Immunity, and Apoptotic Signal
Transduction: Lessons from Knockout Mice
and -12. The CED-3 subfamily includes Casp3, -6, -7, -8, -9, and -10, whereas the third subfamily consists of only one member, Casp2. Within each subfamily, the peptide sequence preferences in the substrates are sim-ilar (Thornberry et al., 1997; Garcia-Calvo et al., 1998). Marek Los, Sebastian Wesselborg,
and Klaus Schulze-Osthoff*† Department of Internal Medicine I Medical Clinics
University of Tu¨bingen
Because at least in some cases different caspases can D72076 Tu¨bingen
cleave the same substrates, this illustrates some degree Germany
of functional redundancy within the caspase family. Caspases exist as latent zymogens that contain an N-terminal prodomain followed by the region that forms Introduction
the two subunits with the catalytic domain. The core of the catalytic center is formed by the conserved penta-Apoptosis is the innate mechanism by which the
organ-peptide sequence QACXG. The proforms of caspases ism eliminates unwanted cells. It is the most common
are activated by proteolytic cleavage at specific aspar-form of cell death and occurs during development,
tis-tate residues. Generally, an initial cleavage event sepa-sue remodeling, cell homeostasis, defense processes,
rates the C-terminal short subunit from the rest of the and immune responses. Dysregulation of apoptosis may
molecule, allowing assembly of an active protease that be directly involved in several human pathologies
in-autocatalytically cleaves off its prodomain. As deduced cluding degenerative and autoimmune diseases,
neo-from the crystal structure of Casp1 and Casp3, the ma-plasia, AIDS, and other viral or bacterial infections
ture enzyme of all caspases is a heterotetrameric com-(Fisher, 1994; Thompson, 1995; Los et al., 1997).
Signifi-plex composed of two large subunits and two small cant progress in our understanding of apoptosis came
subunits (Walker et al., 1994; Wilson et al., 1994; Ro-from genetic studies of developmental cell death in the
tonda et al., 1996; Mittl et al., 1997). Once activated, nematode Caenorhabditis elegans, in which a set of a
some caspases can propagate activation of other family few genes, termed CED for cell death-defective,
regu-members and thus initiate a proteolytic cascade. lates the apoptotic machinery (Ellis et al., 1991). Two of
Based on their structure and order in cell death path-them, ced-3 and ced-4, are required for the execution
ways, caspases can be divided into initiator and effector of cell death. Another gene, ced-9, protects cells from
caspases. Effector caspases generally contain only a undergoing programmed cell death. Loss-of-function
small prodomain and cleave diverse cellular substrates, mutations in ced-9 cause cells that normally live to
un-whereas initiator caspases have a long prodomain and dergo apoptosis (Hengartner et al., 1992). This cell death exert regulatory roles by activating downstream effector requires the activity of ced-3 and ced-4, indicating that
caspases. Activation of initiator caspases is mediated
ced-9 acts by preventing ced-3 and ced-4 from causing by binding of adaptor molecules to protein interaction
cell death. ced-9 is the structural and functional homo- motifs in their prodomains. Two general types of interac-log of Bcl2, one of a large family of genes intimately tion have been identified (Martin et al., 1998; Muzio et involved in prevention of vertebrate apoptosis. CED-4 al., 1998; Srinivasula et al., 1998). Pro-Casp8 and -10 is highly related to an apoptosis regulator in mammals, each contain two tandem death effector domains designated Apaf1 for apoptotic protease-activating fac- (DEDs), while pro-Casp1, -2, -4, and -9 contain a cas-tor-1 (Zou et al., 1997). The mammalian counterpart of pase-recruitment domain (CARD). In each case, the pro-CED-3 has been identified as a member of a family of caspases bind to adaptor molecules containing similar intracellular proteases that form the core of the apo- domains, and they either directly aggregate or interact ptotic machinery (Yuan et al., 1993). Since these are with other molecules (Figure 1). Although not related, cysteine proteases that cleave cellular substrates at CARDs and DEDs have a similar structure composed specific aspartate residues, they are termed caspases of amphipathic, antiparallel a helices, which are also (Alnemri et al., 1996). present in the death domain of death receptors. Upon Mammalian caspases comprise a group of at least ligation, death receptors such as CD95 and TNF-R1 bind fourteen members that can promote apoptosis (Cohen, to and oligomerize adaptor proteins and procaspases. 1997; Nicholson and Thornberry, 1997; Cryns and Yuan, FADD, which contains a DED, is recruited directly to 1998; Stroh and Schulze-Osthoff, 1998). Some cas- CD95 and indirectly to TNF-R1, resulting in the autoacti-pases, in addition, generate mature proinflammatory cy- vation of Casp8. In an analogous manner, the adaptor tokines and thereby regulate immune responses. Based protein RAIDD/CRADD, which contains a CARD, can on phylogenetic analysis and positional scanning stud- associate with TNF-R1 and promote activation of pro-Casp2 (Ahmad et al., 1997; Duan and Dixit, 1997; Li et al., ies of their peptide substrates, caspases are divided into
1997a). Besides death receptor–mediated apoptosis, a three subfamilies: The ICE-like protease family includes
related but different way of caspase activation exists Casp1, -4, -5, -13, and -14 as well as murine Casp11
that is triggered for instance by cytotoxic drugs or p53 and essentially controlled by mitochondria. In early * To whom correspondence should be addressed (e-mail: kso@
phases of apoptosis, mitochondria release cytochrome uni-muenster.de).
c which, together with dATP, activates the adaptor pro-†Present address: Division of Cell Biology, University of Mu¨nster,
Figure 1. Two Principal Signaling Pathways of Apoptosis
One pathway (left) involves ligation of death receptors, resulting in the recruitment of the adaptor protein FADD through interaction be-tween the death domains (DD) of both mole-cules. The death effector domain (DED) of FADD in turn recruits pro-Casp8, which is cleaved and activated at the receptor com-plex. Another pathway (right), which is trig-gered by many apoptotic stimuli, is initiated at the mitochondrion. An early, not well-understood step is the mitochondrial release of cytochrome c into the cytosol which, to-gether with dATP, binds to the CED-4 homolog Apaf1. This event unmasks the CARD motif in Apaf1 and allows binding of procaspase-9 through CARD/CARD interaction. The mito-chondrial but not the death receptor pathway is inhibited by Bcl2. Antiapoptotic members of the Bcl2 family may interfere with the relo-calization of cytochrome c or with the binding of cytochrome c to Apaf1. Following activa-tion of the initiator caspase Casp8 or Casp9, the two pathways converge on the activation of effector Casp3, -6, and -7, which finally cleave various death substrates. Because Casp8 cleaves Bid and generates a trun-cated, proapoptotic BH3-containing frag-ment (tBid) that induces cytochrome c re-lease, both pathways cross-communicate. Casp8, in turn, can be also activated by Casp6 following Casp9 cleavage, thereby amplifying the apoptotic signal.
of pro-Casp9 via CARD–CARD interaction, while a differ- (Ghayur et al., 1997; Gu et al., 1997; Fantuzzi et al., 1998). Since the demonstration that ced-3 of C. elegans ent region of Apaf1 self-associates resulting in Casp9
activation (Figure 1). encodes a protein similar to mammalian Casp1 (Yuan
et al., 1993), and that overexpression of both genes Although caspases are the essential components of
many, if not all apoptotic pathways, their precise physio- promotes apoptosis (Miura et al., 1993; Los et al., 1995), Casp1 has been a subject of intensive research. Point logical role remains controversial. The large number of
members of the caspase family, their overlapping tissue mutations in a region homologous between CED-3 and distribution and similar cleavage specificities present a Casp1 eliminated the ability of both proteins to induce challenge for identifying the function of individual cas- cell death. Apoptosis caused by overexpression could pases in vivo. A number of questions must be raised: be also suppressed by Bcl2 and CrmA, a viral inhibitor First, are caspases redundant or do individual caspases of caspases.
