dependence receptors: a new paradigm in cell signaling and ......differentiation, migration or...

REVIEW

Dependence receptors: a new paradigm in cell signaling and cancer therapy

D Goldschneider and P Mehlen

Apoptosis, Cancer and Development Laboratory- Equipe labellisee ‘La Ligue’, CNRS UMR5238, Centre Leon Berard, Universite deLyon, Lyon, France

Dependence receptors (DRs) now form a family of morethan a dozen membrane receptors that are not linked bytheir structure, but by common functional traits. The mostnotable is their ability to trigger two opposite signalingpathways: in the presence of ligand, these receptorsactivate classic signaling pathways implicated in cellsurvival, migration and differentiation. In the absence ofligand, they do not stay inactive, rather they elicit anapoptotic signal. Thus, cells expressing this kind ofreceptor are dependent on the presence of ligand in theextracellular environment to survive. This review willrecapitulate the increasing data regarding the molecularmechanisms associated with DRs, their potential implica-tion during development, as well as their deregulationduring tumorigenesis and, finally, their emergence as newpossible therapeutic targets for cancer treatment.Oncogene advance online publication, 22 February 2010;doi:10.1038/onc.2010.13

Keywords: dependence receptors; apoptosis; caspase;tumor progression; cancer therapy

Membrane receptors are classically considered as inactiveunless bound to their ligand. However, increasing observa-tions demonstrate that some receptors, in addition to their‘positive’ signaling when their ligand is present, transduce a‘negative’ signal that induces apoptosis in the absence ofligand (Figure 1). Cells expressing these receptors are thusdependent on the presence of ligand to survive. Thesereceptors are named ‘dependence receptors.’ To date, thedependence receptor (DR) family is composed of morethan a dozen members including DCC (deleted in color-ectal carcinoma) (Mehlen et al., 1998), UNC5Hs (un-coordinated 5 homologs), neogenin (Matsunaga et al.,2004), p75NTR (p75 neurotrophin receptor) (Rabizadehet al., 1993), RET (rearranged during transfection)(Bordeaux et al., 2000), TrkC (tyrosine kinase receptor C)(Tauszig-Delamasure et al., 2007), Ptc (patched) (Thibertet al., 2003), EphA4 (ephrin type A receptor 4) (Furneet al., 2009), ALK (anaplastic lymphoma kinase) (Mouraliet al., 2006), MET (Tulasne et al., 2004) and some integrins

(Stupack et al., 2001). All of them are involved in bothnervous system development and cancer progression.

Dependence receptors: a short history

Neurotrophin receptor p75NTR was the first DR to bedescribed. P75NTR was discovered as one of tworeceptors able to bind nerve growth factor (Chaoet al., 1986), the other being TrkA (Kaplan et al.,1991). TrkA was rapidly shown to mediate the knownresponses to NGF, such as neurite outgrowth andneuronal survival (Lee et al., 2001a; Lykissas et al.,2007), whereas the precise biological role of p75NTR

remained misunderstood. p75NTR was shown to collabo-rate with TrkA to form high-affinity sites for NGFbinding (Hempstead et al., 1991). In addition, p75NTR

was shown to alter the ligand specificity of other Trkreceptors. For example, brain-derived neurotrophicfactor, NT3 and NT4/5 can all bind TrkB in the absenceof p75NTR, whereas only brain-derived neurotrophicfactor does so in the presence of p75NTR. In contrast,coexpression of p75NTR with TrkC results in a relaxationin its absolute specificity for NT3 (Hempstead, 2002). Atthe time of its discovery, p75NTR was considered as aunique type of protein but, subsequently, a largesuperfamily of tumor necrosis factor (TNF) receptorswere found to share the overall structure of p75NTR

(Liepinsh et al., 1997). Identification of this superfamilyhelped elucidate some of the biological functions ofp75NTR, including its link to cell death regulation. Therelationship between these TNF death receptors, whichinduce cell death on binding of proapoptotic ligand suchas TNF or FasL, and p75NTR, which binds NGF, atrophic factor known to induce cell survival, led DEBredesen and colleagues to propose that p75NTR inducescell death when unoccupied by NGF, whereas bindingof NGF blocks apoptosis (Rabizadeh et al., 1993)(Figure 2 and Table 1). This finding suggested thatp75NTR expression creates a state of cellular dependenceon NGF. Further studies with knockout mice confirmedthis notion. First, p75NTR-deficient mice have anincreased number of cholinergic neurons, somal hyper-trophy and hyperinnervation in some areas of thehippocampus (Yeo et al., 1997; Naumann et al., 2002).In addition, crossing NGF hemizygous mice, whichdisplay a reduction in cholinergic cell numbers, withp75NTR null mice showed that loss of p75NTR partiallyrestores cholinergic cell numbers (Naumann et al.,

Received 1 October 2009; revised 2 January 2010; accepted 6 January2010

Correspondence: Dr P Mehlen, Apoptosis, Cancer and DevelopmentLaboratory- Equipe labellisee ‘La Ligue’, CNRS UMR5238, CentreLeon Berard, Universite de Lyon, 28 rue Laennec, Lyon, Rhone 69008,France.E-mail: [email protected]

Oncogene (2010), 1–18& 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 $32.00

www.nature.com/onc

http://dx.doi.org/10.1038/onc.2010.13

http://www.nature.com/onc

2002). However, the overall picture of p75NTR functionbecame more complicated when some studies showedthat p75NTR induced apoptosis in response to ligand

binding rather than ligand withdrawal (Casaccia-Bon-nefil et al., 1996; Frade et al., 1996). Particularly, inaddition to its ability to bind mature neurotrophins,

Figure 2 Representation of the DR family. The functional domains present in extra and intracellular domains are represented. DRsare not related to each other according to their structure, but according to their ability to induce apoptosis in the absence of ligand. Allof them are caspase substrates, except p75NTR and integrins. The position of caspase cleavage sites is indicated. Localization of ADD,which has been more or less precisely determined depending on receptor, is indicated by double arrows.

Table 1 DRs and their known ligands

Receptors Ligands

p75NTR NGF, proNGF, BDNF, NT-3, NT-4/5,b-amyloid, prion

DCC netrin-1, netrin-4Neogenin netrin-1, RGMa, RGMb, RGMc, netrin-3, netrin-4Unc50s netrin-1, netrin-4RET GDNF, neurturin, artemin, persephinPtc ShhTrkC NT-3EphA4 ephrinB3, ephrinA1, ephrinA4ALK pleiotrophin, midkin, jebMET HGFAR androgensIntegrin avb3and a5b1

extracellular matrix

The above listed are the DRs and their known ligands. In case ofmultiple ligands, those which were shown to block apoptotic functionare underlined. ALK is a particular case: its putative ligands have notyet been tested for blocking apoptosis, only agonist antibodies havebeen used.

Survival,Migration,

DifferentiationAPOPTOSIS

Figure 1 The DR model. DRs have two faces: in the presence oftheir respective ligand, they transduce a positive signal ofdifferentiation, migration or survival, whereas in the absence oftheir ligand, they do not stay inactive but, rather, induce apoptosis.Thus, cells expressing such receptors are dependent on ligandavailability for survival.

A new paradigm in cell signaling and cancer therapyD Goldschneider and P Mehlen

2

Oncogene

other cell-death-inducing ligands have been proposedfor p75NTR, such as pro-NGF (Lee et al., 2001b),b-amyloid (Yaar et al., 2002) and prion (Della-Biancaet al., 2001) peptides. The decision between ligand-induced apoptosis and ligand-inhibited apoptosismediated by p75NTR likely depends on cell type anddevelopment stage (Barrett and Bartlett, 1994). Becausethe idea of a receptor triggering apoptosis whenunbound to its ligand contradicted the dogma regardingreceptor signaling and the trophic theory, consideringp75NTR as a classic death receptor had more success thanconsidering it as a DR. Consequently, DR p75NTR hasbeen rather forgotten, although it was the first to bedescribed and no definitive evidence has demonstratedthat it is not a DR.

The concept reemerged with DCC. DCC was dis-covered in 1990 as a putative cell-surface receptorencoded by a gene frequently deleted through allelicloss in colorectal carcinoma (Fearon et al., 1990).Observation that DCC expression is reduced or lost incolorectal cancer led to the proposal that DCCexpression represented a constraint for disease progres-sion and is therefore a tumor suppressor. This hypo-thesis was supported by the fact that DCC expression islost or reduced in various cancers (Mehlen and Fearon,2004) and that its loss of expression is associated withpoor prognosis (Shibata et al., 1996; Sun et al., 1999). Inaddition, restoration of DCC expression can suppresstumorigenic property in vitro and in nude mice(Klingelhutz et al., 1993; Velcich et al., 1999). TheDCC extracellular domain shares structural featureswith certain types of cell-adhesion molecules, such asNCAM (Cho et al., 1994) (Figure 2 and Table 1), but itsintracellular domain shows little similarity with knownproteins; hence in spite of the above-mentioned studieson cancers, little was known about the precise biologicalrole of DCC. It was rediscovered as the receptor fornetrin-1 (Keino-Masu et al., 1996), a diffusible moleculeoriginally identified as a chemoattractant for commis-sural axons in the vertebral spinal cord (Serafini et al.,1994). The classic view for netrin-1 is that a gradient ofthis cue diffuses from a ventral spinal cord structure, thefloor plate, and orients the growth of commissural axonsas they extend circumferentially toward the ventralmidline of the embryonic nervous system. The key roleof DCC and netrin-1 in mediating axon outgrowth andpathfinding is supported by a large number of studies,particularly the analysis of DCC and netrin-1 knockoutmice, which display similar defects in the central nervoussystem (Serafini et al., 1996; Fazeli et al., 1997). Such adual role for a receptor, implicated during developmentand functioning as a putative tumor suppressor, seemsto be a common trait for DRs. DCC was thus proposedas a DR when it was shown that its expression in variouscancer cell lines that lacked endogenous DCC expres-sion induced cell death and that addition of netrin-1blocked apoptosis (Mehlen et al., 1998; Chen et al.,1999).

Netrin-1 is in fact the founding member of a family ofextracellular proteins found throughout the animalkingdom and that direct cell and axon migration during

embryogenesis. In vertebrates, besides netrin-1, fourother netrins have been described: netrin-2/3/2like,netrin-4/b, netrin-G1 and netrin-G2 (Puschel, 1999;Mehlen and Mazelin, 2003). Netrins are structurallyrelated to the short arm of laminin (g for netrin 1–3 andb for netrin-4, G1 and G2). Netrins 1–4 are secreted,whereas netrin-G1 and G2 are membrane anchored bymeans of a glycophosphatidylinol tail. Secreted netrinsexert their biological functions by binding to receptorssuch as DCC, UNC5, neogenin and DSCAM, whereasnetrin Gs do not interact with these receptors (Rajase-kharan and Kennedy, 2009). Interestingly, netrin-5 hasrecently appeared in databases, which seems to berelated to the netrin 1–3 group according to its sequence.Netrin-1 is the most studied member of the netrin familyand to date it seems to be the main ligand for DCC, aswell as for UNC5 receptors (see below), although arecent report mentioned that netrin-4 could interact withDCC and UNC5H1 (Qin et al., 2007).

The UNC5 receptor was initially identified inCaenorhabditis elegans as an axonal guidance trans-membrane receptor (Leung-Hagesteijn et al., 1992) and,on the basis of a genetic screen, was predicted to interactwith UNC6 (the C. elegans netrin-1 ortholog) (Hedge-cock et al., 1990). Four homologs of UNC5 have beendescribed in mammals (UNC5H1, 2, 3 and 4 in rodentsand UNC5A, B, C and D in humans) (Leonardo et al.,1997). Although DCC alone is implicated in thechemoattractive effect of netrin-1, it has been proposedthat UNC5, associated with DCCs through theirintracellular domains, is responsible for the repulsiveeffect of netrin-1 (Hong et al., 1999). Besides this role,UNC5 receptors are now known to have critical roles inother cellular processes, such as neuronal migration(Mehlen and Furne, 2005) and embryonic angiogenesis(Lu et al., 2004). Interestingly, UNC5 proteins contain adeath domain related to the death domain of the TNFreceptor superfamily in their cytoplasmic part. AsUNC5 receptors were netrin-1 receptors and containeda death domain, it was suggested that they were possibleDRs (Figure 2 and Table 1). Along this line, UNC5receptors are able to induce apoptosis in the absence ofnetrin-1, whereas addition of ligand efficiently blockedthis effect (Llambi et al., 2001; Wang et al., 2008).

