TNF receptor family

The tumour necrosis factor receptor (TNFR)/nerve growthfactor receptor (NGFR) family of molecules regulate anumber of biological functions, such as growth, differentia-tion and apoptosis in multiple cell types. In the immunesystem, members of this receptor family are involved in thedevelopment of peripheral lymphoid organs, regulation ofinduced inflammatory responses and removal of cells at theend of an immune response.

The TNFR family consists of more than 15 different mol-ecules. Most are type I membrane proteins which resembleeach other largely in their extracellular regions, which allcontain 2–6 characteristic cysteine-rich domains.1

The TNF family receptors are activated upon binding oftheir cognate ligands, most of which are trimers with a struc-ture similar to TNF. Sometimes the ligands are cell boundtype II membrane proteins, but several are cleaved off andappear as soluble trimers. Induction of trimers or higher ordercomplexes of the TNF family of receptors allows their cyto-plasmic domains to aggregate intracytoplasmic signallingmolecules.

Signalling pathways controlled by TNF receptors

The cytoplasmic domains of the TNFR family, which aremore diverse than the extracellular portions, do not have anyintrinsic enzymatic activity, hence they signal by inducingaggregation of intracellular adaptor molecules (Fig. 1).

Death domains

The cytoplasmic domains of TNFR1 (p55), CD95 (Fas/APO-1), NGFR (p75), death receptor (DR) 3, TRAIL-R1 andTRAIL-R2 all bear a motif termed a ‘death domain’ (DD),

so-called because it is required for these receptors to transmitapoptotic signals. The DD is a protein–protein interactionmotif consisting of six alpha helices that allow two proteinswith DD to bind to each other. Structurally the DD is relatedto two other homotypic interaction domains, the death effec-tor domain (DED), and the caspase recruitment domain(CARD).2

Death domain adaptors: TRADD, FADD, RIP and RAIDD

Binding of TNF to TNFR1 induces recruitment of the DD-containing protein TRADD to the DD of TNFR1.3 Over-expression of TRADD alone also induces the TNF-regulatedresponses apoptosis and activation of the transcription factorsNF-κB and Jun kinase (JNK), presumably because TRADDprovides docking sites for downstream signalling proteins tothe receptor complex.4

Two of the proteins that TRADD recruits to the signallingcomplex also bear death domains. One of these, RIP, has anN-terminal DD and a C-terminal kinase domain. Knockoutstudies have shown that RIP is required for induction ofNFκB by TNF.5 The other, Fas-associated protein with deathdomain (FADD), has a C-terminal DD, and an N-terminalDED. The FADD is required for cell death signalling byTNFR1 and also by CD95, to which it binds directly via itsdeath domain.6–8 The DED of FADD allows it to bind to DEDin the pro-domain of caspase 8.

Through these interactions, ligation of TNFR1 or CD95can result in the formation of a death-inducing signallingcomplex, which leads to activation of caspase 8, a cell deatheffector protease. Once activated, caspase 8 cleaves and acti-vates downstream caspases, such as caspase 3, ultimatelyleading to cell death. Because cells from mice lackingcaspase 8 are resistant to death induced by TNF receptors,CD95 and DR3, apoptosis triggered by all of these receptorsmust converge on this caspase.9 However, FADD must haveother functions because FADD knockout mice die duringembryogenesis, and lymphocytes from FADD-dominant neg-ative transgenic mice do not proliferate normally in responseto T cell mitogens in vitro.10–12

Immunology and Cell Biology (1999) 77, 41–46

Special Feature

Signalling by CD95 and TNF receptors: Not only life and death

CARINA MAGNUSSON and DAVID L VAUX

Walter and Eliza Hall Institute of Medical Research, Royal Melbourne Hospital, Parkville, Victoria, Australia

Summary Members of the TNF family of receptors play important roles in normal physiology and in defence.The recent rapid progress in the understanding of the mechanisms of apoptosis has been accompanied by assump-tions that TNF family receptors such as CD95(Fas/APO-1) only have a role in regulating cell survival. While reg-ulation of cell death is one important function of TNF family receptors, they are capable of activating signaltransduction pathways that have many other effects. The present review will focus on signalling of some TNFfamily receptors in the immune system, not only for apoptosis, but also for survival or activation.

Key words: apoptosis, CD95, NF-κB, signal transduction, TNF receptors.

Correspondence: DL Vaux, Walter and Eliza Hall Institute ofMedical Research, Post Office Royal Melbourne Hospital, Parkville,Vic. 3050, Australia.

Received 26 October 1998; accepted 26 October 1998.

