Portal de Periódicos da CAPES

Neurotrophins Live or Let Die: Does p75NTR Decide?

1997; Cell Press; Volume: 18; Issue: 2 Linguagem: Inglês

10.1016/s0896-6273(00)80259-7

1097-4199

Bruce Carter, Gary R. Lewin,

Neuropeptides and Animal Physiology

The neurotrophins NGF, BDNF, NT-3, and NT-4/5 are required for the survival and differentiation of specific neuronal populations during ontogeny (reviewed by16Lewin G.R. Barde Y.-A. Annu. Rev. Neurosci. 1996; 19: 289-317Crossref PubMed Scopus (1734) Google Scholar). The prototypical member of the family, NGF, was shown early on to bind to high and low affinity sites on neurons, subsequently the receptor thought to be responsible for low affinity binding was cloned (reviewed by10Chao M.V. J. Neurobiol. 1994; 25: 1373-1385Crossref PubMed Scopus (390) Google Scholar). Much initial work focused on this receptor, which was initially named the low affinity NGF receptor and later p75NTR, as it became clear that it could bind all of the neurotrophins. The expansion of the neurotrophin family and the discovery of a group of cognate receptor tyrosine kinases (16Lewin G.R. Barde Y.-A. Annu. Rev. Neurosci. 1996; 19: 289-317Crossref PubMed Scopus (1734) Google Scholar) left the p75NTR with a seemingly much diminished role. However, a series of recent reports have highlighted important and intriguing functions of the p75NTR; here, we review these new developments and their implications. The p75NTR appears to bind all of the neurotrophins with more or less equal affinity but different kinetics; however, it was clear early on that it could not account for the survival functions exhibited by the neurotrophins. Later, after the discovery of the trk tyrosine kinase receptors (which have the following receptor specificities: trkA/NGF, trkB/BDNF and NT-4/5, and TrkC/NT-3), it became clear that these receptors mediated, to a large extent, typical growth factor responses (6Bothwell M. Annu. Rev. Neurosci. 1995; 18: 223-253Crossref PubMed Scopus (746) Google Scholar). Thus, neuronal survival in vitro and in vivo, neurite outgrowth from cultured neurons, and differentiation of PC12 cells could all be best accounted for by activation of these receptor tyrosine kinases and their associated signaling pathways (6Bothwell M. Annu. Rev. Neurosci. 1995; 18: 223-253Crossref PubMed Scopus (746) Google Scholar, 16Lewin G.R. Barde Y.-A. Annu. Rev. Neurosci. 1996; 19: 289-317Crossref PubMed Scopus (1734) Google Scholar). It was postulated, however, that the p75NTR played a crucial role as a coreceptor for the trks. Indeed, substantial evidence suggests that this receptor cooperates with trkA to form a high affinity binding site. Several groups, in particular that of Moses Chao and his colleagues, demonstrated a need for expression of both receptors in order to produce a high affinity site (10Chao M.V. J. Neurobiol. 1994; 25: 1373-1385Crossref PubMed Scopus (390) Google Scholar). In addition, experiments in which binding of NGF to p75NTR was blocked using either antibodies to p75NTR or using BDNF (since BDNF can bind p75NTR, but not trkA) demonstrated a decrease in the number of high affinity sites in PC12 cells and sensory neurons (23Weskamp G. Reichardt L.F. Neuron. 1991; 6: 649-663Abstract Full Text PDF PubMed Scopus (247) Google Scholar, 3Barker P.A. Shooter E.P. Neuron. 1994; 13: 203-215Abstract Full Text PDF PubMed Scopus (365) Google Scholar). Thus, p75NTR may facilitate trkA dimerization needed for high affinity binding (see Old Model in Figure 1). However, enhancing neurotrophin binding is not the only role for p75NTR. A number of reports suggest that it also modulates trk signaling. While it is clear that trk signaling can occur in the absence of p75NTR, the responses are increased in its presence. For example, sympathoadrenal cells (MAH cells) expressing only trkA extended neurites and survived in the presence of NGF but when coexpressing p75NTR showed an 8-fold higher tyrosine phosphorylation of trkA (21Verdi J.M. Birren S.J. Ibañez C.F. Persson H. Kaplan D.R. Benedetti M. Chao M.V. Anderson D.J. Neuron. 1994; 12: 733-745Abstract Full Text PDF PubMed Scopus (306) Google Scholar). Analysis of transgenic mice lacking p75NTR also supports a facilitating role for this receptor in trk function: the dose required to elicit a survival response, at least for NGF, in primary neurons is four times higher without p75NTR (11Davies A.M. Lee K.-F. Jaenisch R. Neuron. 