FLIP and the Death Effector Domain Family

JW Yu and Y Shi

Department of Molecular Biology, Lewis Thomas Laboratory, Princeton University, Princeton, NJ, USA

Death effector domains (DEDs) are protein interaction and stress that impinge on the mitochondria or modules found in a number of proteins known to regulate ‘extrinsically’ from extracellular ligands that activate apoptosis from death receptors. The core DED family death receptors at the cell surface. In either scenario, members that orchestrate programmed cell death from each pathway critically relies on the action of a family of death receptors include the adaptor protein FADD, the cysteine proteases known as caspases to initiate and initiator caspases procaspases-8 and -10 and the regulatory execute the cell death process through a two-step protein c-FLIP. Through homotypic DED interactions, cascade (Riedl and Shi, 2004). In the apoptotic these proteins assemble into the death-inducingsignaling proteolytic cascade, the initiator caspases (caspases-2, complex (DISC) to regulate initiator caspase activation -8, -9 and -10) cleave and consequently activate the and launch the apoptotic proteolytic cascade. A consider- effector caspases (caspases-3, -6 and -7), and the effector able body of evidence, however, is revealingthat the same caspases in turn cleave a large array of substrates to core group of DED-containing proteins also paradoxically dismantle and package the cell into apoptotic bodies. promotes survival and proliferation in lymphocytes and Activation of initiator caspases for both the intrinsic possibly other cell types. This review delves into recent and extrinsic pathways is a crucial regulatory step in findings regarding these two opposing functional aspects of inducing cell death and is achieved through adaptor- the core DED proteins. We discuss the current effort induced oligomerization (Bao and Shi, 2007). Assembly expandingour structural and biochemical view of how of the activating platform is mediated by protein–protein DED proteins assemble into the DISC to fully activate interactions driven by small hexa-helical domains that initiator caspases and execute cell death, and finally we include the caspase recruitment domain (CARD), the examine details linkingthe same proteins to proliferation death domain (DD) and the death effector domain and describe how this outcome might be achieved through (DED). These structurally conserved domains are restricted activation of initiator caspases. members of the DD superfamily, and within this large Oncogene (2008) 27, 6216–6227; doi:10.1038/onc.2008.299 group, the DD and DED subfamilies propagate apoptotic and sometimes proliferative signals from death Keywords: apoptosis; extrinsic cell death; caspase-8; receptors in the extrinsic cell death pathway. Although DED; FLIP; NF-kB the DD is dispersed in a wide array of proteins (32 in total in humans) involved in multiple pathways, the DED is predominantly confined to eight proteins that dictate cellular fate from death receptors and possibly other locations (Figures 1 and 2) (Reed et al., 2004). In Introduction recent years, considerable interest has centered on three groups of DED family members, including FADD, Apoptosis is a tightly regulated cell death process procaspases-8/10 and c-FLIP (Long, Short and Raji), actively implemented to ensure proper development and much has been uncovered in the way they assemble during the early stages of life and homeostasis through- through homotypic DED interactions to implement the out the adulthood of all multicellular organisms apoptotic program. Perhaps most interesting of all, a (Rathmell and Thompson, 2002; Danial and Kors- collective body of research is beginning to establish that meyer, 2004). In humans, defects in this cell death the exact same core group of proteins that implement cell program resulting from insufficient or excessive apop- death (FADD, procaspases-8/10 and c-FLIP (L, S and tosis provokes a number of diseases, including auto- R)) can also work together to promote proliferative and immunity, cancer and neurodegenerative disorders survival signals in remarkable ways. Rather than (Hanahan and Weinberg, 2000; Vauxand Flavell, surveying the DED family as a whole, we focus our 2000; Yuan and Yankner, 2000; Green and Evan, attention to these recent developments in this review. 2002). This conserved cellular suicide program is initiated either ‘intrinsically’ through cellular insults Death receptors Correspondence: Dr JW Yu or Professor Y Shi, Department of Molecular Biology, Lewis Thomas Laboratory, Princeton University, Princeton, NJ 08544, USA. As part of the tumor necrosis factor receptor super- E-mails: [email protected] or [email protected] family, the death receptor family so far consists of eight FLIP and the death effector domain family JW Yu and Y Shi 6217 members that possess an extracellular domain consisting and El-Deiry, 2003; Wajant, 2003). Of these identified of up to five cysteine-rich repeats and an intracellular members, only signaling from TNFR1, Fas and TRAIL DD. These include TNFR1, CD95/Fas, DR3, TRAILR1/ (DR4 and DR5) receptors are reasonably well character- DR4, TRAILR2/DR5, DR6, EDAR and NGFR (Ozoren ized and they trigger cell death or in some cases proliferation from the cell surface under varying condi- tions. From a physiological standpoint, Fas-mediated cell death appears to be critical for immune system function by removing self-reacting B and T cells, down-sizing the population of activated lymphocytes after an immune response, restricting immune system access from privileged sites such as the eye and testes, and eliminating cancer cells and virally infected cells (Krammer, 2000; Curtin and Cotter, 2003). In humans, loss of Fas receptor function has been implicated in autoimmune disorders and correlated with cancers of immune cells (and other cell types). In addition to its role in cell death, Fas receptor signaling also has been shown to be critical for T-cell survival and proliferation (Peter et al., 2007). Along this line, the Fas receptor also elicits non-apoptotic signals in diverse tissue Figure 1 The human death effector domain (DED) family types outside the immune system (Peter et al., 2007), members. FADD, procaspases-8/10 and c-FLIP (L, S and R) are core components of the death-inducing signaling complex(DISC). participating in functions such as liver regeneration DEDD, DEDD2 and PEA-15 are not nearly as well characterized, (Desbarats and Newell, 2000) and neurite outgrowth of but nevertheless have also been implicated in regulating extrinsic sensory neurons (Desbarats et al., 2003). The function of cell death in addition to participating in other cellular functions. the TRAIL receptors, on the other hand, are less well PEA-15, when phosphorylated, inhibits death receptor-induced apoptosis through homotyptic DED interactions with FADD to understood, although a number of studies implicate these prevent the recruitment of procaspases-8 and -10 (Condorelli et al., receptors in immune system function as well (Ozoren and 1999; Renganathan et al., 2005). DEDD and DEDD2 conversely El-Deiry, 2003). Most studies have focused on their are thought to function in a proapoptotic manner by interacting anticancer function given the fact that TRAIL ligand with and targeting caspases-8 and -10 to the nucleus, where they has been shown to induce apoptosis in a wide range of might inhibit protein translation (Alcivar et al., 2003). Further details regarding DEDD, DEDD2 and PEA-15 are not discussed in human cancer cell lines, whereas exhibiting no apparent this review. There are also a handful of additional proteins (not adverse effect on normal cells (Walczak et al., 1999; Abe shown in the figure and are also not discussed in this review) that et al., 2000; Wang and El-Deiry, 2003). Apoptotic contain variant DEDs (vDEDs), which are thought to be mere signaling by TNFR1 is less common and activation of coiled-coil regions and nonclassical DED-like (DED-L) domains (Reed et al., 2004). Proteins that contain vDED include Bap31, this receptor appears to be more critical in proinflamma- BAR, Hip and Hippi. Proteins that contain DED-L include tory responses (Ashkenazi and Dixit, 1998; Ozoren and FLASH and Dap3. NLS, nuclear localization signal. El-Deiry, 2003).

Figure 2 Sequence alignment of death effector domains (DEDs) from human and viral proteins. The top panel shows an alignment of the single DED found in FADD, DEDD, DEDD2 and PEA-15. The lower panels show an alignment for the tandem DEDs present in viral FLIPs (MC159 and KSHV (Kaposi’s sarcoma-associated herpesvirus)), c-FLIP, procaspase-8b (C8b) and procaspase-10 (C10). A stretch of sixto ten amino acids that resides between DED1 and DED2 (between a7a and a1b) of the tandem DEDs are omitted from the sequence alignments in the lower panels. Residues that comprise the conserved hydrophobic patch are highlighted in red. The highly conserved RXDL motif found in a6 is denoted in yellow. Secondary structural elements illustrated above the sequences are based on the structures of the FADD (DED) for the top panel and MC159 for the lower panels.