RIP is an adaptor protein with a C-terminal death domainthat can associate with the DD in the cytoplasmic domain ofCD95. Via TRADD, RIP can also associate with the TNFR1.4

Cells from RIP knockout mice show increased susceptibilityto TNF-mediated killing and fail to activate NF-κB inresponse to TNF.5 This indicates that RIP is required for NF-κB activation by TNF. Because RIP is a serine threoninekinase, it is likely to phosphorylate, and thereby activate,kinases that phosphorylate the inhibitor of NF-κB, IκB.13

Interestingly, RIP knockout mice also have abnormaldevelopment of lymph nodes, similar to those in lymphotoxinβ (LTβ) receptor-deficient mice.14,15 Therefore it is possiblethat RIP also takes part in signalling from these receptors.However, because the LTβR lacks a DD, if it does signal viaRIP then it must do so indirectly (see following).

Another DD-bearing adaptor molecule implicated in TNFsignalling of apoptosis is ‘RIP-associated ICH-1/CED-3-homologous protein with a death domain’ (RAIDD). In addi-tion to the DD, RAIDD has a CARD which allows it to bindto the CARD of procaspase 2.16 Overexpression of RAIDD invitro induces apoptosis, suggesting that this interaction isfunctional. However, the significance of this pathway forinduction of cell death is uncertain because neither CD95ligand (CD95L) nor TNF are able to induce apoptosis in micelacking FADD or caspase 8. In these mice, RAIDD andcaspase 2 would presumably be able to function normally.Furthermore, TNF-α was still able to induce cell death in theabsence of caspase 2.17

TRAF

The group of TNF receptor-associated factors (TRAF) inter-act with members of the TNFR family. There are to date sixTRAF proteins identified, TRAF1, TRAF2, TRAF3 (CRAF,LAP-1, CD40-bp), TRAF4 (CART1), TRAF5 and TRAF6(review18). With the exception of TRAF4, TRAF proteinsinteract with receptor molecules either directly, or indirectlythrough binding to other TRAF, or through binding toTRADD. The TNFR2 (p75), CD40, CD30 and lymphotoxin-βreceptor (LTβR) contain conserved, cytoplasmic TRAFbinding motifs and are able to bind directly to TRAF proteins.Because TRAF2 can bind to TRADD, which in turn canassociate with TNFR1, TRAF2 can indirectly participate insignalling from this receptor as well.

The TRAF molecules share similar C-terminal domains,designated the TRAF domain, which is involved inprotein–protein interactions. TRAF2, TRAF3, TRAF5 andTRAF6 also bear an N-terminal RING finger, a zinc bindingmotif found in several types of intracellular proteins.19–23

TNF receptor-associated factor proteins interact as homo-dimers or in heterodimeric complexes. For example, TRAF2binds to TRADD, the TNFR2, LTβR, CD40 or CD30 via itsC-terminal TRAF domain, probably as a heterodimericcomplex with TRAF1 or TRAF5, or as a homodimer.18,19

It has also been shown that TRAF proteins may signal fromother receptors in addition to TNFR family molecules.TRAF6, which binds to CD40, is also involved in IL-1receptor signalling through interaction with IRAK, a serine/threonine kinase that also has a DD.24 Studies of TRAF2 andTRAF3 knockout mice have shown that TRAF proteins arerequired for activation of Jun/AP-1 signalling by TNF recep-tors, and have important roles for normal development, sincethese mice die during early life.25,26

Inhibitor-of-apoptosis proteins

The inhibitor-of-apoptosis (IAP) proteins were first discov-ered as baculovirus proteins, which were able to inhibit viralinduced apoptosis in insect cells.27,28 Several homologueshave been identified in different organisms includingmammals,29–32 Drosophila,33 C. elegans and even yeasts.34

All the IAP contain one-to-three motifs termed ‘baculovirusIAP repeats’ (BIR). In addition, most contain a C-terminalRING finger and two (MIHB (cIAP1/hIAP2) and MIHC(cIAP2/hIAP1)) also contain a CARD. These two IAP pro-teins can also bind to TRAF1 and 2 and thereby becomerecruited into the TNFR complexes.

