Attraction Attractive response silenced
Figure 8 Modulatorsofaxon navigation. Examples of ways in which responses of axons can be modulated during development (see text for description of each example): a, regulated gene expression; b, post-transcriptional control; c, local translation; and d, receptor interaction. Modulation of axon guidance using such methods allows for the generation of diverse outcomes using a small suite of guidance cues.
been identified in the initial axon segment (Steward and Ribak, 1986) and intermittently along the axon shaft (Pannese and Ledda, 1991; Koenig et al., 2000) using both biochemical and immunohistochemical methods (Bassell et al., 1998; Campbell and Holt, 2001). Other components necessary for translation have also been identified in vertebrate axons (reviewed in Alvarez et al., 2000).
What is the role of protein synthesis in axons and does it play a role in aspects of axon guidance? A key finding with respect to these questions is that axon guidance molecules, like netrin-1 and sema-phorin 3A, can trigger protein synthesis in retinal growth cones isolated from their cell bodies within 5-10 min. Moreover, inhibition of translation blocks the chemotropic responses of retinal growth cones to netrin-1 and Sema3A in vitro (Campbell et al., 2001) and local protein synthesis is necessary for the normal turning responses of Xenopus retinal and spinal axons toward gradients of attractants in vitro (Ming et al., 2002). By synthesizing specific receptor/effector proteins locally, an axon can potentially alter its responsiveness to previously encountered signals or gain responsiveness to cues expressed on its subsequent trajectory. Such a mechanism may be used by chick commissural axons, which express high levels of EphA2 on the postcrossing region of the axon. By linking mRNA for the marker protein GFP to the 3' untranslated region (3'UTR) of EphA2 mRNA and driving its expression in commissural neurons, it has been shown that commissural axons specifically translate GFP only in the region extending beyond the mid-line. This suggests that a region in the EphA2 3'UTR promotes selective translation of EphA2 in distal axonal segments. EphA2 may be used later in guidance, conferring sensitivity to ephrinA ligands, though this remains to be shown (Brittis et al., 2002). The key questions in all of these studies are: which proteins are translated in response to various axon guidance cues and how does their translation control the chemotactic turning behavior of a growth cone? A local contact with a guidance cue, for example, through a single filopodium, may induce fast and specific change in protein levels (synthesis and/or degradation) within the growth cone, leading to asymmetric cytoskeletal rearrangements and turning, as has been shown in dendrites (Steward and Schuman, 2003). However, the role of local protein synthesis stimulated by guidance cues in the growth cone is still largely unknown (Figure 8c; reviewed in van Horck et al., 2004).
22.214.171.124 Local Protein Degradation and Endocytosis at the Growth Cone
Xenopus retinal growth cones contain the machinery for proteosomal degradation (proteasome proteins, ubiquitin, and ubiquitinating enzymes). Moreover, inhibitors of proteasome function can block the chemotropic response of growth cone to guidance cues, such as netrin-1 and lysophosphati-dic acid (LPA) (Campbell and Holt, 2001). An increase in ubiquitin-conjugated proteins within 5min is induced by these guidance cues, indicating that they rapidly stimulate the degradation pathway. In addition, the protein caspase-3, a marker of the apoptotic pathway, has been found to be involved in chemotropic guidance in retinal axons (Campbell and Holt, 2003). Of interest is that one of the known cleavage targets of caspase-3 is the translation initiation factor eIF4G (Clemens et al., 1998), raising the possibility that guidance cues, such as netrin-1 and brain-derived neurotrophic factor, which simultaneously activate both translation and caspase pathways, can downregulate the synthesis of proteins that they stimulate using a negativefeedback loop. Furthermore, in Drosophila, the pruning of the g neuron axonal projections requires protein degradation. Indeed, mutations of an ubi-quitin-activating enzyme or proteasome subunits prevent normal pruning (Watts et al., 2003).
Another mechanism involved in growth cone guidance is the removal and/or inactivation of activated receptors. To advance along their way, growth cones have to break previous interactions with the surface. Several mechanisms are involved in this process, including proteolytic cleavage (Hattori et al., 2000), transreceptor silencing (Stein and Tessier-Lavigne, 2001), and endocytosis. Recently, it has been shown that metalloproteases regulate in vivo the growth and guidance of retinal growth cones (Webber et al., 2002). Moreover, it has been shown that endocytosis of functional ephrinB/EphB complexes promote cell detachment of the interacting cells in vitro (Marston et al., 2003; Zimmer et al., 2003). Together, these data suggest that endocytosis might be a fast mechanism for ending the adhesive contacts between growth cones and neighboring cells.
