|Home | About | Journals | Submit | Contact Us | Français|
In neurons, many receptors must be localized correctly to axons or dendrites for proper function. During development, receptors for nerve growth and guidance are targeted to axons and localized to growth cones where receptor activation by ligands results in promotion or inhibition of axon growth. Signaling outcomes downstream of ligand binding are determined by the location, levels, and residence times of receptors on the neuronal plasma membrane. Therefore, the mechanisms controlling the trafficking of these receptors are crucial to the proper wiring of circuits. Membrane proteins accumulate on the axonal surface by multiple routes, including polarized sorting in the trans-Golgi network, sorting in endosomes, and removal by endocytosis. Endosomes also play important roles in the signaling pathways for both growth-promoting and -inhibiting molecules: signaling endosomes derived from endocytosis are important for signaling from growth cones to cell bodies. Growth-promoting neurotrophins and growth-inhibiting Nogo-A can use EHD4/pincher-dependent endocytosis at the growth cone for their respective retrograde signaling. In addition to retrograde transport of endosomes, anterograde transport to axons in endosomes also occurs for several receptors, including the axon-outgrowth promoting cell adhesion molecule L1/NgCAM and TrkA. L1/NgCAM also depends on EHD4/pincher-dependent endocytosis for its axonal polarization. In this review, we will focus on receptors whose trafficking has been reported to be modulated by the EHD4/pincher family of endosomal regulators, namely L1/NgCAM, Trk and Nogo-A. We will first summarize the pathways underlying the axonal transport of these proteins and then discuss the potential roles of EHD4/pincher in mediating their endocytosis.
Neuronal function requires that proteins are spatially segregated to axonal and somatodendritic domains. Regulating which proteins are present at what levels in which locations is a major task for neurons and necessary for modulating neuronal morphology and physiology. One of the major mechanisms of adjusting receptor levels and localization is by membrane trafficking through endosomes (1). Consequently, it is not surprising that a large number of neurological pathologies result from disturbances of membrane traffic, including endosomes (2). Understanding the endosomal system thus goes hand-in-hand with understanding neuronal function. The endosomal system in non-neuronal cells is relatively well understood; after endocytosis into early endosomes, receptors are either trafficked to the late endosome and then to the lysosome for degradation, or sorted to recycling endosomes from which they can return to the plasma membrane. Recycling can also take place from the early endosome. Endosomes in neurons, on the other hand, are not well understood, nor are the neuronal functions of the proteins known to regulate endosomal transport in non-neuronal cells. This review focuses on the roles of endosomal trafficking in regulating the localization and signaling outcomes of several axonal receptors important in axon outgrowth, focusing on those receptors for which endosomal involvement has been strongly implicated. For many other important receptors, trafficking and endosomal transport has not yet been well studied, and we will not discuss those other receptors in this review.
There are multiple routes for accumulation of axonal membrane proteins on the axonal surface: In the direct transport pathway, axonal cargos are first segregated and packaged into different carriers, presumably at the level of the trans-Golgi network, and then separated from those carriers destined for the somatodendritic domain and targeted into the axon. However, experimental evidence for direct sorting from the TGN is still scarce. On the other hand, evidence for indirect axonal transport by transcytosis, whereby newly synthesized axonal proteins are first delivered to the somatodendritic domain, followed by endocytosis and transport to the axon in endosomal carriers, is accumulating. Transcytosis appears to be the primary pathway for axonal targeting of the cell adhesion molecule L1/NgCAM (3), and has also been proposed for cannabinoid receptor CB1R (4), Caspr2 (5), integrins (6), and the neurotrophin receptor TrkA (7).
In addition to long-range anterograde transport from the soma to the axon in endosomes, endosomal transport and recycling also takes place on the retrograde route (axon tip to soma) or locally at the growth cone (see Figure 1). Local endocytosis and recycling can rapidly modify receptor levels in response to local signals and also activate signaling cascades that modify growth cone behavior locally. Retrograde transport of endosomes can lead to degradation of receptors or long-range signaling events in the soma, including changes in gene expression. Recent work trying to identify the molecular machinery for retrograde transport of neurotrophin receptors and for Nogo-A has implicated an endosomal regulator, called EHD4/pincher. Our lab showed that EHD4/pincher also plays a role in endosomal anterograde transport of L1/NgCAM from the soma to the axon. Since EHD4/pincher-mediated uptake is a specialized, regulated, yet incompletely understood pathway, we compare and contrast in this review the known trafficking events for the three currently known EHD4/pincher-cargos, L1/NgCAM, Trk, and NogoΔ20 receptor.