A DCC homolog was discovered and called neogenin(Vielmetter et al., 1994; Meyerhardt et al., 1997).Although there is only one member of the DCC receptorfamily in C. elegans and Drosophila (UNC40 andfrazzled, respectively), vertebrates have evolved twoclosely related orthologs, DCC and neogenin. Owing toits identity of sequence with DCCs, especially in theirectodomain, neogenin was initially considered to be anetrin-1 receptor as well. However, more recently, thepropensity of netrin-1 to function as a ligand forneogenin was challenged when neogenin seemed to havemuch higher affinity for RGM (repulsive guidancemolecule), another guidance molecule, and to mediateits repulsive effect (Rajagopalan et al., 2004). RGM wasfirst identified as a repulsive, membrane-bound cueresponsible for the mapping of temporal retinal axons tothe posterior region of the chick tectum (Monnier et al.,


3

Oncogene

2002). In mammals, three RGMs exist: RGMa, theclosest ortholog of chick RGM, RGMb/DRAGON andRGMc/hemojuvelin. RGMa and b are both expressed inthe central nervous system and follow a complementaryexpression pattern (Niederkofler et al., 2004), whereasRGMc is mainly expressed in striated muscles and liver(Schmidtmer and Engelkamp, 2004). Neogenin is a DRin that it can induce apoptosis when overexpressed inchick neural tube, and its ligand, RGM, but not netrin-1, counteracts this process (Matsunaga et al., 2004)(Figure 2 and Table 1). It should be noted that thequestion of neogenin’s ligand remains tricky: despite itshigher affinity for RGM, neogenin seems to be able tomediate netrin-1 signaling in axon attraction (Wilsonand Key, 2006) or cell adhesion (Srinivasan et al., 2003).Moreover, neogenin has also been proposed to mediatenetrin-3 signaling during myotube formation (Kanget al., 2004), as well as netrin-4 signaling in angiogenesis(Lejmi et al., 2008).

Patched (Ptc) is a 12-transmembrane receptor that ispart of the complex responsible for sonic hedgehog(Shh) morphogen signaling (Figure 2 and Table 1). Shhbinds Ptc and thus abolishes the Ptc-repressing effect onsmoothened (Smo). Smo is a seven-transmembranereceptor that activates downstream Gli transcriptionfactors (Murone et al., 1999). Shh is a glycoproteinsecreted by the notochord and floor plate duringdevelopment, after a ventro-dorsal concentration gra-dient in the ventral neural tube. This gradient deter-mines the induction and specification of ventral neuronsin the vertebrate neural tube (Jessell, 2000). In additionto its morphogen activity, Shh was also shown to be asurvival factor: indeed, Le Douarin and colleaguesdiscovered that experimental withdrawal of Shh inchick embryos by partial destruction of the notochordleads to massive cell death in the developing neural tube(Charrier et al., 1999, 2001). It was subsequently demon-strated that Shh functions as a survival factor byinhibiting the apoptotic function of Ptc (Thibert et al.,2003). Thus, Ptc is a DR and can signal independentlyof Smo.

Integrins are the main receptors that mediate cellularinteractions with extracellular matrix ligands such aslaminins, collagens and fibronectins (Hood and Cher-esh, 2002). They are heterodimeric (ab) type I trans-membrane receptors, and provide a connection betweenthe matrix and the cytoskeleton. Integrins have tradi-tionally been considered as prosurvival receptors, on thebasis of the concept of ‘anchorage dependence’ (Stupackand Cheresh, 2002). Integrin-mediated adhesion sup-ports the formation of cytoskeletal and contractileelements, promotes cellular resistance to exogenousapoptotic stimuli and facilitates signaling by trophicfactor receptors. Most cells require integrin-mediatedadhesion to respond to trophic factors. This has led tothe proposal that controlling cell adhesion and geome-try, thereby permitting responsiveness to survivalfactors, may be the critical function of integrins inmaintaining cell viability. However, expression ofcertain b3 or b1 integrins can also induce apoptosis, ifimmobilized substrates are not available as ligands.

These ‘non-liganded integrins,’ which are either un-ligated or occupied with a soluble antagonist, not onlydisrupt survival signaling but also actively induceapoptosis, hence supporting the view of integrins asDRs (Stupack et al., 2001) (Figure 2 and Table 1).

Beside these receptors, some classical tyrosine kinasereceptors have also emerged as DRs. RET was the firstone (Figure 2 and Table 1). RET is the signalingcomponent of a multisubunit complex that functions asa receptor for glial cell line-derived neurotrophic factor(Jing et al., 1996), neurturin, artemin and persephin(Kotzbauer et al., 1996), four homologous neurotrophicfactors related to the transforming growth factor-bfamily. The receptor complex also includes (GPI)-anchored proteins GFRa1, 2, 3 and 4 that are requiredfor RET dimerization and dictate ligand selectivity(Baloh et al., 2000; Scott and Ibanez, 2001). Afterinteraction with its ligands, RET undergoes autopho-sphorylation and then interacts with multiple effectorssuch as phospholipase C, Shc, enigma, Grb2, Grb7/Grb10, Src kinase and Ras-GAP (Santoro et al., 1994;Arighi et al., 1997; Lorenzo et al., 1997). Gain-of-function mutations of the RET gene have beenassociated with multiple endocrine neoplasia type 2(MEN 2), an autosomal dominant inherited cancersyndrome (Mulligan et al., 1993), whereas loss-of-function mutations of RET have been associated withHirschsprung disease (aganglionosis, HSCR), a frequentcongenital intestinal malformation (1 in 5000 live births)characterized by the absence of neural crest-derivedparasympathetic neurons of the hindgut (Edery et al.,1994; Romeo et al., 1994). In vitro, MEN 2-associatedmutations lead to ligand-independent constitutive acti-vation of RET kinase activity either through covalentdimerization of the receptor (MEN 2A) (Santoro et al.,1995) or through direct structural changes in its kinasedomains (MEN 2B) (Songyang et al., 1995). In contrast,the mechanisms leading to the absence of intramuralganglion cells of the hindgut observed in HSCR remainincompletely understood. The observation that RET isinvolved in both cancer progression and nervous systemdevelopment, similar to previously identified DRs, led tothe question as to whether it could also be one of them.It was then shown that, in different settings, expressionof RET induced apoptosis in the absence, but not in thepresence, of glial cell line-derived neurotrophic factor(Bordeaux et al., 2000; Canibano et al., 2007).

Trk receptors (TrkA, B and C) are the mainneurotrophin receptors. TrkA is the receptor for NGF,TrkB is the receptor for brain-derived neurotrophicfactor and NT4/5, whereas TrkC is the receptor for NT3(Kaplan and Miller, 2000). The classic neurotrophictheory proposes that neuronal survival depends on thepresence of trophic factors such as neurotrophins (Levi-Montalcini and Angeletti, 1963; Huang and Reichardt,2001), and that cell death, which occurs when thesefactors become limited, is strictly due to loss of survivalsignals. On neurotrophin binding, Trk receptors activatePI3K/Akt and Ras/MAP kinase pathways that arethought to inhibit the naturally occurring death pro-gram in neurons (Kaplan and Miller, 2000). However, in


4

Oncogene

light of what was known about DRs, it was tempting tohypothesize that neurotrophin binding served also toblock an active apoptotic signal from Trks. Interest-ingly, when Trk receptors were evaluated as possibleDRs, TrkC was the only one the enforced expression ofwhich induced cell death in the absence of ligand(Tauszig-Delamasure et al., 2007). Thus, TrkC is a DR,whereas TrkA and B are not, suggesting that even aclosely related receptor can acquire a different activitywith regard to cell survival and apoptosis (Figure 2 andTable 1).

Anaplastic lymphoma kinase (ALK) tyrosine kinase isa member of the insulin receptor superfamily (Figure 2and Table 1). It was initially identified as part of theoncogenic nucleophosmin–ALK fusion protein resultingfrom the t(2;5) translocation that is frequently asso-ciated with anaplastic large-cell lymphoma (Morriset al., 1994). Nucleophosmin allows dimerization ofthe fusion protein, causing constitutive activation ofALK kinase and downstream activation of phospholi-pase C-g, PI3K, STATs and pp60c-src (Allouche, 2007).The native full-length ALK receptor is mainly expressedin discrete regions of the developing central andperipheral nervous system (Iwahara et al., 1997).Mourali et al. (2006) forced ALK expression in cells oflymphoid and neuronal origin to investigate wild-typeALK functions. They observed that ALK enhanced ortriggered apoptosis in these cells and that treatment withagonist antibodies mimicking ALK ligand prevented celldeath induction (Mourali et al., 2006).

Eph receptors constitute the largest family of tyrosinekinase receptors and bind ligands called ephrins(Figure 2 and Table 1). Eph receptors regulate a diversearray of cellular processes during development, such asaxon guidance, angiogenesis, or cell migration andpositioning (Pasquale, 2005). More recently, someephrin and Eph receptors have been found to affect celldeath in neurogenic regions. Activation of EphA7 byoverexpressed ephrinA5 in the embryonic cortex re-sulted in neural progenitor apoptosis (Depaepe et al.,2005). On the other hand, lack of ephrinB3 is associatedwith apoptosis in the subventricular zone of adult mice(Ricard et al., 2006), suggesting that its receptor couldfunction as a DR. EphA4 has been shown to function asa DR, the apoptotic activity of which is impaired by itsligand ephrinB3 (Furne et al., 2009).

Another tyrosine kinase receptor should probably beadded to the list: MET, the receptor for hepatocytegrowth factor (Figure 2 and Table 1). MET is wellknown for its essential role in normal development andcell survival. Interestingly, it was reported that METwas cleaved under stress conditions by caspase, therebygenerating an apoptotic fragment (Tulasne et al., 2004).As we will see below, caspase cleavage is one of the mostcommon traits for apoptotic signaling of DRs. More-over, hepatocyte growth factor inhibits this caspasecleavage and concomitantly apoptosis. Therefore,although it is not clear whether MET is able to initiateapoptosis by itself in the absence of ligand, as otherDRs do, this receptor presents some striking similaritieswith DRs.

To close this list, it should be mentioned that anontransmembrane receptor, androgen receptor (AR),has also been proposed to be a DR (Ellerby et al., 1999).AR is a member of the nuclear receptor superfamily ofligand-activated transcription factors. Binding of andro-gens such as testosterone by the AR leads to nucleartranslocation and transcriptional activity. Gene regula-tion by the AR affects widespread processes such asmale gonadal development, cell survival and musculardevelopment, among many others (Lee and Chang,2003). AR displays a profile similar to that of membraneDRs; it is a caspase substrate and its expression inducesapoptosis in the absence of ligand, whereas the additionof ligand inhibits receptor-induced cell death (Ellerbyet al., 1999). Mutations in the AR are associated withboth prostate cancer and neurodegeneration. Neurode-generation-associated mutants give rise to Kennedy’sdisease, a syndrome associated with the degeneration ofmotor neurons in the brainstem and spinal cord,resulting in weakness and muscular atrophy. Interest-ingly, mutations associated with neurodegenerationconsist in expansion (430) of polyglutamine tractspresent in the N-terminal part of the receptor, whereasshortened tracts (p22) are associated with a greaterrisk of developing prostate cancer (Nelson and Witte,2002; Clark et al., 2003). AR, similar to p75NTR, has beenquite forgotten as a DR. Paradoxically, the word‘dependence’ is frequently associated with AR in thefield of prostate cancer. Indeed, studies have led to theconcept that prostate secretory epithelial cells requiretestosterone for survival, and the withdrawal oftestosterone leads to apoptosis in these cells (Craft andSawyers, 1998). Thus, neoplastic prostate epithelial cellsare often treated by hormone deprivation because itleads to apoptosis as a result of their dependence ontestosterone. It is quite regrettable that no link has everbeen made between the dependency phenomenon ofprostate cancer cells and the fact that AR could behaveas a DR.