play a dominant apoptotic role in a tissue- or cell type– Gene targeting revealed that Casp1 plays an impor-specific manner? Second, is any caspase indispensable tant role in the regulation of the immune response, but for the activation of other caspases? Third, does an is presumably dispensable for most apoptotic pathways individual caspase participate in some forms of cell (Li et al., 1995; Kuida et al., 1995). Casp12/2mice had death, but not in others? Finally, does the inhibition of a major defect in the production of mature IL-1b and a certain caspase exert global effects on apoptosis or impaired IL-1a synthesis (Table 1). Secretion of TNF-a only prevent specific forms of death in certain tissues? and IL-6 in response to LPS stimulation was also dimin-Very recent studies using gene targeting and transgene ished. In addition, macrophages from Casp12/2 mice
technologies have provided some answers to these were defective in LPS-induced IFN-g production (Fan-questions and have shed new light on the distinctive tuzzi et al., 1998). Casp12/2mice were highly resistant role of individual caspases in cell death as well as in to the lethal effects of endotoxin. At a dose of LPS that other unexpected biological processes. killed wild-type mice within 30 hr, all Casp12/2 mice
survived (Li et al., 1997b). The proinflammatory role of Casp1 was strengthened by the finding that Casp1-defi-Caspase-1 and Casp1-defi-Caspase-11 Knockout Mice
cient mice revealed decreased necrosis, edema forma-Casp1 was originally identified as interleukin 1
b–con-tion, and serum levels of amylase and lipase during verting enzyme (ICE), the protease that cleaves the
pre-experimentally induced pancreatitis (Norman et al., cursor of IL-1b into the active cytokine (Cerretti et al.,
1997). Pharmacological blockade or genetic deletion of 1992). More recently, Casp1 has been shown to further
Table 1. Phenotypes of Caspase-, FADD- and Apaf1-Deficient Mice
Development and Cytokine Expression Apoptosis
Caspase 1 No developmental defects; Sensitive to most apoptotic inducers; Resistant to LPS-induced septic shock; Reduced ischemic brain injury;
Increased survival in experimental pancreatitis; CD95-induced apoptosis attenuated in thymocytes; No IL-1b and IL-18 processing, impaired production Neurons resistant to trophic factor withdrawal
of IL-1a, IL-6, TNF-a and IFN-g
Caspase 2 Viable, no gross abnormalities; Oocytes resistant to drug-induced death; Excess numbers of female germ cells Defective B cell death in response to granzyme B;
Lymphocytes sensitive to drugs and anti-CD95; Increased susceptibility of sympathetic neurons to
trophic factor withdrawal
Caspase 3 Mice born at lower frequency and with smaller size; No membrane blebbing and nuclear fragmentation Death at 1–3 weeks of age; in hepatocytes and thymocytes;
Disturbed brain development with excessive Cleavage of caspase substrates delayed or absent; numbers of postmitotic cells Reduced apoptosis in diverse settings including
activation-induced T cell death and drug-induced apoptosis of fibroblasts
Caspase 8 Lethal in utero; embryos of smaller size Fibroblasts resistant to TNF-R1, CD95, and DR3, but Impaired heart muscle development; sensitive to drug-induced apoptosis;
Congested accumulation of erythrocytes and Normal JNK and NF-kB activation massive hemorrhage
Decreased number of hematopoietic stem cells
Caspase 9 Perinatal lethality; Embryonic stem cells and fibroblasts resistant to Enlarged and malformed cerebrum due to reduced several apoptotic stimuli;
apoptosis during brain development; Thymocytes resistant to dexamethasone- and Lack of Casp3 activation in embryonic brains g-irradiation-induced apoptosis, but sensitive
to UV irradiation and CD95;
Splenocytes not protected against drug-induced apoptosis
Caspase 11 No developmental defects, similar to Casp1 Cells resistant to apoptosis induced by
deficiency overexpression of Casp1
Resistant to endotoxic shock;
Lack of IL-1a and IL-1b production due to blocked Casp1 activation
FADD Cardiac failure and massive hemorrhage; Fibroblasts resistant to death receptor, but sensitive Phenotype similar to Casp82/2mice; to drug-, E1A-, and c-Myc-induced apoptosis; IL-2 expression of thymocytes intact Thymocytes of RAG1 chimeras show impaired
survival and proliferation
Apaf1 Lethal at E16.5; Embryonic fibroblasts exhibit reduced response to Brain overgrowth, exencephaly; various apoptotic stimuli;
Severe craniofacial and ossification defects; Thymocytes sensitive to CD95-, but resistant to Strong alterations of the lens and retina; drug- and irradiation-induced apoptosis Persistence of interdigital webs
proinflammatory cytokines and was associated with An involvement of Casp1 in some forms of apoptosis was shown in experimental models of neuronal cell dramatic survival benefits.
In contrast to inflammatory responses, Casp1 null death. Mutant Casp1, in which the active site cysteine was changed into a glycine, inhibited trophic factor with-mice did not show any gross abnormalities in
develop-ment or profound defects in apoptosis. For instance, drawal-induced apoptosis in dorsal root ganglial cells (Friedlander et al., 1997a). Transgenic mice expressing ATP-induced apoptosis of peritoneal macrophages was
not affected (Li et al., 1995). Thymocytes from young this mutant Casp1 under control of a neuron-specific promoter appeared normal. However, brain injury in-Casp12/2mice equally underwent apoptosis, when
trig-gered by dexamethasone or g-irradiation. However, duced by middle cerebral artery occlusion, a mouse model of stroke, was significantly reduced in Casp1 these thymocytes revealed a partial resistance toward
CD95-induced apoptosis (Kuida et al., 1995). Studies mutant and deficient mice (Friedlander et al., 1997a; Schielke et al., 1998). In addition, expression of mutant from knockout (KO) mice indicated that Casp1 mediates
partially apoptosis induced by IFN-g. In this scenario, a Casp1 in neurons of superoxide dismutase-deficient mice, a model of amyotrophic lateral sclerosis, was able positive amplification loop between IFN-g and Casp1
seems to exist. IFN-g is able to induce Casp1 expression to delay disease progression (Friedlander et al., 1997b). Recently, it has been proposed that Casp1 is activated through activation of the STAT (signal transducer and
activator of transcription) signaling pathway (Chin et by a direct physical interaction with murine Casp11 (Ich-3), which is most homologous to human Casp4 and al., 1997). IFN-g did not induce Casp1 expression and
apoptosis in STAT1-deficient cells. Therefore, activation Casp5 (Wang et al., 1998). Casp11 transcription and translation is strongly induced upon stimulation with of the STAT pathway may induce apoptosis in some
which could be inhibited by CrmA and Bcl2 (Wang et for granzyme B-, but not CD95-mediated apoptosis. Thymocytes and T lymphocytes were also equally sensi-al., 1996). Casp11-deficient cells were resistant to
apo-ptosis induced by Casp1 overexpression, suggesting tive to apoptosis caused by doxorubicin, etoposide, staurosporine, org-irradiation. Interestingly, no differ-that activation of Casp1 requires Casp11. Analogously
to Casp12/2mice, Casp11-deficient animals developed ences in TNF-induced apoptosis were observed in Casp2-deficient embryonic fibroblasts. Casp2 has been normally and were resistant to endotoxic shock.