The L1 cell adhesion molecule belongs to the immunoglobulin superfamily of adhesion molecules. NgCAM is the name for the chick homolog of L1. L1 is highly enriched on the axonal surface of developing and regenerating axons, and is involved in axonal pathfinding and guidance, axonal branching, and myelination. The importance of L1 is underscored by the fact that mutations in human L1 are responsible for an X-linked recessive neurological disorder (CRASH) that presents with varying penetrance of hydrocephalus, mental retardation, and hypoplasia of long axonal tracts (8–10). L1 is a single-pass transmembrane protein containing a large extracellular domain and a short cytoplasmic tail. The extracellular region mediates homophilic and heterophilic binding and elicits signaling essential for L1-dependent neurite outgrowth (11–13). The cytoplasmic tail of L1 binds a number of intracellular partners, including the clathrin endocytosis adaptor AP-2, ankyrin, and members of the ezrin-radixinmoesin (ERM) family, among others (14–17). Binding to AP-2 and subsequent endocytosis is required for growth cone advance on L1 substrates. Importantly, binding to both AP-2 and ankyrin, and likely also to ERM, is regulated by phosphorylation. Surprisingly, in vivo many of the roles of the cytoplasmic tail can be compensated for by unknown mechanisms (18).
Studies by our lab showed that L1/NgCAM travels to the axons via an indirect transcytosis pathway: newly synthesized L1/NgCAM is first delivered to the somatodendritic domain, followed by endocytosis and transport to the axon from somatodendritic endosomes (3, 19). Sorting of L1/NgCAM from the TGN to the somatodendritic domain is mediated by a previously described tyrosine-based endocytosis motif (YRSLE) (14, 20) in the cytoplasmic tail (21). The YRSLE motif is therefore critical for somatodendritic targeting and for subsequent endocytosis. In addition, a 15 amino acid glycine- and serine-rich stretch in the cytoplasmic tail of L1/NgCAM serves as an axonal targeting signal (21). Surprisingly, L1/NgCAM contains a second sufficient axonal targeting signal in the extracellular domain (22). The presence of the second signal improves axonal targeting. A biological function for L1/NgCAM transcytosis to the axon is not immediately apparent. It is possible that the transient appearance of L1/NgCAM on the somatodendritic surface serves a purpose, such as signaling in dendrites, binding of a ligand, or a guidance or feedback function that coordinates dendrite and axonal growth. Given the fact that the YRSLE motif is subject to regulation by phosphorylation (23), transcytotic routing might be regulated. To what ultimate end is an intriguing question to be pursued (1).
The neurotrophins are a family of closely related proteins which promote survival of neurons and regulate many aspects of neuronal development and function, including synapse formation and synaptic plasticity (24). There are four neurotrophins expressed in mammals, i.e. NGF, BDNF, NT-3 and NT-4. In mammals, there are three members of Trk receptor tyrosine kinases that bind to specific neurotrophins, i.e. TrkA, Trk B and TrkC (24, 25). Trk receptors are single transmembrane proteins harboring a tyrosine kinase domain in their cytoplasmic region. Activation of Trk receptors is stimulated by neurotrophin-mediated dimerization and trans-phosphorylation of an activation loop tyrosine (26). Trk auto-phosphorylation leads to association with proteins such as Shc, FRS, raps, phospholipase C-g, and activation of downstream signaling cascades (24). Upon ligand binding, Trk receptors at axon terminals undergo endocytosis and are transported retrogradely along the axon into the cell soma, a transport process that is critical for mediating retrograde survival signaling (27). Each neurotrophin also activates the p75 neurotrophin receptor (p75), but with lower affinity. p75, the first identified neurotrophin receptor, is a member of the tumor necrosis receptor superfamily (TNFR). Engagement of p75 by neurotrophins and its association with signaling adaptor proteins, including Traf6, NRAGE and RhoGDI, activates signaling pathways that are important for regulating neuronal survival, differentiation and synaptic plasticity (24). In addition to interacting with Trk receptors, p75 serves as a co-receptor for multiple other signaling receptors (reviewed in (28, 29)).