Overexpression of IAP protects cells from apoptosisinduced by a variety of different pro-apoptotic stimuli.30,31,35,36

While the mechanism of action of IAP has not been deter-mined with certainty, genetic and biochemical evidence fromDrosophila suggest that they act upstream to prevent caspaseactivation, whereas biochemistry of mammalian IAP in vitrosuggests that IAP proteins bind to activated caspases andthereby inhibit their action directly.37,38

Daxx

In some cell types in vitro, ligation of CD95 is able to acti-vate the JNK/SAPK pathway. A candidate for mediating this

C Magnusson and DL Vaux42

Figure 1 Structure and signalling from some tumour necrosisfactor receptor (TNFR) family members and their intracellularadaptor proteins. Homologous motifs that interact with each otherare shown with the same patterns. Due to limited space, allpossible interactions are not shown. Some pathways that are not established to date are indicated with question marks; see textfor discussion. death domain; death effector domain; caspase recruitment domain. JNK, Jun kinase.

v

activity is the CD95 ‘death domain-associated protein’ Daxx,which was identified in yeast two-hybrid experiments usingthe cytoplasmic tail of CD95, containing its DD, as bait.39

Daxx was also reported to bind to the death domain ofTNFR1, although Daxx itself lacks a DD. When over-expressed in cell lines, Daxx activates the JNK pathway,possibly through activation of the ASK-1.40 The precise roleof Daxx remains uncertain, because human Daxx is confinedto the nucleus, where it would not be available for CD95signalling.41 Furthermore, Daxx-stimulated apoptosis iscompletely inhibitable by Bcl-2, whereas CD95 and TNFR-stimulated death is not.

FLIP

Another factor involved in regulating signals from CD95(and possibly other DD containing receptors) is FLICE-inhibitory protein (FLIP). FLIP resembles caspase 8 becauseit has two DED and a caspase-like domain, but it does nothave any protease activity. FLIP interacts with FADD andthereby prevents FADD from activating caspase 8.42,43

NF-κB activation

Activation of NF-κB is a frequent outcome of signalling byTNFR family members. NF-κB directs transcription for alarge number of genes involved in regulation of cell growth,transcription factors, cytokines and cell surface receptors(review44). The IAP proteins, which inhibit apoptosis, havebeen suggested as targets of NF-κB activation. Transcrip-tion of IAP proteins TRAF1 and 2 was found to be depen-dent on NF-κB, and expression of these proteins was alsoable to protect NF-κB-deficient cells from TNF-α-inducedapoptosis.45 In another study, expression of the X-linked IAPprotein Xiap was found to be suppressed by overexpressionof IκBα, an inhibitor of NF-κB activity, rendering endo-thelial cells sensitive to TNF-induced cell death.46 Theseresults are in agreement with the finding that induction ofapoptosis by TNF is much greater in the presence of cyclo-heximide or actinomycin D or if NF-κB is inactivated, andare consistent with models of TNF signalling in which celldeath and protection from cell death are stimulated simul-taneously. While this explains why cells from mice lackingcomponents of NF-κB are more sensitive to TNF-inducedapoptosis, it does not explain how pathways requiring geneexpression stimulated by NF-κB manage to operate in timeto block signalling by TRADD, FADD and caspase 8, thatpresumably should be able to act more rapidly.

JNK activation

Because Jun/AP-1 are also activated by TNF it is possiblethat these pathways could influence susceptibility to celldeath. The Jun-N-terminal kinase (JNK)/stress-activatedkinase (SAPK) is part of the mitogen-activated kinase(MAPK) pathway, which leads to activation of the trans-cription factor AP-1. Knockout studies have shown thatTRAF2 is required for JNK activation in response to TNF,but is not required for NF-κB activation.25,47 It is not knownhow TRAF proteins connect with JNK, but one candidatemolecule that may link TRAF to JNK is the kinase apop-

tosis stimulating kinase-1 (ASK-1).40 However, becausethymocytes from TRAF2 knockout animals showedincreased susceptibility to TNF-induced apoptosis, TRAFare not likely to be involved in passing signals that inducecell death. In other studies, TNF-induced JNK activationwas also not found to be involved in signalling to cell death,which suggests that JNK activation may preferentiallypromote cell survival.48,49

Roles of TNFR family members in vivo

TNF receptors

TNFR1 is expressed on most cell types.50,51 It has beenregarded as the primary signalling receptor for TNF-αinflammatory responses, because TNFR1 knockout mice are resistant to endotoxic shock.52,53 In addition, it has alsobeen suggested that TNFR1 mediates signals required forproper formation of lymphoid structures such as primary B cell follicles and germinal centres, possibly through adefect in the development of follicular dendritic cells(FDC).54–57

TNFR2, like TNFR1, is widely expressed. TNFR2 bindssoluble TNF poorly, and may only be activated by membranebound TNFα.58 Unlike TNFR1 it has no cytoplasmic deathdomain, but it can directly bind TRAF1 and TRAF2, and inthis way give activation signals and/or modulator activities ofTNFR1.59–61 However, TNFR2 knockout mice have increasedresistance to TNF-α-induced cell death and tissue necro-sis,62,63 raising the possibility that some death signals aremediated also by this receptor.