Produc-tion of both IL-1a and IL-1b after LPS stimulation, a shown to interact physically with RAIDD, an adaptor protein containing a region homologous to the death critical event during septic shock, was blocked in
Casp112/2mice. Thus, Casp1 and Casp11 seem to play domain and prodomain of Casp2 (Ahmad et al., 1997;
Duan and Dixit, 1997). The death domain of RAIDD inter-an importinter-ant role in inflammation by activating
cyto-kines, but may not mediate apoptosis in development acts with RIP, a death domain containing serine/threo-nine kinase that is part of the TNF death pathway. The and most circumstances.
results with Casp2 mutant mice therefore suggest that Casp2 and RAIDD may be redundant in TNF signaling. Targeted Disruption of Caspase-2
In conclusion, Casp2 may be essential for apoptosis in Casp2 (Ich-1, Nedd2) was initially identified by a
sub-female germ cells, whereas in other situations it may tractive cloning approach as a gene highly expressed
function to delay cell death. The ultimate action of Casp2 in embryonic brain and downregulated in adult brain
appears to be dependent on the tissue type, cell lineage, (Kumar et al., 1992). Overexpression of Casp2 in
mam-developmental stage, differential splicing of its mRNA, malian cells induced apoptosis that could be blocked
and the presence and absence of other caspases. by Bcl2 (Kumar et al., 1994; Wang et al., 1994). Using
antisense technology, decreasing Casp2 levels have
been shown to delay cell death induced by trophic factor Caspase-3-Deficient Mice
Casp3 (CPP32, YAMA, Mch2, apopain) is regarded as deprivation in hematopoietic and neuronal cell lines.
Gene targeting revealed that Casp22/2mice reached a prototype effector caspase able to degrade the vast
majority of “death substrates.” The three-dimensional adulthood and were devoid of severe phenotypic
abnor-malities (Bergeron et al., 1998). However, Casp2 ap- structure of a complex of Casp3 with a tetrapeptide inhibitor has been determined (Rotonda et al., 1996; Mittl peared to be required for female germ cell death. During
development, over one-half of the ovarian germ cells et al., 1997). Its architecture resembles Casp1 in overall structure, but its P4 substrate residue (four amino acids undergo apoptosis. This massive germ cell death occurs
at later stages of fetal life and persists through day 3 N-terminal of the cleavage site) is different. These differ-ences account for the distinct specificity of the two postpartum, at which time remaining oocytes are
en-closed by granulosa cells to form primordial follicles. proteases and may enable the design of selective inhib-itors.
Casp2-deficient mice contained significantly higher
numbers of newly formed primordial follicles than wild- Gene targeting of Casp3 generated the first mice with profound defects in apoptosis (Kuida et al., 1996; Woo type mice. Oocytes of Casp2 KO mice were also
resis-tant to cell death following exposure to chemotherapeu- et al., 1998). The mice, born at a frequency lower than expected by Mendelian genetics, were smaller than their tic drugs.
Unlike apoptosis of germ cells, cell death of facial littermates and died at 1–3 weeks of age. Their pheno-type was remarkably restricted. The most pronounced motor neurons was not suppressed but even
acceler-ated in Casp2-deficient mice (Bergeron et al., 1998). effect was seen in the central nervous system, while no discernible abnormalities were found in embryonic Similarly, Casp2-deficient sympathetic neurons
under-went apoptosis more effectively than wild-type neurons heart, lung, liver, or kidney. Casp3-deficient animals showed massive hyperplasia and ectopic cell masses when deprived of nerve growth factor. Thus, Casp2 not
only acts as a positive effector of apoptosis, but in some of the brain with a variety of disorganized cellular struc-tures. Also protrusions of the neuroepithelium in the cells it may also delay cell death. The pro- versus
anti-apoptotic effect of Casp2 gene targeting may be caused retina, which caused compression of the lens, could be seen. Pyknotic clusters, which are normally detected by the existence of splice variants that are expressed to
variant degrees in different tissues and developmental at sites of major morphogenetic change during normal brain development, were not observed in the mutant stages. Alternative splicing generates two Casp2
mes-sages that encode Casp2L, which induces cell death, embryos. In contrast, the excessive cells were
postmi-totic and terminally differentiated. Hence, Casp3 plays and Casp2S, a truncated protein that antagonizes cell
death (Wang et al., 1994). a critical role during morphogenetic cell death in the mammalian brain.
Unlike Casp1, Casp2 is not involved in ischemic brain
injury caused experimentally by middle artery occlusion, The central role of Casp3 for induction of apoptosis has been shown also in hepatocytes and thymocytes as the extent of ischemic injury was unaffected in mutant
mice. There were also no differences in the rate of motor derived from KO mice (Zheng et al., 1998). Although Casp32/2 hepatocytes or thymocytes were killed at a
neuron death after denervation in vivo. Furthermore,
Casp2 does not obviously contribute to cell death of similar rate as control cells, when cocultured with CD95 ligand expressing 3T3 cells, wild-type cells displayed sympathetic neurons induced by trophic factor
with-drawal, although it has previously been reported that typical apoptotic features such as cytoplasmic blebbing and nuclear fragmentation within 6 hr, but neither event Casp2 antisense constructs inhibited this form of cell
death (Kumar et al., 1994). was observed for Casp32/2cells. The cleavage of vari-ous caspase substrates implicated in apoptotic events, B lymphoblasts from Casp2 KO mice were deficient
including gelsolin, fodrin, lamin B, and DFF/ICAD, an Casp8 (Slee et al., 1999). A Jurkat T cell line deficient in Casp8 was not only completely resistant to death endonuclease inhibitor, was delayed or absent in the
receptor–induced apoptosis, but also partially resistant Casp32/2cells. Thus, the altered cleavage of these key
to cell death induced by UV irradiation, adriamycin, and substrates is likely to be responsible for the aberrant
etoposide (Juo et al., 1998). Complementation of these apoptotic phenotype in Casp3-deficient cells.
cells with Casp8 restored apoptosis sensitivity. In addi-Although Casp3 is highly expressed in cells of
hema-tion, adenoviral E1A induced pro-Casp8 processing and topoietic origin, lack of Casp3 did not affect immature
apoptosis in cells deleted of FADD, indicating the exis-T and B cell development (Kuida et al., 1996).
Gene-tence of alternative activation pathways (Nguyen et al., targeted mice contained fewer thymocytes, related to
1998). A possible candidate in this context could be an their overall smaller size. Surprisingly, thymocytes from
apoptosome-like complex in the endoplasmic reticulum Casp3 null and wild-type mice were equally sensitive to
(ER). It is composed of the integral ER membrane protein induction of apoptosis by anti-CD95, dexamethasone,
p28 Bap31, Casp8, and possibly Apaf1 or a related mole-ceramide, staurosporin, andg-irradiation. Using Casp32/2
cule (Ng and Shore, 1998). Activation of this complex lymphocytes from RAG1 chimeric mice, it was shown
might be induced upon viral infection or in other situa-that peripheral T cells were resistant to
activation-tions associated with abnormal ER function. Hence, induced cell death and apoptosis triggered by anti-CD3
Bap31 could play a similar function in the ER as the and anti-CD95 (Woo et al., 1998). The requirement for
CD95 death-inducing signaling complex at the surface intact Casp3 appeared to dependent on the apoptotic
membrane. stimulus. In embryonic stem cells, Casp3 was necessary
While Casp8 can be activated indirectly in the mito-for efficient apoptosis following UV irradiation, but not
chondrial pathway, it has also been found that,
con-g-irradiation or cytotoxic killing of target cells.