In neurons, Trk receptors are localized to diverse subcellular compartments, including the plasma membrane of dendritic spines, axon terminals, dendritic shafts and cell bodies, and are also localized intracellularly. Trk receptors are synthesized in the cell body, then transported into the axon via anterograde transport (30). Some of the molecular machinery responsible for anterograde transport of TrkB has been identified. Arimura et al (31) demonstrated that CRMP2, collapsin response mediator protein-2, is involved in the anterograde transport of vesicles containing TrkB. TrkB forms a complex with Slp1/Rab27B/CRMP2 in a Slp1/Rab27-dependent manner. Slp1 is an effector of Rab27 and a CRMP-2 interacting molecule and binds directly to the cytoplasmic tail of TrkB. Slp1 thus serves as an adaptor linking TrkB and Rab27B to CRMP2. Since CRMP2 interacts with kinesin light chain, it is capable of directly linking the TrkB/Slp1/Rab27B complex to kinesin 1 for anterograde axonal transport (31). In addition, sortilin has been implicated in mediating anterograde transport of Trk receptors to the axon (32).
A recent study by Ascano et al (7) suggests that Trk receptors are sorted anterogradely to the axon terminals in endosomes via a transcytotic pathway. Using compartmentalized sympathetic neuronal cultures, they demonstrated that TrkA receptors on the surface of dendrites and cell soma undergo constitutive endocytosis and local recycling. In addition to stimulating endocytosis of Trk in the cell soma, neurotrophins also promote anterograde transcytosis of Trk receptors to the axon and rapid exocytosis of Trk to the cell surface of growth cones. Live cell imaging showed that endocytosed Trk receptors are transported via Rab11-positive recycling endosomes in axons. Most importantly, endocytic recycling of TrkA is required for modulating the neuronal sensitivity to NGF-dependent signaling and axon growth. In what ways the rab11-dependent transcytotic route and the Slp1/rab27 pathway for anterograde transport of Trk are related to each other, remains to be established.
Anterograde axonal transport in endosomes is not yet well understood and many questions are still unanswered. One open question is whether L1/NgCAM and Trk share endocytic machinery during transcytosis to the axon. One common feature of the transcytotic pathway shared by both molecules is that their endocytosis at the somatodendritic domain is dynamin-dependent. Although dynamin is known to be involved in clathrin-mediated endocytosis, it also participates in clathrin-independent endocytic pathways. Whereas clathrin and dynamin can be recruited to L1/NgCAM vesicles via the clathrin-adaptor AP-2 (14), it is unknown whether TrkA endocytosis from the somatodendritic domain (7) takes place via the AP-2/clathrin pathway since no apparent direct interaction between TrkA and AP2/clathrin has been reported. The neuronal-specific endosomal regulator NEEP21 (neuronal early endosomal protein 21kd) has been implicated in L1/NgCAM trafficking to axons (see below)(19) and the recycling regulator rab11 in TrkA trafficking to axons, but to what extent the two cargos use the same pathway and machinery is currently not known.
