凋亡新途径

Nature 419, 634 - 637 (2002); Apoptosis initiated by Bcl-2-regulated caspase activation independently of the c ytochrome c/Apaf-1/caspase-9 apoptosome VANESSA S. MARSDEN*, LIAM O'CONNOR*†, LORRAINE A. O'REILLY*, JOHN SILKE*, DONALD METCALF*, PAUL G. EKERT*‡, DAVID C. S. HUANG*, FRANCESCO CECCONI§, KEISUKE KUIDA¶, KEVIN J. TOMASELLI#, SOPHIE ROY, DON W. NICHOLSON, DAVID L . VAUX*, PHILIPPE BOUILLET*, JERRY M. ADAMS*†† & ANDREAS STRASSER*&# 8224;† * The Walter and Eliza Hall Institute, Melbourne, 3050, Australia ‡ Murdoch Children's Research Institute, Melbourne, 3050, Australia § Department of Biology, Universita Tor Vergata, Rome, Italy ¶ Genomic Pharmacology, Vertex Pharmaceuticals, Cambridge, Massachusetts 02 139, USA # Idun Pharmaceuticals, San Diego, California 92121, USA Merck-Frosst, Pointe-Claire-Dorval H9H 3L1, Canada †† These authors contributed equally to this work † Present address: Incyte Genomics, Palo Alto, California 94304, USA Correspondence and requests for materials should be addressed to A.S. (e-mail: s trasser@wehi.edu.au). Apoptosis is an evolutionarily conserved cell suicide process executed by cystei ne proteases (caspases) and regulated by the opposing factions of the Bcl-2 prot ein family1, 2. Mammalian caspase-9 and its activator Apaf-1 were thought to be essential, because mice lacking either of them display neuronal hyperplasia and their lymphocytes and fibroblasts seem resistant to certain apoptotic stimuli3-6 . Because Apaf-1 requires cytochrome c to activate caspase-9, and Bcl-2 prevents mitochondrial cytochrome c release, Bcl-2 is widely believed to inhibit apoptos is by safeguarding mitochondrial membrane integrity7-9. Our results suggest a di fferent, broader role, because Bcl-2 overexpression increased lymphocyte numbers in mice and inhibited many apoptotic stimuli, but the absence of Apaf-1 or casp ase-9 did not. Caspase activity was still discernible in cells lacking Apaf-1 or caspase-9, and a potent caspase antagonist both inhibited apoptosis and retarde d cytochrome c release. We conclude that Bcl-2 regulates a caspase activation pr ogramme independently of the cytochrome c/Apaf-1/caspase-9 'apoptosome', which s eems to amplify rather than initiate the caspase cascade. How Bcl-2 prevents apoptosis remains hotly debated2, 7-10. Its Caenorhabditis el egans homologue CED-9 binds to the adapter CED-4 to prevent it from activating t he caspase CED-3 (ref. 1). Because the mammalian adaptor Apaf-1 does not bind to Bcl-2 (ref. 11) and requires cytochrome c to activate caspase-9, it is widely b elieved that the Bcl-2 family regulates mitochondrial membrane pores that releas e cytochrome c and other apoptogenic factors7-9. However, because mammalian Bcl- 2 can inhibit cell death in C. elegans1, 12, where no evidence for mitochondrial release of apoptogenic factors has emerged1, we and others have hypothesized th at Bcl-2 can regulate activation of initiator caspases independently of the cyto chrome c/Apaf-1/caspase-9 'apoptosome'2, 10, 11. To investigate whether the apoptosome is required for all apoptosis regulated by the Bcl-2 protein family, we have compared the impact of loss of Apaf-1 or casp ase-9 with that caused by Bcl-2 over-expression or loss of its antagonist Bim on programmed cell death in the haematopoietic system, where Apaf-1 and caspase-9 are widely expressed (Supplementary Fig. 1). In cytokine-supported cultures, fet al liver cells from wild-type, Apaf-1-/-, caspase-9-/- and vav–bcl-2 transgenic mice produced colonies of similar number, size and composition (Fig. 1a and data not shown). Without cytokine support, Bcl-2 overexpression protected the colony-forming cells up to 100-fold, but progenitors lacking caspase-9 or Apaf-1 died as rapidly as those from wild-type mice (Fig. 1a). Figure 1 Cytokine withdrawal-induced and programmed death of haematopoietic cel ls lacking Apaf-1 or caspase-9. Full legend High resolution image and legend (55k) Because most Apaf-1-deficient or caspase-9-deficient