concentration, the total concentration of caspase-9 was kept constant by adding a C285A catalytic mutant (B). Activation depends on the second power of the enzyme concentration (n = 1.8). The control plot shows a similar pattern (n = 2.1) meaning that the nonlinear dependence of activity versus concentration is not due to Kd limitations between caspase-9 and Apaf-1 interaction during a variation in caspase concentration. White squares represent activity of caspase-9 in absence of Apaf-1. The error bars represent the standard deviation of five independent experiments. to the apoptosome. We do not know the origin of the enhanced activity on Ac-LEHD-AFC, but it may be related to the partial first-order component of the activation by this kosmotrope (see above) in that Na-citrate may stabilize the mobile active site loops of caspase-9 more efficiently than the apoptosome. Activation of a Hybrid Form of Caspase-8 by the Apoptosome The allosteric model for caspase-9 activation (see Figure 1A) predicts the Apaf-1 to have a surface that specifically interacts with the catalytic domain of caspase-9. In contrast, the induced proximity hypothesis predicts no specific interactions between the catalytic domain and Apaf-1. To test these predictions we generated a hybrid caspase in which the catalytic domain of caspase-8 was fused to the CARD of caspase-9. This hybrid, which for convenience we call caspase-9/8 (Figure 1B), is recruited via the typical CARD/CARD homotypic interaction to the apoptosome, but its catalytic domain should have no surface compatible with Apaf-1 binding. We expressed caspase-9/8 in E. coli and obtained a two-chain version, and by mutating D297 and D314 (caspase-8 isoform a numbering) to Ala we obtained a one-chain version. Both hybrids were robust enzymes with essentially the same catalytic parameters as the parental caspase-8 constructs from which they derive (Table S1). This means that fusion of the caspase-9 CARD to the caspase-8 catalytic domain is of minor consequence for caspase-8 catalytic activity. To determine whether caspase-9/8 could substitute for caspase-9, we utilized cytosolic extracts where addition of cytochrome c (cyt c) and dATP results in formation of the apoptosome and activation of caspase-9 (Li et al., 1997). We then immunodepleted caspase-9 from the extracts and asked which caspase forms could compensate for this depletion. As a readout of caspase activation we monitored the activity of executioner caspases (Stennicke et al., 1999). Full activity of extracts depleted of endogenous caspase-9 can be restored by addition of recombinant caspase-9 produced in E. coli (Stennicke et al., 1999). In Figure 4 we show that twochain caspase-9/8 is able to restore about 70% of the full activity in caspase-9-depleted extracts. Significantly this is dependent on the CARD, since caspase-8 was unable to substitute, and occurred in a cyt c- and dATPdependent manner. Interestingly, one-chain caspase9/8 only restores about 20% maximal activity, indicating the importance of cleavage in the stability of the caspase-8 catalytic domain. To rule out the possibility of artifactual or unknown mechanisms of executioner caspase activation, we employed naturally specific caspase-8 and caspase-9 inhibitors (Stennicke et al., 2002). Figure 4C demonstrates that the BIR3 domain of XIAP, a caspase-9-specific inhibitor (Salvesen and Duckett, 2002), selectively abolished activation of executioner caspases (DEVDase activity) in undepleted extracts. Significantly, BIR3 was unable to prevent activation of executioner caspases when depleted extracts were reconstituted with two-chain caspase-9/8, whereas the somewhat specific caspase-8 inhibitor, CrmA (Zhou et al., 1997), showed an anticipated reduction in activity. Thus the activation of the caspase-8 catalytic domain of the hybrid is achieved in an apoptosome-dependent