Some of the endosomal compartments and regulatory machinery involved in transcytosis of endocytosed L1/NgCAM in neurons have been identified. Internalization of L1/NgCAM for trafficking to axons from the soma appears to utilize a specialized internalization pathway mediated by EHD1 (RME1) and EHD4 (pincher) in neurons (33). The EHD family, comprised of EHD1-EHD4, plays crucial roles in endosomal trafficking. In non-polarized cells, EHD1, EHD3, and EHD4/pincher function in endosomes and regulate trafficking through early and recycling endosomes (34). EHD1 might also regulate endocytosis of IGF1 receptor (35). In neurons, EHD4 (36) has been implicated in Trk receptors endocytosis, rather than (or in addition to) recycling (37). We found that overexpression of EHD1 and EHD4 impairs axonal polarization of L1/NgCAM and leads to impairment of L1/NgCAM internalization in neurons, but not in fibroblasts. Transferrin internalization is unaffected by overexpression of EHD1, suggesting cell-type and cargo-specific roles for EHD1 in endocytosis. At longer overexpression times of EHD1, L1/NgCAM endocytosis returns to normal levels due to the rapid up-regulation of compensatory endocytic pathways. Rapid compensation of L1 endocytosis has been observed previously (38), but the compensatory pathway is not known. L1/NgCAM endocytosis requires oligomerization of EHD1. Furthermore, co-expression of EHD4 and EHD1 rescues the L1/NgCAM endocytosis defect observed by expression of each alone, suggesting they act as hetero-oligomeric complexes in this pathway. It is still unclear whether EHD1/EHD4-mediated L1/NgCAM endocytosis also involves AP-2/clathrin and/or dynamin. Interestingly, the AP-2 binding protein numb has been shown to bind EHD proteins (39), but whether EHDs and numb are on the same endocytic pathway for L1/NgCAM at the same locations needs to be tested directly.
Once L1/NgCAM is internalized, it colocalizes transiently with transferrin in EEA1-containing somatodendritic early endosomes, and subsequently recycles preferentially to the axonal plasma membrane. Interestingly, transcytosis of L1/NgCAM requires a neuronal-specific early endosomal compartment, marked by NEEP21 (19). NEEP21 is highly expressed in neurons, localizes to the somatodendritic domain and is absent from axons (40). Endocytosed L1/NgCAM accumulates in large, stationary NEEP21-endosomes, but is rarely transported in NEEP21-containing vesicles (19). The roles of NEEP21-endosomes and regulation of transport via NEEP21-endosomes are not yet well understood, but NEEP21 has been implicated in trafficking of GluR2, neurotensin receptors, and βAPP (41–43).
EHD4-dependent endocytosis pathways are not well characterized, but seem to be used by only a subset of receptors. Intriguingly, EHD4-dependent endocytosis has so far only been observed in neurons, but not in fibroblasts where EHD4 is implicated instead in transport from early endosomes to recycling endosomes and lysosomes. The first discovery of EHD4/pincher in endocytosis pathways in neurons was made by Simon Halegoua’s group by identifying EHD4/pincher as an NGF-induced protein in PC12 cells (36, 44). They showed that TrkA-mediated survival signaling depends on EHD4/pincher-dependent endocytosis. The EHD4/pincher pathway appears distinct from clathrin-mediated endocytosis, but also from macropinocytosis, leading to the term “macroendocytosis” to designate the pincher-dependent uptake pathway. Trk receptors were thus the first identified cargo for EHD4/pincher-endocytosis. Recently, two more cargos were described, L1/NgCAM (as detailed above) and Nogo-A (45) (see below). Nogo-A is a myelin-derived inhibitory molecule and mediates growth cone collapse, whereas L1/NgCAM and Trk signaling promote axon outgrowth. The EHD4-machinery is thus shared among Trk, L1/NgCAM and Nogo-A, which raises intriguing questions about potential cross-talk and co-regulation of these distinct receptor systems and their associated signaling cascades to regulate the extent and timing of axon outgrowth.
he location, levels, and residence time of adhesion and guidance receptors crucially influence their functional activity (46) and hence axon guidance and growth. In neurons, many trafficking events are not constitutive, but regulated by extracellular signals or electrical activity. Regulating fusion and endocytosis of guidance receptors is likely important for locally controlling the levels of guidance receptors in growth cones. These local mechanisms operate in addition to the long-range trafficking events for targeting to the axon, described above. The occurrence of local regulation of exocytosis, endocytosis and tethering has been shown for L1/NgCAM in axonal growth cones, but we are only at the beginning of unraveling the molecular pathways by which signaling through guidance receptors and through other pathways regulate insertion, diffusion, and removal of guidance receptors locally (1).