mice die before birth3-6, w e assessed the role of the apoptosome in haematopoietic homeostasis by reconstit uting this compartment of C57BL/6-Ly5.1 mice with fetal liver stem cells from th e mutant embryos, or from wild-type, Bim-deficient or vav–bcl-2 embryos. Lymphocytes of Apaf-1-/- or caspase-9-/- origin were no more numerous than those of wild-type origin in the thymus (Fig. 1b), spleen (Fig. 1c), lymph nodes or bone marrow (data not shown). In contrast, mice reconstituted with Bim-deficient or vav?Cbcl-2 stem cells had 3- to 5-fold more splenic B and T cells (Fig. 1c), as do u nmanipulated bim-/- and vav–bcl-2 mice13, 14. Thus, unlike Bim and Bcl-2, caspase-9 and Apaf-1 are not critical determinants of programmed cell death in haematopoiesis. To compare the role of these molecules in stress-induced apoptosis, we isolated donor-derived CD4+8+ thymocytes, pro-B cells and mature T and B cells from the r econstituted mice and tested their sensitivity to cytokine withdrawal, -radiatio n, etoposide and dexamethasone. None of these death responses required Apaf-1 or caspase-9. In thymocytes and activated B and T lymphoblasts, absence of either cell death protein afforded at most modest (<2-fold) protection, whereas in matu re T cells and pro-B cells, lack of caspase-9 or Apaf-1 had no impact (Fig. 2a, b, Supplementary Fig. 3 and data not shown). In contrast, Bcl-2 markedly protect ed against all these apoptotic stimuli, and, as reported13, Bim was required for death provoked by cytokine withdrawal but was largely dispensable for that indu ced by dexamethasone or etoposide. These findings were not restricted to haemato poietic cell types, because Apaf-1-deficient and caspase-9-deficient embryonic f ibroblasts were as sensitive as wild-type cells to loss of attachment (anoikis) (Fig. 2c). Figure 2 Caspase-9 and Apaf-1 are largely dispensable for death induced by grow th-factor withdrawal- and stress-induced death of lymphocytes and fibroblasts. Full legend High resolution image and legend (70k) The absence of Apaf-1 or caspase-9 has previously been reported to delay the dea th of lymphocytes and fibroblasts induced by DNA damage or cytotoxic drugs3-5, b ut those lymphocytes came from the few mutant mice that survived to birth, and i ll health may have selected for stress-resistance. Whereas our fibroblast experi ments used a physiological death stimulus, the strong cytotoxic agents used prev iously4, 5 may have triggered apoptosome activation by directly damaging mitocho ndria. Also, some discrepancies may reflect the mixed genetic background used pr eviously, because this affects the Apaf-1-/- phenotype (ref. 15 and F.C., unpubl ished results). Although Apaf-1-deficient embryonic stem cells and fibroblasts treated with cyto toxic agents were reported to die by a non-apoptotic mechanism16, 17, we found t hat the mutant cells underwent apoptosis. Mutant T lymphoblasts deprived of cyto kine exhibited condensed chromatin and plasma membrane blebbing (Fig. 3a), and d ying thymocytes and fibroblasts exhibited hallmarks of apoptosis, including cell -surface exposure of phosphatidylserine (Supplementary Fig. 4c, d), loss of mito chondrial membrane potential (Supplementary Fig. 4a) and inter-nucleosomal DNA c leavage, as revealed by DNA laddering (Fig. 3b) and the TUNEL reaction (TdT-medi ated dUTP nick end labelling; see Supplementary Fig. 4b). Figure 3 Apoptotic hallmarks in dying caspase-9-deficient and Apaf-1-deficient lymphocytes. Full legend High resolution image and legend (70k) Are caspases activated in the absence of the apoptosome? In the -irradiated muta nt cells, the two classical caspase substrates PARP (poly(ADP-ribose) polymerase ) and ICAD (inhibitor of the caspase-activated DNAse, also known as DFF45; refs 18, 19) yielded fragments of the sizes seen in wild-type cells, albeit less effi ciently (Fig. 4a). The ICAD cleavage is consistent with the inter-nucleosomal DN A degradation (Fig. 3c), because ICAD proteolysis is necessary to free CAD, the