Oncogene FLIP and the death effector domain family JW Yu and Y Shi 6218 Apoptosis mediated by assembly of DED-containing In this case, allosteric activation of the protease domain proteins of procaspase-8 is thought to occur through hetero- dimerization with the catalytically inactive protease In the case of Fas or TRAIL (DR4 and DR5) receptors, domain of c-FLIPL (Chang et al., 2002; Micheau apoptosis is initiated by binding of a trimeric ligand to et al., 2002; Boatright et al., 2004). preassociated receptor complexes at the cell surface In addition to modulating apoptosis, c-FLIPL is also (Chan et al., 2000; Siegel et al., 2000). This in turn capable of redirecting death signals towards survival results in receptor oligomerization (Algeciras-Schimnich and proliferative pathways. Although this path is not et al., 2002; Holler et al., 2003) and subsequent well defined, current evidence indicates that at moderate recruitment of the adaptor protein FADD and initiator expression levels c-FLIPL allosterically activates procas- caspases (procaspase-8 or -10) to form the death- pase-8, but restricts its activity below a threshold such inducing signaling complex(DISC) (Kischkel et al., that the mature and highly active form is not generated. 1995; Medema et al., 1997). It is now well accepted that Instead, active procaspase-8 is only partially processed assembly of DISC at activated death receptors is (or unprocessed) and restricted to the DISC, where it mediated by homotypic DD and DED interactions. cleaves substrates in close proximity, such as c-FLIPL For example, the carboxy-terminal DD of FADD has itself, to recruit specific proteins and initiate non- been shown to bind to the intracellular DD of death apoptotic signaling pathways. This dynamic regulation receptors, whereas the amino-terminal DED of FADD of procaspase-8 activation at the DISC to elicit diverse has been shown to bind the amino-terminal prodomain cellular responses is supported by the fact that the (tandem DED) of procaspase-8 or -10 (Reed et al., expression level of c-FLIPL, but not FADD or 2004). Through these homotypic interactions, the procaspase-8, is tightly regulated by a number of initiator caspases are clustered together in a large different pathways including NF-kB (nuclear factor- oligomeric complexat the cell surface, allowing dimer- kB), MAPK/ERK and AKT (Krueger et al., 2001). ization of their carboxy-terminal protease domains Finally, it should be noted that details associated with through a specific interface to elicit proteolytic activity two different cell types can also affect the proficiency of (Boatright et al., 2003; Chang et al., 2003; Donepudi DISC assembly and propagation of apoptosis, specifi- et al., 2003). These caspases then undergo autoproteo- cally with regard to Fas receptor signaling (Scaffidi lysis to generate a mature and highly active dimeric form et al., 1998). In type I cells, Fas receptors are associated that is released from the DISC to initiate the apoptotic with lipid rafts and consequently the DISC is efficiently proteolytic cascade. assembled and procaspase-8 is robustly activated to In addition to FADD and the initiator caspases, a facilitate cell death (Muppidi and Siegel, 2004). Con- group of viral and cellular DED-containing proteins versely, Fas receptors from type II cells are excluded named FLIP also appears to be an integral component from lipid rafts and inefficiently assemble DISC upon of DISC. The cellular version of FLIP is predominantly activation of death receptors. These cells activate only a expressed in three forms: the long form (c-FLIPL)is small fraction of procaspase-8 and therefore require an highly homologous to procaspases-8 and -10, possessing amplification step through cleavage of the BH3-only tandem DEDs at its amino-terminus and a catalytically protein Bid (by caspase-8) to initiate the intrinsic cell inactive protease domain at its carboxy-terminus death pathway and promote cell death (Li et al., 1998; (Figure 1). The shorter forms (c-FLIPS and c-FLIPR) Luo et al., 1998). consist of only the amino-terminal tandem DEDs followed by a short carboxy-terminal stretch and is very similar in architecture to the viral FLIPs (Krueger et al., 2001; Golks et al., 2005). Although v-FLIP and c- Homotypic DED interactions amongFADD, FLIPS have been established as potent inhibitors of procaspase-8 and c-FLIP in DISC assembly procaspase-8 activation, the function of c-FLIPL was initially reported to be either proapoptosis (Goltsev As described above, DISC assembly is mediated by et al., 1997; Han et al., 1997; Inohara et al., 1997; Shu homotypic DD interactions between Fas receptor and et al., 1997) or antiapoptosis (Hu et al., 1997; Irmler FADD and homotypic DED interactions between et al., 1997; Srinivasula et al., 1997; Rasper et al., 1998). FADD and the prodomain of procaspases-8 and -10 Despite the conflicting biological roles initially reported, (or FLIP). Although these interactions are well docu- further study since then supports the notion that mented, the stoichiometry of each component at the c-FLIPL can either promote or inhibit apoptosis DISC and the precise mode of domain assembly are depending on its expression levels. When highly largely unclear. This is in part due to the fact that most expressed, c-FLIPL (and c-FLIPS) appears to compete but not all isolated DEDs (of FADD and procaspases-8 with procaspase-8 for recruitment into the DISC and -10 in particular) and isolated DDs exhibit a strong through its own amino-terminal tandem DEDs thereby tendency to aggregate and consequently exhibit poor preventing initiator caspase activation. However, at solution behavior, making biochemical and biophysical physiological expression levels (B1% of procaspase-8 characterization treacherous. Despite this technical expression levels; Scaffidi et al., 1999), c-FLIPL can still difficulty, NMR structures of the isolated DD of Fas be targeted to the DISC, where it substantially promotes receptor (Huang et al., 1996) and the DD and DED procaspase-8 activation and enhances apoptotic signaling. (F25Y mutant with significantly improved solution

Oncogene FLIP and the death effector domain family JW Yu and Y Shi 6219 properties) of FADD (Eberstadt et al., 1998; Jeong interface between both domains and contacts a1 and a4 et al., 1999; Berglund et al., 2000; Carrington et al., of DED2. Although the structures of the tandem DEDs 2006) have been determined. These structures reveal a of c-FLIP and procaspases-8 and -10 are yet to be shared overall fold, with sixantiparallel a-helices typical determined, conservation of specific residues at the of all DD superfamily members, but exhibit strikingly interface between both domains strongly indicate that different surface properties. The DD of Fas and FADD the overall arrangement observed for MC159 is pre- appears to be extensively charged at their surface and served also in procaspases-8 and -10 and the various mutagenesis data suggest that they associate with each FLIPs. More importantly, the packing between the other through electrostatic interactions (Jeong et al., DEDs of MC159 is similar in arrangement to the 1999; Bang et al., 2000; Hill et al., 2004). By contrast, heterodimeric CARD–CARD complexfrom Apaf-1 the DED of FADD possesses a key surface-exposed and procaspase-9 (proteins critical for apoptosis in the hydrophobic patch (localized to a2 and a5) that appears intrinsic pathway), supporting the notion that the DED to be conserved in virtually all other DEDs including the assembly involving the conserved hydrophobic patch tandem DEDs of procaspase-8 and FLIP proteins might represent one potential mode whereby FADD (Figures 2 and 3). Notably, this region is thought and procaspase-8 interact (see below). to be critical in mediating DED–DED interactions Although the conserved hydrophobic patch on DED1 (Eberstadt et al., 1998). of procaspase-8 (and FLIP) is involved in tandem DED Recently, the crystal structure of the tandem DED of interactions, the same conserved hydrophobic patch on the viral FLIP MC159 has been determined (Yang et al., DED2 is solvent exposed. Recent mutagenesis analysis 2005; Li et al., 2006). DED1 and DED2 were found to suggests that the DED2 hydrophobic patch is required tightly associate with each other through a hydrophobic for interaction with the FADD–DED (Yang et al., interface, forming a single rigid unit rather than 2005). Conversely, residues within the a1anda4 helices individual domains linked together by a flexible linker of FADD–DED have been shown to be essential for (Figure 3). Significantly, this structure has yielded interaction with procaspases-8 and -10 (Carrington detailed insight into how DEDs between two different et al., 2006). On the basis of these findings, it was proteins might associate in a homotypic manner. In this proposed that the DED of FADD (a1 and a4) engages case, the conserved hydrophobic patch of MC159 DED2 of procaspase-8 (a2 and a5) in a manner similar DED1 that centers on the a2anda5 is integral to the to that observed between the tandem DEDs of MC159