Con-versely, CD95-mediated activation of Casp8 can deliver versely, the same stimulus can show a tissue specificity:
a signal to mitochondria (Scaffidi et al., 1998). This event TNF-a treatment induced normal levels of apoptosis in
is initiated by Casp8-mediated cleavage of the Bcl2 pro-Casp32/2thymocytes, but defective apoptosis in
onco-tein Bid, which generates a proapoptotic fragment that gene-transformed fibroblasts (Woo et al., 1998). Hence,
triggers cytochrome c release (Li et al., 1998; Luo et al., the consequences of Casp3 deficiency appear to be
1998). Thus, both the death receptor and mitochondrial remarkably context-dependent with apoptotic defects
pathway of caspase activation may be interconnected being both cell type– and stimulus-specific.
leading to the amplification of an apoptotic signal. Homozygous disruption of the mouse Casp8 gene Gene Targeting of Death Receptor Pathways
was found to be lethal in utero (Varfolomeev et al., 1998). Casp8 (FLICE, Mach, Mch5) was originally identified as
However, whereas Casp32/2 mice exhibited profound the proximal caspase in the death-inducing signaling
brain defects, Casp82/2embryos presumably died from
complex of CD95 (Boldin et al., 1996; Muzio et al., 1996;
cardiac failure. Two salient features of Casp8 null mice Srinivasula et al., 1996). The prodomain of Casp8 as well
were impaired heart muscle development and abdomi-as that of Cabdomi-asp10 contain duplicates of a death effector
nal hemorrhage. Extensive erythrocytosis was present domain, which enable Casp8 to interact with FADD, an
also in other organs such as liver and lung. The reason adaptor molecule associating with the CD95 receptor
for this is not understood but may be caused by a defect (Figure 1). It is assumed that besides CD95 a similar
in angiogenesis, disturbed hematopoiesis, or heart fail-pathway is activated by other death receptors including
ure. Although the heart of Casp82/2mice was not appre-TNF-R1, DR3, and TRAIL receptors. Unlike most other
ciably larger than normal, the trabeculae and ventricular caspases, Casp8 is believed to autoproteolytically
acti-musculature were thin and resembled early mesen-vate itself upon recruitment and oligomerization at the chyme. The fact that the heart was hypotrophic rather vicinity of death receptors (Martin et al., 1998; Muzio et than enlarged strongly suggests that Casp8 may be al., 1998). Very recently, a novel molecule called FLASH involved in the transmission of survival rather than death has been identified to associate with Casp8 in the CD95 signals. This is supported by the fact that hematopoietic death-inducing signaling complex through a death ef- precursor cells from KO mice revealed a strongly im-fector domain related motif (Imai et al., 1999). FLASH, paired colony-forming activity. Disruption of Casp8 thus in addition, contains a region with homology to Apaf1, appears to result in a primary or secondary depletion which binds ATP and which is required for self-associa- of the hematopoietic precursor pool (Varfolomeev et al., tion of FLASH molecules. However, although this would 1998).
suggest similarities between FLASH and Apaf1, it is cur- Embryonic fibroblasts deficient in Casp8 were com-rently unknown whether FLASH is essentially required pletely resistant to apoptosis mediated by death recep-for or just facilitates death receptor–mediated apo- tors, such as CD95, TNF-R1 and DR3, whereas they
ptosis. retained sensitivity to a wide range of apoptotic stimuli
While Casp8 plays a central role in propagating death including UV irradiation, ceramide, chemotherapeutic receptor-mediated signals, proteolytic processing of drugs, and infection with the cytopathic vesicular sto-Casp8 has been found also in other apoptosis settings, matitis virus. Yet in contrast, embryonic fibroblasts re-such as chemotherapeutic drug-induced apoptosis sponded normally to nonapoptotic signals emanating (Ferrari et al., 1998; Wesselborg et al., 1999). In this from death receptors and activated Jun N-terminal ki-pathway, Casp8 activation is presumably independent nases as well as transcription factor NF-kB equally as of death receptors and mediated by prior cytochrome wild-type cells. Thus, these findings indicate that Casp8 c release and Apaf1 activation. It has been shown that plays a necessary and nonredundant role in apoptosis
induction by death receptors. Casp9 cleaves Casp6, which in turn can directly activate
Like Casp8 KO mice, gene targeting of the adaptor to a variety of apoptotic stimuli is associated with the protein FADD resulted in a lethal phenotype with pro- early mitochondrial release of cytochrome c, an event found signs of cardiac failure and hemorrhage (Yeh et that is blocked by antiapoptotic members of the Bcl2 al., 1998). Overexpression of CD95, TNF-R1 and DR-3 family. In the cytosol, cytochrome c, together with dATP, could not induce apoptosis in embryonic fibroblasts forms a complex with Apaf1 that results in the cleavage from FADD2/2mice, whereas DR4 and chemotherapeu- of pro-Casp9 and subsequent activation of downstream tic drugs, as well as oncogene expression of c-Myc caspases (Li et al., 1997c). It has been further demon-and E-1A, did. Since FADD2/2mice die in utero, T cell strated that, upon ligation of death receptors, activation maturation has been analyzed in chimeric mice deficient of Casp8 can result in the proapoptotic cleavage of for the recombination activating gene product RAG1, Bid, which induces cytochrome c translocation (Li et al., which activates rearrangement of immunoglobulin and 1998; Luo et al., 1998). Thus, the death receptor and the T cell receptor genes (Zhang et al., 1998). T lymphocytes mitochondrial pathway can cross-communicate in order from FADD2/2 chimera were completely resistant to
to increase the speed and efficiency of the death CD95-induced apoptosis. Thymocyte populations were process.
apparently normal in newborn chimeras. As these mice Recently both Casp9- and Apaf1-deficient mice have age, their thymocytes, however, decreased to undetect- been generated (Cecconi et al., 1998; Hakem et al., 1998; able levels, although peripheral T cells were present.
Kuida et al., 1998; Yoshida et al., 1998). The phenotypes Thus, FADD2/2mice seem to be inefficient in maintaining
of these mice resembled Casp3-deficient animals in that thymic cellularity possibly due to an intrinsic survival
they primarily developed brain malformations and over-defect.
growth with excess cells in the central nervous system. Whereas mice lacking a functional CD95 system
de-The ectopic cell masses consisted of differentiated velop lymphadenopathy and splenomegaly as a result
postmitotic cells that had escaped apoptosis. A nearly of the accumulation of an abnormal T cell population in
10-fold reduction of TUNEL-positive cells was found in the periphery, this population was not detected in FADD
the brain of Casp92/2mice at E12.5. In situ
immunostain-KO mice 5 months old. The absence of
lymphoprolifera-ing showed the absence of Casp3 activation in vivo. tive disease and the T cell population characteristic of
Both Casp9- and Apaf1-deficient mice died at about day CD95 deficiency in FADD2/2 mice may result from a
16.5 of development. Unlike the brain, other nonneural defect in the proliferation of T cells. Indeed,
activation-organs such as heart, lung, and liver or spinal cord were induced proliferation was impaired in FADD2/2T cells,
unaffected and appeared remarkably normal. The simi-despite normal production of IL-2. Thus, this suggests a
lar phenotypes of Casp3-, Casp9-, and Apaf1-deficient rather unexpected connection between cell proliferation
mice therefore indicate that the molecules act in line in and apoptosis (Zhang et al., 1998).
a common apoptotic pathway. A growth-promoting activity of FADD is supported by
It is interesting to note that the phenotypes of the the phenotype of transgenic mice expressing a
domi-three KO mice are similar, but do not accurately mimic nant-negative FADD mutant (FADD-DN) under control
each other. Although disruption of either Casp9 or of a T cell–specific promoter (Newton et al., 1998).
Ex-pression of FADD-DN enhanced negative selection of Casp3 caused brain malformations, the abnormalities self-reactive thymic lymphocytes. FADD-DN mice dis- were more severe in mice lacking Casp9. Casp3-defi-played increased apoptosis and reduced proliferation cient cells rarely exhibited an overt exencephaly abnor-and clonogenic growth of mitogen-activated T cells. In- mality. One possibility to account for this difference in terestingly, this impaired T cell proliferation was not severity of the malformation is that Casp9 may also observed in CD95-deficient mice or animals overex- activate other effector caspases such as Casp6 or pressing the viral caspase inhibitor CrmA (Smith et al., Casp7.