DRs are caspase substrates

The molecular hallmark of programmed cell death(apoptosis) is the activation of caspases (Thornberryand Lazebnik, 1998). During apoptosis, caspases, whichform a family of cysteine-dependent aspartate-directedproteases, can cleave a wide range of substrates, therebyinactivating survival and activating proapoptotic me-chanisms. Except for the fact that they induce apoptosisin the absence of their ligand, the other most commoncharacteristic of DR is that they are cleaved by caspases.Receptor cleavage is important for apoptotic function,as mutation of the cleavage site abolishes cell deathinduction. DCC, neogenin, Ptc, ALK, EphA4 arecleaved once, roughly in the middle of their intracellulardomain (Mehlen et al., 1998; Thibert et al., 2003;Matsunaga et al., 2004; Mourali et al., 2006; Furneet al., 2009). The cleavage site of UNC5 receptors is veryclose to the plasma membrane (Llambi et al., 2001),whereas RET, Trkc and MET have two cleavage sites


5

Oncogene

(Bordeaux et al., 2000; Foveau et al., 2007; Tauszig-Delamasure et al., 2007). The cases of p75NTR andintegrins are less clear, as there is no solid evidence thatcaspase cleavage is needed for their proapoptotic effect.Interestingly, caspase cleavage sites of DRs seem to beconserved in mammals, variably in other vertebrates butnever in orthologs in lower organisms, such as nematodeor Drosophila (Table 2) (Mehlen and Thibert, 2004).These findings may suggest that the appearance as acaspase substrate, and therefore the mediation of thedependence state, is a relatively late event in theevolution of these proteins. This may make sense, giventhe greater plasticity of the mammalian nervous systemcompared with those of invertebrates and the necessityfor more complex and higher lifespan mammals todevelop antitumor mechanisms.

Role of ligand binding

Another common feature of DRs is the inhibition ofapoptosis induction on ligand binding. The mainhypothesis with regard to the role of ligand binding isinhibition of the caspase cleavage. This hypothesis hasbeen partly confirmed for some DRs, UNC5H2(Tanikawa et al., 2003), TrkC (Tauszig-Delamasureet al., 2007), EphA4 (Furne et al., 2009) and MET(Tulasne et al., 2004; Foveau et al., 2007). For otherreceptors, the effect of ligand on caspase cleavagecannot be assessed because cleavage products haverelatively short half-lives and are thus hardly detectablein vivo. On the other hand, ectopic expression of atruncated receptor, mimicking the caspase cleavedreceptor, leads to apoptosis induction even in thepresence of ligand (Mehlen et al., 1998; Thibert et al.,2003; Matsunaga et al., 2004). This is an indirectargument in favor of a role for the ligand in caspase

cleavage inhibition. In addition, ligand binding has beenproposed to have other structural effects, such asreceptor multimerization. In fact, with the exception ofPtc and integrins, most DRs display homo-multimeriza-tion properties in the presence of their respective ligand(Wang et al., 2000; Manie et al., 2001; Stein et al., 2001;Mille et al., 2009a). Initially described as beingimportant for positive signaling, receptor multimeriza-tion also seems to have a role in blocking apoptosisinduction. P75NTR exerts its proapoptotic effect as amonomer, whereas multimerization abrogates this effect(Rabizadeh et al., 2000). In the same way, resultsobtained with DCC and UNC5H2 showed that thesereceptors trigger cell death when their ligand-inducedmultimerization is hindered (Mille et al., 2009a). Netrin-1 has also been proposed to suppress the apoptoticfunction of UNC5H2 by inducing interaction of thereceptor with the GTPase PIKE-L (Tang et al., 2008).This interaction triggers the activation of PI3 kinasesignaling and consequently the inhibition of UNC5H2proapoptotic function. Furthermore, recent data fromcristallography evidenced that the UNC5H2 intra-cellular domain adopts an autoinhibited closed con-formation. In this conformation, ZU5 and deathdomains bind to each other and are thus unable toinduce cell death. Netrin-1 is unable to preventapoptosis induced by the UNC5H2 mutant that has aconstitutive open conformation, which leads the authorsto suggest that netrin-1 somehow stabilizes the auto-inhibited conformation of UNC5H2 (Wang et al., 2009).

It must be noted that, for receptors accepting morethan one ligand, the antiapoptotic effect has not beendemonstrated for all ligands. For example, only glial cellline-derived neurotrophic factor was shown to blockRET-induced apoptosis (Bordeaux et al., 2000), no dataare available for neurturin, artemin and persephin,although neurturin, similar to glial cell line-derivedneurotrophic factor, was recently proposed to be a

Table 2 Conservation of caspase cleavage sites of DR among species

Homosapiens

Pantroglodytes

Bostaurus

Canisfamiliaris

Musmusculus

Rattusnorvegicus

Gallusgallus

Xenopustropicalis

Xenopuslaevis

Daniorerio

DCC LSVD LSVD NA LSVD LSVD LSVD NA NA LTVD NoUNC5B DITD DITD DITD DITD DITD DITD DITD NA DITD EITDNeogenin CCTD CCTD CCTD CCTD CCTD CCTD PCAD? GPED? NA CTTDPtc PETD PETD PETD PETD PETD PETD NEDD? HEND? HEND? NoRET VSVD VSVD VSVD VSVD VPVD VSVD VSVD NA MSVD VAID

DYLD DYLD DYLD DYLD DYLD DYLD DYLD DYLD DYLDTrkC SSLD SSLD SSLD SSLD SSLD SSLD SSLD NA NA NA

ILVD ILVD ILVD ILVD ILVD ILVD ILVDEphA4 LEDD LEDD LEDD LEDD LEDD LEDD LEDD LEDD LEDD LEEDALK DELD DELD DELD DELD DELD DELD DELD DEMD NA DELDMET ESVD ESVD ESVD ESVD ESVD ESVD ESVD ESVD ESVD ESVD

DNAD DNAD No DNID DNID DNID DNTD No No SNLD?

Caspase cleavage sites of DR have been mapped by experiment on human or rodent proteins. This table lists these sites for various DRs and theamino acids found at the same position in sequence from other species. Caspase cleavage sites are well conserved in mammals and variablyconserved in other vertebrates: for example, cleavage site of EphA4 is better conserved than those of neogenin or patched. In some cases, sequenceis not perfectly conserved but only a slight variation is observed between species (amino acid is shown in bold). In other cases an aspartate residue isfound at the correct position, but the surrounding residues are not conserved (shown in italic). No: means that no asparte residue has been found bysequence alignment. NA: means not analyzed or not found in databases. Alignments have been performed using sequence from Ensembl or NCBIdatabases.


6

Oncogene

survival factor for parasympathetic neurons (Peterzielet al., 2007). In the case of neogenin, the ligandantiapoptotic effect was reported only for chick RGM/RGMa (Matsunaga et al., 2004) but not for othermembers of the RGM family or for netrin-1/3/4. EphA4-associated cell death induction is hindered by ephrinB3,but not by its other known ligands, ephrinA1 andephrinA4 (Furne et al., 2009). With regard to ALK, onlyligand-mimicking antibodies have been used because theidentity of the receptor’s physiological ligand is still amatter of debate. Pleiotrophin and midkine, two heparin-binding growth factors, have been proposed to functionas ALK ligand but these proteins fail to make a generalconsensus. This controversy is further supported by thegenetic identification in Drosophila melanogaster of anALK ligand, jelly belly (jeb), which has no similaritieswith pleiotrophin or midkine (Englund et al., 2003; Leeet al., 2003; Loren et al., 2003).

Finally, in the particular case of integrins, the ligandmust be an immobilized substrate ligand to blockapoptosis as, contrary to other DRs, soluble ligandis not sufficient to counteract cell death induction(Stupack, 2005).

Proapoptotic signaling by DRs

DRs share the property of being caspase amplifiers;indeed most of them fail to induce apoptosis in thepresence of general caspase inhibitors such as zVADfmk(Mehlen and Thibert, 2004). The way that leads tocaspase activation and amplification has begun to bedecrypted specifically for some of them. First, all DRscontain, in their intracellular part, a domain required forapoptosis induction. This domain, the ADD (addiction/dependence domain) (Bredesen et al., 1998), is requiredand often sufficient for cell death induction. Caspasecleavage, except for p75NTR and integrins, is thought to beresponsible for unmasking the ADD. In most cases,ADD is borne by the remaining membrane-anchoredfragment. In the case of UNC5H, RET, TrkC and METreceptors, however, it is the cytosolic generated fragmentthat is proapoptotic. ADDs are usually unique regionsthat are not structurally related to known proteinfunctional domains. Two notable exceptions are p75NTR

and UNC5H receptors: in those two cases, two regionscorresponding to known functional domains are respon-sible for apoptosis induction. The first region is theirdeath domain, which is structurally related to the deathdomain of receptors of the TNF receptor superfamily(Hofmann and Tschopp, 1995; Bredesen et al., 1998;Llambi et al., 2001). The second region is the chopperdomain for p75NTR (Coulson et al., 2000) and the ZU5domain for UNC5H (Williams et al., 2003).

After ADD release/exposition, caspase amplificationseems to be more or less direct, depending on receptors.In some cases, ADD recruits caspase-activating com-plexes that are different from those implicated in deathreceptors and in intrinsic mitochondrial classicalapoptotic pathways. For example, in the absence of

netrin-1, DCC recruits and activates caspase 9, therebyallowing caspase 3 activation, but this process does notrequire cytochrome c release and subsequent formationof an apoptosome (cytochrome c/apaf-1/caspase 9)complex, as is the case in the classic mitochondrialpathway (Forcet et al., 2001). DCC does not interactdirectly with caspase 9, hence it may recruit one or moreadaptor proteins (Figure 3). One of them could beDIP13a (DCC-interacting protein 13-a), a proteinidentified as an interactor of DCC ADD, and shownto be important for DCC-induced cell death (Liu et al.,2002). However, the precise role of DIP13a in DCC-triggered apoptosis remains quite obscure, as it does notseem to mediate interaction of DCC with caspase 9 andfurther studies performed on DIP13a, also known asAPPL1 (adaptor protein containing PH domain, PTBdomain and leucine zipper motif 1), have not providedclear evidence for a role of this protein in apoptosisinduction. In addition, palmitoylation and lipid raftlocalization were reported to be a prerequisite for DCCproapoptotic activity, both in vitro and in primarycommissural neurons (Furne et al., 2006). Lipid rafts areordered membrane microdomains enriched in sphingo-lipids and cholesterol, and are proposed to have animportant role in cell signaling, in particular through theorganization of surface receptors, signaling enzymes andadaptor molecules into complexes at specific sites in themembrane (Simons and Toomre, 2000; Hueber, 2003). Itwas then shown that DCC presence in lipid rafts isrequired to allow caspase-9/DCC interaction, suggestingthat this caspase-activating complex occurs in and takesadvantage of lipid rafts.

The discovery of a caspase-activating complex re-cruited to a DR was recently made for Patched. Ptc wasindeed found to interact through its ADD, and only inthe absence of its ligand Shh, with DRAL/FHL2 (Milleet al., 2009b). DRAL was already known to promoteapoptosis through an unknown mechanism in a widevariety of cells when overexpressed (Scholl et al., 2000)and to interact with TUCAN, a CARD-containingadaptor protein for caspases 1 and 9 (Stilo et al., 2002).Mille and colleagues showed that the Ptc–DRALassociation serves as a platform for recruiting TUCAN(and/or NALP1, a protein closely related to TUCAN)and caspase 9 (Figure 4). This then allows caspase 9recruitment to Ptc and caspase 9 activation. Thecomplex involving DRAL, TUCAN and caspase 9 wasnamed dependosome, by analogy to other knowncaspase-activating complexes such as the apoptosome(comprising Apaf-1, cytochrome c and caspase 9), DISC(comprising Fas, FADD and caspase 8), the PIDDo-some (comprising PIDD, RAID and caspase 2) and theinflammosome (comprising NALPs, ASC, caspases 1and 5). Additional studies are required to determinewhether this dependosome is a common platform forother DRs.

In any case, if such a dependosome existed and werecommon to DRs, the initiator recruited caspase couldnot always be caspase 9. Indeed, Stupack et al. (2001)elegantly showed that integrins, as DRs, triggerapoptosis through recruitment of caspase 8.


7

Oncogene

The formation of a caspase-activating complex doesnot seem to be the only mechanism used by DRs totrigger apoptosis. As an example, the mechanism ofUNC5H-induced apoptosis has been documented.Despite their structural homology, members of theUNC5H family seem to mediate their apoptotic signalby interacting preferentially with distinct partners. In

fact, apoptotic pathways downstream of UNC5H3 and4 have not yet been documented, and functional data areavailable only for UNC5H1 and 2. UNC5H1 inducesapoptosis by interacting with NRAGE (neurotrophinreceptor-interacting MAGE homolog) through its ZU5domain (Williams et al., 2003). NRAGE possiblytransduces UNC5H1-induced apoptosis through at least

Netrin-1

Cleavage

Caspase 3

CASPASE9

Apoptosis

DIP13ααX

Amplification ofreceptor cleavage

Caspase 3activation

Figure 3 Model of cell death induction by DCC. In the presence of netrin-1, DCC is dimerized and interacts with procaspase 3.Following ligand withdrawal, DCC becomes a monomer and is cleaved, possibly by bound caspase 3 or another activated protease.Cleavage leads to ADD exposure and to its direct interaction with apoptotic partners such as DIP13a, the role of which remainsunclear, or to indirect interaction with caspase 9. Consequently, these interactions lead to caspase 9 activation, which in turn activatescaspase 3.