Figure 3 Structure of the single and tandem death effector domains (DEDs). The NMR structure of FADD (DEDF25Y) and the crystal structure of MC159 are represented as ribbon and surface models. DEDs, as exempliﬁed by FADD (DED), typically possess six antiparallel-a helices; however, DED1 within the context of tandem DEDs of MC159 appear to slightly diverge from this arrangement as a3 is noticeably absent, and an additional a helix( a7) forms part of the linker leading to DED2. The highly conserved surface- exposed hydrophobic patch for FADD and MC159 is illustrated as red areas encompassing a2/a5.

Oncogene FLIP and the death effector domain family JW Yu and Y Shi 6220

Figure 4 Model of FADD assembly with procaspases-8/10 or c-FLIP through homotypic death effector domain (DED) interactions. Mutagenesis data suggest that a2/a5 (the conserved hydrophobic patch) of procaspase-8 (DED2) binds a1/a4 of FADD (DED), indicating that both proteins might interact as observed between DED1 and DED2 of MC159 (see Figure 3). The tandem DED region of procaspases-8/10 or c-FLIP, represented here by the structure of MC159, was docked onto the NMR structure of FADD (DEDF25Y þ DD) accordingly.

(Carrington et al., 2006) to facilitate DISC assembly and self-associate into higher order oligomers. In vitro, these induction of apoptosis (Figure 4). Because of its ability domains are highly aggregated and exhibit extremely to effectively compete away procaspase-8 from FADD poor solution properties. Similarly, non-physiological in pull-down experiments from cell lysates, c-FLIP is overexpression of the isolated DED of FADD or the also thought to engage FADD in the same manner tandem DEDs of procaspase-8 as GFP fusions in (Yang et al., 2005). The viral FLIP MC159, on the other human cell lines induces the formation of long hand, cannot compete procaspase-8 away in the same cytoplasmic filaments designated death effector fila- assay and appears to bind FADD–DED through an ments, which are capable of activating endogenous extensive area outside the surface-exposed conserved initiator caspases and inducing apoptosis independent of hydrophobic patch that includes the well-conserved death receptors (Siegel et al., 1998). In contrast to these RXDL motif (in a6) present in virtually all other DEDs proapoptotic DEDs, the isolated tandem DEDs from (Yang et al., 2005). In this regard, MC159 is proposed to the antiapoptotic viral FLIPs of E8 or MC159 are disrupt DISC formation by preventing FADD–DED monomeric in vitro (Yang et al., 2005; Li et al., 2006) self-association rather than competing away procaspase-8 and cannot form these filaments in cell culture; in fact, (see below). Thus, perhaps not all viral FLIPs assemble these viral DEDs can effectively blunt the ability of the into the DISC in the same manner as the cellular FLIPs DEDs of FADD and procaspase-8 to form death and procaspases-8 and -10 as suggested above. These effector filaments, thus preventing apoptosis (Siegel observations and analyses now provide an important et al., 1998). At a simplistic level, regulated DED self- glimpse into a key homotypic DED interaction required assembly likely promotes apoptosis by perhaps merely for DISC assembly. Nonetheless, the validity of the facilitating initiator caspase dimerization and activation predictions outlined here is still uncertain and remains in the context of a large oligomeric complex. Beyond to be confirmed by structural data in the form of protein this generalized view, recent findings discussed below complexes. indicate that DED self-assembly may play a more intricate role at different stages of death receptor signaling. With regard to FADD, oligomerization of its amino- DED-dependent self-association of FADD in DISC terminal DED has been recently implicated in driving assembly homotypic DD interactions between the cytoplasmic tails of death receptors and the carboxy-terminus of Along with linking FADD to procaspases-8 or 10, the FADD itself. The possible role of FADD–DED in this DEDs within both proteins also have the propensity to aspect of DISC assembly was initially uncovered in a

Oncogene FLIP and the death effector domain family JW Yu and Y Shi 6221 random screen that identified FADD mutants in the were originally proposed to interact with the tandem context of full-length protein (the majority of which DEDs of procaspase-8 and c-FLIP (Eberstadt et al., localized to the DED), which prevent association with 1998; Kaufmann et al., 2002), current mutagenesis data the DD of Fas and TRAIL receptors without disrupting imply that self-association of FADD–DED in this interaction with procaspase-8 or TRADD (another DD- region enables binding to procaspase-8 and FLIP containing adaptor protein involved in TNFR1 signal- perhaps through a cooperative assembly mechanism as ing) (Thomas et al., 2002, 2004). Since then, it has been described above (Sandu et al., 2006). By contrast, the proposed that FADD self-association through its DED region within a1/a4 of FADD–DED appears to directly functions by enhancing binding affinity to death bind procaspase-8 without affecting DED self-associa- receptors through formation of multivalent homotypic tion at all (as mentioned above) (Sandu et al., 2006). DD interactions afforded by oligomeric FADD mole- Another group has recently proposed that the highly cules and oligomeric ligand-bound receptors at the cell conserved RXDL motif in a6 of FADD–DED mediates surface (Sandu et al., 2006). In support of this view, self-assembly (Muppidi et al., 2006), but the validity of it has been demonstrated that enforced FADD–DD this observation remains in question, as mutations in dimers or trimers (through leucine zipper motifs) this motif disrupt structural integrity (as assessed by interacts robustly with death receptors in cells and NMR) as well as the ability to self-associate (Carrington efficiently blocks apoptosis as a dominant-negative et al., 2006). inhibitor, whereas the isolated FADD–DD at similar expression levels does so rather inefficiently in both regards (Sandu et al., 2006). A plausible interpretation of these findings is that weak DED interactions (which The core DISC components are also involved facilitate FADD self association) and weak DD inter- in proliferation actions (between Fas and FADD) may be cooperatively assembled to form a stable signaling complexat the In addition to executing cell death, the core DED- plasma membrane when death receptors bind their containing DISC components (FADD, procaspase-8 specific ligands (Sandu et al., 2006). Alternatively, and c-FLIP) are also critical mediators of cell survival aggregation-prone FADD may be kept in a monomeric and proliferation, particularly in lymphocytes. Although state by chaperones, and initiation of the extrinsic disruption of any of these genes yields embryonic pathway could release the chaperone and hence allow lethality in mice, conditional or targeted deletion of spontaneous oligomerization and translocation to the each component in murine T cells reveals an inability to cytoplasmic tails of death receptors through an as yet proliferate upon T-cell receptor (TCR) stimulation—a unidentified mechanism (Stilo et al., 2003). It should critical step required to expand the population of also be noted that after DISC assembly, FADD–DED lymphocytes and mount the immune response (Zhang self-association is also proposed to crosslink an array of et al., 1998; Salmena et al., 2003; Chau et al., 2005; trimeric receptor complexes at the cell surface to Zhang and He, 2005). Earlier studies examining the generate microscopically visible clusters of receptor effects of caspase inhibitors on T-cell activation para- termed SPOTS (signaling protein oligomeric transduc- doxically showed that protease activity (likely from tion structures) (Siegel et al., 2004; Muppidi et al., 2006). procaspase-8 normally thought to induce death) is The formation of SPOTS absolutely requires FADD indeed required for proliferation (Kennedy et al., (Siegel et al., 2004) and likely amplifies the apoptotic 1999). More importantly, procaspase-8 deficiency in signal by perhaps augmenting initiator caspase recruit- humans, as a result of an inactivating mutation, also ment and promoting efficient dimerization of their manifests antigen receptor activation and proliferative protease domain to induce catalytic activity. This notion defects in T cells as well as B cells and natural killer cells, is supported by a previous study demonstrating that suggesting perhaps a common non-apoptotic role for maximal recruitment of FADD and procaspase-8 occurs DED proteins in different immunological cell types within the time frame of SPOTS formation (B10– (Chun et al., 2002). 15 min) (Kischkel et al., 1995). Furthermore, given the How then do these DED proteins transmit a perpendicular arrangement of procaspase-8 relative to proliferative signal from antigen receptors in T cells FADD modeled in Figure 4, it is conceivable that at (and possibly other lymphocytes) rather than execute least two groups of trimeric receptors need to be cell death? In the case of FADD, some phenotypic crosslinked together to optimally dimerize two mole- differences compared with procaspase-8 and FLIP- cules of procaspase-8 in the specific antiparallel orienta- deficient murine T cells suggest that its involvement in tion required to stimulate proteolytic activity. proliferation may be attributed to its unique and From a structural perspective, there is currently little ubiquitous role in cell cycle progression. During the insight into how FADD–DED self-associates in the early phase of mitosis, casein kinase Ia interacts with context of the DISC. Nonetheless, surface-exposed FADD at the mitotic spindle and spindle poles, where it residues involved in FADD self-assembly have been phosphorylates FADD at its flexible carboxy-terminal identified and interestingly map to the conserved tail (Ser194 in humans) and consequently interferes with hydrophobic patch centering on a2 and a5 (as described cell cycle progression at the G2/M checkpoint through above) and a small basic region encompassing a3 an unknown mechanism; it is thought that regulated (Figures 3 and 4). Although these regions of FADD dephosphorylation or degradation of FADD might then