manner simply by providing a caspase-9 recruitment domain (CARD). Discussion Previously we demonstrated that an artificial dimerization domain could activate recombinant caspase-8 Molecular Cell 272 Figure 3. Cleavage of Procaspase-3 and Ac-LEHD-AFC by Differentially Activated Caspase-9 Procaspase-3(C285A) was incubated at 23ºC with caspase-9 (15–500 nM), preactivated by either Apaf-1 (8 mM) or 1 M sodium Na-citrate. The arrows indicate the protein components, and the black arrow indicates the most readily visible cleavage product of procaspase-3. LS/SS: large/small subunit. The asterisk represents cleavage of Apaf-1 by caspase-9, which does not influence its activity (not shown). The estimated kcat/KM value for gel reactions was derived according to Equation 2. The catalytic parameters for Ac-LEHD-AFC cleavage by caspase-9 determined under the same conditions are shown on the right (kcat in s21; KM in mM; kcat/KM in M21 s21). (Muzio et al., 1998), and that kosmotropic salts such as Na-citrate show a marked Hofmeister effect on the activation of caspases-8 and -9 (Boatright et al., 2003). By analyzing the activation of caspase-9 by its physiologic activator Apaf-1 we were able to show that Na-citrate activates caspase-9 to the same extent as the apoptosome, when measured on its natural substrate procaspase-3. Caspase-9 in association with the apoptosome is considered to most closely represent the physiologic form of the enzyme, sometimes called holo-caspase-9 (Rodriguez and Lazebnik, 1999). Thus we answer with an affirmative the question of whether the mechanism of activation by Na-citrate simulates the activation by the physiologic activator (Boatright and Salvesen, 2003; Shi, 2004). Another previously unresolved question was whether the apoptosome is an allosteric regulator or a dimerization machine (Shi, 2004). Support that the apoptosome activates caspase-9 by dimerization comes from our observation that it follows a second-order process, with only a minor, if any, first-order component. We do not rule out that the apoptosome has an additional ability to further stabilize active caspase-9 by allosteric mechanisms, but our experimental data show no evidence for this. Indeed, Na-citrate seems capable of an additional first-order (allosteric) process, but not the apoptosome. Perhaps the most physiologically relevant system that is amenable to biochemical dissection is to use cytosolic extracts that can be programmed to activate executioner caspases in a cyt c-dependent manner (Li et al., 1997). We now show that addition of the hybrid caspase-9/8 to caspase-9-depleted extracts recapitulates executioner caspase activation, meaning that the hybrid can substitute for caspase-9. This demonstrates that recruitment to the apoptosome via the CARD is sufficient to activate caspase-8, and by inference this is probably all that is required to activate caspase-9. It would be far-fetched to propose that Apaf-1 or the apoptosome had evolved a surface capable of allosteric activation of monomeric caspase-8 since the catalytic domain of this caspase is only 34% identical to caspase-9 (Denault and Salvesen, 2002). Consequently, we propose that the mechanism of activation of caspase-9/8 and caspase-9 at the apoptosome is dimerization, and that the apoptosome provides a platform that recruits caspase-9 molecules, thereby overcoming the kinetic barrier to dimerization. The most obvious difference between the zymogens of executioner caspases-3, -6, and -7 and the apical caspases-8 and -9 is that the former are dimers, and the latter monomers (Boatright et al., 2003). Inspection of the available atomic resolution structures suggests that this is primarily due to the weak hydrophobic character of the dimer interface in caspases-8 and -9, in comparison with the relatively hydrophobic nature of the dimer interface in caspases-3 and -7. Thus, under relatively physiologic buffer conditions the Kd of the caspase-3 dimer is less than 50 nM (Pop et al., 2001), whereas caspase-8 and caspase-9 have a Kd value situated in the low micromolar range (Donepudi et al., 2003; Renatus et al., 2001). We show that for caspase-9, 0.7 M Na-citrate decreases the Kd to low nanomolar range. A dimer of caspase-9 can be forced by replacing some of the residues from caspase-3 that constitute its dimer interface, but activity is drastically reduced compared with apoptosome-activated caspase-9 (Chao et al., 2005). This is likely due to alterations in the ability of the active site loops to adopt a productive substrate binding conformation since alterations in this region have the same deleterious effects in caspase-3 (Pop et al., 2003). Eventually, conversion from the single- to two-chain form provides a stabilizing influence to caspase-9, and such Caspase-9 Activation 273 Figure 4. The Caspase-9/8 Hybrid Activates Executioner Caspases in an Apoptosome-Dependent Manner (A) DEVD-ase activity of lysates activated by the addition of cyt c and dATP, depleted of caspase-9 and reconstituted with recombinant caspases (* = one chain, or ** = two-chain caspases). The assay reads out the activity of executioner caspases activated by endogenous caspase-9 or reconstituted with the indicated caspases. The linear portion of the slope was used to compute the amount of activity of each of the apical caspases (see B). (B) Caspase-9/8 activated in lysates by cyt c and dATP can be inhibited by recombinant CrmA, suggesting that the DEVD-ase activity comes either directly from caspase-9/8 or from cleaved caspase-3. BIR-3 inhibits caspase-9 but does not inhibit caspase-9/8. The error bars represent the standard deviation of three independent experiments. (C) Caspase-9/8 is activated only in the presence of cyt c. Caspase-8 (catalytic domain only) is the negative control to demonstrate that the activation of caspase-9/8 is dependent on the CARD. The data represent the mean and standard deviation (error bars) of five independent experiments. The buffer controls represent: I, caspase-8*; II, caspase-8**; III, caspase-9/8*; IV, caspase-9/8**; V, caspase-9**—all added to the hypotonic buffer in the absence of cell lysate. cleavage has a 2-fold enhancement on the efficiency of executioner caspase activation (Stennicke et al., 1999). It is now fairly clear that cofactor-driven protease activation is a common theme in triggering the first proteolytic event in a proteolytic pathway or cascade (reviewed in Fuentes-Prior and Salvesen, 2004; Khan and James, 1998). Once a catalytic site has been generated in an apical protease, then downstream activations simply require limited proteolysis to promote the pathway. Our data firmly support the hypothesis that the most wellknown caspase cofactors, the DISC and the apoptosome, operate as recruitment platforms to promote the natural dimerization of their respective apical caspases. It is primarily by this dimerization, and not by proteolysis or allosteric mechanisms, that apical caspase activity is achieved for caspase-9, and we expect the same for caspase-8 given available evidence (Boatright et al., 2003; Donepudi et al., 2003; Yang et al., 2005). The dimer-induced activation of apical caspases may be a general theme since evidence can be found for a monomer-dimer transition in caspase-1 (Talanian et al., 1996) and caspase-2 (Baliga et al., 2004). Moreover, the driving force for apical caspase activation is also apparently conserved since the cofactors of C. elegans CED-3 and the Drosophila apical caspase Dronc—CED-4 and Dark respectively—have recently been shown to occur as multimeric recruitment platforms in the active form (Yan et al., 2005; Yu et al., 2006). Together, these observations lead to the testable prediction that most apical caspases, if