Lymphotoxin-β receptor

The lymphotoxins are homo- or heterotrimeric moleculesthat bind to both TNFR (LTα3), TNFR1 (LTα2β1) or LTβR(LTα1β2) (review64). The cytoplasmic part of LTβR binds to TRAF3 and TRAF5 and these interactions have beenreported to activate NF-κB.22,65 Lymphotoxin-β receptorknockout mice have substantial defects in peripheral lym-phoid organs, because they lack lymph nodes and Peyer’spatches and fail to develop B cell follicles and germinalcentres.14 Defective development of peripheral lymphoidorgans and/or micro-anatomic compartments within organshave also been reported for LTα66–68 and LTβ69 knockoutmice. It is possible that chemokines and cell adhesionmolecules, which both may function in directing and/ormaintaining cells to a specific site, are important factors inlymphoid organogenesis. Tumour necrosis factor-α andlymphotoxins are potent inducers of adhesion moleculesand it has been reported that ligation of the LTβR inducesproduction of chemokines.64,70 Moreover, transgenic expres-sion of LTα in the pancreas induces inflammation andstructures resembling lymph nodes.71 Considering thesimilar phenotypes found in the LTβR knockout and the RIPknockout mice, one may speculate that both these proteinsparticipate in signalling pathways leading to expression ofchemokines and/or adhesion molecules, which are involvedin lymphoid organogenesis.

CD95 and TNF receptor signalling 43

CD95 (Fas/APO-1)

CD95 triggers physiological cell death in many cell types andis involved in activation-induced cell death in mature lym-phocytes. CD95-induced apoptosis requires the binding of theDD containing adaptor molecule FADD to the death domainof CD95. The importance of the CD95 or the CD95L mole-cules in regulating lymphocyte homeostasis is illustrated bylpr and gld mice, respectively (review72). Although lympho-cyte development in bone marrow and thymus is fairlynormal, these mice show massive peripheral lymphadeno-pathy, splenomegaly and autoimmune disorders when aged.Cells lacking genes for FADD and caspase 8 are just asresistant to induction of apoptosis as cells from lpr mice, indi-cating that all apoptosis signals from CD95 require FADD andcaspase 8. Because lpr and gld mice develop lymphoidaccumulations and autoimmune disease, but FADD knockout,caspase 8 knockout and crmA transgenic mice do not, theseeffects cannot be due to loss of cell death triggered by CD95,but must be due to some other activity signalled by it.10,73,74

The first suggestion that CD95 may also mediate activa-tion or mitogenic signals came from experiments usingmonoclonal antibodies against CD95 on human T lympho-cytes or thymocytes in vitro.75,76 That FADD does not onlyfunction in death-stimulating pathways was shown by studiesusing chimeric Rag-1 knockout/FADD knockout mice andmice carrying a dominant negative interfering FADD trans-gene, which lacks the DED and therefore is unable totransduce a death signal.11,12 In these animals, thymocytesand mature T lymphocytes were resistant to CD95-mediatedkilling. However, thymocyte numbers were reduced andnegative selection appeared to be unaffected by disruption ofFADD-mediated cell death. In addition, mature T cells wereunable to proliferate normally in response to mitogens invitro. While these results show that FADD must participate insome mitogenic pathways, they did not indicate whether thesignals stemmed from CD95, TNFR1 or some other recep-tors. Interestingly, the effect of FADD deficiency on B lym-phocyte development appears to be even more severe than forT cells, because an almost complete absence of peripheral B cells was seen in chimeric Rag-1 knockout/FADD knock-out mice.

It is possible that in quiescent T cells, signals transmittedby FADD are directed towards an activation/proliferationpathway. Consistent with this, resting lymphocytes arerelatively resistant to CD95-induced apoptosis, while IL-2-activated T lymphocytes become susceptible.77 It is possiblethat activation (or transformation) of lymphocytes shifts the CD95 signalling pathways towards induction of celldeath, for example if levels of FLIP decline. Indeed, it wasrecently shown that IL-2 treatment suppresses transcriptionof FLIP.78

Conclusion

Tumour necrosis factor was first identified as a cytokine thatinduced cell death; namely necrosis of tumour cells. CD95was first found as the target of monoclonal antibodies,chosen because they could induce apoptosis in lymphoidtumour lines. Death domains received their name becausethey were found on the cytoplasmic tails of both CD95 and

TNFR1, and were required for them to transmit death signals.However, it is now clear that CD95 and TNFR1 and othermembers of the TNFR family can also stimulate other cellu-lar responses, such as proliferation, activation or even lym-phoid organogenesis. Similarly FADD, which contains onlyDD and DED, has important roles in addition to the activationof apoptosis. Other molecules, such as IRAK, myd88 andNF-κB p100 bear bonafide death domains79 and yet appar-ently do not play a role in induction of cell death. There is anatural tendency towards simplification, and in the absenceof contradictory data, the simplest explanation is the one tochoose. Despite our wishes, proteins resist simple behavioursand simple classifications, and the recent knowledge aboutthe TNFR family of molecules reminds us not to stick toover-simplistic explanations.