Oncogene FLIP and the death effector domain family JW Yu and Y Shi 6222 allow transition to the next phase of the cell cycle (Alappat et al., 2005). In contrast to this ubiquitous role, recent analysis has shown that catalytically active procaspase-8 promotes cell survival/proliferation in human T, B and natural killer cells by specifically activating the transcription factor NF-kB(Suet al., 2005). Closer examination of the proteolytic activity from caspase-8 in this setting revealed two interesting observations (Su et al., 2005; Misra et al., 2007): (1) the ‘active’ enzyme is unprocessed (normally recognized as the latent form) as opposed to the fully auto-processed form (two-chain), which is associated with apoptosis, and (2) only a small fraction of total procaspase-8 is activated compared with the significantly higher degree of activation observed in cells undergoing death Figure 5 Proteolytic cleavage of c-FLIP by procaspase-8 enhances its interaction with key signaling proteins to activate nuclear factor receptor-mediated apoptosis. (NF)-kB. Procaspase-8 (and possibly mature caspase-8) cleaves In light of these observations, the regulatory protein c-FLIP at D376 and D198 to generate the products as shown. c-FLIPL appears to be an ideal candidate that can Processing at D376 of c-FLIP merely severs the flexible inter- function as a procaspase-8 activator capable of restrict- subunit linker (located between p20 and p12 of its caspase-like domain) without drastically altering the p20/p12 caspase-like unit. ing the degree of activation such that procaspase-8 is not It is still unclear how scission of this loop translates into enhanced auto-processed to the fully active form and subsequently direct (or possibly indirect) interaction with TRAF2 and RIP1. serve as a proteolytic substrate to initiate NF-kB Moreover, the specific regions of c-FLIP that bind TRAF2 or RIP1 signaling. Indeed, a previous study has shown that are also still undefined. Cleavage of c-FLIP at D198 yields the p22 overexpression of c-FLIP enhances T-cell proliferation unit, which has been shown to interact with the regulatory subunit L (IKKg) of the IKK complex. The role of p22 c-FLIP in activating in transgenic mice (Lens et al., 2002) and promotes TCR IKK and hence NF-kB signaling is strongly supported by recent (and TCR-Fas receptor co-stimulation)-induced activa- structural data showing that the viral FLIP KSHV (Kaposi’s tion of ERK and NF-kB signaling in tissue culture sarcoma-associated herpesvirus) can directly associate with IKKg as well (see Figure 6). (Kataoka et al., 2000). In further support, c-FLIPL has also been shown to interact with Raf1, a member of the ERK pathway, and with other signaling proteins that considered on the basis that Fas ligand can induce activate NF-kB including TNFR-associated factor 1 T-cell proliferation when co-treated with suboptimal (TRAF1), TRAF2 and receptor-interacting protein 1 doses of TCR ligand (Alderson et al., 1993; Kataoka (RIP1). Perhaps more importantly, subsequent effort et al., 2000). Moreover, c-FLIP was shown to recruit has demonstrated that proteolytic cleavage of the TRAF1, TRAF2, RIP1 and Raf-1 to Fas receptor flexible intersubunit linker within FLIP’s protease-like complexes in human embryonic kidney cells (Kataoka domain (at D376) by procaspase-8 promotes the et al., 2000), implying that the same may be true in recruitment of TRAF2 (Kataoka and Tschopp, 2004) lymphocytes. However, the likelihood of this possibility and RIP1 (Dohrman et al., 2005) to receptor complexes. is confounded by the fact that murine T cells with Additional evidence also suggests that c-FLIP (both the defective Fas receptors (lpr mice) still undergo normal long and short forms) can be further cleaved at D198 by TCR-dependent proliferation (Boissonnas et al., 2002). procaspase-8 to generate a p22 form that only consists Alternatively, it has been recently proposed that of its amino-terminal tandem DEDs, and this form procaspase-8 can associate with the Carma1–Bcl10– directly interacts with and recruits the regulatory Malt1 (CBM) complex(Su et al., 2005; Misra et al., subunit IKKg of the IkB kinase (IKK) complexto 2007), a critical platform that mediates NF-kB signaling potently activate NF-kB (Golks et al., 2006). Thus based downstream of antigen receptors in lymphocytes. In this on the available evidence, cleavage of c-FLIPL and context, ligand-bound antigen receptors trigger a possibly c-FLIPS by procaspase-8 may be a key number of events that lead to the activation of PKC-y signaling event that facilitates recruitment of the (for T cells) and ensuing phosphorylation of membrane- relevant proteins described above to initiate cellular associated Carma1 culminates in the recruitment of survival and proliferation (Figure 5). A prominent role Bcl10–Malt1 to lipid rafts. The assembly of CBM for c-FLIP in the induction of NF-kB is further activates the IKK complexthrough a mechanism that supported by the observation that signaling from this is not completely understood but includes K63-linked pathway is potently activated by the viral FLIP KSHV ubiquitylation of the regulatory subunit IKKg to (Kaposi’s sarcoma-associated herpesvirus) (Chaudhary possibly facilitate additional protein interactions (K48- et al., 1999). Similar to c-FLIPL, this viral FLIP can also linked ubiquitylation by contrast targets proteins for directly interact with TRAF2 (Guasparri et al., 2006) proteasomal degradation) (Rawlings et al., 2006). The and the IKKg regulatory subunit (Liu et al., 2002; Field IKK complexsuccessively phosphorylates I kB (inhibi- et al., 2003). tor of NF-kB) and signals IkB degradation through Do FADD, procaspase-8 and c-FLIP assemble at the poly-ubiquitylation and proteasomal digestion, which cytoplasmic face of Fas receptors to activate NF-kB frees NF-kB from cytosolic sequestration by IkB so that upon stimulation of TCR? This idea was initially it can enter the nucleus and induce gene expression.