L1/NgCAM-mediated axon outgrowth, at least in culture, depends on endocytosis at growth cones, which leads to activation of the MAPK pathway, PI3-kinase, and Rac for cytoskeleton rearrangements in the growth cone (47, 48). Internalization of L1/NgCAM occurs preferentially in the central domain of migrating growth cones, followed by anterograde transport of L1/NgCAM-containing vesicles, and subsequent recycling near the leading edge (48, 49). The neuronal form of L1/NgCAM binds AP-2, presumably mediating the clathrin-mediated endocytosis of L1/NgCAM. A study by Nishimura (50) showed that the monomeric adaptor Numb co-localizes with L1/NgCAM at the central region of growth cones and co-immunoprecipitates with L1/NgCAM and AP-2 from brain. CRMP2, which is critical for regulating axon growth, regulates Numb-mediated endoyctosis of L1/NgCAM and L1/NgCAM-mediated axonal outgrowth. These results suggest that L1/NgCAM might utilize a specific endocytic pathway locally at axonal growth cones. Interestingly, CRMP-2 is not co-immunoprecipitated with L1/NgCAM, suggesting that CRMP-2 interacts with Numb–AP-2, but not with the Numb–AP-2–L1/NgCAM complex. This finding raises the question if the dissociation of CRMP2 from Numb might be a key step for regulating local L1/NgCAM internalization for growth cone advance (50).
Removal of a receptor from the surface is not the only consequence of endocytosis, but rather elaborate signaling cascades are activated downstream of endocytosis (51). Little is known about the endosomes from which adhesion receptors signal, but signaling outcomes can differ in important ways depending on the endocytic route taken (clathrin-dependent or – independent) and the subsequent post-endocytic trafficking (recycling endosome, late endosome, etc.) (for examples see (52, 53)). For L1/NgCAM-mediated axon growth in vitro, proper sorting and recycling of L1/NgCAM after endocytosis is critical. Recycling of L1/NgCAM in growth cones is regulated by phosphorylation to ensure endocytosed L1/NgCAM is sorted into recycling endosomes, but not into lysosome/late endosomes. Inhibition of casein kinase II or mutation of serine1181 causes mislocalization of endocytosed L1/NgCAM into unidentified transferrin-negative endosomes and leads to a defect in axon growth on L1/NgCAM substrate, but not on other substrates (54). Furthermore, the residence time in endosomes can also be regulated and greatly influence signaling outcomes (55). These events downstream of the endocytosis event per se thus will be important topics of study in the future.
Endocytosis and retrograde transport of Trk-neurotrophin complexes play major roles in conveying Trk signals from distal axons to the cell body. Upon activation by neurotrophins at growth cones, activated Trk receptors can be internalized via several routes: clathrin-mediated, caveolin-dependent (53, 56, 57), and EHD4/pincher-mediated macroendocytosis (36, 44, 58). The exact contribution of each of these endocytic pathways to Trk signaling is not entirely clear and multiple pathways might be used in different neurons. In addition, different Trk receptors as well as p75NTR might be regulated differently. In addition to potential cell type-, and receptor type-differences, endocytic pathways are capable of rapid and extensive compensation. Therefore, pinpointing the exact contribution of different endocytic pathways can be challenging.
The Halegoua lab identified the EHD4/pincher pathway (see above) necessary for sustained endosomal Trk signaling in neurons (58). This novel endocytic process is dynamin-, but not clathrin-dependent. Trk association with clathrin structures has been reported in other studies, and might be subject to regulation (53, 59). Overexpression of EHD4/Pincher increases the internalization of TrkA, and enhances recruitment of phospho-ERK5 but not Erk1/2, upon addition of NGF (36, 44). Immunofluorescence and immuno-EM analysis after expression of HA-tagged pincher demonstrated that HA-EHD4/Pincher mediates the formation of Trk-containing macroendosomes in soma, dendrites and axons of both sympathetic and hippocampal neurons. These endosomes are likely derived from membrane ruffles and might be directly processed to multivesicular bodies, bypassing lysosomal degradation. Since reliable antibodies against EHD4/pincher suitable for immunolocalization are difficult to obtain, the localization of endogenous pincher on the ultrastructural level remains to be determined.