Acknowledgements

This work was supported by the NH&MRC and WEHI RegKey 973002. CM is supported by the Swedish MedicalResearch Council.

References

1 Beutler B, van Huffel C. Unravelling function in the TNF ligandand receptor families. Science 1994; 264: 667–8.

2 Chou JJ, Matsuo H, Duan H, Wagner G. Solution structure of theRAIDD CARD and model for CARD/CARD interaction incaspase-2 and caspase-9 recruitment. Cell 1998; 94: 171–80.

3 Hsu H, Xiong J, Goeddel DV. The TNF receptor 1-associatedprotein TRADD signals cell death and NF-kappa B activation.Cell 1995; 81: 495–504.

4 Hsu H, Huang J, Shu HB, Baichwal V, Goeddel DV. TNF-dependent recruitment of the protein kinase RIP to the TNFreceptor-1 signaling complex. Immunity 1996; 4: 387–96.

5 Kelliher MA, Grimm S, Ishida Y, Kuo F, Stanger BZ, Leder P.The death domain kinase RIP mediates the TNF-induced NF-kappaB signal. Immunity 1998; 8: 297–303.

6 Chinnaiyan AM, O’Rourke K, Tewari M, Dixit VM. FADD, anovel death domain-containing protein, interacts with the deathdomain of Fas and initiates apoptosis. Cell 1995; 81: 505–12.

7 Boldin MP, Goncharov TM, Goltsev YV, Wallach D. Involve-ment of MACH, a novel MORT1/FADD-interacting protease, inFas/APO-1- and TNF receptor-induced cell death. Cell 1996; 85:803–15.

8 Muzio M, Chinnaiyan AM, Kischkel FC et al. FLICE, a novelFADD-homologous ICE/CED-3-like protease, is recruited to theCD95 (Fas/APO-1) death-inducing signaling complex. Cell1996; 85: 817–27.

9 Varfolomeev EE, Schuchmann M, Luria V et al. Targeteddisruption of the mouse caspase 8 gene ablates cell death induc-tion by the TNF receptors, Fas/Apo1, and DR3 and is lethalprenatally. Immunity 1998; 9: 267–76.

10 Yeh WC, Pompa JL, McCurrach ME et al. FADD. essential forembryo development and signaling from some, but not all,inducers of apoptosis. Science 1998; 279: 1954–8.

11 Zhang J, Cado D, Chen A, Kabra NH, Winoto A. Fas-mediatedapoptosis and activation-induced T-cell proliferation are defec-tive in mice lacking FADD/Mort1. Nature 1998; 392: 296–300.

12 Newton K, Harris AW, Bath ML, Smith KGC, Strasser A. Adominant interfering mutant of FADD/MORT1 enhances dele-tion of autoreactive thymocytes and inhibits proliferation ofmature T lymphocytes. EMBO J. 1998; 17: 706–18.

C Magnusson and DL Vaux44

13 Stanger BZ, Leder P, Lee TH, Kim E, Seed B. RIP: A novelprotein containing a death domain that interacts with Fas/APO-1(CD95) in yeast and causes cell death. Cell 1995; 81: 513–23.

14 Futterer A, Mink K, Luz A, Kosco-Vilbois MH, Pfeffer K. Thelymphotoxin b receptor controls organogenesis and affinitymaturation in peripheral lymphoid tissues. Immunity 1998; 9:59–70.

15 Rennert PD, James D, Mackay F, Browning JL, Hochman PS.Lymph node genesis is induced by signaling through the lym-photoxin b receptor. Immunity 1998; 9: 71–9.

16 Duan H, Dixit VM. RAIDD is a new ‘death’ adaptor molecule.Nature 1997; 385: 86–9.

17 Bergeron L, Perez GI, Macdonald G et al. Defects in regulationof apoptosis in caspase-2-deficient mice. Genes Dev. 1998; 12:1304–14.

18 Arch RH, Gedrich RW, Thompson CH. Tumor necrosis factorreceptor-associated factors (TRAFs): A family of adapter pro-teins that regulate life and death. Genes Dev. 1998; 12: 2821–30.

19 Rothe M, Wong SC, Henzel WJ, Goeddel DV. A novel family ofputative signal transducers associated with the cytoplasmicdomain of the 75 kDa tumor necrosis factor receptor. Cell 1994;78: 681–92.