1996). Thus, signaling through FADD does not lead ex- While Casp92/2 mice exhibited a more pronounced
clusively to cell death, but under certain circumstances phenotype than mice with a Casp3 null mutation, the can promote cell survival and proliferation. Since T lym- most severe morphogenetic distortions were found in phocytes from CrmA transgenic mice responded nor- Apaf1 KO mice (Cecconi et al., 1998; Yoshida et al., mally to mitogens, it may be speculated that, unlike the
1998). Apaf12/2mice showed strong craniofacial
abnor-pathway leading to apoptosis, the growth-promoting
malities such as ossification defects, deficient midline signal induced by FADD in lymphocytes does not require
fusion of the palatal shelves, as well as dramatic alter-activation of Casp8. Furthermore, because
CD95-defi-ations of the lens and retina. Also cauliflower-like cient mice do not show any cardiovascular or
hemato-masses on the face present in Apaf12/2mice were not
poietic abnormalities, another receptor may exist that
observed in either Casp3 or Casp9 gene–targeted mice. employs the FADD/Casp8 pathway in order to induce
Another obvious difference was the formation of inter-proliferative or morphogenetic signals during the
devel-digital space. Whereas removal of interdevel-digital webs was opment. The molecular nature of the putative growth
normal, although slightly delayed, in Casp92/2mice,
in-signals generated by FADD/Casp8 action, however, is
terdigital webs persisted in Apaf1 KO mice. Although it entirely unknown.
has not formerly been excluded that this cell death oc-curs independently of caspases, the intact retraction of Knocking Out Mitochondria-Controlled
interdigital webs in Casp92/2mice indicates that Apaf1
Caspase Activation
may either interact with another caspase or play addi-Besides death receptors that couple to FADD and
tional roles beyond that of caspase activation. It has Casp8, another pathway is essentially controlled by
Figure 2. Apoptotic Pathways with Different Requirements of Caspase-9 and Caspase-3
Apaf1 in overexpression systems. A more detailed anal- required for apoptosis in brain development. In contrast, there also exists a Casp9-dependent pathway that ysis, however, showed that Casp9 is the only
Apaf1-interacting protease in living cells (Slee et al., 1999). works independently of Casp3: Whereas Casp92/2
thy-mocytes are resistant to several apoptotic stimuli includ-Thus, the idea that another caspase may substitute
Casp9 is appealing, but currently none of the other ing dexamethasone andg-irradiation, Casp32/2 thymo-cytes remain sensitive to these inducers, implying that known caspases appears to be activated by Apaf1.
Despite a normal development of the thymus, thymo- in thymocytes these stimuli trigger the activation of an apoptotic pathway that is Casp3 independent but Casp9 cytes from Casp9- and Apaf1-deficient mice were
resis-tant to a number of apoptotic stimuli. Nevertheless, both dependent. Presumaby, in this pathway other effector caspases such as Casp6 or Casp7 may be dominant. mice contained a comparable number of
double-posi-tive thymocytes, indicating that Apaf1 and Casp9 are Third, an apoptotic pathway independent of Casp9 and Casp3: Although Casp92/2 embryonic stem cells are not involved in negative selection. In thymocytes and
embryonic stem cells from KO mice, cytochrome c was resistant to a broad range of apoptotic stimuli, Casp9 deficiency does not protect other cell types from apo-translocated into the cytosol, confirming that Apaf1 and
Casp9 act downstream of cytochrome c. Thymocytes ptosis induced by the same inducers. Unlike embryonic stem cells, however, Casp92/2thymocytes and
spleno-from Apaf12/2mice were also resistant to a loss of
mito-chondrial transmembrane potential (Yoshida et al., cytes undergo apoptosis in response to UV irradiation. Thus, these finding suggest a functional diversification 1998). It has been suggested that opening of the
mito-chondrial permeability pore and loss of transmembrane of caspase cascades in vivo that may be highly depen-dent on the cell type and apoptotic stimulus. Certainly, potential may be an initial event for cytochrome c release
and subsequent caspase activation (Kroemer, 1997). Apaf1 and Casp9 play a central role in events of mito-chondria-dependent pathways of apoptosis that are The results observed in KO mice, however, strongly
sug-gest that the decrease of the transmembrane potential critical for several histogenetic and morphogenetic cell deaths during development. The difference in the degree is downstream of Apaf1-mediated caspase activation.
This assumption is supported by the fact that in several of resistance between different cell types, however, il-lustrates a distinct requirement of Apaf1 and Casp9 in apoptotic systems cytochrome c release preceded
mi-tochondrial membrane depolarization by many hours. the apoptotic machinery. Therefore, it is highly conceiv-able that other hitherto nonidentified pathways exist A comparison of the so far established KO mice allows
the separation of distinct in vivo apoptotic pathways which may be relevant for certain apoptotic stimuli in particular tissues.
that are strikingly cell type–specific and differentially utilized in response to certain stimuli (Figure 2). First,
the classical death receptor pathway that is Casp3- What Do We Learn from Knockout Mice?
Previously, most of the information regarding the differ-dependent but indiffer-dependent of Casp9: The existence of
this pathway is exemplified in activated T cells. Casp3 ent roles of caspases and their signaling cascades has been obtained from overexpression of proteins and non-deficiency protects activated T cells from
apoptosis-induced anti-CD95, TRAIL, and anti-CD3, whereas Casp9 functional mutants in transfected cells or from yeast two-hybrid assays. Because supraphysiological overex-deficiency has no obvious effect. Second, the classical
mitochondrial pathway that is dependent on Apaf1 and pression of proteins may not reflect the in vivo situation, our knowledge about the functional significance of Casp9: For example, embryonic stem cells from both
Apaf12/2 and Casp92/2 mice are resistant to several apoptotic cascades is still very limited. An apparent
example of a discrepancy between data obtained in vitro apoptotic signals including UV andg-irradiation or
che-motherapeutic drugs. The mitochondrial pathway may with cell lines and with gene-targeted mice is apoptotic pathways that have been suggested to involve CD95 or be further distinguished based on the requirement of
certain effector caspases, such as Casp3. Mutation of other death receptors. Some reports demonstrated that anticancer drug–induced cell death may involve the Casp9 or Casp3 produces similar effects on brain
Different drugs used in chemotherapy have been shown (Izquierdo et al., 1999). This suggests either that a cas-pase other than Casp1, -2, -3, -8, or -9 is involved in to induce CD95 ligand expression in certain tumor cells.
Because drug-induced cell death was inhibited by CD95 negative selection or that, due to a redundant role of certain caspases, gene targeting of a single caspase neutralizing reagents, ligation of CD95 has been
pro-posed to trigger the apoptosis cascade in chemosensi- may not be sufficient to block this process.