1993; 11: 565-574Abstract Full Text PDF PubMed Scopus (316) Google Scholar). Clearly, during development, when neurons are competing for limited amounts of target-derived trophic factor, a higher sensitivity provides a significant advantage. The catalytic forms of the trk receptors are expressed primarily in neurotrophin-responsive neuronal populations. p75NTR is often but not always coexpressed in neurons with full-length trk receptors. However, cellular responses to neurotrophins have been seen in cells expressing only p75NTR and no full-length trk (e.g., Schwann cells1Anton E.S. Weskamp G. Reichardt L.F. Matthew W.D. Proc. Natl. Acad. Sci. USA. 1994; 91: 2795-2799Crossref PubMed Scopus (296) Google Scholar; Carter et al., 1996), implying the existence of a p75NTR-specific signal. Recent reports have begun to characterize such a signaling pathway and its possible functional significance. p75NTR is a member of a family of structurally related cytokine receptors, including type I and II TNF receptors (TNFRI and TNFRII), the Fas antigen (also referred to as Apo1), CD40, CD30, CD27, DR3, and the lymphotoxin β receptor. These receptors are characterized by an extracellular domain containing repeated cysteine clusters. However, the intracellular domains are substantially divergent and lack any catalytic domain that could provide clues about their mechanisms of action. The only obvious intracellular homology is a short segment of similarity in Fas and TNFRI referred to as the death domain; mutations in this domain abrogate the activation of apoptosis by these receptors (7Brakebusch C. Nophar Y. Kemper O. Engelmann H. Wallach D. EMBO J. 1992; 11: 943-950Crossref PubMed Scopus (145) Google Scholar). Recently, careful alignment analysis has revealed weakly homologous regions in other family members, including p75NTR, as well as other nonreceptor proteins (14Feinstein E. Kimchi A. Wallach D. Boldin M. Varfolomeev E. Trends Biochem. Sci. 1995; 20: 342-344Abstract Full Text PDF PubMed Scopus (261) Google Scholar). The cellular responses to activation of this family of receptors so far characterized are the activation of gene transcription via nuclear factor κ B (NFκB) and the regulation of cell survival/apoptosis. Several of the receptors have been shown to produce the lipid messenger ceramide following agonist binding through stimulation of sphingomyelinase. The first clear biochemical evidence for p75NTR signaling came from 12Dobrowsky R.T. Werner M.H. Castellino A.M. Chao M.V. Hannun Y.A. Science. 1994; 265: 1596-1599Crossref PubMed Scopus (540) Google Scholar who found that binding of any of the neurotrophins to p75NTR in T9 glioma cells or NIH3T3 cells also activated sphingomyelinase and subsequently increased ceramide production. Recently, p75NTR-mediated NFκB activation in Schwann cells was investigated. These cells are well suited to study p75NTR signaling for several reasons: they express the receptor in a regulated manner, they do not express trkA, and NGF binding to p75NTR elicits their migration (1Anton E.S. Weskamp G. Reichardt L.F. Matthew W.D. Proc. Natl. Acad. Sci. USA. 1994; 91: 2795-2799Crossref PubMed Scopus (296) Google Scholar). It was demonstrated that NGF binding to p75NTR activated nuclear translocation of the p65 subunit of NFκB and stimulated DNA binding. Surprisingly, this activation was specific for NGF; neither NT3 nor BDNF produced a signal (8Carter B.D. Kaltschmidt C. Kaltschmidt B. Offenhäuser N. Böhm-Matthaei R. Baeuerle P.A. Barde Y.-A. Science. 1996; 272: 542-545Crossref PubMed Scopus (599) Google Scholar). The mechanism for this selectivity is not clear since all of the neurotrophins bind to the receptor with similar affinities. Moreover, different signaling pathways appear to be activated by the different neurotrophins; all have been shown to activate ceramide production, yet only NGF led to NFκB stimulation. The similarity of p75NTR/NGF receptor signaling in certain cell types to cytokine signaling receptors whose activation often leads to apoptotic events led to the rather heretical notion that the p75NTR can also initiate apoptosis. However, a number of cytokines can function as trophic or pro-apoptotic signals in the immune system. One of the first indications that NGF could mediate neuronal cell death had been suggested by the experiments of 22Von Bartheld C.S. Kinoshita Y. Prevette D. Yin Q.W. Oppenheim R.W. Bothwell M. Neuron. 1994; 12: 639-654Abstract Full Text PDF PubMed Scopus (126) Google Scholar who found that while BDNF retrogradely transported to chick isthmo-optic nuclei could promote the survival of these neurons, NGF increased the number of dying cells. However, this result was interpreted as NGF decreasing binding of endogenous BDNF to trkB by blocking the p75NTR coreceptor, thus resulting in cell death of some BDNF-dependent neurons. Recently, direct evidence for mediation of cell death by p75NTR upon NGF binding has been reported by two groups. 15Frade J.M. Rodríguez-Tébar A. Barde Y.-A. Nature. 