20 Cheng G, Cleary AM, Ye ZS, Hong DI, Lederman S, BaltimoreD. Involvement of CRAF1, a relative of TRAF, in CD40 signal-ing. Science 1995; 267: 1494–8.

21 Ishida TK, Tojo T, Aoki T et al. TRAF5, a novel tumor necrosisfactor receptor-associated factor family protein, mediates CD40signaling. Proc. Natl Acad. Sci. USA 1996; 93: 9437–42.

22 Nakano H, Oshima H, Chung W et al. TRAF5, an activator ofNF-kappaB and putative signal transducer for the lymphotoxin-beta receptor. J. Biol. Chem. 1996; 271: 14 661–4.

23 Ishida T, Mizushima S, Azuma S et al. Identification of TRAF6,a novel tumor necrosis factor receptor-associated factor proteinthat mediates signaling from an amino-terminal domain of theCD40 cytoplasmic region. J. Biol. Chem. 1996; 271: 28 745–8.

24 Cao Z, Xiong J, Takeuchi M, Kurama T, Goeddel DV. TRAF6 isa signal transducer for interleukin-1. Nature 1996; 383: 443–6.

25 Yeh WC, Shahinian A, Speiser D et al. Early lethality, functionalNF-kappaB activation, and increased sensitivity to TNF-inducedcell death in TRAF2-deficient mice. Immunity 1997; 7: 715–25.

26 Xu Y, Cheng G, Baltimore D. Targeted disruption of TRAF3leads to postnatal lethality and defective T-dependent immuneresponses. Immunity 1996; 5: 407–15.

27 Crook NE, Clem RJ, Miller LK. An apoptosis inhibitingbaculovirus gene with a zinc finger like motif. J. Virol. 1993; 67:2168–74.

28 Birnbaum MJ, Clem RJ, Miller LK. An apoptosis inhibitinggene from a nuclear polyhedrosis virus encoding a polypeptidewith cys/his sequence motif. J. Virol. 1994; 68: 2521–8.

29 Rothe M, Pan MG, Henzel WJ, Ayres TM, Goeddel DV. TheTNFR2-TRAF signaling complex contains two novel proteinsrelated to baculoviral inhibitor of apoptosis proteins. Cell 1995;83: 1243–52.

30 Uren AG, Pakusch M, Hawkins CJ, Puls KL, Vaux DL. Cloningand expression of apoptosis inhibitory protein homologs thatfunction to inhibit apoptosis and/or bind tumor necrosis factorreceptor-associated factors. Proc. Natl Acad. Sci. USA 1996; 93:4974–8.

31 Liston P, Roy N, Tamai K et al. Suppression of apoptosis inmammalian cells by NAIP and a related family of IAP genes.Nature 1996; 379: 349–53.

32 Ambrosini G, Adida C, Altieri DC. A novel anti-apoptosis gene,survivin, expressed in cancer and lymphoma. Nat. Med. 1997; 3:917–21.

33 Hay BA, Wassarman DA, Rubin GM. Drosophila homologs ofbaculovirus inhibitor of apoptosis proteins function to block celldeath. Cell 1995; 83: 1253–62.

34 Uren AG, Coulson EJ, Vaux DL. Conservation of baculovirusinhibitor of apoptosis repeat proteins (BIRPs) in viruses, nema-todes, vertebrates and yeasts. Trends Biochem. Sci. 1998; 23:159–62.

35 Hawkins CJ, Uren AG, Hacker G, Medcalf RL, Vaux DL. Inhi-bition of interleukin 1 beta-converting enzyme-mediatedapoptosis of mammalian cells by baculovirus IAP. Proc. NatlAcad. Sci. USA 1996; 93: 13 786–90.

36 Vucic D, Kaiser WJ, Miller LK. Inhibitor of apoptosis proteinsphysically interact with and block apoptosis induced byDrosophila proteins HID and GRIM. Mol. Cell. Biol. 1998; 18:3300–9.

37 Deveraux QL, Takahashi R, Salvesen GS, Reed JC. X-linkedIAP is a direct inhibitor of cell-death proteases. Nature 1997;388: 300–4.

38 Roy N, Deveraux QL, Takahashi R, Salvesen GS, Reed JC. Thec-IAP-1 and c-IAP-2 proteins are direct inhibitors of specificcaspases. EMBO J. 1997; 16: 6914–25.

39 Yang XL, Khosravifar R, Chang HY, Baltimore D. Daxx, a novelfas-binding protein that activates JNK and apoptosis. Cell 1997;89: 1067–76.

40 Chang HY, Nishitoh H, Yang X, Ichijo H, Baltimore D. Activa-tion of apoptosis signal-regulating kinase 1 (ASK1) by theadapter protein Daxx. Science 1998; 281: 1860–3.