To mediate a sustained retrograde signal, the internalized Trk endosomes must bypass degradation or local recycling, and selectively recruit specific signaling complexes to regulate gene expression (44, 58). Rab proteins regulate the fate of many internalized receptors, including neurotrophin receptors. Rab5 and APPL proteins (Rab5 effectors) have been implicated in Trk intracellular trafficking and signaling (60–62). Activated Trk, Rap, and Erk1/2 were recovered from the same endosomal fraction enriched with Rab5 and EEA1 from NGF-treated PC12 cells (63). Rab5 was found to mediate Trk macroendosomal processing, co-localizing with EHD4/Pincher and TrkA at membrane ruffles and at the peripheral membrane of immature multivesicular bodies (MVB). Studies using PC12 cells showed that NGF treatment activated Rab5 GTPase-activating protein, Rab5GAP, to suppress the cellular levels of active GTP-Rab5, thereby prevented subsequent conversion of early endosomes into late endosomes (i.e. Rab5 to Rab7)(64, 65). Thus, it has been proposed that the persistent association of Rab5 might help prevent Trk-containing MVBs from Rab7-mediated lysosomal processing, thereby allowing for sustained long term Trk endosomal signaling (58, 66).
In work from the Schiavo lab, Rab7 GTPase, which regulates cargo progression from early to late endosome, has been implicated in regulating p75 and TrkB signaling and axonal transport of neurotrophin signals (53, 67). This study showed by elegant live imaging that Rab5 regulates an early sorting step preceding axonal transport, whereas Rab7 regulates fast long-range retrograde axonal transport in motoneurons. Rab signaling is therefore required to engage long-range transport over local recycling. The exact nature of Rab5 vs Rab7 regulation is currently still unclear. It is not certain under what circumstances Rab5 remains associated with retrograde signaling endosomal carriers. Also the exact nature of the retrograde carrier(s) in not clear, since some groups describe vesiculotubular moving compartments (67) whereas others observe multi-vesicular bodies by electron microscopy (66). Since these multivesicular carriers still contain rab5, they likely represent early endosomal intermediates or immature MVBs. It is possible that different neurons differentially rely on Rab5 vs Rab7 for generating the retrograde carrier, or that multiple diverse populations of carriers are generated, some of these still containing Rab5 whereas others do not (66). Multiple factors may influence the detected differences, for instance the type of neurons used, the types and abundance of receptors and co-receptors present, the type and concentration of neurotrophin used, as well as developmental differences relating to the maturity of the neurons. What is clear is that neurotrophin-containing signaling endosomes escape acidification (68) and degradation (66) but the exact molecular machinery and regulation for this event is still to be fleshed out.
Although p75 has been indicated to be involved in retrograde transport of neurotrophins, the molecular mechanisms underlying the axonal trafficking of p75 remain unclear. Endocytosis of p75 complex upon ligand stimulus has been observed in various cell types. Depending on the cell-type and ligands present, p75can be internalized via clathrin-dependent or – independent pathways (53, 69–72). In PC12 cells, p75 endocytosed independently of TrkA into early endosomes, indicating the existence of p75-containing signaling endosomes that are distinct from those of Trk receptors (70). Internalized p75 complex has been shown to travel retrogradely in rab7-containing endosomes that are shared with tetanus neurotoxin (67, 68). However, detailed information on what endosomal compartments these p75 endosomes traverse to reach the cell body is still lacking. In addition, whether these p75-endosomes act as signaling endosomes to mediate retrograde signaling pathways that are important for neuronal survival/apoptosis remains to be examined.
Nogo-A (reticulon 4) is a member of the reticulon family, closely associated with the endoplasmic reticulum (73, 74). By alternative splicing, the Nogo gene gives rise to three major isoforms, i.e. Nogo-A, B, and C (74–76). Nogo-A is the best-characterized of the Nogo proteins. It is expressed in oligodendrocytes in the adult and in neurons during development, localizing mostly to the ER (77, 78). Only a small fraction of Nogo-A translocates to the plasma membrane in myelin forming oligodendrocytes (78). The cell surface-associated Nogo-A and soluble Nogo-A fragments are believed to bind to receptors in trans and elicit signaling leading to growth cone collapse and inhibition of neurite outgrowth (79). Nogo-A exerts its inhibitory activity at growth cones by upregulating RhoA activity. High RhoA activity confers a block in neurite outgrowth, whereas inhibiting RhoA activation facilitates nerve regeneration (80).