A very remarkable feature of the currently studied tive tumor cells. Experiments from FADD- and
Casp8-deficient mice, however, convincingly demonstrate that KO mice is their rather restricted phenotype. Whereas Casp3, Casp9, and Apaf1 null mice revealed massive a death receptor–mediated pathway may not be the
principal mechanism of drug-induced apoptosis (Yeh et malformations of neural tissues, FADD and Casp8 KO animals died primarily from defective heart develop-al., 1998; Varfolomeev et develop-al., 1998). Although embryonic
fibroblasts from both mutant mice were resistant to apo- ment. Both types of mutant mice, however, had a normal development of other organs including lung, liver, and ptosis mediated by CD95 and other death receptors,
they retained sensitivity to anticancer drugs. Like drug- thymus. Interestingly, several cell types of these mice were still sensitive to a variety of apoptotic stimuli, sug-mediated apoptosis, FADD or Casp8 gene–targeted
cells did not exhibit an altered apoptosis sensitivity fol- gesting that presumably still other nonidentified apo-ptotic cascades exist. Whether these are mediated by lowing irradiation or overexpression of c-Myc. Thus,
despite the fact that cell type–specific variations may caspases or by other apoptotic signal transducers, such as the recently identified apoptosis-inducing factor account for these discrepancies, FADD and Casp8,
al-though strictly required for death receptor–mediated (AIF), a mitochondrial flavoprotein (Susin et al., 1999), remains to be determined. Unlike in mice, targeted dis-pathways, are not a prerequisite in other apoptosis
set-tings. ruption of the caspase DCP-1 in Drosophila resulted in
multiple developmental defects associated with wide-The finding that fibroblasts from Casp8 KO mice are
resistant to death receptor–mediated apoptosis further spread melanotic tumors and larval death (Song et al., 1997).
indicates that other receptor-associated pathways, in-cluding activation of Daxx or RAIDD/Casp2, have little importance (Varfolomeev et al., 1998). This is supported
by the observation that cells from Casp2 null mice were Are There Nonapoptotic Functions of Caspases? There are some evidences that both receptor- and mito-still sensitive to death receptor–mediated apoptosis
(Bergeron et al., 1998). Several previous reports have chondria-controlled caspase cascades may not only control apoptosis but play additional roles beyond that emphasized a role of stress-activated protein kinases
in apoptosis mediated by CD95 (Goillot et al., 1997; of cell death. It is noteworthy that for instance Casp3 and Casp8 mutant mice have a smaller size, although Lenczowski et al., 1997). In contrast, fibroblasts from
Casp82/2mice retained kinase activation, indicating that the opposite phenotype may be expected. Why Casp3
and Casp8 deficiency primarily leads to neuronal hyper-at least in fibroblasts these protein kinases have no
considerable impact for the propagation of the apo- plasia is not clear. In contrast, mice deficient for p27/ Kip1, an inhibitor of cyclin-dependent kinases, reveal ptotic signal (Varfolomeev et al., 1998).
Disruption of the CD95 system results in pronounced hyperplasia of all organs, and not only neuronal tissues (Nakayama et al., 1996).
effects on immune functions (Nagata 1997;
Schulze-Osthoff et al., 1998; Krammer, 1999). lpr (for lymphopro- Thymic lymphocytes from FADD null chimeric mice or from animals expressing a dominant-negative mutant liferation) mice that lack a functional CD95 receptor, as
well as gld (for generalized lymphoproliferative disease) display increased apoptosis and reduced proliferation in response to mitogens (Newton et al., 1998; Woo et mice that bear a mutant CD95 ligand, exhibit various
autoimmune phenomena resembling systemic lupus er- al., 1998). This suggests that the FADD/Casp8 cascade may be linked to an intrinsic survival pathway. A growth-ythematosus in men. Both mouse strains produce
auto-antibodies and accumulate abnormal CD42CD82T cells promoting effect of caspases is also evident from the observation that hematopoietic precursors of Casp8 null leading to lymphadenopathy, splenomegaly, and other
autoimmune symptoms. It is interesting to note that mice show a reduced colony-forming activity (Varfolo-meev et al., 1998). In addition, caspases may not neces-autoimmune disorders have not been detected in KO
embryos of either FADD, Casp8, or other caspases nor sarily trigger cell death as transient activation of Casp3 was observed during T cell stimulation and allogenic in deficient lymphocytes of RAG1 chimeric mice. In
addi-tion, neither T cell hyperplasia nor serum autoantibodies mixed lymphocyte reaction, which is not linked to apo-ptosis (Miossec et al., 1997).
were observed in transgenic animals overexpressing
the viral caspase inhibitor CrmA or dominant-negative The impaired heart muscle development and massive hemorrhage caused by Casp8 deficiency provides fur-FADD in T lymphocytes (Smith et al., 1996; Newton et
al., 1998). Therefore, the lpr phenotype may not be sim- ther evidence for a morphogenetic role of caspases, which may not only be associated with apoptotic events ply due to a disturbed apoptotic pathway, but rather
may involve other signaling events of CD95. All mutant but also involve other biological processes including proliferation or differentiation of cells. Gene targeting of mice studied so far contained normal populations of
thymocytes indicating also that negative selection in- Braf, a member of the Raf kinase family, results in a phenotype with alterations similar to Casp8-deficient volves neither Apaf1/Casp9- nor death
receptor/Casp8-linked pathways. Interestingly, transgenic mice overex- mice (Wojnowski et al., 1997). It will be interesting to investigate whether erythrocytosis observed in both pressing the baculovirus protein p35, a powerful, broad
Boldin, M.P., Goncharov, T.M., Goltseve, Y.V., and Wallach, D. to activate Casp8 may be linked to a disturbed
differenti-(1996). Involvement of MACH, a novel MORT1/FADD-interacting ation of angioblasts, a decreased proliferation of
endo-protease, in Fas/APO-1- and TNF receptor–induced cell death. Cell thelial cells, or dysregulation of extracellular matrix
for-85, 803–815.
mation. In this respect, it has been found in vitro that
Cecconi, F., Alvarez-Bolado, G., Meyer, B.I., Roth, K.A., and Gruss, spreading of cells on extracellular matrix may be regu- P. (1998). Apaf1 (CED-4 homolog) regulates programmed cell death lated by caspase activation (Watanabe and Akaike, in mammalian development. Cell 94, 727–737.
1999). Certainly, future investigations are needed to elu- Cerretti, D.P., Kozlosky, C.J., Mosley, B., Nelson, N., Van Ness, K., cidate whether, apart from execution of apoptosis, cas- Greenstreet, T.A., March, C.J., Kronheim, S.R., Druck, T., Canniz-zaro, L.A., et al. (1992). Molecular cloning of the interleukin-1b con-pases participate in other cellular pathways.
verting enzyme. Science 256, 97–100.
Chin, Y.E., Kitagawa, M., Kuida, K., Flavell, R.A., and Fu, X.Y. (1997). Concluding Remarks Activation of the STAT signaling pathway can cause expression of The rapid discovery of a great number of caspases, caspase 1 and apoptosis. Mol. Cell. Biol. 17, 5328–5337. together with multiple control points of their activation, Cohen, G.M. (1997). Caspases: the executioners of apoptosis. Bio-proceeds well ahead of our knowledge of their physio- chem. J. 326, 1–16.
logical roles within the organism. There are still major Cryns, V., and Yuan, J. (1998). Proteases to die for. Genes Dev. 12, 1551–1570.
gaps, but recent gene targeting of caspases provides
us with several new and fundamental aspects of their Duan, H., and Dixit, V.M. (1997). RAIDD is a new ‘death’ adaptor molecule. Nature 385, 86–89.