Oncogene FLIP and the death effector domain family JW Yu and Y Shi 6223 Activated and unprocessed procaspase-8, in this setting, modeling data suggests that the mode of interaction is has been shown to interact with Bcl10–Malt1 before likely to be very similar. Thus, c-FLIP-mediated recruitment and activation of the IKK complex, and recruitment and allosteric activation of IKKg (see conversely recruitment and activation of the IKK Figure 6 legend) in conjunction with Malt1 (and/or complexdoes not occur in the absence of procaspase-8 TRAF6)-dependent ubiquitylation of IKKg at lipid (Su et al., 2005). Furthermore, c-FLIP has also been rafts are probably necessary for optimal activation of shown to associate with the CBM complexin lipid rafts the IKK complex, although the exact mechanistic details during T-cell activation (Misra et al., 2007). So how of this pathway remain to be elucidated. might active procaspase-8 and c-FLIP function in this Other major questions that emerge from this current regard? On the basis of the information outlined above, view of TCR signaling centers on how procaspase-8 and we know that c-FLIPL can activate procaspase-8 below c-FLIP are recruited to and interact with the CBM the threshold necessary for apoptosis (depending on complexand what role, if any, does FADD play in this expression level), and following that, proteolytic proces- capacity. Unlike the situation with DISC assembly in sing of c-FLIP by procaspase-8 may be a central step apoptotic signaling from death receptors, the role of the towards initiation of NF-kB through the recruitment of isolated DEDs in this situation is completely unknown. the IKK complexand other key proteins. This scenario FADD was shown to only transiently associate with is reinforced by recent structural data revealing that the CBM during the very early stages of TCR signaling (Su viral FLIP KSHV can indeed directly bind the et al., 2005). By contrast, procaspase-8 and c-FLIP were regulatory subunit of the IKK complex(Figure 6) demonstrated to associate with CBM more stably (Su (Bagneris et al., 2008). More importantly, the cleaved et al., 2005; Misra et al., 2007). On the basis of these p22 form of c-FLIP also binds IKKg, and homology- preliminary ﬁndings, we can only speculate for now that

Figure 6 Activation of IKK by FLIP. The crystal structure of v-FLIP KSHV (Kaposi’s sarcoma-associated herpesvirus) in complex with a fragment of IKKg (helical region of HLX2) is represented as a ribbon model. The schematic domain organization of full-length IKKg is also shown to the left of the structure. The listed domains are as follows: HLX1 and HLX2: helical domains, CC1 and CC2: coiled-coil regions, LZ: leucine zipper and ZF: zinc ﬁnger. The IKK complexconsists of two distinct kinases (IKK a and IKKb) and a regulatory subunit (IKKg). In unstimulated cells, the regulatory subunit (IKKg) is thought to exist as a monomer in an auto-inhibited conformation with the HLX2 domain packed against CC2 and LZ (Bagneris et al., 2008). In the presence of KSHV FLIP or p22 c-FLIP, IKKg is proposed to unravel into an active oligomeric conformation through direct interaction with FLIP revealed in the structure shown here. This in turn would allow direct recruitment of IKKa and IKKb to facilitate their phosphorylation and subsequent activation.

Oncogene FLIP and the death effector domain family JW Yu and Y Shi 6224 FADD may not significantly influence procaspase-8 and placing c-FLIPL as the central switch between prolifera- c-FLIP recruitment to CBM. Yet the possibility that tion (through ERK or NF-kB) and death at the DISC. FADD and Fas receptors (which also are concentrated Recent examples that strongly support this notion in the in lipid rafts in T cells) may somehow participate in context of Fas signaling have been observed in thyroid TCR signaling cannot be excluded, especially given the carcinoma cell lines (Mitsiades et al., 2006) and fact that Fas ligand can induce T-cell proliferation when pancreatic b cells (Maedler et al., 2002). Interestingly, co-treated with suboptimal doses of TCR ligand. The a critical role for c-FLIPL in transmitting ERK- only protein implicated in recruiting procaspase-8 into dependent proliferation from TRAIL receptors in lipid rafts and the CBM complexso far is the ubiquitin human glioma cell lines has also been recently demon- ligase TRAF6, although exactly how this happens and strated (Vilimanovich and Bumbasirevic, 2008). what additional proteins, if any, are needed is still in Precisely how FADD, c-FLIPL and procaspase-8 question (Bidere et al., 2006). modulate cell fate from different death receptors in The observations outlined above certainly provide a different cellular/environmental contexts still remains to direct link from procaspase-8 and c-FLIP to the IKK be elucidated. machinery in TCR signaling within the confines of the CBM signalosome; however, there are many other members that participate in this pathway and it is Concludingremarks conceivable that other players may play a more central or parallel role. In this regard, it is also worthwhile to As highlighted in this review, detailed insights into the mention that for many arguments made for the well-established capacity of the core DED family contributions of procaspase-8 and c-FLIP to lympho- members to induce apoptosis from death receptors at cyte proliferation and NF-kB signaling, there are also the DISC, particularly from a structural and biochem- other data that suggest otherwise. For example, ical viewpoint, are still in high demand. Furthermore, an procaspase-8 or c-FLIP deficiency in murine T cells intriguing yet preliminary map outlining how the same does not appear to overtly affect TCR-induced NF-kB proteins participate in cellular survival and proliferation activation (Salmena et al., 2003; Zhang and He, 2005), is beginning to come into view. Despite this remarkable but in stark contrast procaspase-8 deficiency in humans progress, significantly more questions (in addition to abrogates TCR-mediated NF-kB signaling (Su et al., those posed throughout the review) arise and central to 2005). Similarly, procaspase-8 deletion in murine B cells these include understanding how the two opposing also does not impair NF-kB signaling upon activation of cellular fates emanating from DED family members are various antigen receptors (Beisner et al., 2005), whereas precisely regulated and executed in the context of the same deficiency in human B cells indicates otherwise different cell types, different stages of development (Su et al., 2005). These observations and a number of and during the course of alternate environmental and others highlight the fact that there is likely to be much signaling cues. At a simple level, regulation of c-FLIP greater complexity than is appreciated in this review and protein levels (which normally has a short half-life in that the theme presented here might not apply to all cells) through signal-dependent expression and protea- lymphocytes. Nevertheless, the circumstantial evidence somal degradation is believed to be a crucial factor that in total suggests that this model involving c-FLIP might modulates the continuum between life and death signals reflect one path by which proliferative signals in from procaspase-8, and thus represents at least one lymphocytes are transmitted by procaspase-8, a protease important avenue of continued and further exploration. normally associated with death. Recent progress on this front has been made and indeed Although extensive effort regarding the proliferative a Jun kinase (JNK)-regulated E3 ubiquitin ligase named role for the core DED proteins has centered on T-cell Itch has been linked to c-FLIPL turnover (Chang et al., homeostasis, evidence supporting a similar function 2006). At another level, undoubtedly more proteins will for these proteins (particularly c-FLIPL) in cell types be added to the mixthat comprise DED-mediated outside the immune system are beginning to emerge. signaling platforms (such as TRAF1, TRAF2 and Raf1 As mentioned above, a previous study has already as mentioned above and others), and elucidating if and revealed that c-FLIPL can recruit signaling proteins how these proteins integrate with any of the core DED (including TRAF1, TRAF2, RIP1 and Raf-1) to Fas- family members to activate various signaling pathways induced DISC and activate NF-kB and ERK in human will represent another initial yet crucial step in answer- embryonic kidney cells (Kataoka et al., 2000). In the ing the posed questions. Along with these challenges, same cell line, cleaved FLIP (both the p43/p12 and the new and exciting frontiers involving DED family p22 forms) preferentially interacts with TRAF2, RIP1 members are being discovered and a few examples and IKKg (as described above, Figure 5) (Kataoka and include: induction of programmed cell death from Tschopp, 2004; Golks et al., 2006). Thus, it is highly locations outside of death receptors including unligated probable that in the context of the DISC, proteolytic integrins (Stupack et al., 2001, 2006) and the mitochon- processing of c-FLIPL by partially activated procaspase- dria (Lee et al., 2007), regulation of autophagy (and 8 will also be an important non-apoptotic signaling cue autophagic cell death) by caspase-8 and FADD (Yu in other cell types. Detailed delineation of proliferative et al., 2004; Pyo et al., 2005; Thorburn et al., 2005), signaling pathways from death receptors in other cell regulation of JNK (Nakajima et al., 2006) and WNT types is not yet available, but growing evidence is indeed signaling by c-FLIPL (Naito et al., 2004; Nakagiri et al.,

Oncogene FLIP and the death effector domain family JW Yu and Y Shi 6225 2005), and modulation of gene expression from the Acknowledgements androgen receptor (Qi et al., 2007) and p53 by procaspase-8 (Yao et al., 2007). The increasingly Work from Yigong Shi’s laboratory in this area of research complexinvolvement of DED proteins in different was supported by a grant from the National Institutes of pathways and cellular processes is indeed unfolding Health (2 R01 CA90269). Jong W Yu is a Damon Runyon from ongoing effort, and additional answers that verify Fellow supported by the Damon Runyon Cancer Research and map out these new connections lie ahead. Foundation (DRG-1905-06).