1996; 383: 166-168Crossref PubMed Scopus (653) Google Scholar showed that early retinal cells expressing p75NTR, and not trkA, undergo cell death that could be prevented by application of antibodies to NGF, thereby demonstrating naturally occurring NGF-mediated apoptosis during chick development via p75NTR. In a different system9Casaccia-Bonnefil P. Carter B.D. Dobrowsky R.T. Chao M.V. Nature. 1996; 383: 716-719Crossref PubMed Scopus (704) Google Scholar showed NGF induced apoptosis again through the p75NTR in cultured rat oligodendrocytes not expressing trkA. These authors found that the cell death correlated with a sustained increase in ceramide production and activation of the c-jun kinase pathway. Strikingly, the effect was highly specific for NGF and for mature oligodendrocytes. As was the case for NFκB activation in Schwann cells, neither NT3 nor BDNF activated apoptosis. In contrast to the oligodendrocytes, NGF did not stimulate c-jun kinase or programmed cell death in NIH3T3 cells expressing p75NTR. The notion that the p75 neurotrophin receptor could initiate an apoptotic signal was actually first put forward several years ago by Bredesen and his colleagues when they observed that expression of p75NTR in specific neuroblastoma cell types increased cell death upon withdrawal of neurotrophin (18Rabizadeh S. Oh J. Zhong L. Yang J. Bitler C.M. Butcher L.L. Bredesen D.E. Science. 1993; 261: 345-348Crossref PubMed Scopus (739) Google Scholar). These findings led to the proposal of a ligand-independent signal by the receptor, which was blocked by neurotrophin binding. Subsequently4Barrett G.L. Bartlett P.F. Proc. Natl. Acad. Sci. USA. 1994; 91: 6501-6505Crossref PubMed Scopus (353) Google Scholar showed that cultured postnatal mouse dorsal root ganglion neurons treated with antisense oligonucleotides to p75NTR survived longer than controls in the absence of neurotrophin, suggesting that the presence of p75NTR is responsible for their apoptosis. The interpretation of these earlier experiments was, however, complicated by the possible presence of full-length Trk receptors, a complication more explicitly avoided in recent experiments. The two-hybrid approach has been successfully used with a number of receptors in the p75NTR family to identify unique signaling molecules interacting with their intracellular domains. The group of Goeddel initially isolated proteins associating with the TNF receptors: TRADD and TRAF1, 2, and 3 (reviewed by2Baker S.J. Reddy E.P. Oncogene. 1996; 12: 1-9PubMed Google Scholar). The TRAF family has continued to grow with three additional members now identified as receptor interactors. A number of the TRAFs interact with other members of the p75NTR family, and several have been shown to activate NFκB. TRADD appears to be a branch point, either transducing a death signal to FADD (also called MORT1), a protein originally isolated as a Fas antigen interactor, or activating NFκB via TRAF2. FADD initiates cell death via MACH (Boldin et al., 1996; also referred to as FLICE17Muzio M. Chinnaiyan A.M. Kischkel F.C. O'Rourke K. Shevchenko A. Ni J. Scaffidi C. Bretz J.D. Zhang M. Gentz R. Mann M. Krammer P. Peter M.E. Dixit V.M. Cell. 1996; 85: 817-827Abstract Full Text Full Text PDF PubMed Scopus (2656) Google Scholar), which contains a domain with homology to the ICE family of apoptosis-inducing proteases. The multiplicity of these interactors for the other members of the p75NTR receptor family of course suggests that the p75NTR may also have access to its own set of interactors. To investigate the intracellular mechanisms through which p75NTR acts, the yeast two-hybrid system has also been used to identify a potential signal-transducing protein (NRIF) associated with the cytoplasmic domain of the p75NTR (8Carter B.D. Kaltschmidt C. Kaltschmidt B. Offenhäuser N. Böhm-Matthaei R. Baeuerle P.A. Barde Y.-A. Science. 1996; 272: 542-545Crossref PubMed Scopus (599) Google ScholarSoc. Neurosci. abstract). This novel zinc finger protein associates with p75NTR in a ligand-dependent manner, which is, interestingly, specific for NGF. The newly identified functions for the p75NTR in a cellular context invite one to speculate whether novel physiological roles during normal development and maturity for NGF/p75NTR signaling have been hitherto overlooked. Although a new model is suggested by this data (Figure 1), the old model almost certainly still applies in most situations where p75NTR and full-length trk receptors are present. So far, the data suggests that cells can transduce an NGF signal via p75NTR, which could lead to apoptosis in a very cell type–specific manner. This is obviously not a prominent feature of target-dependent cell death as NGF is well known to rescue peripheral NGF-responsive neurons during the period of naturally occurring cell death and not to kill them. This fact, together with reduced neurotrophin binding in the absence of p75NTR, probably accounts in large part for the substantial loss of sensory and sympathetic neurons in the p75NTR knockout animals (11Davies A.M. Lee K.