not all, should be activated on oligomeric platforms, and the mechanism is dimerization, not allostery. Experimental Procedures Expression Constructs Caspase-9/8 was cloned in pET23b as a C-terminal His-tag protein using the DNA sequence that corresponds to caspase-9 residues 1 to 152 (CARD and the prodomain linker), and the catalytic domain of caspase-8 starting at residue E151. Caspase-9/8 one chain was obtained by mutating D297 and D314 (caspase-8 isoform a numbering) to alanine using the PCR quick-change method. Protein Purification Full-length caspase-9, full-length caspase-9(C285A), DDED-caspase-8 constructs, and caspase-3(C285A) were expressed with a C-terminal His-tag and purified as shown before (Stennicke et al., 1999). Caspase-9/8 construct purification was done as described above, except that following induction, the cells were grown at 25ºC for 6 hr. All proteins were dialyzed against buffer A (50 mM Tris, 150 mM NaCl, pH 7.5, 1 mM DTT) prior to use and the concentration was determined using absorbance at 280 nm (Edelhoch, 1967). Monomeric or dimeric caspases were obtained by separation by gel filtration in buffer A without DTT as described (Boatright et al., 2003), or by ion-exchange using buffer A and a 0–300 mM NaCl gradient on Mono-Q (Pharmacia). A full-length Apaf-1 clone was a kind gift of Gabriel Nunez. The coding sequence was modified to produce Apaf-1(1–591) in pET21b as a C-terminal His-tag and purified as described before (Riedl et al., 2005). The BIR3 domain of XIAP and Molecular Cell 274 CrmA recombinant proteins were purified as described elsewhere (Quan et al., 1995; Sun et al., 2000). Apoptosome Reconstitution in Cytosolic Extracts Hypotonic cytosolic extracts of human HEK293A cells and immunodepletion of caspase-9 from these extracts were performed as described before (Stennicke et al., 1999). Recombinant caspases were added to the caspase-9-depleted extracts at final concentration of 40 nM, roughly equivalent to the endogenous concentration (Stennicke et al., 1999). Lysates were activated with 10 mM cyt c and 0.5 mM dATP, and DEVD-pNA-ase activity was read at 405 nm (1 hr, 37ºC). For caspase inhibition assays, CrmA or XIAP-BIR3 (375 nM) was added to the lysate prior to activation. Apoptosome Reconstitution from Recombinant Proteins The apoptosome was reconstituted as described before (Riedl et al., 2005) in reconstitution buffer containing 50 mM Hepes, pH 7.5, 5 mM DTT, 100 mM dATP, 2.5 mM MgCl2. Depending on the protein batch, the maximum efficiency was reached for caspase-9/Apaf-1 ratio of 0.2 mM/ 4-5.2 mM. For activation of caspase-9 by the apoptosome in presence of caspase-9(C285A), the total concentration of caspase-9 was kept constant at 200 nM, and the ratio wt/catalytic mutant was varied as shown in Figure 2B. The dependence of activity, A, versus monomeric enzyme concentration, E, was fit to the polynomial equation: n A = k½E
(1) where n defines the number of caspase monomers from the active species. Procaspase-3 Cleavage by Caspase-9 Procaspase-3(C285A) (2 mM) was diluted in assay buffer (50 mM Na-phosphate, 150 mM NaCl, 1.5% sucrose, 10 mM DTT, pH 7.4) containing 1.0 M Na-citrate, or in Apaf-1 reconstitution buffer as appropriate. Caspase-9 was preactivated in either 1.0 M Na-citrate or by the apoptosome (25ºC, 15 min), and then added to the procaspase-3(C285A) at final concentrations between 15 and 500 nM. Samples were incubated for 90 min at 25ºC, then precipitated with 10% TCA and run in 8%–18% SDS-PAGE. Measurement of Kinetic Parameters A crude measure of kcat/KM for the cleavage of procaspase-3 by caspase-9 was obtained by determining the enzyme concentration at which 50% conversion is obtained within the timeframe used, according to the equation: k = In 2=Et (2) where E is the active caspase concentration required to deplete half of the substrate in the incubation