References

Abe K, Kurakin A, Mohseni-Maybodi M, Kay B, Khosravi-Far R. for caspase-8 activation and CD95-mediated apoptosis. EMBO J (2000). The complexity of TNF-related apoptosis-inducing ligand. 21: 3704–3714. Ann NY Acad Sci 926: 52–63. Chang L, Kamata H, Solinas G, Luo JL, Maeda S, Venuprasad K Alappat EC, Feig C, Boyerinas B, Volkland J, Samuels M, Murmann et al. (2006). The E3 ubiquitin ligase itch couples JNK activation to AE et al. (2005). Phosphorylation of FADD at serine 194 TNFalpha-induced cell death by inducing c-FLIP(L) turnover. Cell by CKIalpha regulates its nonapoptotic activities. Mol Cell 19: 124: 601–613. 321–332. Chau H, Wong V, Chen NJ, Huang HL, Lin WJ, Mirtsos C et al. Alcivar A, Hu S, Tang J, Yang X. (2003). DEDD and DEDD2 (2005). Cellular FLICE-inhibitory protein is required for T cell associate with caspase-8/10 and signal cell death. Oncogene 22: survival and cycling. J Exp Med 202: 405–413. 291–297. Chaudhary PM, Jasmin A, Eby MT, Hood L. (1999). Modulation of Alderson MR, Armitage RJ, Maraskovsky E, Tough TW, RouxE, the NF-kappa B pathway by virally encoded death effector Schooley K et al. (1993). Fas transduces activation signals in normal domains-containing proteins. Oncogene 18: 5738–5746. human T lymphocytes. J Exp Med 178: 2231–2235. Chun HJ, Zheng L, Ahmad M, Wang J, Speirs CK, Siegel RM et al. Algeciras-Schimnich A, Shen L, Barnhart BC, Murmann AE, (2002). Pleiotropic defects in lymphocyte activation caused by Burkhardt JK, Peter ME. (2002). Molecular ordering of the initial caspase-8 mutations lead to human immunodeficiency. Nature 419: signaling events of CD95. Mol Cell Biol 22: 207–220. 395–399. Ashkenazi A, Dixit VM. (1998). Death receptors: signaling and Condorelli G, Vigliotta G, Cafieri A, Trencia A, Andalo P, Oriente F modulation. Science 281: 1305–1308. et al. (1999). PED/PEA-15: an anti-apoptotic molecule Bagneris C, Ageichik AV, Cronin N, Wallace B, Collins M, Boshoff C that regulates FAS/TNFR1-induced apoptosis. Oncogene 18: et al. (2008). Crystal structure of a vFlip-IKKgamma complex: 4409–4415. insights into viral activation of the IKK signalosome. Mol Cell 30: Curtin JF, Cotter TG. (2003). Live and let die: regulatory mechanisms 620–631. in Fas-mediated apoptosis. Cell Signal 15: 983–992. Bang S, Jeong EJ, Kim IK, Jung YK, Kim KS. (2000). Fas- and tumor Danial NN, Korsmeyer SJ. (2004). Cell death: critical control points. necrosis factor-mediated apoptosis uses the same binding Cell 116: 205–219. surface of FADD to trigger signal transduction. A typical Desbarats J, Birge RB, Mimouni-Rongy M, Weinstein DE, Palerme model for convergent signal transduction. J Biol Chem 275: JS, Newell MK. (2003). Fas engagement induces neurite 36217–36222. growth through ERK activation and p35 upregulation. Nat Cell Bao Q, Shi Y. (2007). Apoptosome: a platform for the activation of Biol 5: 118–125. initiator caspases. Cell Death Differ 14: 56–65. Desbarats J, Newell MK. (2000). Fas engagement accelerates liver Beisner DR, Ch’en IL, Kolla RV, Hoffmann A, Hedrick SM. (2005). regeneration after partial hepatectomy. Nat Med 6: 920–923. Cutting edge: innate immunity conferred by B cells is regulated by Dohrman A, Kataoka T, Cuenin S, Russell JQ, Tschopp J, Budd RC. caspase-8. J Immunol 175: 3469–3473. (2005). Cellular FLIP (long form) regulates CD8+ T cell activation Berglund H, Olerenshaw D, Sankar A, Federwisch M, McDonald NQ, through caspase-8-dependent NF-kappa B activation. J Immunol Driscoll PC. (2000). The three-dimensional solution structure and 174: 5270–5278. dynamic properties of the human FADD death domain. J Mol Biol Donepudi M, Mac Sweeney A, Briand C, Grutter MG. (2003). Insights 302: 171–188. into the regulatory mechanism for caspase-8 activation. Mol Cell 11: Bidere N, Snow AL, Sakai K, Zheng L, Lenardo MJ. (2006). Caspase- 543–549. 8 regulation by direct interaction with TRAF6 in T cell receptor- Eberstadt M, Huang B, Chen Z, Meadows RP, Ng SC, Zheng L et al. induced NF-kappaB activation. Curr Biol 16: 1666–1671. (1998). NMR structure and mutagenesis of the FADD (Mort1) Boatright KM, Deis C, Denault JB, Sutherlin DP, Salvesen GS. death-effector domain. Nature 392: 941–945. (2004). Activation of caspases-8 and -10 by FLIP(L). Biochem J 382: Field N, Low W, Daniels M, Howell S, Daviet L, Boshoff C et al. 651–657. (2003). KSHV vFLIP binds to IKK-gamma to activate IKK. J Cell Boatright KM, Renatus M, Scott FL, Sperandio S, Shin H, Pedersen Sci 116: 3721–3728. IM et al. (2003). A unified model for apical caspase activation. Mol Golks A, Brenner D, Fritsch C, Krammer PH, Lavrik IN. (2005). Cell 11: 529–541. c-FLIPR, a new regulator of death receptor-induced apoptosis. Boissonnas A, Bonduelle O, Lucas B, Debre P, Autran B, Combadiere J Biol Chem 280: 14507–14513. B. (2002). Differential requirement of caspases during naive T cell Golks A, Brenner D, Krammer PH, Lavrik IN. (2006). The c-FLIP- proliferation. Eur J Immunol 32: 3007–3015. NH2 terminus (p22-FLIP) induces NF-kappaB activation. J Exp Carrington PE, Sandu C, Wei Y, Hill JM, Morisawa G, Huang T et al. Med 203: 1295–1305. (2006). The structure of FADD and its mode of interaction with Goltsev YV, Kovalenko AV, Arnold E, Varfolomeev EE, Brodianskii procaspase-8. Mol Cell 22: 599–610. VM, Wallach D. (1997). CASH, a novel caspase homologue with Chan FK, Chun HJ, Zheng L, Siegel RM, Bui KL, Lenardo MJ. (2000). death effector domains. J Biol Chem 272: 19641–19644. A domain in TNF receptors that mediates ligand-independent Green DR, Evan GI. (2002). A matter of life and death. Cancer Cell 1: receptor assembly and signaling. Science 288: 2351–2354. 19–30. Chang DW, Xing Z, Capacio VL, Peter ME, Yang X. (2003). Guasparri I, Wu H, Cesarman E. (2006). The KSHV oncoprotein Interdimer processing mechanism of procaspase-8 activation. vFLIP contains a TRAF-interacting motif and requires TRAF2 and EMBO J 22: 4132–4142. TRAF3 for signalling. EMBO Rep 7: 114–119. Chang DW, Xing Z, Pan Y, Algeciras-Schimnich A, Barnhart BC, Han DK, Chaudhary PM, Wright ME, Friedman C, Trask BJ, Riedel Yaish-Ohad S et al. (2002). c-FLIP(L) is a dual function regulator RT et al. (1997). MRIT, a novel death-effector domain-containing