physiological functions. The fact that different lines of
Ellis, R.E., Yuan, J.Y., and Horvitz, H.R. (1991). Mechanisms and KO mice exhibit preferential apoptosis defects rather
functions of cell death. Annu. Rev. Cell Biol. 7, 663–698. than a global suppression of apoptosis indicates that
Fantuzzi G., Puren, A.J., Hardig, M.W., Livingston, D.J., and Dina-caspases play a largely nonredundant apoptotic role in
rello, C.A. (1998). Interleukin-18 regulation of interferong production a tissue- and stimulus-dependent manner. Furthermore,
and cell proliferation as shown in interleukin-1b-converting enzyme despite compelling evidence for a key role of Casp8
(caspase-1)-deficient mice. Blood 91, 21118–21125. and Casp9, the restricted phenotype of both
Casp8-Fernandes-Alnemri, T., Armstrong, R.C., Krebs, J., Srinivasula, S.M., and Casp9-deficient mice suggests that other apical Wang, L., Bullrich, F., Fritz, L.C., Trapani, J.A., Tomaselli, K.J., Lit-caspases must exist regulating apoptotic processes. wack, G., and Alnemri, E.S. (1996). In vitro activation of CPP32 Casp10, for instance, which is very similar in its structure and Mch3 by Mch4, a novel human apoptotic cysteine protease containing two FADD-like domains. Proc. Natl. Acad. Sci. USA 93, to Casp8, has been shown to be recruited to death
7464–7469. receptors (Fernandes-Almemri et al., 1996; Vincenz and
Ferrari, D., Stepczynska, A., Los, M., Wesselborg, S., and Schulze-Dixit, 1997). Although fibroblasts from Casp8 null mice
Osthoff, K. (1998). Differential regulation and ATP requirement for are almost completely resistant to death receptor–
caspase-8 and caspase-3 activation during CD95- and anticancer mediated apoptosis, it cannot be excluded that Casp10,
drug-induced apoptosis. J. Exp. Med. 188, 979–984. having little importance in fibroblasts, exerts crucial
Fisher, D.E. (1994). Apoptosis in cancer therapy: crossing the thresh-functions in other cell types. In addition, since certain
old. Cell 78, 539–542. cell types such as embryonic fibroblasts from Apaf1
Friedlander, R.M., Gagliardini, V., Hara, H., Fink, K.B., Li, W., Mac-KO mice are still considerably sensitive to a variety of Donald, G., Fishman, M.C., Greenberg, A.H., Moskowitz, M.A., and apoptosis inducers, it is very likely that other yet undis- Yuan, J. (1997a). Expression of a dominant negative mutant of in-covered key regulators exist. We also have to be aware terleukin-1b converting enzyme in transgenic mice prevents neu-ronal cell death induced by trophic factor withdrawal and ischemic that the restricted phenotype of most KO mice may
brain injury. J. Exp. Med. 185, 933–940. underestimate the role of the targeted caspases,
be-Friedlander, R.M., Brown, R.H., Gagliardini, V., Wang, J., and Yuan, cause single caspases may substitute other family
mem-J. (1997b). Inhibition of ICE slows ALS in mice. Nature 388, 31. bers. A major obstacle of most KO mice is their prenatal
Friesen, C., Herr, I., Krammer, P.H., and Debatin, K.M. (1996). lethality, which precludes manifestations of caspase
Involvement of the CD95 (APO-1/Fas) receptor/ligand system in drug functions in the adult organism. Therefore, in future
re-induced apoptosis in leukemia cells. Nat. Med. 2, 574–577. search, conditional disruption in a cell type–specific
Garcia-Calvo, M., Peterson, E.P., Leiting, B., Ruel, R., Nicholson, manner or in certain developmental stages will be re- D.W., and Thornberry, N.A. (1998). Inhibition of human caspases by quired to elucidate more precisely the in vivo functional peptide-based and macromolecular inhibitors. J. Biol. Chem. 273, significance of individual caspases in development, im- 32608–32613.
mune functions, and pathological forms of apoptosis. Ghayur, T., Banerjee, S., Hugunin, M., Butler, D., Herzog, L., Carter, A., Quintal, L., Sekut, L., Talanian, R., Paskind, M., et al. (1997). Caspase-1 processes IFN-g-inducing factor and regulates LPS-References
induced IFN-g production. Nature 386, 619–623.
Goillot, E., Raingeaud, J., Ranger, J., Tepper, R.I., Davis, R.J., Har-Ahmad, M., Srinivasula, S.M., Wang, L., Talanian, R.V., Litwack, G.,
low, E., and Sanchez, I. (1997). Mitogen-activated protein kinase-Fernandes-Alnemri, T., and Alnemri. E.S. (1997). CRADD, a novel
mediated Fas apoptotic signaling pathway. Proc. Natl. Acad. Sci. human apoptotic adaptor molecule for caspase-2, and FasL/tumor
USA 94, 3302–3307. necrosis factor receptor-interacting protein RIP. Cancer Res. 57,
615–619. Gu, Y., Kuida, K., Tsutsui, H., Ku, G., Hsiao, K., Fleming, M.A., Ha-yashi, N., Higashino, K., Okamura, H., Nakanishi, K., et al. (1997). Alnemri, E.S., Livingston, D.J., Nicholson, D.W., Salvesen, G.,
Activation of interferon-g inducing factor mediated by interleukin-Thornberry, N.A., Wong, W.W., and Yuan, J. (1996). Human ICE/
1b converting enzyme. Science 275, 206–209. CED-3 protease nomenclature. Cell 87, 171.
Hakem, R., Hakem, A., Duncan, G.S., Henderson, J.T., Woo, M., Bergeron, L., Perez, G.I., Macdonald, G., Shi, L., Sun, Y., Jurisicova,
Soengas, M. S., Elia, A., de la Pompa, J.L., Kagi, D., Khoo, W., A., Varmuza, S., Latham, K.E., Flaws, J.A., Salter, J.C., et al. (1998).
et al. (1998). Differential requirement for caspase 9 in apoptotic Defects in regulation of apoptosis in caspase-2-deficient mice.
Hengartner, M.O., Ellis, R.E., and Horvitz, H.R. (1992). Caenorhab- Martin, D.A., Siegel, R.M., Zheng, L., and Lenardo, M.J. (1998). Mem-brane oligomerization and cleavage activates the caspase-8 (FLICE/ ditis elegans gene ced-9 protects cells from programmed cell death.
Nature 356, 494–499. MACHalpha1) death signal. J. Biol. Chem. 273, 4345–4349. Miossec, C., Dutilleul, V., Fassy, F., and Diu Hercend, A. (1997). Imai, Y., Kimura, T., Murakami, A, Yajima, N., Sakamaki, K., and
Yonehara, S. (1999). The CED-4 homologous protein FLASH is in- Evidence for CPP32 activation in the absence of apoptosis during T lymphocyte stimulation. J. Biol. Chem. 272, 13459–13462. volved in Fas-mediated activation of caspase-8 during apoptosis.
Nature 398, 777–785. Mittl, P.R., Di Marco, S., Krebs, J.F., Bai, X., Karanewsky, D,S., Priestle, J.P., Tomaselli, K.J., and Grutter, M.G. (1997). Structure of Izquierdo, M., Grandien, A., Criado, L.M., Robles, S., Leonardo E.,
Albar, J.P., Gonzalez de Buitrago, G., and Martinez-A, C. (1999). recombinant human CPP32 in complex with the tetrapeptide acetyl-Asp-Val-Ala-Asp fluoromethyl ketone. J. Biol. Chem. 272, 6539– Blocked negative selection in developing T cells in mice expressing
the baculovirus p35 caspase inhibitor. EMBO J. 18, 156–166. 6547.
Miura, M., Zhu, H., Rotello, R., Hartwieg, E.A., and Yuan, J. (1993). Juo, P., Kuo, C.J., Yuan, J., and Blenis, J. (1998). Essential
require-ment for caspase-8/FLICE in the initiation of the Fas- induced apo- Induction of apoptosis in fibroblasts by IL-1b-converting enzyme, a mammalian homolog of the C. elegans cell death gene ced-3. Cell ptotic cascade. Curr. Biol. 8, 1001–1008.
75, 653–660.
Krammer, P.H. (1999). CD95 (APO-1/Fas)-mediated apoptosis: live
and let die. Adv. Immunol. 71, 163–210. Mu¨ller, M., Strand, S., Hug, H., Heinemann, E.M., Walczak, H., Hof-mann, W.J., Stremmel, W., Krammer, P.H., and Galle, P.R. (1997). Kroemer, G. (1997). Mitochondrial control of apoptosis. Immunol.
Drug-induced apoptosis in hepatoma cells is mediated by the CD95 Today 18, 44–51.