Cleavage

Apoptosis

Shh

Shh


Caspase 3activation

Figure 4 Model of cell death induction by Ptc. Following ligand withdrawal, Ptc is cleaved by caspase (or by another activatedprotease), thus allowing exposure of its ADD. Ptc recruits DRAL, which in turns recruits TUCAN (or NALP1) and caspase 9.Formation of this complex leads to caspase 9 activation and consequently to caspase 3 activation.


8

Oncogene

two pathways: one implicating the degradation of thecaspase inhibitor XIAP, and the other implicatingactivation of the proapoptotic JNK signaling pathway.NRAGE can also interact with UNC5H2 and 3, butwith a much weaker binding affinity. On the other hand,UNC5H2 triggers apoptosis mainly by recruiting theserine/threonine death-associated protein kinase(DAPK) (Llambi et al., 2005). UNC5H2 and DAPKbind each other in part through their respective deathdomain, but not only so, as deletion of these domains isnot sufficient to abrogate their association. Surprisingly,although UNC5H2-mediated induction of DAPK ac-tivity is observed only in the absence of netrin-1 andrequires UNC5H2 caspase cleavage, DAPK seems tointeract constitutively with UNC5H2, that is, not onlyin the absence of netrin-1. Indeed, DAPK is known tobe capable of autophosphorylation, which inhibitsits activity by inducing a conformational change(Shohat et al., 2001). On this basis, and according totheir observations, Llambi and colleagues proposedthat, in the presence of netrin-1, DAPK is in an inactiveautophosphorylated state, and it interacts with

UNC5H2 through their non-‘death domain’-interactingregions, whereas, in the absence of netrin-1, DAPKinteracts with the death domain of UNC5H2,which allows DAPK activation (Figure 5). This hypoth-esis strongly fits with the more recent study byWang et al. (2009), arguing that UNC5H2 can adopt aclosed conformation, preventing association of thedeath domain with other proteins. Downstreameffectors of DAPK in UNC5H2-mediated cell deathremain to be identified, but DAPK is already known totrigger cell death through p53-dependent and -indepen-dent mechanisms. Moreover, phosphorylation of themyosin light chain by DAPK leads to membraneblebbing, a hallmark of programmed cell death(Gozuacik and Kimchi, 2006). Interestingly, it wasrecently suggested that UNC5H2 was not the only DRto trigger apoptosis through DAPK, as neogenin wasalso shown to interact with DAPK and to requireDAPK for apoptosis induction (Fujita et al., 2008), inthe same way that DCC, UNC5H1, 2 and 3 also requirelipid raft association to induce cell death (Maisse et al.,2008).

P

DD

P

DAPK

DD

Closedconformation

Unc5H2

Netrin-1


Cleavage

DAPK

DAPK

ZU5

DD

ZU5 ZU5

ZU5

Caspase 3activation

Apoptosis

Figure 5 Model of cell death induction by UNC5H2/UNC5B. In the presence of netrin-1, UNC5H2 and UNC5B are dimerized, andtheir intracellular domain adopts a close conformation in which ZU5 and death domains interact with each other. DAPK interactswith the closed intracellular domain, but is in an inactive autophosphorylated state. Following ligand withdrawal, UNC5H2 receptorbecomes a monomer, whereas the intracellular domain undergoes both caspase (or another protease) cleavage and opening/dissociation of ZU5 and DD in parallel. Relation and chronology between cleavage and opening remain unclear, but thesemodifications should allow interaction between the DD of UNC5H and DAPK, thus leading to DAPK activation and initiation of anapoptotic program. Precisely how activated DAPK induces apoptosis remains to be determined.


9

Oncogene

The fact that DRs are at the same time caspasesubstrates and caspase activators/amplifiers points out aparadox. How can a receptor that requires caspasecleavage to be a proapoptotic molecule participate inapoptosis induction? One possibility could be theinitiation of DR cleavage by a noncaspase protease,which would be sufficient to generate a caspaseamplification loop. Another view could be that caspasesare never completely inactive, even in nonapoptoticcells, and that this residual caspase activity is sufficientto detect receptors disengaged from their ligand.Interestingly, caspase 3 was observed to interact withDCC, downstream of its cleavage site, only in non-apoptotic conditions, that is, in the presence of netrin-1,or when the DCC caspase cleavage site is mutated(Forcet et al., 2001). It is tempting to speculate that,when dimerized in the presence of its ligand, DCCadopts a conformation that prevents its cleavage bycaspases, whereas it can be efficiently cleaved as amonomer in the absence of ligand. This event results inunmasking its ADD, thus allowing capase 9 activationand caspase 3 amplification.

Role of DRs during embryonic development

Before being identified as DRs, all members of thisfunctional family were already known to be implicatedin nervous system development. Implications of theirpositive signaling in the presence of their ligand arelargely documented, whereas a role for negativeproapoptotic signaling long remained essentially spec-ulative. However, several lines of evidence now accu-mulate to further support the view that DRs participatein nervous system development through their proapop-totic function as well. Because of this ability to triggerapoptosis in settings of ligand absence or limitation,they were hypothesized to control cell numbers in somespecific areas of the developing brain and to dictateadequate territories of neuron migration and axonprojection by eliminating those localized out of regionsof ligand availability. For example, netrin-1 receptors donot only mediate the chemotropic effect of netrin-1 inthe developing nervous system but also seem to regulatethe survival of olivary neurons, as these cells, known toexpress DCC and UNC5H receptors, display increasedapoptosis in netrin-1�/� mice (Llambi et al., 2001).Moreover, netrin-1 functions as a survival factor forspinal cord commissural neurons, which was shown inboth primary neuron cultures and animal models (Furneet al., 2008).

Similarly, it was shown that Shh, the ligand of DR Ptc,is not only a morphogen but also a survival factor(Charrier et al., 1999, 2001). Interestingly, it wasdemonstrated that Shh functions as a survival factor byinhibiting the apoptotic function of Ptc and that thisproapoptotic function is crucial for adequate neural tubedevelopment (Thibert et al., 2003; Mille et al., 2009b).

The recent observation that neurotrophin receptorTrkC is a DR also led to questioning the implication of

TrkC-induced apoptosis in the classic neurotrophintheory. The classic dogma suggests that each neuron ismoving toward death unless a survival/neurotrophinsignal is provided. The integration of the DR notionwithin this neurotrophin theory would then be that eachneuron is actively pushed toward apoptosis by a DR,such as TrkC, in the context of ligand limitation,whereas ligand presence not only activates survivalsignals but also blocks the active process of cell death.The idea of TrkC being a DR is particularly attractivewhile analyzing data from knockout mice for neuro-trophins and their respective receptors. Indeed, inactiva-tion of TrkA or NGF in mice results in the same amountof sensory neuron loss at birth (that is, nociceptiveneurons) (Crowley et al., 1994). Similarly, inactivationof either TrkB or brain-derived neurotrophic factorresults in an equivalent loss of mechanoceptive neurons(Ernfors et al., 1994; Minichiello et al., 1995). On theother hand, neonates invalidated for TrkC present a lossof 30% DRG neurons, whereas NT-3�/� neonates havelost 70% of them (Tessarollo et al., 1994, 1997). This isin agreement with the view that TrkC, contrary to TrkAand B, can trigger, when deprived of ligand, an activeapoptotic signaling. A confirmation that TrkC apopto-tic signaling controls the fate of sensory neurons isprovided by an experience in which microinjection insensory neurons of a mutated intracellular domain ofTrkC, known to interfere with TrkC apoptotic function,dramatically enhanced survival of NT-3-deprived neu-rons (Tauszig-Delamasure et al., 2007).

The ligand/DR pair ephrinB3/EphA4 was alsorecently demonstrated, by studying knockout mice, tohave an important role in regulating cell number in anadult mouse subventricular zone through apoptosismodulation. Indeed, extinction of eprhinB3 results inincreased apoptosis in subventricular zone, whereas theabsence of EphA4 results in an excessive number ofneuroblasts in this zone (Furne et al., 2009). In addition,infusion of soluble ephrinB3 into the lateral ventriclereduced cell death. Thus, it is tempting to speculate thatEphA4, as a DR, is important in regulating the fate ofneuronal stem cells during brain development.

An elegant study from Palmer and colleaguesevidenced a role for the ALK/jeb pair in the Drosophi-la-developing visual system. ALK is expressed andrequired in target neurons in the optic lobe, whereasjeb is primarily generated by photoreceptor axons andfunctions in the eye to control target selection of specificphotoreceptor cell axons. Interestingly, the level ofneuronal cell death (measured by active caspase 3 level)in the ALK expressing optic lobe medulla increases inmutants lacking an expression of jeb. Moreover,caspase-dependent neuronal apoptosis dramaticallydecreases in mutants overexpressing jeb (Bazigouet al., 2007). These results suggest that ALK could havea role in the physiological negative selection of neuronsshaping the optic lobe in the Drosophila nervous systemby favoring apoptosis in the absence of the ALK ligand.

Interestingly, even though the data reported so far onthe role of DRs during development seem to be linked tonervous system development, it has recently been


10

Oncogene

proposed that these DRs may also be involved outsidethe developing nervous system. Along this line, netrin-1was recently shown to control the survival of endothelialcells and to promote angiogenesis, at least in part, byblocking apoptosis triggered by its unbound UNC5H2receptor (Castets et al., 2009). The DR activity ofUNC5H2 can indeed conciliate conflicting resultsregarding the implication of netrin-1 in angiogenesis(Lu et al., 2004; Wilson et al., 2006).

DRs are altered during tumor progression

In addition to a role during embryonic development, theDR model also predicts a role for such receptors astumor suppressors, because of their ability to promotecell death when disengaged from their ligand. A tumorcell submitted to an abnormal environment (highlyproliferative cells in an environment with limited andconstant ligand concentration or metastatic cells migrat-ing to sites in which ligand is absent) would displayunbound DRs and thus undergo apoptosis (Figure 6).This mechanism would represent an alternative safe-guard mechanism to limit tumor progression. It is thenexpected that in aggressive tumors, tumor cells have toturndown this DR pathway. Consistent with this view, aloss of receptor expression would then represent aselective advantage for tumor cells and seem to be aprimary method to overcome this safeguard mechanism.

Since its discovery in the 1990s, DCC has beensuspected to be a tumor suppressor gene, even thoughno definitive evidence has been proposed so far. Located

on chromosome 18q, DCC is submitted to loss ofheterozygosity in over 70% of colorectal cancers(Fearon et al., 1990). DCC is also submitted to loss ofheterozygosity and/or to decreased expression in variousother cancers including gastric, prostate, endometrial,ovarian, esophageal, breast and testicular cancer, as wellas neuroblastoma and hematological malignancies(Mehlen and Fearon, 2004). Loss of DCC expressionis often associated with poor prognosis and advancedcancer or metastasis (Shibata et al., 1996; Saito et al.,1999), suggesting a role of DCC loss in cancerprogression rather than in cancer initiation. Moreover,restoration of DCC expression can suppress tumorigenicgrowth properties in vitro or in nude mice (Klingelhutzet al., 1993; Velcich et al., 1999; Kato et al., 2000;Rodrigues et al., 2007). On the other hand, the fact thatonly 10–15% of colon cancers carry mutations in DCCand the lack of a tumor predisposing effect of DCCinactivation in mouse models (Fazeli et al., 1997) ledsome investigators to conclude that DCC had little or nobiological role in colon cancer, and that its inactivationwas an epiphenomenon. However, such a categoricjudgment was partly due to the lack of understanding ofthe biological roles of DCC, which has since beencompensated by its characterization as a DR (Grady,2007).

In the same way, the UNC5H receptor family isdownregulated in human cancers, including colorectal,breast, ovary, uterus, stomach, lung or kidney cancers(Thiebault et al., 2003). Recently, two studies havefocused on UNC5H3/UNC5C alteration in colorectalcarcinoma. UNC5C is indeed the most downregulatedmember of the UNC5H family (74–77% of cases,

normal cell

cancer cell

apoptotic cancer cell

bound receptor

unbound receptor

Figure 6 DRs and tumor suppression. Normal cells located in their normal environment express DRs that are bound to ligandssupplied by local source. In contrast, in this limited and constant ligand concentration environment, highly proliferative tumor cellsdisplay some unbound DRs and thus undergo apoptosis. In the case of metastatic tumor cells escaping from the primary site, migrationaway from locally supplied ligands leads to apoptosis because of unbound receptors. DRs could thus counteract tumor growth and/orinvasiveness. Loss of expression of these receptors or overexpression of ligands would represent a selective advantage for cancer cells.