Oncogene FLIP and the death effector domain family JW Yu and Y Shi 6226 protein, interacts with caspases and BclXL and initiates cell death. dria in response to activation of cell surface death receptors. Cell 94: Proc Natl Acad Sci USA 94: 11333–11338. 481–490. Hanahan D, Weinberg RA. (2000). The hallmarks of cancer. Cell 100: Maedler K, Fontana A, Ris F, Sergeev P, Toso C, Oberholzer J et al. 57–70. (2002). FLIP switches Fas-mediated glucose signaling in human Hill JM, Morisawa G, Kim T, Huang T, Wei Y, Werner MH. (2004). pancreatic beta cells from apoptosis to cell replication. Proc Natl Identification of an expanded binding surface on the FADD death Acad Sci USA 99: 8236–8241. domain responsible for interaction with CD95/Fas. J Biol Chem Medema JP, Scaffidi C, Kischkel FC, Shevchenko A, Mann M, 279: 1474–1481. Krammer PH et al. (1997). FLICE is activated by association with Holler N, Tardivel A, Kovacsovics-Bankowski M, Hertig S, Gaide O, the CD95 death-inducing signaling complex(DISC). EMBO J 16: Martinon F et al. (2003). Two adjacent trimeric Fas ligands are 2794–2804. required for Fas signaling and formation of a death-inducing Micheau O, Thome M, Schneider P, Holler N, Tschopp J, Nicholson signaling complex. Mol Cell Biol 23: 1428–1440. DW et al. (2002). The long form of FLIP is an activator of caspase-8 Hu S, Vincenz C, Ni J, Gentz R, Dixit VM. (1997). I-FLICE, a novel at the Fas death-inducing signaling complex. J Biol Chem 277: inhibitor of tumor necrosis factor receptor-1- and CD-95-induced 45162–45171. e-pub ahead of print 4 September 2002. apoptosis. J Biol Chem 272: 17255–17257. Misra RS, Russell JQ, Koenig A, Hinshaw-Makepeace JA, Wen R, Huang B, Eberstadt M, Olejniczak ET, Meadows RP, Fesik SW. Wang D et al. (2007). Caspase-8 and c-FLIPL associate in lipid rafts (1996). NMR structure and mutagenesis of the Fas (APO-1/CD95) with NF-kappaB adaptors during T cell activation. J Biol Chem 282: death domain. Nature 384: 638–641. 19365–19374. Inohara N, Koseki T, Hu Y, Chen S, Nunez G. (1997). CLARP, a Mitsiades CS, Poulaki V, Fanourakis G, Sozopoulos E, McMillin D, death effector domain-containing protein interacts with caspase-8 Wen Z et al. (2006). Fas signaling in thyroid carcinomas is and regulates apoptosis. Proc Natl Acad Sci USA 94: 10717–10722. diverted from apoptosis to proliferation. Clin Cancer Res 12: Irmler M, Thome M, Hahne M, Schneider P, Hofmann K, Steiner V 3705–3712. et al. (1997). Inhibition of death receptor signals by cellular FLIP. Muppidi JR, Lobito AA, Ramaswamy M, Yang JK, Wang L, Wu H Nature 388: 190–195. et al. (2006). Homotypic FADD interactions through a conserved Jeong EJ, Bang S, Lee TH, Park YI, Sim WS, Kim KS. (1999). The RXDLL motif are required for death receptor-induced apoptosis. solution structure of FADD death domain. Structural basis of Cell Death Differ 13: 1641–1650. death domain interactions of Fas and FADD. J Biol Chem 274: Muppidi JR, Siegel RM. (2004). Ligand-independent redistribution of 16337–16342. Fas (CD95) into lipid rafts mediates clonotypic T cell death. Nat Kataoka T, Budd RC, Holler N, Thome M, Martinon F, Irmler M Immunol 5: 182–189. et al. (2000). The caspase-8 inhibitor FLIP promotes activation of Naito M, Katayama R, Ishioka T, Suga A, Takubo K, Nanjo M et al. NF-kappaB and Erk signaling pathways. Curr Biol 10: 640–648. (2004). Cellular FLIP inhibits beta-catenin ubiquitylation and Kataoka T, Tschopp J. (2004). N-terminal fragment of c-FLIP(L) enhances Wnt signaling. Mol Cell Biol 24: 8418–8427. processed by caspase 8 specifically interacts with TRAF2 and Nakagiri S, Murakami A, Takada S, Akiyama T, Yonehara S. (2005). induces activation of the NF-kappaB signaling pathway. Mol Cell Viral FLIP enhances Wnt signaling downstream of stabilized Biol 24: 2627–2636. beta-catenin, leading to control of cell growth. Mol Cell Biol 25: Kaufmann M, Bozic D, Briand C, Bodmer JL, Zerbe O, Kohl A et al. 9249–9258. (2002). Identification of a basic surface area of the FADD death Nakajima A, Komazawa-Sakon S, Takekawa M, Sasazuki T, Yeh effector domain critical for apoptotic signaling. FEBS Lett 527: WC, Yagita H et al. (2006). An antiapoptotic protein, c-FLIPL, 250–254. directly binds to MKK7 and inhibits the JNK pathway. EMBO J Kennedy NJ, Kataoka T, Tschopp J, Budd RC. (1999). Caspase 25: 5549–5559. activation is required for T cell proliferation. J Exp Med 190: Ozoren N, El-Deiry WS. (2003). Cell surface death receptor signaling 1891–1896. in normal and cancer cells. Semin Cancer Biol 13: 135–147. Kischkel FC, Hellbardt S, Behrmann I, Germer M, Pawlita M, Peter ME, Budd RC, Desbarats J, Hedrick SM, Hueber AO, Newell Krammer PH et al. (1995). Cytotoxicity-dependent APO-1 (Fas/ MK et al. (2007). The CD95 receptor: apoptosis revisited. Cell 129: CD95)-associated proteins form a death-inducing signaling complex 447–450. (DISC) with the receptor. EMBO J 14: 5579–5588. Pyo JO, Jang MH, Kwon YK, Lee HJ, Jun JI, Woo HN et al. (2005). Krammer PH. (2000). CD95’s deadly mission in the immune system. Essential roles of Atg5 and FADD in autophagic cell death: Nature 407: 789–795. dissection of autophagic cell death into vacuole formation and cell Krueger A, Baumann S, Krammer PH, Kirchhoff S. (2001). FLICE- death. J Biol Chem 280: 20722–20729. inhibitory proteins: regulators of death receptor-mediated apopto- Qi W, Wu H, Yang L, Boyd DD, Wang Z. (2007). A novel function of sis. Mol Cell Biol 21: 8247–8254. caspase-8 in the regulation of androgen-receptor-driven gene Lee HJ, Pyo JO, Oh Y, Kim HJ, Hong SH, Jeon YJ et al. (2007). AK2 expression. EMBO J 26: 65–75. activates a novel apoptotic pathway through formation of a Rasper DM, Vaillancourt JP, Hadano S, Houtzager VM, Seiden I, complexwith FADD and caspase-10. Nat Cell Biol 9: 1303–1310. Keen SL et al. (1998). Cell death attenuation by ‘Usurpin’, a Lens SM, Kataoka T, Fortner KA, Tinel A, Ferrero I, MacDonald mammalian DED-caspase homologue that precludes caspase-8 RH et al. (2002). The caspase 8 inhibitor c-FLIP(L) modulates recruitment and activation by the CD-95 (Fas, APO-1) receptor T-cell receptor-induced proliferation but not activation-induced cell complex. Cell Death Differ 5: 271–288. death of lymphocytes. Mol Cell Biol 22: 5419–5433. Rathmell JC, Thompson CB. (2002). Pathways of apoptosis in Li FY, Jeffrey PD, Yu JW, Shi Y. (2006). Crystal structure of a viral lymphocyte development, homeostasis, and disease. Cell 109 FLIP: insights into FLIP-mediated inhibition of death receptor (Suppl): S97–107. signaling. J Biol Chem 281: 2960–2968. Rawlings DJ, Sommer K, Moreno-Garcia ME. (2006). The CARMA1 Li H, Zhu H, Xu CJ, Yuan J. (1998). Cleavage of BID by caspase 8 signalosome links the signalling machinery of adaptive and innate mediates the mitochondrial damage in the Fas pathway of immunity in lymphocytes. Nat Rev Immunol 6: 799–812. apoptosis. Cell 94: 491–501. Reed JC, Doctor KS, Godzik A. (2004). The domains of apoptosis: a Liu L, Eby MT, Rathore N, Sinha SK, Kumar A, Chaudhary PM. genomics perspective. Sci STKE 2004: re9. (2002). The human herpes virus 8-encoded viral FLICE inhibitory Renganathan H, Vaidyanathan H, Knapinska A, Ramos JW. (2005). protein physically associates with and persistently activates the Phosphorylation of PEA-15 switches its binding specificity from Ikappa B kinase complex. J Biol Chem 277: 13745–13751. ERK/MAPK to FADD. Biochem J 390: 729–735. Luo X, Budihardjo I, Zou H, Slaughter C, Wang X. (1998). Bid, a Bcl2 Riedl SJ, Shi Y. (2004). Molecular mechanisms of caspase regulation interacting protein, mediates cytochrome c release from mitochon- during apoptosis. Nat Rev Mol Cell Biol 5: 897–907.