(APO-1/Fas) receptor/ligand system and involves activation of wild-Kuida, K., Lippke, J.A., Ku, G., Harding, M.W., Livingston, D.J., Su,
type p53. J. Clin. Invest. 99, 403–413. M.S., and Flavell, R.A. (1995). Altered cytokine export and apoptosis
Muzio, M., Chinnaiyan, A.M., Kischkel, F.C., O’Rourke, K., Shev-in mice deficient Shev-in Shev-interleukShev-in-1b converting enzyme. Science 267,
chenko, A., Ni, J., Scaffidi, C., Bretz, J.D., Zhang, M., Gentz, R., et 2000–2003.
al. (1996). FLICE, a novel FADD-homologous ICE/CED-3-like prote-Kuida, K., Zheng, T.S., Na, S., Kuan, C., Yang, D., Karasuyama, H.,
ase, is recruited to the CD95 (Fas/APO-1) death-inducing signaling Rakic, P., and Flavell, R.A. (1996). Decreased apoptosis in the brain
complex. Cell 85, 817–827. and premature lethality in CPP32-deficient mice. Nature 384,
Muzio, M., Stockwell, B.R., Stennicke, H.R., Salvesen, G.S., and 368–372.
Dixit, V.M. (1998). An induced proximity model for caspase-8 activa-Kuida, K., Haydar, T.F., Kuan, C.-Y., Gu, Y., Taya, C., Karasuyama,
tion. J. Biol. Chem. 273, 2926–2930. H., Su, M.S., Rakic, P., and Flavell, R.A. (1998). Reduced apoptosis
Nagata, S. (1997). Apoptosis by death factor. Cell 88, 355–365. and cytochrome c–mediated caspase activation in mice lacking
cas-pase 9. Cell 94, 325–337. Nakayama, K., Ishida, N., Shirane, M., Inomata, A., Inoue, T., Shi-shido, N., Horii, I., Loh, D.Y., and Nakayama, K. (1996). Mice lacking Kumar, S., Tomooka, Y., and Noda, M. (1992). Identification of a set
p27Kip1 display increased body size, multiple organ hyperplasia, of genes with developmentally down-regulated expression in the
retinal dysplasia, and pituitary tumors. Cell 85, 707–720. mouse brain. Biochem. Biophys. Res. Commun. 185, 1155–1161.
Newton, K., Harris, A.W., Bath, M.L., Smith, K.G.C., and Strasser, Kumar, S., Kinoshita, M., Noda, M., Copeland, N.G., and Jenkins,
N.A. (1994). Induction of apoptosis by the mouse Nedd2 gene, which A. (1998). A dominant interfering mutant of FADD/MORT1 enhances encodes a protein similar to the product of the Caenorhabditis ele- deletion of autoreactive thymocytes and inhibits proliferation of ma-gans cell death gene ced-3 and the mammalian IL-1b-converting ture T lymphocytes. EMBO J. 17, 706–718.
enzyme. Genes Dev. 8, 1613–1626. Ng, F.W., and Shore, G.C. (1998). Bcl-XL cooperatively associates Lenczowski, J.M., Dominguez, L., Eder, A.M., King, L.B., Zachar- with the Bap31 complex in the endoplasmic reticulum, dependent chuk, C. M., and Ashwell, J.D. (1997). Lack of a role for Jun kinase on procaspase-8 and Ced-4 adaptor. J. Biol. Chem. 273, 3140–3143. and AP-1 in Fas-induced apoptosis. Mol. Cell. Biol. 17, 170–181. Nguyen, M., Branton, P.E., Roy, S., Nicholson, D.W., Alnemri, E.S., Li, P., Allen, H., Banerjee, S., Franklin, S., Herzog, L., Johnston, C., Yeh, W.C., Mak, T.W., and Shore, G.C. (1998). E1A-induced pro-McDowell, J., Paskind, M., Rodman, L., and Salfeld, J. (1995). Mice cessing of procaspase-8 can occur independently of FADD and is deficient in IL-1b-converting enzyme are defective in production of inhibited by Bcl-2. J. Biol. Chem. 273, 33099–33102.
mature IL-1b and resistant to endotoxic shock. Cell 80, 401–411.
Nicholson, D.W., and Thornberry, N.A. (1997). Caspases: killer prote-Li, H., Bergeron, L., Cryns, V., Pasternack, M.S., Zhu, H., Shi, L., ases. Trends Biochem. Sci. 22, 299–306.
Greenberg, A., and Yuan, J. (1997a). Activation of caspase-2 in
Norman, J., Yang, J., Fink, G., Carter, G., Ku, G., Denham, W., and apoptosis. J. Biol. Chem. 272, 21010–21017.
Livingston, D. (1997). Severity and mortality of experimental pancre-Li, P., Allen, H., Banerjee, S., and Seshadri, T. (1997b). Characteriza- atitis are dependent on interleukin-1b converting enzyme (ICE). J. tion of mice deficient in interleukin-1b converting enzyme. J Cell. Interferon Cytokine Res. 17, 113–118.
Biochem. 64, 27–32.
Rotonda, J., Nicholson, D.W., Fazil, K.M., Gallant, M., Gareau, Y., Li, P., Nijhawan, D., Budihardjo, I., Srinivasula, S.M., Ahmad, M., Labelle, M., Peterson, E.P., Rasper, D.M., Ruel, R., Vaillancourt, J.P., Alnemri, E.S., and Wang, X. (1997c). Cytochrome c and
dATP-depen-et al. (1996). The three-dimensional structure of apopain/CPP32, a dent formation of Apaf-1/Caspase-9 complex initiates an apoptotic key mediator of apoptosis. Nat. Struct. Biol. 3, 619–625. protease cascade. Cell 91, 479–489.
Scaffidi, C., Fulda, S., Srinivasan, A., Friesen, C., Li, F., Tomaselli, Li, H., Zhu, H., Xu, C.J., and Yuan, J. (1998). Cleavage of BID by
K.J., Debatin, K.-M., Krammer, P.H., and Peter M.E. (1998). Two caspase 8 mediates the mitochondrial damage in the Fas pathway
CD95 (APO-1/Fas) signaling pathways, EMBO J. 17, 1675–1687. of apoptosis. Cell 94, 491–501.
Schielke, G.P., Yang, G.Y., Shivers, B.D., and Betz, A.L. (1998). Re-Los, M., van de Craen, M., Penning, C.L., Schenk, H., Westendorp,
duced ischemic brain injury in interleukin-1b converting enzyme-M., Baeuerle, P.A., Dro¨ge, W., Krammer, P.H., Fiers, W., and
Schulze-deficient mice. J. Cereb. Blood Flow Metab. 18, 180–185. Osthoff, K. (1995). Requirement of an ICE/Ced-3 protease for Fas/
Schulze-Osthoff, K., Ferrari, D., Los, M., Wesselborg, S., and Peter, Apo-1 mediated apoptosis. Nature 375, 81–83.
M.E. (1998). Apoptosis signaling by death receptors. Eur. J. Bio-Los, M., Herr, I., Friesen, C., Fulda, S., Schulze-Osthoff, K., and
chem. 254, 439–459. Debatin, K.M. (1997). Cross-resistance of CD95- and drug-induced
Slee, E.A., Harte, M.T., Kluck, R.M., Wolf, B.B., Casiano, C.A., New-apoptosis as a consequence of deficient activation of caspases
meyer, D.D., Wang, H.G., Reed, J.C., Nicholson, D.W., Alnemri, E.S., (ICE/Ced-3 proteases). Blood 90, 3118–3129.
et al. (1999). Ordering the cytochrome c-initiated caspase cascade: Luo, X., Budihardjo, I., Zou, H., Slaughter, C., and Wang, X. (1998).
hierarchical activation of caspases-2, -3, -6, -7, -8, and -10 in a Bid, a Bcl2 interacting protein, mediates cytochrome c release fom
caspase-9-dependent manner. J. Cell Biol. 144, 281–292. mitochondria in response to activation of cell surface death