Nogo-A is a large protein consisting of three major domains, NiR, NiG, and reticulon homology domain (RHD). There are at least two distinct regions in Nogo-A that confer inhibitory activity on neurite growth. The NiG region in Nogo-A mediates its inhibitory activity by binding to integrins (81). The 66 amino acid loop domain Nogo-66, spanning the two hydrophobic regions within the RHD, binds to two neuronal receptor proteins, NgR1 (82) and PirB (83). Binding of Nogo-A to NgR1 via its Nogo-66 domain elicits Nogo-signaling and inhibits neuronal growth. Upon ligand binding, NgR1 forms signaling complexes with LINGO-1 and either p75NTR or TROY (84–86). This signaling complex is primarily responsible for mediating RhoA activation. Two other myelin-associated inhibitors, myelin associated glycoprotein (MAG) and the oligodendrocyte myelin glycoprotein (OMG) exert their inhibitory activity through interaction with NgR (87, 88). Nogo-A knockout mice show a moderate increase in regeneration and compensatory fiber growth after lesion, while neutralizing Nogo-A by function-blocking antibodies induces neuronal growth after spinal cord injury in rodents and primates (89–91). Nogo-A knockout mouse lines vary significantly with regard to regeneration of severed axons, so Nogo- A appears to be not the only potent inhibitor of nerve regeneration to be overcome. In contrast to its inhibitory properties, Nogo-A plays important roles in neuronal migration and regulation of synaptic plasticity during development when NgR1 is absent.
So far, little is known about the intracellular trafficking of either Nogo-A or its receptors. Similarly, the molecular mechanism underlying Nogo-A signaling within growth cones is only beginning to be unraveled. A recent study by Joset et al (45) showed that an active Nogo-A fragment, NogoΔ20, endocytoses rapidly in neurons and colocalizes with EEA1-positive early endosomes. Interestingly, NogoΔ20 internalization occurs independently of dynamin, clathrin, and cholesterol. However, overexpression of the dominant-negative EHD4G68E mutant inhibits the internalization of NogoΔ20, similar to the inhibition of Trk endocytosis. Interestingly, EHD4G68E blocks endocytosis of NogoΔ20 and prevents Nogo-induced, but not semaphorin3A-induced growth cone collapse. This suggests that Nogo-induced growth inhibition requires EHD4/pincher-mediated endocytosis, whereas semaphorin3A-induced growth cone collapse uses different endocytic machinery (see for example also (92)). Similar to EHD4/Pincher-dependent Trk endocytosis, Rac is also involved in the internalization of Nogo-A.
Importantly, the authors find that NogoΔ20 signaling appears to occur by retrograde transport of signaling endosomes and changes in nuclear transcription programs: internalized NogoΔ20 is transported retrogradely in vesicles from distal axons to the cell bodies in compartmentalized cultures. The retrograde transport of NogoΔ20-containing vesicles is EHD4/Pincher dependent, and activated RhoA is co-transported. Subsequent retrograde signaling of Nogo then results in a marked decrease of pCREB in the cell body and growth inhibition. Since NogoΔ20 internalization in this study is NgR1-independent, it is important to address if EHD4/Pincher-dependent endocytosis is also involved in the inhibitory signaling pathways via NgR1. Another study showed that Nogo-A mediates its inhibitory activity by binding to integrins and triggers inhibition of integrin signaling (81). Whether EHD4/Pincher-mediated NogoΔ20 endosomes contain integrin remains to be investigated. In addition, it is unknown if MAG-, or OMGp-mediated growth inhibition require EHD4/Pincher-mediated endocytosis.