relevant DNAse18, 19. PARP and ICAD are substrates of the closely related effect or caspases-3 and -7, which have a preferred DEVD specificity18, 19, and lysates of the dying cells exhibited DEVDase activity with a fluorogenic substrate, alb eit about tenfold less than in the wild-type cells (Fig. 4b). This activity was due to caspases, because the inhibitor DEVD-CHO blocked it (Fig. 4b). To determi ne which effector caspases were activated, immunoblots were examined for process ing of the zymogens (inactive precursors). In -irradiated mutant thymocytes, no processing of pro-caspase-3 or -6 was seen (Supplementary Fig. 5). The relevant effector caspase is probably caspase-7, because its precursor was processed disc ernibly, albeit about tenfold less than in wild-type cells (Fig. 4a), and becaus e extracts from dying caspase-9-/- cells cleaved wild-type ICAD but not ICAD wit h a mutation (D224E) in the known caspase-7 or -3 recognition site (DAVD224)18, 19 (Fig. 4c). The ICAD cleavage is linked to apoptosis, because none was detecta ble in irradiated cells protected by Bcl-2 overexpression (Fig. 5c). Figure 4 Caspase activity and cleavage of caspase substrates in absence of Apaf -1 or caspase-9. Full legend High resolution image and legend (85k) Figure 5 Death of Apaf-1-deficient and caspase-9-deficient cells and ICAD prote olysis is caspase-dependent. Full legend High resolution image and legend (74k) To determine whether caspase activity caused the death of the mutant cells, we t ested two recently described peptido-mimetic inhibitors that block multiple casp ases with high specificity (refs 20, 21 and Supplementary Table 1). IDN-1965 str ongly inhibited dexamethasone-induced apoptosis in wild-type as well as mutant T cells (Fig. 5a). It blocked apoptosis and ICAD cleavage nearly as well as Bcl-2 overexpression and did not further enhance survival of the bcl-2 transgenic cel ls (Fig. 5). The structurally distinct inhibitor IDN-6275 (refs 20, 21) was as e ffective (data not shown), and the widely used peptide-based caspase inhibitor z VAD-fmk also inhibited apoptosis, although less well, presumably owing to its li mited membrane permeability and short half-life. Notably, in -irradiated wild-ty pe and Apaf-1-deficient thymocytes, IDN-1965 also retarded cytochrome c release from mitochondria (Fig. 6a). Thus, caspase activity may be required for mitochon drial membrane disruption. Figure 6 Role of caspases and mitochondrial disruption in apoptosis. Full leg end High resolution image and legend (50k) To purify the active caspases responsible for apoptosis, cell extracts were incu bated with the irreversible pseudo-substrate biotin-DEVD-aomk (at a concentratio n high enough to react with most caspases) and biotinylated polypeptides were re covered on a streptavidin column22. Blots of polypeptides from immortalized Apaf -1-/- myeloid cells treated with etoposide yielded at least three polypeptides ( p18, p26 and p35) absent from untreated cells (Fig. 6b). In accord with Fig. 4, immunoblotting identified the p18 fragment as the large subunit of caspase-7, wh ereas the p26 fragment represented processed caspase-1 (Fig. 6b). Although the p 35 fragment did not react with antibodies to caspases 1, 2, 3, 6, 7, 8 or 9, it has the size expected for a partially processed initiator caspase such as caspas e-11 or -12. Our results establish that the cell-death pathway controlled by Bcl-2 does not r equire caspase-9 or its activator Apaf-1. In keeping with our evidence (Fig. 1b, c) that neither is required for haematopoietic homeostasis, in which the Bcl-2 family has major roles10, deletion of thymocytes with self-reactivity depends on Bim23 but not on Apaf-1 (ref. 24). Because apoptosis was at most only slightly delayed by the absence of Apaf-1 or caspase-9 (Figs 2, 3 and Supplementary Figs 3, 4), we conclude that the apoptosome is not an essential trigger for apoptosis but is rather a machine for amplifying the caspase cascade (Fig. 6c). Presumabl y this amplification