-F. Jaenisch R. Neuron. 1993; 11: 565-574Abstract Full Text PDF PubMed Scopus (316) Google Scholar). On the other hand, the data from the retina (15Frade J.M. Rodríguez-Tébar A. Barde Y.-A. Nature. 1996; 383: 166-168Crossref PubMed Scopus (653) Google Scholar) suggests that very early cells, the precise identity of which is not clear (i.e., postmitotic neurons or neuronal precursors), are susceptible to NGF/p75NTR-induced death. On the other end of the developmental time scale, the observations of 9Casaccia-Bonnefil P. Carter B.D. Dobrowsky R.T. Chao M.V. Nature. 1996; 383: 716-719Crossref PubMed Scopus (704) Google Scholar show that mature but not immature oligodendrocytes are susceptible to NGF-induced death. As these cells do not express significant levels of p75NTR under normal in vivo conditions, but do in lesion models, this finding suggests an apoptotic response function for this receptor in adult animals. In vivo data regarding central neurons expressing the p75NTR (specifically, projection neurons of the septohippocampal system) suggest that the absence of p75NTR can be neuroprotective during development (20Van der Zee C.E.E.M. Ross G.M. Riopelle R.J. Hagg T. Science. 1996; 274: 1729-1732Crossref PubMed Scopus (172) Google Scholar). These authors observed an increase in cholinergic forebrain neurons in mice lacking p75NTR or in wild-type mice injected with a peptide that blocks NGF binding to p75NTR. The neurons lost appeared to be those lacking trkA, suggesting that the presence of both receptors results in an overall survival response. The authors suggest that cholinergic neurons possessing the p75NTR receptor without trkA (reported to be 25% of the total) are killed during postnatal development. TrkA may inhibit the apoptotic signal from p75NTR. Indeed, an inhibitory effect of trkA on p75NTR-mediated ceramide production has been reported (13Dobrowsky R.T. Jenkins G.M. Hannun Y.A. J. Biol. Chem. 1995; 270: 22135-22142Crossref PubMed Scopus (219) Google Scholar). It is notable that NGF is present in the targets of these cholinergic neurons and perhaps acts through the p75NTR to kill those lacking trkA. What appears to be remarkably consistent from the published experiments is that the newly discovered signals mediated by p75NTR appear to be specific for NGF and are absent with the other neurotrophins. This specificity has yet to be explained and was unexpected based on previous binding studies with this receptor. An intriguing possibility remains that a second messenger activated only after NGF binding (such as NRIF, for example) could partly account for such specificity. It is still quite possible that the other neurotrophins might initiate an apoptotic signal through p75NTR in different cellular contexts where a different intracellular signal is engaged. In the above description, we have concentrated on the new pro-apoptotic role for p75NTR. However, the results of 8Carter B.D. Kaltschmidt C. Kaltschmidt B. Offenhäuser N. Böhm-Matthaei R. Baeuerle P.A. Barde Y.-A. Science. 1996; 272: 542-545Crossref PubMed Scopus (599) Google Scholar with Schwann cells show that NGF can switch on genes via the activation of p75NTR and NFκB without leading to cell death. NFκB is a master switch regulating the expression of a variety of genes, some of which may control cell survival. In addition, this transcription factor is known to upregulate many adhesion molecules, which could account for the NGF-mediated migration of Schwann cells (Anton et al., 1995). Interestingly, the expression of the adhesion molecule L1 was reported to be increased by NGF addition to Schwann cells several years ago (19Seilheimer B. Schachner M. EMBO J. 1987; 6: 1611-1616Crossref PubMed Scopus (149) Google Scholar). Thus, one can envisage cases where p75NTR transduces a more subtle signal in certain cellular contexts. No doubt more p75NTR-specific regulatory events will be found in the next phase of neurotrophin research. The new-found Jekyll and Hyde nature of p75NTR is intriguing as it can promote apoptosis in some circumstances, while in other situations it is a neuroprotective molecule. In addition, one particularly surprising aspect of recent reports is that p75NTR has found a niche independent of its neurotrophin receptor partners, the trks. Future research will undoubtedly uncover more intricacies of p75NTR signaling and place them in their proper cellular and physiological contexts during development and maturity.