period t. One can make the assumption that the concentration of protein substrate is below its KM, and in this case, with the proviso that proteolysis follows the normal Michaelis relationship, the value of k is equal to kcat/KM (Stennicke and Salvesen, 1999). Activity Assays Catalytic parameters of caspases in citrate were determined as described (Boatright et al., 2003) with the following exceptions. Caspases (12–50 nM) were preincubated in assay buffer containing 0.05% CHAPS and 1.0 M Na-citrate for 15 min at 25ºC (caspase-9) or for 4 hr at 37ºC (caspase-8, 9/8) to allow maximal activation. For activation by Apaf-1, caspase-9 (160 nM) was incubated with 5.2 mM Apaf-1 in the reconstitution buffer for 5 min at room temperature prior to Ac-LEHD-AFC addition. All caspases were titrated with Z-VAD-FMK in the appropriate buffer (Stennicke et al., 1999). For activation in citrate, caspase-9 was diluted at 3–100 nM in 0.7 M Na-citrate in assay buffer, incubated for 10 min at 25ºC, followed by addition of 50 mM substrate Ac-LEHD-AFC. Caspase-3 activity was measured at 1–15 nM in either citrate or assay buffer, against Ac-DEVD-AFC. Supplemental Data Supplemental Data include Experimental Procedures, two figures, and one table and can be found with this article online at http:// www.molecell.org/cgi/content/full/22/2/269/DC1/. Acknowledgments We thank Fiona Scott and Jean-Bernard Denault for supplying constructs and for helpful suggestions, Doug Green for enlightening discussions, Scott Snipas and Annamarie Price for expert technical assistance, and Gabriel Nunez for supplying an Apaf-1 plasmid. This work was supported by NIH grants CA69381 and NS37878. Received: December 31, 2005 Revised: February 2, 2006 Accepted: March 3, 2006 Published: April 20, 2006 References Acehan, D., Jiang, X., Morgan, D.G., Heuser, J.E., Wang, X., and Akey, C.W. (2002). Three-dimensional structure of the apoptosome: Implications for assembly, procaspase-9 binding and activation. Mol. Cell 9, 423–432. Baliga, B.C., Read, S.H., and Kumar, S. (2004). The biochemical mechanism of caspase-2 activation. Cell Death Differ. 11, 1234– 1241. Boatright, K.M., and Salvesen, G.S. (2003). Mechanisms of caspase activation. Curr. Opin. Cell Biol. 15, 725–731. Boatright, K.M., Renatus, M., Scott, F.L., Sperandio, S., Shin, H., Pedersen, I., Ricci, J.-E., Edris, W.A., Sutherlin, D.P., Green, D.R., and Salvesen, G.S. (2003). A unified model for apical caspase activation. Mol. Cell 11, 529–541. Chao, Y., Shiozaki, E.N., Srinivasula, S.M., Rigotti, D.J., Fairman, R., and Shi, Y. (2005). Engineering a dimeric caspase-9: a re-evaluation of the induced proximity model for caspase activation. PLoS Biol. 3, e183. 10.1371/journal.pbio.0030381. Debatin, K.M., and Krammer, P.H. (2004). Death receptors in chemotherapy and cancer. Oncogene 23, 2950–2966. Denault, J.B., and Salvesen, G.S. (2002). Caspases: keys in the ignition of cell death. Chem. Rev. 102, 4489–4500. Donepudi, M., Mac Sweeney, A., Briand, C., and Gruetter, M.G. (2003). Insights into the regulatory mechanism for caspase-8 activation. Mol. Cell 11, 543–549. Edelhoch, H. (1967). Spectroscopic determination of tryptophan and tyrosine in proteins. Biochemistry 6, 1948–1954. Fuentes-Prior, P., and Salvesen, G.S. (2004). The protein structures that shape caspase activity, specificity, activation and inhibition. Biochem. J. 384, 201–232. Jiang, X., and Wang, X. (2004). Cytochrome C-mediated apoptosis. Annu. Rev. Biochem. 