A major open question concerns the singularity or multiplicity of EHD4/pincher pathways. Several observations to date differ with respect to the three EHD4/pincher cargos, raising the possibility that there are multiple, distinct EHD4/pincher-mediated entry routes. For instance, we recently found that EHD4/Pincher regulates the internalization of L1/NgCAM at dendrites. EHD4/pincher-mediated endocytosis of Trk and NogoΔ20, on the other hand, is proposed to play a role as well at axonal terminals. It is not known if EHD4/pincher also contributes to transcytosing Trk from the somatodendritic domain to the axon. Similarly, the roles of EHD4/Pincher in L1/NgCAM endocytosis at axonal growth cones for adhesion and activation of MAPK pathway for neurite outgrowth is still unknown. Whether EHD4/pincher-dependent macroendocytosis takes place all over the neuronal plasma membrane, or is differentially regulated at different locations, therefore, is unknown. It is also not entirely resolved to what extent the EHD4/pincher pathway uses clathrin, caveolin, or neither.
Another open question is whether other EHD proteins co-regulate EHD4/pincher endocytosis via different hetero-oligomeric complexes. We reported that either EHD4 or EHD1 overexpression impairs L1/NgCAM internalization, which can be rescued by co-expression of EHD1 and EHD4. This observation suggests that overexpression of either EHD1 or EHD4 might disrupt the balance of the levels of the two proteins leading to mis-assembly of an EHD1/4 hetero-oligomer complex. Thus, EHD1/4 hetero-oligomers might be involved in mediating L1/NgCAM endocytosis. In contrast, overexpression of EHD4 promotes, rather than inhibits, internalization of both TrkA and Nogo-A proteins, suggesting that EHD4 in Trk- and Nogo-macroendocytosis might function as homo-oligomers. Thus, we speculate that L1/NgCAM endocytosis involves hetero-oligomers of EHD1/4, whereas EHD4 homo-oligomers might be responsible for Trk and Nogo-A internalization. It should be noted that TrkA actually contains an NPF motif which could in principle bind directly to the EH domain of EHD4. L1/NgCAM on the other hand does not have an NPF motif, so the mechanisms of regulation likely differ for TrkA and L1/NgCAM. It is likely that the functions and roles of EHD proteins in endocytic trafficking are cell-type-, cargo-, and oligomer – dependent and need to be worked out for each specific cargo and cell type.
During injury, damage to myelinated fibers leads to the release of growth-inhibitors, including Nogo, MAG and OMgp, and results in an unfavorable growth environment that impedes axon regeneration. Growing evidence showed that there are several potential ways to overcome the inhibitory effect of growth inhibitors. Treatment with a soluble peptide fragment of NgR1 that binds and blocks all NgR ligands promotes functional regeneration of sensory neurons after dorsal root crush (93). Overexpression of GAP-43 in transgenic mice enhances axon growth in injured Purkinje axons by modifying the subcellular localization and exposure of NgR on the neuritic membrane (94). Interestingly, if first primed with neurotrophins, primary sensory neurons and cerebellar granular cells are capable of extending their axons across purified MAG (95, 96). L1/NgCAM has also been implicated in CNS regeneration in adult rodents (97, 98).
Although EHD4/Pincher-mediated endocytosis of TrkA and NogoΔ20 requires Rac, the resulting retrograde signaling gives rise to two opposing responses, i.e. an increase in pCREB for growth induction or a decrease in pCREB for growth inhibition. Via Rac, Nogo-A activates RhoA and leads to decreased pCREB levels, whereas Trk-signaling activates Ras-Erk pathway and increases pCREB levels in the cell body. Activation of CREB by elevating cAMP level is sufficient to overcome the effects of Nogo/myelin inhibitors in adult neurons (95, 99). In developing DRG neurons, MAG was seen as attractant cue for growth due to high levels of cAMP (100). On the other hand, decreases in cAMP levels convert neurotrophins from attractant to repulsive cues, indicating the level of cAMP is a key signal in determining the downstream biological outcomes (101, 102). Understanding the mechanisms and conditions that promote EHD4/pincher-mediated ligand-receptor complex endocytosis or other endocytic pathways might provide links to improving treatments for CNS injury. In addition, identification of associated intracellular signals, and understanding how signals from growth inhibitors are integrated with outgrowth promoting signals might further facilitate the development of interventions to improve the outcome of CNS injury.
We thank expert reviewers for insightful comments on the manuscript. Research in our lab has been supported by National Institutes of Health Grant GM086913 and American Heart Association AHA 10GRNT3770063.