is needed much more in some cell types (for example, neuron al precursors) than others (for example, lymphocytes), perhaps depending on thei r complement of caspases, level of caspase inhibitors (for example, IAPs) or con tent of vital substrates. Although mitochondrial release of apoptogenic molecule s contributes to cell destruction7-9, apoptosis in the absence of the apoptosome seems to depend on caspase activity (Fig. 5a) and caspase-7 was identified as t he probable effector (Figs 4, 6b). The low level of caspase-7 activity observed (Figs 4a, b) would be expected in the absence of amplification by the apoptosome 3-6 but appears to be sufficient to demolish many cell types. Although all caspa se activation has been thought to require mitochondrial disruption7-9, efficient methods of caspase inhibition (Fig. 5 and ref. 25) reveal that caspase activati on may instead be required for mitochondrial membrane disruption. Cell death in Drosophila also seems not to require cytochrome c release26, 27. Which upstream caspase(s) might be responsible? In certain transformed cells cas pase-2 has been implicated25, but it cannot be the sole initiator, because the c aspase-2 knockout phenotype is unremarkable28 and caspase-2-/- lymphocytes and n eurons retain normal sensitivity to apoptotic stimuli29. In dying Apaf-1-deficie nt cells, we identified active caspase-1 and a polypeptide that may be processed caspase-11 or -12 (Fig. 6b). Because mice lacking caspases 1, 2, 11 or 12 displ ay no gross defects in apoptosis10, we propose that developmental and stress-ind uced apoptosis can be triggered by several initiator caspases acting in a redund ant manner and that, like CED-9 in C. elegans, Bcl-2 controls their activators ( Fig. 6c). These initiator caspases could either directly activate caspase-7, byp assing the mitochondrion, or mediate disruption of its outer membrane, to amplif y the caspase cascade through the apoptosome. Supplementary information accompanies this paper. Received 4 July 2002;accepted 3 September 2002 References 1. Hengartner, M. O. The biochemistry of apoptosis. Nature 407, 770-7 76 (2000) | Article | PubMed | 2. Cory, S. & Adams, J. M. The Bcl-2 family: Regulators of the cellular life-or- death switch. Nature Rev. Cancer 2, 647-656 (2002) | Article | PubMed | 3. Kuida, K. et al. Reduced apoptosis and cytochrome c-mediated caspase activati on in mice lacking caspase 9. Cell 94, 325-337 (1998) | PubMed | 4. Hakem, R. et al. Differential requirement for caspase 9 in apoptotic pathways in vivo. Cell 94, 339-352 (1998) | PubMed | 5. Yoshida, H. et al. Apaf1 is required for mitochondrial pathways of apoptosis and brain development. Cell 94, 739-750 (1998) | PubMed | 6. Cecconi, F., Alvarez-Bolado, G., Meyer, B. I., Roth, K. A. & Gruss, P. Apaf-1 (CED-4 homologue) regulates programmed cell death in mammalian development. Cel l 94, 727-737 (1998) | PubMed | 7. Green, D. R. & Reed, J. C. Mitochondria and apoptosis. Science 281, 1309-1311 (1998) | Article | PubMed | 8. Gross, A., McDonnell, J. M. & Korsmeyer, S. J. Bcl-2 family members and the m itochondria in apoptosis. Genes Dev. 13, 1899-1911 (1999) | PubMed | 9. Wang, X. The expanding role of mitochondria in apoptosis. Genes Dev. 15, 2922 -2933 (2001) | PubMed | 10. Strasser, A., O'Connor, L. & Dixit, V. M. Apoptosis signaling. Annu. Rev. Bi ochem. 69, 217-245 (2000) | Article | PubMed | 11. Hausmann, G. et al. Pro-apoptotic apoptosis protease-activating Factor 1 (Ap af-1) has a cytoplasmic localization distinct from Bcl-2 or Bcl-xL. J. Cell Biol . 