Oncogene FLIP and the death effector domain family JW Yu and Y Shi 6227 Salmena L, Lemmers B, Hakem A, Matysiak-Zablocki E, Murakami Thomas LR, Henson A, Reed JC, Salsbury FR, Thorburn A. (2004). K, Au PY et al. (2003). Essential role for caspase 8 in T-cell Direct binding of Fas-associated death domain (FADD) to the homeostasis and T-cell-mediated immunity. Genes Dev 17: 883–895. tumor necrosis factor-related apoptosis-inducing ligand receptor Sandu C, Morisawa G, Wegorzewska I, Huang T, Arechiga AF, Hill DR5 is regulated by the death effector domain of FADD. J Biol JM et al. (2006). FADD self-association is required for stable Chem 279: 32780–32785. interaction with an activated death receptor. Cell Death Differ 13: Thomas LR, Stillman DJ, Thorburn A. (2002). Regulation of Fas- 2052–2061. associated death domain interactions by the death effector domain Scaffidi C, Fulda S, Srinivasan A, Friesen C, Li F, Tomaselli KJ et al. identified by a modified reverse two-hybrid screen. J Biol Chem 277: (1998). Two CD95 (APO-1/Fas) signaling pathways. EMBO J 17: 34343–34348. 1675–1687. Thorburn J, Moore F, Rao A, Barclay WW, Thomas LR, Grant KW Scaffidi C, Schmitz I, Krammer PH, Peter ME. (1999). The role of et al. (2005). Selective inactivation of a Fas-associated death domain c-FLIP in modulation of CD95-induced apoptosis. J Biol Chem 274: protein (FADD)-dependent apoptosis and autophagy pathway in 1541–1548. immortal epithelial cells. Mol Biol Cell 16: 1189–1199. Shu HB, Halpin DR, Goeddel DV. (1997). Casper is a FADD- and VauxDL, Flavell RA. (2000). Apoptosis genes and autoimmunity. caspase-related inducer of apoptosis. Immunity 6: 751–763. Curr Opin Immunol 12: 719–724. Siegel RM, Frederiksen JK, Zacharias DA, Chan FK, Johnson M, Vilimanovich U, Bumbasirevic V. (2008). TRAIL induces proliferation Lynch D et al. (2000). Fas preassociation required for apoptosis of human glioma cells by c-FLIPL-mediated activation of ERK1/2. signaling and dominant inhibition by pathogenic mutations. Science Cell Mol Life Sci 65: 814–826. 288: 2354–2357. Wajant H. (2003). Death receptors. Essays Biochem 39: 53–71. Siegel RM, Martin DA, Zheng L, Ng SY, Bertin J, Cohen J et al. Walczak H, Miller RE, Ariail K, Gliniak B, Griffith TS, Kubin M (1998). Death-effector filaments: novel cytoplasmic structures that et al. (1999). Tumoricidal activity of tumor necrosis factor-related recruit caspases and trigger apoptosis. J Cell Biol 141: 1243–1253. apoptosis-inducing ligand in vivo. Nat Med 5: 157–163. Siegel RM, Muppidi JR, Sarker M, Lobito A, Jen M, Martin D et al. Wang S, El-Deiry WS. (2003). TRAIL and apoptosis induction by (2004). SPOTS: signaling protein oligomeric transduction structures TNF-family death receptors. Oncogene 22: 8628–8633. are early mediators of death receptor-induced apoptosis at the Yang JK, Wang L, Zheng L, Wan F, Ahmed M, Lenardo MJ et al. plasma membrane. J Cell Biol 167: 735–744. (2005). Crystal structure of MC159 reveals molecular mechanism of Srinivasula SM, Ahmad M, Ottilie S, Bullrich F, Banks S, Wang Y DISC assembly and FLIP inhibition. Mol Cell 20: 939–949. et al. (1997). FLAME-1, a novel FADD-like anti-apoptotic Yao Z, Duan S, Hou D, Heese K, Wu M. (2007). Death effector molecule that regulates Fas/TNFR1-induced apoptosis. J Biol domain DEDa, a self-cleaved product of caspase-8/Mch5, translo- Chem 272: 18542–18545. cates to the nucleus by binding to ERK1/2 and upregulates Stilo R, Liguoro D, di Jeso B, Leonardi A, Vito P. (2003). The alpha- procaspase-8 expression via a p53-dependent mechanism. EMBO chain of the nascent polypeptide-associated complexbinds to and J 26: 1068–1080. regulates FADD function. Biochem Biophys Res Commun 303: Yu L, Alva A, Su H, Dutt P, Freundt E, Welsh S et al. (2004). 1034–1041. Regulation of an ATG7-beclin 1 program of autophagic cell death Stupack DG, Puente XS, Boutsaboualoy S, Storgard CM, Cheresh by caspase-8. Science 304: 1500–1502. DA. (2001). Apoptosis of adherent cells by recruitment of caspase-8 Yuan J, Yankner BA. (2000). Apoptosis in the nervous system. Nature to unligated integrins. J Cell Biol 155: 459–470. 407: 802–809. Stupack DG, Teitz T, Potter MD, Mikolon D, Houghton PJ, Kidd VJ Zhang J, Cado D, Chen A, Kabra NH, Winoto A. (1998). Fas- et al. (2006). Potentiation of neuroblastoma metastasis by loss of mediated apoptosis and activation-induced T-cell proliferation are caspase-8. Nature 439: 95–99. defective in mice lacking FADD/Mort1. Nature 392: 296–300. Su H, Bidere N, Zheng L, Cubre A, Sakai K, Dale J et al. (2005). Zhang N, He YW. (2005). An essential role for c-FLIP in the Requirement for caspase-8 in NF-kappaB activation by antigen efficient development of mature T lymphocytes. J Exp Med 202: receptor. Science 307: 1465–1468. 395–404.

Oncogene