11

Oncogene

whereas UNC5H1/UNC5A and H2/B show a reducedexpression in 48 and 27% of the cases, respectively).This loss of expression observed in human primarytumors, as well as in cell lines, is essentially due topromoter methylation (Bernet et al., 2007; Shin et al.,2007). Furthermore, Bernet and colleagues took advan-tage of a natural UNC5H3 loss-of-function occurring inmice (rcm, rostral cerebellar malformation) to demon-strate that UNC5H3 loss of function is associated withtumor progression: mice that carry the APC1638N germ-line mutation, known to predispose mice to thedevelopment of low-grade adenoma (Sieber et al.,2000), and are heterozygous or homozygous for mutantUNC5H3, develop adenomas that progress to adeno-carcinoma at a higher frequency than what is seen inAPC1638N mice. Interestingly, loss of UNC5H3 functionin mice also correlates with apoptosis reduction in micetumors. Another clue in favor of a role of the UNC5Hfamily in cancer is that UNC5B and D are p53 targetgenes, able to mediate part of p53 proapoptotic activity(Tanikawa et al., 2003; Wang et al., 2008).

Patched is also a known tumor suppressor (Stoneet al., 1996). Inactive mutations of Ptc, as well as loss ofexpression, are found in basal cell carcinoma andmedulloblastoma (Wicking and McGlinn, 2001). Ptcexpression inhibits the hallmarks of cell transformationin vitro (Koike et al., 2002) and, interestingly, Ptc alsoinhibits growth in soft agar of transformed cells. This islinked to its proapoptotic function, as growth inhibitiondoes not occur in the presence of Shh, or of a generalcaspase inhibitor, or when Ptc is mutated on its caspasecleavage site (Thibert et al., 2003). However, there is nostrong in vivo evidence so far that Ptc functions as atumor suppressor because it triggers apoptosis.

There is a wide spectrum of data supporting the roleof most DRs as tumor suppressors. As an example,EphA4 is downregulated in invasive forms of breastcancers (Fox and Kandpal, 2004), in liver and kidneycancers (Hafner et al., 2004) and in metastatic melano-mas (Easty et al., 1997), whereas a progressive decreasein p75NTR expression is described in prostate cancers(Pflug and Djakiew, 1998). TrkC is associated with goodprognosis in several cancers (Yamashiro et al., 1997).However, the tumor suppressive functions of thesereceptors are yet to be demonstrated per se.

Whereas DR loss during tumorigenesis occurs in awide fraction of cancers, another selective advantage fortumor cells would be to constitutively overexpressligand. There is now accumulating evidence with regardto netrin-1 to support this idea. Indeed, forced expres-sion of netrin-1 in the digestive tract of transgenic micehas been associated with decreased apoptosis in theintestine, development of advanced adenomas andtumor progression to adenomacarcinoma in a settingof adenoma predisposition (Mazelin et al., 2004). Morerecently, high levels of netrin-1 were detected in a largepanel of human cancers from distinct organs, andnetrin-1 overexpression was correlated with a blockingof the proapoptotic functions of netrin-1 receptors.First, in breast cancer, netrin-1 was shown to be amarker of metastatic disease: decrease in netrin-1

expression by small interfering RNA or netrin-1titration by decoy soluble receptor ectodomain causesapoptosis and prevents metastasis formation both in asyngenic mouse model and in a xenograft model(Fitamant et al., 2008). In the same way, high levels ofnetrin-1 were detected in almost 50% of non-small-celllung cancer and in a large fraction of aggressiveneuroblastoma. As in the breast cancer study, strategiesdisrupting the netrin-1 autocrine loop led to apoptosisinduction and tumor growth inhibition in xenograftedmodels (Delloye-Bourgeois et al., 2009a, b). In thesethree cases, apoptosis resulting from netrin-1 inhibitionseems to be mediated by UNC5H receptors, rather thanby DCC. Interestingly, in the case of aggressiveneuroblastomas, netrin-1 expression levels were foundto have prognosis significance. Aggressive stage 4metastatic neuroblastoma is divided into three groups:stage 4 in children aged more than 1 year has the worseprognosis, whereas stage 4S and stage 4 in children lessthan 1-year old generally have a more favorableprognosis, even though many infants succumb todisease. High levels of netrin-1 were shown to correlatewith adverse outcome of stage 4S and stage 4 (o1 year).A study by Link et al. (2007), reporting that netrin-1expression has significant impact on the overall survivalof patients with poorly differentiated pancreas tumors,completes this description of netrin-1 upregulation inhuman neoplasias. Although the mechanism for auto-crine production of netrin-1 remains to be determined, itcould be at least in part a result of nuclear factor-kBactivation, as netrin-1 is a direct target gene of thistranscription factor (Paradisi et al., 2008). Moreover,according to the well-admitted link between inflamma-tion and colorectal cancer predisposition, it has beensuggested that nuclear factor-kB activation resultingfrom inflammatory stimulus could lead to local netrin-1production, and thus to tumor promotion by apoptosisinhibition. Along this line, colorectal tumor formationin an animal model for chronic inflammation wasinhibited by treatment with netrin-1 titrating agents(Paradisi et al., 2009).

Finally, a third possible way for tumor cells to escapethe proapoptotic activity of DRs would be to inactivatetheir downstream signaling pathways. Notably, threeeffectors of DRs display functional inactivation inhuman cancers: DAPK, DRAL and caspase 8, whichtransduce UNC5H2-, Ptc- and integrin-mediated apop-tosis, respectively. DAPK loss of expression, essentiallythrough promoter methylation, has now been describedin a wide variety of cancers, including lymphoma,leukemia, brain tumors, bladder, breast, renal, cervix,prostate and colorectal carcinomas (Kissil et al., 1997;Raveh and Kimchi, 2001; Gozuacik and Kimchi, 2006).Moreover, DAPK loss of expression is associated with amore malignant tumor phenotype and increased meta-static capacity. DAPK is absent in highly metastaticvariants of mouse lung cancer cell lines, and is present inthe low metastatic variants of those same cell lines(Inbal et al., 1997). In lung and head and neck cancers,DAPK promoter methylation was associated withaggressive disease and poor survival (Sanchez-Cespedes


12

Oncogene

et al., 2000; Tang et al., 2000; Kim et al., 2001; Konget al., 2005). DRAL was initially characterized as adownregulated gene in rhabdomyosarcoma (Geniniet al., 1997). Low or no transcript levels were observedin lymphoblastic leukemia, promyelocytic leukemia andBurkitt’s lymphoma cells (Johannessen et al., 2006;Desmond et al., 2007). Finally, caspase 8 expression isselectively lost during establishment of neuroblastomametastases, rendering these cells refractory to unligatedintegrin-mediated cell death (Stupack et al., 2006).

DRs as new therapeutic targets

Reactivating proapoptotic pathways in human cancercells is one of the main strategies that is being developedin oncology nowadays. Therefore, DRs can join the listof promising targets for cancer treatment. Whereasrestoring the expression of an extinguished receptor-coding gene does not seem conceivable, interfering withligand binding seems to be a more pertinent approach,further supported by recent studies reporting an over-expression of netrin-1 in various cancers (Link et al.,2007; Fitamant et al., 2008; Delloye-Bourgeois et al.,2009a, b). These studies provide an idea of what thefuture therapeutic molecules used in clinical studiescould be; for example, a decoy receptor correspondingto the entire ectodomain of DCC receptor that wouldtitrate the ligand, thus leading to unbound membranereceptors (Fitamant et al., 2008). However, the large size(1100 aa) of the complete extracellular domain of DCCmay complicate its use in vivo. An alternative could be touse shorter polypeptides, corresponding to the fourth orfifth fibronectin-like domain of DCC (Fitamant et al.,2008; Delloye-Bourgeois et al., 2009a, b), which are bothknown to interact with netrin-1 (Geisbrecht et al., 2003).Unlike the complete ectodomain, the fifth and fourthfibronectin-like domains do not counteract interactionbetween netrin-1 and its receptors, but rather blockreceptor multimerization (Mille et al., 2009a). Alterna-tively, other interfering molecules, such as blockingantibodies or peptidomimetic compounds, could blockthe interaction between netrin-1 and its receptors.Interestingly, cancers that show high levels of netrin-1,that is, metastatic breast cancers, neuroblastoma, non-small-cell lung cancer and pancreatic cancers, are alltumors in which there is a need for effective newtherapies. Moreover, and importantly, therapy resis-tance in cancer is frequently associated with deregula-tion in the mechanisms that control apoptosis, andcancer cells are often reliant on these molecularaberrations for survival. Inducing a novel apoptosispathway may add one new component to the combinedtreatment of cancer, thereby increasing the therapeutic

window. However, it remains to be determined whetherinterfering with DR/ligand pairs will stand the test ofclinical tests. The situation in human patients is oftenmore complex than that in murine experimental systems,and potential secondary effects due to perturbation ofnetrin-1 in normal tissues need to be evaluated. Finally,this therapeutic approach could be extended to otherDR ligands besides netrin-1, if further investigationsshow abnormal expression levels in cancers.

Future members of the family?

The list of DRs is of course not closed, and futurestudies will probably reveal the unexpected ability ofsome receptors to induce apoptosis in the absence ofligand. Some receptors already share some intriguingcharacteristics with DRs. For example, ErbB2, areceptor tyrosine kinase related to the EGF receptorfamily, can be cleaved by caspase, and the generatedcytosolic fragment triggers apoptosis (Tikhomirov et al.,2005; Strohecker et al., 2008). Interestingly, this frag-ment contains a BH3-like domain related to the Bclprotein family involved in mitochondrial apoptosis. Animportant issue would be to check whether ErbB2cleavage is affected by the presence of ligand. Unfortu-nately, ErbB2 remains, to date, an orphan receptor. Theepidermal growth factor receptor and Notch1 receptorsare also caspase substrates, but no effect of ligandbinding has been reported, and caspase cleavage hasonly been proposed as a mechanism for turning downsignaling by these receptors (He et al., 2003, 2006;Cohen et al., 2005). An even more promising candidateis the b-amyloid precursor protein (APP), a transmem-brane receptor, the biological function of which remainslargely unknown. APP abnormal processing by secretaseproteases led to accumulation and extracellular depositof the small Ab peptide associated with the developmentof Alzheimer’s disease, a common neurodegenerativedisorder (Koo, 2002). Until recently, APP was anorphan receptor, yet it was recently proposed as anetrin-1 receptor (Lourenco et al., 2009). In addition,APP is known to be cleaved in its intracellular domainby caspase, thus yielding a toxic fragment (Nguyenet al., 2008). Whether APP is a netrin-1 DR remains anintriguing question. In addition, if APP is a DR, onewould expect it to function as a tumor suppressor. Sofar, no convincing data link APP to cancer, but thisremains to be carefully assessed.

Conflict of interest

The authors declare no conflict of interest.

References

Allouche M. (2007). ALK is a novel dependence receptor:potential implications in development and cancer. Cell Cycle 6:1533–1538.

Arighi E, Alberti L, Torriti F, Ghizzoni S, Rizzetti MG, Pelicci G et al.(1997). Identification of Shc docking site on Ret tyrosine kinase.Oncogene 14: 773–782.


13

Oncogene

Baloh RH, Enomoto H, Johnson Jr EM, Milbrandt J. (2000). TheGDNF family ligands and receptors—implications for neuraldevelopment. Curr Opin Neurobiol 10: 103–110.

Barrett GL, Bartlett PF. (1994). The p75 nerve growth factor receptormediates survival or death depending on the stage of sensory neurondevelopment. Proc Natl Acad Sci USA 91: 6501–6505.

Bazigou E, Apitz H, Johansson J, Loren CE, Hirst EM, Chen PL et al.(2007). Anterograde Jelly belly and Alk receptor tyrosine kinasesignaling mediates retinal axon targeting in Drosophila. Cell 128:961–975.

Bernet A, Mazelin L, Coissieux MM, Gadot N, Ackerman SL, ScoazecJY et al. (2007). Inactivation of the UNC5C Netrin-1 receptor isassociated with tumor progression in colorectal malignancies.Gastroenterology 133: 1840–1848.

Bordeaux MC, Forcet C, Granger L, Corset V, Bidaud C, Billaud Met al. (2000). The RET proto-oncogene induces apoptosis: a novelmechanism for Hirschsprung disease. EMBO J 19: 4056–4063.