73, 87–106. Khan, A.R., and James, M.N. (1998). Molecular mechanisms for the conversion of zymogens to active proteolytic enzymes. Protein Sci. 7, 815–836. Li, P., Nijhawan, D., Budihardjo, I., Srinivasula, S.M., Ahmad, M., Alnemri, E.S., and Wang, X. (1997). Cytochrome c and dATP-dependent formation of Apaf-1/Caspase-9 complex initiates an apoptotic protease cascade. Cell 91, 479–489. Muzio, M., Stockwell, B.R., Stennicke, H.R., Salvesen, G.S., and Dixit, V.M. (1998). An induced proximity model for caspase-8 activation. J. Biol. Chem. 273, 2926–2930. Pop, C., Chen, Y.R., Smith, B., Bose, K., Bobay, B., Tripathy, A., Franzen, S., and Clark, A.C. (2001). Removal of the pro-domain does not affect the conformation of the procaspase-3 dimer. Biochemistry 40, 14224–14235. Pop, C., Feeney, B., Tripathy, A., and Clark, A.C. (2003). Mutations in the procaspase-3 dimer interface affect the activity of the zymogen. Biochemistry 42, 12311–12320. Caspase-9 Activation 275 Quan, L.T., Caputo, A., Bleackley, R.C., Pickup, D.J., and Salvesen, G.S. (1995). Granzyme B is inhibited by the cowpox virus serpin CrmA. J. Biol. Chem. 270, 10377–10379. Renatus, M., Stennicke, H.R., Scott, F.L., Liddington, R.C., and Salvesen, G.S. (2001). Dimer formation drives the activation of the cell death protease caspase 9. Proc. Natl. Acad. Sci. USA 98, 14250– 14255. Riedl, S.J., and Shi, Y. (2004). Molecular mechanisms of caspase regulation during apoptosis. Nat. Rev. Mol. Cell Biol. 5, 897–907. Riedl, S.J., Li, W., Chao, Y., Schwarzenbacher, R., and Shi, Y. (2005). Structure of the apoptotic protease-activating factor 1 bound to ADP. Nature 434, 926–933. Rodriguez, J., and Lazebnik, Y. (1999). Caspase-9 and APAF-1 form an active holoenzyme. Genes Dev. 13, 3179–3184. Salvesen, G.S., and Duckett, C.S. (2002). IAP proteins: blocking the road to death’s door. Nat. Rev. Mol. Cell Biol. 3, 401–410. Shi, Y. (2004). Caspase activation: revisiting the induced proximity model. Cell 117, 855–858. Shiozaki, E.N., Chai, J., and Shi, Y. (2002). Oligomerization and activation of caspase-9, induced by Apaf-1 CARD. Proc. Natl. Acad. Sci. USA 99, 4197–4202. Shiozaki, E.N., Chai, J., Rigotti, D.J., Riedl, S.J., Li, P., Srinivasula, S.M., Alnemri, E.S., Fairman, R., and Shi, Y. (2003). Mechanism of XIAP-mediated inhibition of caspase-9. Mol. Cell 11, 519–527. Stennicke, H.R., and Salvesen, G.S. (1999). Caspases: preparation and characterization. Methods 17, 313–319. Stennicke, H.R., Deveraux, Q.L., Humke, E.W., Reed, J.C., Dixit, V.M., and Salvesen, G.S. (1999). Caspase-9 can be activated without proteolytic processing. J. Biol. Chem. 274, 8359–8362. Stennicke, H.R., Ryan, C.A., and Salvesen, G.S. (2002). Reprieval from execution: The molecular basis of caspase inhibition. Trends Biochem. Sci. 27, 94–101. Sun, C., Cai, M., Meadows, R.P., Xu, N., Gunasekera, A.H., Herrmann, J., Wu, J.C., and Fesik, S.W. (2000). NMR structure and mutagenesis of the third BIR domain of the inhibitor of apoptosis protein XIAP. J. Biol. Chem. 275, 33777–33781. Talanian, R.V., Dang, L.C., Ferenz, C.R., Hackett, M.C., Mankovich, J.A., Welch, J.P., Wong, W.W., and Brady, K.D. (1996). Stability and oligomeric equilibria of refolded interleukin-1beta converting enzyme. J. Biol. Chem. 271, 21853–21858. Yan, N., Jijie Chai, J., Lee, E.S., Gu, L., Liu, Q., He, J., Wu, J.-W., Kokel, D., Li, H., Hao, Q., et al. (2005). Structure of the CED-4/ CED-9 complex reveals insights into programmed cell death in Caenorhabditis elegans. Nature 437, 831–837. Yang, J.K., Wang, L., Zheng, L., Wan, F., Ahmed, M., Lenardo, M.J., and Wu, H. (2005). Crystal structure of MC159 reveals molecular mechanism of DISC assembly and FLIP inhibition. Mol. Cell 20, 939–949. Yu, X., Wang, L., Acehan, D., Wang, X., and Akey, C.W. (2006). Threedimensional structure of a double apoptosome formed by the Drosophila Apaf-1 related killer. J. Mol. Biol. 355, 577–589. Zhou, Q., Snipas, S., Orth, K., Dixit, V.M., and Salvesen, G.S. (1997). Target protease specificity of the viral serpin CrmA: analysis of five caspases. J. Biol. Chem. 273, 7797–7800. Zou, H., Li, Y., Liu, X., and Wang, X. (1999). An APAF-1.cytochrome c multimeric complex is a functional apoptosome that activates procaspase-9. J. Biol. Chem. 274, 11549–11556.