149, 623-634 (2000) | Article | PubMed | 12. Vaux, D. L., Weissman, I. L. & Kim, S. K. Prevention of programmed cell deat h in Caenorhabditis elegans by human bcl-2. Science 258, 1955-1957 (1992) | PubM ed | 13. Bouillet, P. et al. Proapoptotic Bcl-2 relative Bim required for certain apo ptotic responses, leukocyte homeostasis, and to preclude autoimmunity. Science 2 86, 1735-1738 (1999) | Article | PubMed | 14. Ogilvy, S. et al. Constitutive bcl-2 expression throughout the hematopoietic compartment affects multiple lineages and enhances progenitor cell survival. Pr oc. Natl Acad. Sci. USA 96, 14943-14948 (1999) | Article | PubMed | 15. Honarpour, N. et al. Adult Apaf-1-deficient mice exhibit male infertility. D ev. Biol. 218, 248-258 (2000) | Article | PubMed | 16. Haraguchi, M. et al. Apoptotic protease activating factor 1 (Apaf-1)-indepen dent cell death suppression by Bcl-2. J. Exp. Med. 191, 1709-1720 (2000) | Artic le | PubMed | 17. Susin, S. A. et al. Two distinct pathways leading to nuclear apoptosis. J. E xp. Med. 192, 571-580 (2000) | Article | PubMed | 18. Liu, X., Zou, H., Slaughter, C. & Wang, X. DFF, a heterodimeric protein that functions downstream of caspase-3 to trigger DNA fragmentation during apoptosis . Cell 89, 175-184 (1997) | PubMed | 19. Enari, M. et al. A caspase-activated DNase that degrades DNA during apoptosi s, and its inhibitor ICAD. Nature 391, 43-50 (1998) | Article | PubMed | 20. Wu, J. C. & Fritz, L. C. Irreversible caspase inhibitors: tools for studying apoptosis. Methods Enzymol. 17, 320-328 (1999) | Article | 21. Hoglen, N. C. et al. Characterization of the caspase inhibitor IDN-1965 in a model of apoptosis-associated liver injury. J. Pharmacol. Exp. Therapeut. 297, 811-818 (2001) 22. Nicholson, D. W. et al. Identification and inhibition of the ICE/CED-3 prote ase necessary for mammalian apoptosis. Nature 376, 37-43 (1995) | PubMed | 23. Bouillet, P. et al. BH3-only Bcl-2 family member Bim is required for apoptos is of autoreactive thymocytes. Nature 415, 922-926 (2002) | Article | PubMed | 24. Hara, H. et al. The apoptotic protease-activating factor 1-mediated pathway of apoptosis is dispensable for negative selection of thymocytes. J. Immunol. 16 8, 2288-2295 (2002) | PubMed | 25. Lassus, P., Opitz-Araya, X. & Lazebnik, Y. Requirement for caspase-2 in stre ss-induced apoptosis before mitochondrial permeabilization. Science 297, 1352-13 54 (2002) | Article | PubMed | 26. Dorstyn, L. et al. The role of cytochrome c in caspase activation in Drosoph ila melanogaster cells. J. Cell Biol. 156, 1089-1098 (2002) | Article | PubMed | 27. Zimmermann, K. C., Ricci, J. E., Droin, N. M. & Green, D. R. The role of ARK in stress-induced apoptosis in Drosophila cells. J. Cell Biol. 156, 1077-1087 ( 2002) | Article | PubMed | 28. Bergeron, L. et al. Defects in regulation of apoptosis in caspase-2-deficien t mice. Genes Dev. 12, 1304-1314 (1998) | PubMed | 29. O'Reilly, L. A. et al. Caspase-2 is not required for thymocyte or neuronal a poptosis even though cleavage of caspase-2 is dependent on both Apaf-1 and caspa se-9. Cell Death Differ. 9, 832-841 (2002) | Article | PubMed | 30. Adams, J. M. & Cory, S. Apoptosomes: engines for caspase activation. Curr. O pin. Cell Biol. (in the press) Acknowledgements. We thank Y. Lazebnik and P. Lassus for sharing unpublished res ults. We also thank P. Gruss for Apaf-1+/- mice, Y. Lazebnik and X. Opitz-Araya for monoclonal antibodies to caspases 3, 7 and 9, P. Vandenabeele and M. Kalai f or the anti-mouse caspase-1 antibody, R. Anderson for the anti-HSP70 antibody an d S. Nagata for the ICAD constructs. We thank E. Loza, A. Milligan, C. Tilbrook, A. Naughton and J. Merryful for animal care, F. Battye, D. Kaminaris, V. Lapati s, J. Chan and C. Tarlinton for cell sorting, S. Mifsud, L. DiRago, L.-C. Zhang and L. Tai for expert technical help and G. Filby for editorial assistance. We a re grateful to S. Cory, A. Harris, K. Newton and H. Puthalakath for discussions and critical reading of this manuscript. This work was supported by fellowships and grants from the Dr Josef Steiner Cancer Research Foundation, the NHMRC, the Leukemia and Lymphoma Society (SCOR Center), the Anti-Cancer Council of Victoria , the Sylvia and Charles Viertel Charitable Foundation, the NIH, the AIRC and th e Commonwealth Department of Education, Science and Training. F.C. is an Assista nt Telethon Scientist. Competing interests statement. Th 等几年看王晓东实验室对这篇的反应