Bredesen DE, Ye X, Tasinato A, Sperandio S, Wang JJ, Assa-Munt Net al. (1998). p75NTR and the concept of cellular dependence: seeinghow the other half die. Cell Death Differ 5: 365–371.

Canibano C, Rodriguez NL, Saez C, Tovar S, Garcia-Lavandeira M,Borrello MG et al. (2007). The dependence receptor Ret inducesapoptosis in somatotrophs through a Pit-1/p53 pathway, preventingtumor growth. EMBO J 26: 2015–2028.

Casaccia-Bonnefil P, Carter BD, Dobrowsky RT, Chao MV. (1996).Death of oligodendrocytes mediated by the interaction of nervegrowth factor with its receptor p75. Nature 383: 716–719.

Castets M, Coissieux MM, Delloye-Bourgeois C, Bernard L, DelcrosJG, Bernet A et al. (2009). Inhibition of endothelial cell apoptosis bynetrin-1 during angiogenesis. Dev Cell 16: 614–620.

Chao MV, Bothwell MA, Ross AH, Koprowski H, Lanahan AA,Buck CR et al. (1986). Gene transfer and molecular cloning of thehuman NGF receptor. Science 232: 518–521.

Charrier JB, Lapointe F, Le Douarin NM, Teillet MA. (2001). Anti-apoptotic role of Sonic hedgehog protein at the early stages ofnervous system organogenesis. Development 128: 4011–4020.

Charrier JB, Teillet MA, Lapointe F, Le Douarin NM. (1999).Defining subregions of Hensen’s node essential for caudalwardmovement, midline development and cell survival. Development 126:4771–4783.

Chen YQ, Hsieh JT, Yao F, Fang B, Pong RC, Cipriano SC et al.(1999). Induction of apoptosis and G2/M cell cycle arrest by DCC.Oncogene 18: 2747–2754.

Cho KR, Oliner JD, Simons JW, Hedrick L, Fearon ER, PreisingerAC et al. (1994). The DCC gene: structural analysis and mutationsin colorectal carcinomas. Genomics 19: 525–531.

Clark PE, Irvine RA, Coetzee GA. (2003). The androgen receptorCAG repeat and prostate cancer risk. Methods Mol Med 81:255–266.

Cohen LY, Bourbonniere M, Sabbagh L, Bouchard A, Chew T,Jeannequin P et al. (2005). Notch1 antiapoptotic activity isabrogated by caspase cleavage in dying T lymphocytes. Cell Death

Differ 12: 243–254.Coulson EJ, Reid K, Baca M, Shipham KA, Hulett SM, Kilpatrick TJ

et al. (2000). Chopper, a new death domain of the p75 neurotrophinreceptor that mediates rapid neuronal cell death. J Biol Chem 275:30537–30545.

Craft N, Sawyers CL. (1998). Mechanistic concepts in androgen-dependence of prostate cancer. Cancer Metastasis Rev 17: 421–427.

Crowley C, Spencer SD, Nishimura MC, Chen KS, Pitts-Meek S,Armanini MP et al. (1994). Mice lacking nerve growth factor displayperinatal loss of sensory and sympathetic neurons yet develop basalforebrain cholinergic neurons. Cell 76: 1001–1011.

Della-Bianca V, Rossi F, Armato U, Dal-Pra I, Costantini C, Perini Get al. (2001). Neurotrophin p75 receptor is involved in neuronaldamage by prion peptide-(106-126). J Biol Chem 276: 38929–38933.

Delloye-Bourgeois C, Brambilla E, Coissieux MM, Guenebeaud C,Pedeux R, Firlej V et al. (2009a). Interference with netrin-1 andtumor cell death in non-small cell lung cancer. J Natl Cancer Inst

101: 237–247.

Delloye-Bourgeois C, Fitamant J, Paradisi A, Cappellen D, Douc-Rasy S, Raquin MA et al. (2009b). Netrin-1 acts as a survival factorfor aggressive neuroblastoma. J Exp Med 206: 833–847.

Depaepe V, Suarez-Gonzalez N, Dufour A, Passante L, Gorski JA,Jones KR et al. (2005). Ephrin signalling controls brainsize by regulating apoptosis of neural progenitors. Nature 435:1244–1250.

Desmond JC, Raynaud S, Tung E, Hofmann WK, Haferlach T,Koeffler HP. (2007). Discovery of epigenetically silenced genes inacute myeloid leukemias. Leukemia 21: 1026–1034.

Easty DJ, Mitchell PJ, Patel K, Florenes VA, Spritz RA, Bennett DC.(1997). Loss of expression of receptor tyrosine kinase familygenes PTK7 and SEK in metastatic melanoma. Int J Cancer 71:1061–1065.

Edery P, Lyonnet S, Mulligan LM, Pelet A, Dow E, Abel L et al.(1994). Mutations of the RET proto-oncogene in Hirschsprung’sdisease. Nature 367: 378–380.

Ellerby LM, Hackam AS, Propp SS, Ellerby HM, Rabizadeh S,Cashman NR et al. (1999). Kennedy’s disease: caspase cleavage ofthe androgen receptor is a crucial event in cytotoxicity. J Neurochem

72: 185–195.Englund C, Loren CE, Grabbe C, Varshney GK, Deleuil F, Hallberg B

et al. (2003). Jeb signals through the Alk receptor tyrosine kinase todrive visceral muscle fusion. Nature 425: 512–516.

Ernfors P, Lee KF, Jaenisch R. (1994). Mice lacking brain-derivedneurotrophic factor develop with sensory deficits. Nature 368:147–150.

Fazeli A, Dickinson SL, Hermiston ML, Tighe RV, Steen RG, SmallCG et al. (1997). Phenotype of mice lacking functional Deleted incolorectal cancer (Dcc) gene. Nature 386: 796–804.

Fearon ER, Cho KR, Nigro JM, Kern SE, Simons JW, Ruppert JMet al. (1990). Identification of a chromosome 18q gene that is alteredin colorectal cancers. Science 247: 49–56.

Fitamant J, Guenebeaud C, Coissieux MM, Guix C, Treilleux I,Scoazec JY et al. (2008). Netrin-1 expression confers a selectiveadvantage for tumor cell survival in metastatic breast cancer. Proc

Natl Acad Sci USA 105: 4850–4855.Forcet C, Ye X, Granger L, Corset V, Shin H, Bredesen DE et al.

(2001). The dependence receptor DCC (deleted in colorectal cancer)defines an alternative mechanism for caspase activation. Proc Natl

Acad Sci USA 98: 3416–3421.Foveau B, Leroy C, Ancot F, Deheuninck J, Ji Z, Fafeur V et al.

(2007). Amplification of apoptosis through sequential caspasecleavage of the MET tyrosine kinase receptor. Cell Death Differ

14: 752–764.Fox BP, Kandpal RP. (2004). Invasiveness of breast carcinoma cells

and transcript profile: Eph receptors and ephrin ligands asmolecular markers of potential diagnostic and prognostic applica-tion. Biochem Biophys Res Commun 318: 882–892.

Frade JM, Rodriguez-Tebar A, Barde YA. (1996). Induction of celldeath by endogenous nerve growth factor through its p75 receptor.Nature 383: 166–168.

Fujita Y, Taniguchi J, Uchikawa M, Endo M, Hata K, Kubo T et al.(2008). Neogenin regulates neuronal survival through DAP kinase.Cell Death Differ 15: 1593–1608.

Furne C, Corset V, Herincs Z, Cahuzac N, Hueber AO, Mehlen P.(2006). The dependence receptor DCC requires lipid raftlocalization for cell death signaling. Proc Natl Acad Sci USA 103:4128–4133.

Furne C, Rama N, Corset V, Chedotal A, Mehlen P. (2008). Netrin-1is a survival factor during commissural neuron navigation. Proc

Natl Acad Sci USA 105: 14465–14470.Furne C, Ricard J, Cabrera JR, Pays L, Bethea JR, Mehlen P et al.

(2009). EphrinB3 is an anti-apoptotic ligand that inhibits thedependence receptor functions of EphA4 receptors during adultneurogenesis. Biochim Biophys Acta 1793: 231–238.

Geisbrecht BV, Dowd KA, Barfield RW, Longo PA, Leahy DJ.(2003). Netrin binds discrete subdomains of DCC and UNC5 andmediates interactions between DCC and heparin. J Biol Chem 278:32561–32568.


14

Oncogene

Genini M, Schwalbe P, Scholl FA, Remppis A, Mattei MG, SchaferBW. (1997). Subtractive cloning and characterization of DRAL, anovel LIM-domain protein down-regulated in rhabdomyosarcoma.DNA Cell Biol 16: 433–442.

Gozuacik D, Kimchi A. (2006). DAPk protein family and cancer.Autophagy 2: 74–79.

Grady WM. (2007). Making the case for DCC and UNC5C as tumor-suppressor genes in the colon. Gastroenterology 133: 2045–2049.

Hafner C, Schmitz G, Meyer S, Bataille F, Hau P, Langmann T et al.(2004). Differential gene expression of Eph receptors and ephrins inbenign human tissues and cancers. Clin Chem 50: 490–499.

He YY, Huang JL, Chignell CF. (2006). Cleavage of epidermal growthfactor receptor by caspase during apoptosis is independent of itsinternalization. Oncogene 25: 1521–1531.

He YY, Huang JL, Gentry JB, Chignell CF. (2003). Epidermal growthfactor receptor down-regulation induced by UVA in humankeratinocytes does not require the receptor kinase activity. J Biol

Chem 278: 42457–42465.Hedgecock EM, Culotti JG, Hall DH. (1990). The unc-5, unc-6, and

unc-40 genes guide circumferential migrations of pioneer axons andmesodermal cells on the epidermis in C elegans. Neuron 4: 61–85.

Hempstead BL. (2002). The many faces of p75NTR. Curr Opin

Neurobiol 12: 260–267.Hempstead BL, Martin-Zanca D, Kaplan DR, Parada LF, Chao MV.

(1991). High-affinity NGF binding requires coexpression of the trkproto-oncogene and the low-affinity NGF receptor. Nature 350:678–683.

Hofmann K, Tschopp J. (1995). The death domain motif found in Fas(Apo-1) and TNF receptor is present in proteins involved inapoptosis and axonal guidance. FEBS Lett 371: 321–323.

Hong K, Hinck L, Nishiyama M, Poo MM, Tessier-Lavigne M, SteinE. (1999). A ligand-gated association between cytoplasmic domainsof UNC5 and DCC family receptors converts netrin-induced growthcone attraction to repulsion. Cell 97: 927–941.

Hood JD, Cheresh DA. (2002). Role of integrins in cell invasion andmigration. Nat Rev Cancer 2: 91–100.

Huang EJ, Reichardt LF. (2001). Neurotrophins: roles in neuronaldevelopment and function. Annu Rev Neurosci 24: 677–736.

Hueber AO. (2003). Role of membrane microdomain rafts in TNFR-mediated signal transduction. Cell Death Differ 10: 7–9.

Inbal B, Cohen O, Polak-Charcon S, Kopolovic J, Vadai E, EisenbachL et al. (1997). DAP kinase links the control of apoptosis tometastasis. Nature 390: 180–184.

Iwahara T, Fujimoto J, Wen D, Cupples R, Bucay N, Arakawa T et al.(1997). Molecular characterization of ALK, a receptor tyrosinekinase expressed specifically in the nervous system. Oncogene 14:439–449.

Jessell TM. (2000). Neuronal specification in the spinal cord: inductivesignals and transcriptional codes. Nat Rev Genet 1: 20–29.

Jing S, Wen D, Yu Y, Holst PL, Luo Y, Fang M et al. (1996). GDNF-induced activation of the ret protein tyrosine kinase is mediated byGDNFR-alpha, a novel receptor for GDNF. Cell 85: 1113–1124.

Johannessen M, Moller S, Hansen T, Moens U, Van Ghelue M.(2006). The multifunctional roles of the four-and-a-half-LIM onlyprotein FHL2. Cell Mol Life Sci 63: 268–284.

Kang JS, Yi MJ, Zhang W, Feinleib JL, Cole F, Krauss RS. (2004).Netrins and neogenin promote myotube formation. J Cell Biol 167:493–504.

Kaplan DR, Martin-Zanca D, Parada LF. (1991). Tyrosine phosphor-ylation and tyrosine kinase activity of the trk proto-oncogeneproduct induced by NGF. Nature 350: 158–160.

Kaplan DR, Miller FD. (2000). Neurotrophin signal transduction inthe nervous system. Curr Opin Neurobiol 10: 381–391.