41 Pluta AF, Earnshaw WC, Goldberg IG. Interphase-specificassociation of intrinsic centromere protein CENP-C withHDaxx. A death domain-binding protein implicated in Fas-mediated cell death. J. Cell Sci. 1998; 111: 2029–41.

42 Irmler M, Thome M, Hahne M et al. Inhibition of death recep-tor signals by cellular FLIP. Nature 1997; 388: 190–5.

43 Shu HB, Halpin DR, Goeddel DV. Casper is a FADD- andcaspase-related inducer of apoptosis. Immunity 1997; 6: 751–63.

44 Baeuerle PA, Henkel T. Function and activation of NF-kB in theimmune system. Annu. Rev. Immunol. 1994; 12: 141–79.

45 Wang C-Y, Mayo MW, Korneluk RG, Goeddel DV, Baldwin JrAS. NF-κB antiapoptosis: Induction of TRAF1 and TRAF2 andc-IAP1 and c-IAP2 to supress caspase-8 activation. Science1998; 281: 1680–3.

46 Stehlik C, de Martin R, Kumabashiri I, Schmid JA, Binder BR,Lipp J. Nuclear factor (NF)-kB-regulated X-chromosome-linkedIAP gene expression protects endothelial cells from Tumor Necro-sis Factor α-induced apoptosis. J. Exp. Med. 1998; 188: 211–16.

47 Lee SY, Reichlin A, Santana A, Sokol KA, Nussenzweig MC,Choi Y. TRAF2 is essential for JNK but not NF-kappaB activa-tion and regulates lymphocyte proliferation and survival.Immunity 1997; 7: 703–13.

48 Liu ZG, Hsu H, Goeddel DV, Karin M. Dissection of TNF recep-tor 1 effector functions: JNK activation is not linked to apoptosiswhile NF-kappaB activation prevents cell death. Cell 1996; 87:565–76.

49 Natoli G, Costanzo A, Ianni A et al. Activation of SAPK/JNK byTNF receptor 1 through a noncytotoxic TRAF2-dependentpathway. Science 1997; 275: 200–3.

50 Loetscher H, Pan YC, Lahm HW et al. Molecular cloning andexpression of the human 55 kD tumor necrosis factor receptor.Cell 1990; 61: 351–9.

51 Schall TJ, Lewis M, Koller KJ et al. Molecular cloning andexpression of a receptor for human tumor necrosis factor. Cell1990; 61: 361–70.

52 Rothe J, Lesslauer W, Lotscher H et al. Mice lacking the tumournecrosis factor receptor 1 are resistant to TNF-mediated toxicity

CD95 and TNF receptor signalling 45

C Magnusson and DL Vaux46

but highly susceptible to infection by Listeria monocytogenes.Nature 1993; 364: 798–802.

53 Pfeffer K, Matsuyama T, Kundig TM et al. Mice deficient forthe 55 kD tumor necrosis factor receptor are resistant to endo-toxic shock, yet succumb to L. monocytogenes infection. Cell1993; 73: 457–67.

54 Pasparakis M, Alexopoulou L, Episkopou V, Kollias G. Immuneand inflammatory responses in TNF alpha-deficient mice: A critical requirement for TNF alpha in the formation ofprimary B cell follicles, follicular dendritic cell networks andgerminal centers, and in the maturation of the humoral immuneresponse. J. Exp. Med. 1996; 184: 1397–411.

55 Pasparakis M, Alexopoulou L, Grell M, Pfizenmaier K, Blueth-mann H, Kollias G. Peyer’s patch organogenesis is intact yetformation of B lymphocyte follicles is defective in peripherallymphoid organs of mice deficient for tumor necrosis factor andits 55-kDa receptor. Proc. Natl Acad. Sci. USA 1997; 94:6319–23.

56 Le Hir M, Bluethmann H, Kosco VM et al. Differentiation offollicular dendritic cells and full antibody responses requiretumor necrosis factor receptor-1 signaling. J. Exp. Med. 1996;183: 2367–72.

57 Tkachuk M, Bolliger S, Ryffel B et al. Crucial role of tumornecrosis factor receptor 1 expression on nonhematopoietic cellsfor B cell localization within the splenic white pulp. J. Exp.Med. 1998; 187: 469–77.

58 Grell M, Douni E, Wajant H et al. The transmembrane form oftumor necrosis factor is the prime activating ligand of the 80kDa tumor necrosis factor receptor. Cell 1995; 83: 793–802.

59 Tartaglia LA, Pennica D, Goeddel DV. Ligand passing: The 75-kDa tumor necrosis factor (TNF) receptor recruits TNF forsignaling by the 55-kDa TNF receptor. J. Biol. Chem. 1993; 268:18 542–8.