Kato H, Zhou Y, Asanoma K, Kondo H, Yoshikawa Y, Watanabe Ket al. (2000). Suppressed tumorigenicity of human endometrialcancer cells by the restored expression of the DCC gene. Br J Cancer

82: 459–466.Keino-Masu K, Masu M, Hinck L, Leonardo ED, Chan SS, Culotti

JG et al. (1996). Deleted in Colorectal Cancer (DCC) encodes anetrin receptor. Cell 87: 175–185.

Kim DH, Nelson HH, Wiencke JK, Christiani DC, Wain JC, Mark EJet al. (2001). Promoter methylation of DAP-kinase: associationwith advanced stage in non-small cell lung cancer. Oncogene 20:1765–1770.

Kissil JL, Feinstein E, Cohen O, Jones PA, Tsai YC, Knowles MAet al. (1997). DAP-kinase loss of expression in various carcinomaand B-cell lymphoma cell lines: possible implications for role astumor suppressor gene. Oncogene 15: 403–407.

Klingelhutz AJ, Smith PP, Garrett LR, McDougall JK. (1993).Alteration of the DCC tumor-suppressor gene in tumorigenicHPV-18 immortalized human keratinocytes transformed by nitro-somethylurea. Oncogene 8: 95–99.

Koike C, Mizutani T, Ito T, Shimizu Y, Yamamichi N, Kameda Tet al. (2002). Introduction of wild-type patched gene suppresses theoncogenic potential of human squamous cell carcinoma cell linesincluding A431. Oncogene 21: 2670–2678.

Kong WJ, Zhang S, Guo C, Zhang S, Wang Y, Zhang D. (2005).Methylation-associated silencing of death-associated proteinkinase gene in laryngeal squamous cell cancer. Laryngoscope 115:1395–1401.

Koo EH. (2002). The beta-amyloid precursor protein (APP) andAlzheimer’s disease: does the tail wag the dog? Traffic 3: 763–770.

Kotzbauer PT, Lampe PA, Heuckeroth RO, Golden JP, Creedon DJ,Johnson Jr EM et al. (1996). Neurturin, a relative of glial-cell-line-derived neurotrophic factor. Nature 384: 467–470.

Lee DK, Chang C. (2003). Molecular communication betweenandrogen receptor and general transcription machinery. J Steroid

Biochem Mol Biol 84: 41–49.Lee FS, Kim AH, Khursigara G, Chao MV. (2001a). The uniqueness

of being a neurotrophin receptor. Curr Opin Neurobiol 11: 281–286.Lee HH, Norris A, Weiss JB, Frasch M. (2003). Jelly belly protein

activates the receptor tyrosine kinase Alk to specify visceral musclepioneers. Nature 425: 507–512.

Lee R, Kermani P, Teng KK, Hempstead BL. (2001b). Regulation ofcell survival by secreted proneurotrophins. Science 294: 1945–1948.

Lejmi E, Leconte L, Pedron-Mazoyer S, Ropert S, Raoul W, LavaletteS et al. (2008). Netrin-4 inhibits angiogenesis via binding toneogenin and recruitment of Unc5B. Proc Natl Acad Sci USA

105: 12491–12496.Leonardo ED, Hinck L, Masu M, Keino-Masu K, Ackerman SL,

Tessier-Lavigne M. (1997). Vertebrate homologues of C. elegansUNC-5 are candidate netrin receptors. Nature 386: 833–838.

Leung-Hagesteijn C, Spence AM, Stern BD, Zhou Y, Su MW,Hedgecock EM et al. (1992). UNC-5, a transmembrane protein withimmunoglobulin and thrombospondin type 1 domains, guides celland pioneer axon migrations in C elegans. Cell 71: 289–299.

Levi-Montalcini R, Angeletti PU. (1963). Essential role of the nervegrowth factor in the survival and maintenance of dissociatedsensory and sympathetic embryonic nerve cells in vitro. Dev Biol 7:653–659.

Liepinsh E, Ilag LL, Otting G, Ibanez CF. (1997). NMR structure ofthe death domain of the p75 neurotrophin receptor. EMBO J 16:4999–5005.

Link BC, Reichelt U, Schreiber M, Kaifi JT, Wachowiak R, BogoevskiD et al. (2007). Prognostic implications of netrin-1 expression andits receptors in patients with adenocarcinoma of the pancreas. Ann

Surg Oncol 14: 2591–2599.Liu J, Yao F, Wu R, Morgan M, Thorburn A, Finley Jr RL et al.

(2002). Mediation of the DCC apoptotic signal by DIP13 alpha. J

Biol Chem 277: 26281–26285.Llambi F, Causeret F, Bloch-Gallego E, Mehlen P. (2001). Netrin-1

acts as a survival factor via its receptors UNC5H and DCC. EMBO

J 20: 2715–2722.Llambi F, Lourenco FC, Gozuacik D, Guix C, Pays L, Del Rio G

et al. (2005). The dependence receptor UNC5H2 mediates apoptosisthrough DAP-kinase. EMBO J 24: 1192–1201.

Loren CE, Englund C, Grabbe C, Hallberg B, Hunter T, Palmer RH.(2003). A crucial role for the anaplastic lymphoma kinase receptortyrosine kinase in gut development in Drosophila melanogaster.EMBO Rep 4: 781–786.


15

Oncogene

Lorenzo MJ, Gish GD, Houghton C, Stonehouse TJ, Pawson T,Ponder BA et al. (1997). Ret alternate splicing influences theinteraction of activated RET with the SH2 and PTB domains ofShc, and the SH2 domain of Grb2. Oncogene 14: 763–771.

Lourenco FC, Galvan V, Fombonne J, Corset V, Llambi F, Muller Uet al. (2009). Netrin-1 interacts with amyloid precursor protein andregulates amyloid-beta production. Cell Death Differ 16: 655–663.

Lu X, Le Noble F, Yuan L, Jiang Q, De Lafarge B, Sugiyama D et al.(2004). The netrin receptor UNC5B mediates guidance events con-trolling morphogenesis of the vascular system. Nature 432: 179–186.

Lykissas MG, Batistatou AK, Charalabopoulos KA, Beris AE. (2007).The role of neurotrophins in axonal growth, guidance, andregeneration. Curr Neurovasc Res 4: 143–151.

Maisse C, Rossin A, Cahuzac N, Paradisi A, Klein C, Haillot MLet al. (2008). Lipid raft localization and palmitoylation: identifica-tion of two requirements for cell death induction by the tumorsuppressors UNC5H. Exp Cell Res 314: 2544–2552.

Manie S, Santoro M, Fusco A, Billaud M. (2001). The RET receptor:function in development and dysfunction in congenital malforma-tion. Trends Genet 17: 580–589.

Matsunaga E, Tauszig-Delamasure S, Monnier PP, Mueller BK,Strittmatter SM, Mehlen P et al. (2004). RGM and its receptorneogenin regulate neuronal survival. Nat Cell Biol 6: 749–755.

Mazelin L, Bernet A, Bonod-Bidaud C, Pays L, Arnaud S, Gespach Cet al. (2004). Netrin-1 controls colorectal tumorigenesis by regulat-ing apoptosis. Nature 431: 80–84.

Mehlen P, Fearon ER. (2004). Role of the dependence receptor DCCin colorectal cancer pathogenesis. J Clin Oncol 22: 3420–3428.

Mehlen P, Furne C. (2005). Netrin-1: when a neuronal guidance cueturns out to be a regulator of tumorigenesis. Cell Mol Life Sci 62:2599–2616.

Mehlen P, Mazelin L. (2003). The dependence receptors DCC andUNC5H as a link between neuronal guidance and survival. Biol Cell

95: 425–436.Mehlen P, Rabizadeh S, Snipas SJ, Assa-Munt N, Salvesen GS,

Bredesen DE. (1998). The DCC gene product induces apoptosis by amechanism requiring receptor proteolysis. Nature 395: 801–804.

Mehlen P, Thibert C. (2004). Dependence receptors: between life anddeath. Cell Mol Life Sci 61: 1854–1866.

Meyerhardt JA, Look AT, Bigner SH, Fearon ER. (1997). Identifica-tion and characterization of neogenin, a DCC-related gene.Oncogene 14: 1129–1136.

Mille F, Llambi F, Guix C, Delloye-Bourgeois C, Guenebeaud C,Castro-Obregon S et al. (2009a). Interfering with multimerization ofnetrin-1 receptors triggers tumor cell death. Cell Death Differ 16:1344–1351.

Mille F, Thibert C, Fombonne J, Rama N, Guix C, Hayashi H et al.(2009b). The Patched dependence receptor triggers apoptosisthrough a DRAL-caspase-9 complex. Nat Cell Biol 11: 739–746.

Minichiello L, Piehl F, Vazquez E, Schimmang T, Hokfelt T, RepresaJ et al. (1995). Differential effects of combined trk receptormutations on dorsal root ganglion and inner ear sensory neurons.Development 121: 4067–4075.

Monnier PP, Sierra A, Macchi P, Deitinghoff L, Andersen JS, MannM et al. (2002). RGM is a repulsive guidance molecule for retinalaxons. Nature 419: 392–395.

Morris SW, Kirstein MN, Valentine MB, Dittmer KG, Shapiro DN,Saltman DL et al. (1994). Fusion of a kinase gene, ALK, to anucleolar protein gene, NPM, in non-Hodgkin’s lymphoma. Science

263: 1281–1284.Mourali J, Benard A, Lourenco FC, Monnet C, Greenland C, Moog-

Lutz C et al. (2006). Anaplastic lymphoma kinase is a dependencereceptor whose proapoptotic functions are activated by caspasecleavage. Mol Cell Biol 26: 6209–6222.

Mulligan LM, Kwok JB, Healey CS, Elsdon MJ, Eng C, Gardner Eet al. (1993). Germ-line mutations of the RET proto-oncogene inmultiple endocrine neoplasia type 2A. Nature 363: 458–460.

Murone M, Rosenthal A, de Sauvage FJ. (1999). Sonic hedgehogsignaling by the patched-smoothened receptor complex. Curr Biol 9:76–84.

Naumann T, Casademunt E, Hollerbach E, Hofmann J, Dechant G,Frotscher M et al. (2002). Complete deletion of the neurotrophinreceptor p75NTR leads to long-lasting increases in the number ofbasal forebrain cholinergic neurons. J Neurosci 22: 2409–2418.

Nelson KA, Witte JS. (2002). Androgen receptor CAG repeats andprostate cancer. Am J Epidemiol 155: 883–890.

Nguyen TV, Galvan V, Huang W, Banwait S, Tang H, Zhang J et al.(2008). Signal transduction in Alzheimer disease: p21-activatedkinase signaling requires C-terminal cleavage of APP at Asp664. JNeurochem 104: 1065–1080.

Niederkofler V, Salie R, Sigrist M, Arber S. (2004). Repulsive guidancemolecule (RGM) gene function is required for neural tube closurebut not retinal topography in the mouse visual system. J Neurosci

24: 808–818.Paradisi A, Maisse C, Bernet A, Coissieux MM, Maccarrone M,

Scoazec JY et al. (2008). NF-kappaB regulates netrin-1 expressionand affects the conditional tumor suppressive activity of the netrin-1receptors. Gastroenterology 135: 1248–1257.

Paradisi A, Maisse C, Coissieux MM, Gadot N, Lepinasse F, Delloye-Bourgeois C et al. (2009). Netrin-1 up-regulation in inflammatorybowel diseases is required for colorectal cancer progression. Proc

Natl Acad Sci USA 106: 17146–17151.Pasquale EB. (2005). Eph receptor signalling casts a wide net on cell

behaviour. Nat Rev Mol Cell Biol 6: 462–475.Peterziel H, Paech T, Strelau J, Unsicker K, Krieglstein K. (2007).

Specificity in the crosstalk of TGFbeta/GDNF family members isdetermined by distinct GFR alpha receptors. J Neurochem 103:2491–2504.

Pflug B, Djakiew D. (1998). Expression of p75NTR in a humanprostate epithelial tumor cell line reduces nerve growth factor-induced cell growth by activation of programmed cell death. Mol

Carcinog 23: 106–114.PuschelPuschel AW. (1999). Divergent properties of mouse netrins.

Mech Dev 83: 65–75.Qin S, Yu L, Gao Y, Zhou R, Zhang C. (2007). Characterization of the

receptors for axon guidance factor netrin-4 and identification of thebinding domains. Mol Cell Neurosci 34: 243–250.