60 Peschon JJ, Torrance DS, Stocking KL et al. TNF receptor-deficient mice reveal divergent roles for p55 and p75 in severalmodels of inflammation. J. Immunol. 1998; 160: 943–52.

61 Grell M, Becke FM, Wajant H, Mannel DN, Scheurich P. TNFreceptor type 2 mediates thymocyte proliferation independentlyof TNF receptor type 1. Eur. J. Immunol. 1998; 28: 257–63.

62 Erickson SL, de Sauvage FJ, Kikly K et al. Decreased sensitiv-ity to tumour-necrosis factor but normal T-cell development inTNF receptor-2-deficient mice. Nature 1994; 372: 560–3.

63 Sheehan KC, Pinckard JK, Arthur CD, Dehner LP, Goeddel DV,Schreiber RD. Monoclonal antibodies specific for murine p55and p75 tumor necrosis factor receptors: Identification of anovel in vivo role for p75. J. Exp. Med. 1995; 181: 607–17.

64 von Boehmer H. Lymphotoxins: From cytotoxicity to lymphoidorganogenesis. Proc. Natl Acad. Sci. USA 1997; 94: 8926–7.

65 VanArsdale TL, VanArsdale SL, Force WR et al. Lymphotoxin-beta receptor signaling complex: Role of tumor necrosis factor

receptor-associated factor 3 recruitment in cell death andactivation of nuclear factor kappaB. Proc. Natl Acad. Sci. USA1997; 94: 2460–5.

66 De Togni P, Goellner J, Ruddle NH et al. Abnormal developmentof peripheral lymphoid organs in mice deficient in lymphotoxin.Science 1994; 264: 703–7.

67 Matsumoto M, Mariathasan S, Nahm MH, Baranyay F, PeschonJJ, Chaplin DD. Role of lymphotoxin and the type I TNF recep-tor in the formation of germinal centers. Science 1996; 271:1289–91.

68 Korner H, Cook M, Riminton DS et al. Distinct roles forlymphotoxin-alpha and tumor necrosis factor in organogenesisand spatial organization of lymphoid tissue. Eur. J. Immunol.1997; 27: 2600–9.

69 Koni PA, Sacca R, Lawton P, Browning JL, Ruddle NH, FlavellRA. Distinct roles in lymphoid organogenesis for lymphotoxinsalpha and beta revealed in lymphotoxin beta-deficient mice.Immunity 1997; 6: 491–500.

70 Degli-Esposti MA, Davis-Smith T, Din WS, Smolak PJ, GoodwinRG, Smith CA. Activation of the lymphotoxin beta receptor bycross-linking induces chemokine production and growth arrest inA375 melanoma cells. J. Immunol. 1997; 158: 1756–62.

71 Kratz A, Campos-Neto A, Hanson MS, Ruddle NH. Chronicinflammation caused by lymphotoxin is lymphoid neogenesis.J. Exp. Med. 1996; 183: 1461–72.

72 Nagata S. Apoptosis by death factor. Cell 1997; 88: 355–65.73 Varfolomeev EE, Boldin MP, Goncharov TM, Wallach D. A

potential mechanism of ‘cross-talk’ between the p55 tumornecrosis factor receptor and Fas/APO1: Proteins binding to thedeath domains of the two receptors also bind to each other.J. Exp. Med. 1996; 183: 1271–5.

74 Smith KG, Strasser A, Vaux DL. CrmA expression in T lympho-cytes of transgenic mice inhibits CD95 (Fas/APO-1)-transducedapoptosis, but does not cause lymphadenopathy or autoimmunedisease. EMBO J. 1996; 15: 5167–76.

75 Alderson MR, Tough TW, Davis ST et al. Fas ligand mediatesactivation-induced cell death in human T lymphocytes. J. Exp.Med. 1995; 181: 71–7.

76 Alderson MR, Tough TW, Braddy S et al. Regulation of apopto-sis and T cell activation by Fas-specific mAb. Int. Immunol.1994; 6: 1799–806.

77 Lenardo MJ. Interleukin-2 programs mouse alpha beta T lym-phocytes for apoptosis. Nature 1991; 353: 858–61.

78 Refaeli Y, Van Parijs L, London CA, Tschopp J, Abbas AK.Biochemical mechanisms of IL-2-regulated Fas-mediated T cellapoptosis. Immunity 1998; 8: 615–23.

79 Feinstein E, Kimchi A, Wallach D, Boldin M, Varfolomeev E.The death domain: A module shared by proteins with diversecellular functions. Trends Biochem. Sci. 1995; 20: 342–4.