Rabizadeh S, Oh J, Zhong LT, Yang J, Bitler CM, Butcher LL et al.(1993). Induction of apoptosis by the low-affinity NGF receptor.Science 261: 345–348.

Rabizadeh S, Ye X, Sperandio S, Wang JJ, Ellerby HM, Ellerby LMet al. (2000). Neurotrophin dependence domain: a domain requiredfor the mediation of apoptosis by the p75 neurotrophin receptor.J Mol Neurosci 15: 215–229.

Rajagopalan S, Deitinghoff L, Davis D, Conrad S, Skutella T,Chedotal A et al. (2004). Neogenin mediates the action of repulsiveguidance molecule. Nat Cell Biol 6: 756–762.

Rajasekharan S, Kennedy TE. (2009). The netrin protein family.Genome Biol 10: 239.

Raveh T, Kimchi A. (2001). DAP kinase-a proapoptotic gene thatfunctions as a tumor suppressor. Exp Cell Res 264: 185–192.

Ricard J, Salinas J, Garcia L, Liebl DJ. (2006). EphrinB3 regulates cellproliferation and survival in adult neurogenesis. Mol Cell Neurosci

31: 713–722.Rodrigues S, De Wever O, Bruyneel E, Rooney RJ, Gespach C. (2007).

Opposing roles of netrin-1 and the dependence receptor DCC incancer cell invasion, tumor growth and metastasis. Oncogene 26:5615–5625.

Romeo G, Ronchetto P, Luo Y, Barone V, Seri M, Ceccherini I et al.(1994). Point mutations affecting the tyrosine kinase domain of theRET proto-oncogene in Hirschsprung’s disease. Nature 367:377–378.

Saito M, Yamaguchi A, Goi T, Tsuchiyama T, Nakagawara G, UranoT et al. (1999). Expression of DCC protein in colorectal tumors andits relationship to tumor progression and metastasis. Oncology 56:134–141.

Sanchez-Cespedes M, Esteller M, Wu L, Nawroz-Danish H, Yoo GH,Koch WM et al. (2000). Gene promoter hypermethylation in tumorsand serum of head and neck cancer patients. Cancer Res 60:892–895.


16

Oncogene

Santoro M, Carlomagno F, Romano A, Bottaro DP, Dathan NA,Grieco M et al. (1995). Activation of RET as a dominanttransforming gene by germline mutations of MEN2A and MEN2B.Science 267: 381–383.

Santoro M, Dathan NA, Berlingieri MT, Bongarzone I, Paulin C,Grieco M et al. (1994). Molecular characterization of RET/PTC3; anovel rearranged version of the RETproto-oncogene in a humanthyroid papillary carcinoma. Oncogene 9: 509–516.

Schmidtmer J, Engelkamp D. (2004). Isolation and expression patternof three mouse homologues of chick Rgm. Gene Expr Patterns 4:105–110.

Scholl FA, McLoughlin P, Ehler E, de Giovanni C, Schafer BW.(2000). DRAL is a p53-responsive gene whose four and a half LIMdomain protein product induces apoptosis. J Cell Biol 151: 495–506.

Scott RP, Ibanez CF. (2001). Determinants of ligand bindingspecificity in the glial cell line-derived neurotrophic factor familyreceptor alpha S. J Biol Chem 276: 1450–1458.

Serafini T, Colamarino SA, Leonardo ED, Wang H, Beddington R,Skarnes WC et al. (1996). Netrin-1 is required for commissuralaxon guidance in the developing vertebrate nervous system. Cell 87:1001–1014.

Serafini T, Kennedy TE, Galko MJ, Mirzayan C, Jessell TM, Tessier-Lavigne M. (1994). The netrins define a family of axon outgrowth-promoting proteins homologous to C. elegans UNC-. Cell 78: 409–424.

Shibata D, Reale MA, Lavin P, Silverman M, Fearon ER, Steele Jr Get al. (1996). The DCC protein and prognosis in colorectal cancer. N

Engl J Med 335: 1727–1732.Shin SK, Nagasaka T, Jung BH, Matsubara N, Kim WH, Carethers

JM et al. (2007). Epigenetic and genetic alterations in Netrin-1receptors UNC5C and DCC in human colon cancer. Gastroenter-

ology 133: 1849–1857.Shohat G, Spivak-Kroizman T, Cohen O, Bialik S, Shani G, Berrisi H

et al. (2001). The pro-apoptotic function of death-associated proteinkinase is controlled by a unique inhibitory autophosphorylation-based mechanism. J Biol Chem 276: 47460–47467.

Sieber OM, Tomlinson IP, Lamlum H. (2000). The adenomatouspolyposis coli (APC) tumour suppressor—genetics, function anddisease. Mol Med Today 6: 462–469.

Simons K, Toomre D. (2000). Lipid rafts and signal transduction. Nat

Rev Mol Cell Biol 1: 31–39.Songyang Z, Carraway IIIrd KL, Eck MJ, Harrison SC, Feldman RA,

Mohammadi M et al. (1995). Catalytic specificity of protein-tyrosinekinases is critical for selective signaling. Nature 373: 536–539.

Srinivasan K, Strickland P, Valdes A, Shin GC, Hinck L. (2003).Netrin-1/neogenin interaction stabilizes multipotent progenitor capcells during mammary gland morphogenesis. Dev Cell 4: 371–382.

Stein E, Zou Y, Poo M, Tessier-Lavigne M. (2001). Binding of DCCby netrin-1 to mediate axon guidance independent of adenosineA2B receptor activation. Science 291: 1976–1982.

Stilo R, Leonardi A, Formisano L, Di Jeso B, Vito P, Liguoro D.(2002). TUCAN/CARDINAL and DRAL participate in a commonpathway for modulation of NF-kappaB activation. FEBS Lett 521:165–169.

Stone DM, Hynes M, Armanini M, Swanson TA, Gu Q, Johnson RLet al. (1996). The tumour-suppressor gene patched encodes acandidate receptor for Sonic hedgehog. Nature 384: 129–134.

Strohecker AM, Yehiely F, Chen F, Cryns VL. (2008). Caspasecleavage of HER-2 releases a Bad-like cell death effector. J Biol

Chem 283: 18269–18282.Stupack DG. (2005). Integrins as a distinct subtype of dependence

receptors. Cell Death Differ 12: 1021–1030.Stupack DG, Cheresh DA. (2002). ECM remodeling regulates

angiogenesis: endothelial integrins look for new ligands. Sci STKE

2002: PE7.Stupack DG, Puente XS, Boutsaboualoy S, Storgard CM, Cheresh

DA. (2001). Apoptosis of adherent cells by recruitment of caspase-8to unligated integrins. J Cell Biol 155: 459–470.

Stupack DG, Teitz T, Potter MD, Mikolon D, Houghton PJ, Kidd VJet al. (2006). Potentiation of neuroblastoma metastasis by loss ofcaspase-8. Nature 439: 95–99.

Sun XF, Rutten S, Zhang H, Nordenskjold B. (1999). Expression ofthe deleted in colorectal cancer gene is related to prognosis in DNAdiploid and low proliferative colorectal adenocarcinoma. J Clin

Oncol 17: 1745–1750.Tang X, Jang SW, Okada M, Chan CB, Feng Y, Liu Y et al. (2008).

Netrin-1 mediates neuronal survival through PIKE-L interactionwith the dependence receptor UNC5B. Nat Cell Biol 10: 698–706.

Tang X, Khuri FR, Lee JJ, Kemp BL, Liu D, Hong WK et al. (2000).Hypermethylation of the death-associated protein (DAP) kinasepromoter and aggressiveness in stage I non-small-cell lung cancer.J Natl Cancer Inst 92: 1511–1516.

Tanikawa C, Matsuda K, Fukuda S, Nakamura Y, Arakawa H.(2003). p53RDL1 regulates p53-dependent apoptosis. Nat Cell Biol

5: 216–223.Tauszig-Delamasure S, Yu LY, Cabrera JR, Bouzas-Rodriguez J,

Mermet-Bouvier C, Guix C et al. (2007). The TrkC receptor inducesapoptosis when the dependence receptor notion meets the neuro-trophin paradigm. Proc Natl Acad Sci USA 104: 13361–13366.

Tessarollo L, Tsoulfas P, Donovan MJ, Palko ME, Blair-Flynn J,Hempstead BL et al. (1997). Targeted deletion of all isoforms of thetrkC gene suggests the use of alternate receptors by its ligandneurotrophin-3 in neuronal development and implicates trkC innormal cardiogenesis. Proc Natl Acad Sci USA 94: 14776–14781.

Tessarollo L, Vogel KS, Palko ME, Reid SW, Parada LF. (1994).Targeted mutation in the neurotrophin-3 gene results in loss ofmuscle sensory neurons. Proc Natl Acad Sci USA 91: 11844–11848.

Thibert C, Teillet MA, Lapointe F, Mazelin L, Le Douarin NM,Mehlen P. (2003). Inhibition of neuroepithelial patched-inducedapoptosis by sonic hedgehog. Science 301: 843–846.

Thiebault K, Mazelin L, Pays L, Llambi F, Joly MO, Scoazec JY et al.(2003). The netrin-1 receptors UNC5H are putative tumorsuppressors controlling cell death commitment. Proc Natl Acad

Sci USA 100: 4173–4178.Thornberry NA, Lazebnik Y. (1998). Caspases: enemies within.

Science 281: 1312–1316.Tikhomirov O, Dikov M, Carpenter G. (2005). Identification of

proteolytic fragments from ErbB-2 that induce apoptosis. Oncogene

24: 3906–3913.Tulasne D, Deheuninck J, Lourenco FC, Lamballe F, Ji Z, Leroy C

et al. (2004). Proapoptotic function of the MET tyrosine kinasereceptor through caspase cleavage. Mol Cell Biol 24: 10328–10339.

Velcich A, Corner G, Palumbo L, Augenlicht L. (1999). Alteredphenotype of HT29 colonic adenocarcinoma cells following expres-sion of the DCC gene. Oncogene 18: 2599–2606.

Vielmetter J, Kayyem JF, Roman JM, Dreyer WJ. (1994). Neogenin,an avian cell surface protein expressed during terminal neuronaldifferentiation, is closely related to the human tumor suppressormolecule deleted in colorectal cancer. J Cell Biol 127: 2009–2020.

Wang H, Ozaki T, Shamim Hossain M, Nakamura Y, Kamijo T, XueX et al. (2008). A newly identified dependence receptor UNC5H4 isinduced during DNA damage-mediated apoptosis and transcrip-tional target of tumor suppressor p53. Biochem Biophys Res

Commun 370: 594–598.Wang JJ, Rabizadeh S, Tasinato A, Sperandio S, Ye X, Green M et al.

(2000). Dimerization-dependent block of the proapoptotic effect ofp75(NTR). J Neurosci Res 60: 587–593.

Wang R, Wei Z, Jin H, Wu H, Yu C, Wen W et al. (2009).Autoinhibition of UNC5b revealed by the cytoplasmic domainstructure of the receptor. Mol Cell 33: 692–703.

Wicking C, McGlinn E. (2001). The role of hedgehog signalling intumorigenesis. Cancer Lett 173: 1–7.

Williams ME, Strickland P, Watanabe K, Hinck L. (2003). UNC5H1induces apoptosis via its juxtamembrane region through aninteraction with NRAGE. J Biol Chem 278: 17483–17490.

Wilson BD, Ii M, Park KW, Suli A, Sorensen LK, Larrieu-Lahargue Fet al. (2006). Netrins promote developmental and therapeuticangiogenesis. Science 313: 640–644.

Wilson NH, Key B. (2006). Neogenin interacts with RGMa andnetrin-1 to guide axons within the embryonic vertebrate forebrain.Dev Biol 296: 485–498.


17

Oncogene

Yaar M, Zhai S, Fine RE, Eisenhauer PB, Arble BL, Stewart KB et al.(2002). Amyloid beta binds trimers as well as monomers of the 75-kDa neurotrophin receptor and activates receptor signaling. J Biol

Chem 277: 7720–7725.Yamashiro DJ, Liu XG, Lee CP, Nakagawara A, Ikegaki N,

McGregor LM et al. (1997). Expression and function of

Trk-C in favourable human neuroblastomas. Eur J Cancer 33:2054–2057.

Yeo TT, Chua-Couzens J, Butcher LL, Bredesen DE, Cooper JD,Valletta JS et al. (1997). Absence of p75NTR causes increased basalforebrain cholinergic neuron size, choline acetyltransferase activity,and target innervation. J Neurosci 17: 7594–7605.


18

Oncogene

dependence receptors: a new paradigm in cell signaling and ......differentiation, migration or...

Documents