Interaction between receptor proteins at the surface of the growth cone has been shown to be a remarkably efficient means by which the complexity of axonal responses to guidance cues can be magnified. The combinatorial assembly of heteromeric receptor complexes has a number of potential benefits, such as receptors with distinct signaling properties being used to potentiate the activity of other guidance ligands. The ability to modulate receptor activity via the expression of coreceptors is evident in instances of Semaphorin signaling. Sema receptors are often complexes of different proteins, and this confers the potential to modulate responses to a small number of cues with great subtlety. For example, the responses of dorsal root ganglion (DRG) neurons to Sema3A are modulated by the CAM, L1-CAM, which is also expressed on DRG axons (Castellani et al., 2000).
As well as modulating axonal responses, co-receptor expression can directly convert responses to guidance cues. As mentioned previously, netrin-1s are bifunctional. When signaling via UNC40/ DCC, netrin-ls act as attractants. However, when axons express UNC-5, this response is converted to repulsion. Netrin-1 can interact with UNC-5 directly to mediate repulsion, and UNC-5 can also bind DCC though its cytoplasmic domain, which essentially silences DCC-mediated attraction.
As a growth cone navigates through the terrain of the developing nervous system, it undoubtedly encounters multiple guidance cues simultaneously. At the choice point of the spinal cord floorplate, for example, a commissural growth cone will 'see' both netrin-1 and Slit, each of which is promoting a completely opposite reaction. How is such information successfully processed within the axon? Here too, interactions between receptors may be pivotal, as in vitro experiments with Xenopus spinal neurons suggest. Axons from these neurons turn and grow toward a local source of netrin-1 in culture. Application of Slit blocks netrin-1-mediated turning, yet axon growth mediated by netrin-1 is not affected. Silencing of attraction occurs via a direct interaction between the cytoplasmic domains of DCC and Robo. This specific silencing of netrin-1-mediated attraction by the Robo receptor is an elegant example of signal integration within the growth cone, demonstrating how encounters with concurrent signals may be ordered and prioritized to generate a directional response (Figure 8d; Stein and Tessier-Lavigne, 2001).
The degree of plasticity endowed by the above mechanisms enables developing axons to regulate their sensitivity to extrinsic guidance cues in a highly specific manner. Thus, in addition to responding in a stereotyped fashion to attractants or repellents, growth cones can navigate in environments which may not intuitively appear to be conducive for guidance, such as extending toward repulsive targets (e.g., Slit-expressing midline) or moving on from attractive intermediate targets (e.g., netrin-1-expressing ONH). Such changes in axon sensitivity to guidance cues have been described in many different neuronal populations (see above). Modulation of axonal responses has been particularly well documented for netrin-1. For instance, commissural axons from the rodent metencephalon lose responsiveness to this che-moattractant after crossing the midline (Shirasaki et al., 1998), and Xenopus retinal axons advancing along the visual pathway switch netrin-1-respon-siveness from attraction to repulsion over time (Shewan et al., 2002).
As well as gaining (or losing) responsiveness to guidance molecules, growth cones can also adjust their sensitivity to changing concentrations of such cues. This process, known as adaptation, enables growth cones to navigate in gradients of chemotro-pic molecules. Adaptation has been described in Xenopus retinal (Piper et al., 2005) and spinal (Ming et al., 2002) growth cones in vitro, where exposure to a low level of a guidance cue elicits an initial desensitization to additional exposure to the cue, subsequently followed by resensitization and a resumption of responsiveness. Growth cone adaptation is thought to increase the sensitivity of axons when in gradients of guidance cues, allowing them to respond to subtle differences in the environmental concentration of the cue as they proceed. For example, retinal axons in vitro are able to grow further up a gradient of a repulsive guidance cue when initially exposed to the cue as compared to those axons not exposed to the cue at the start of the assay (Rosentreter et al., 1998).
In summary, axon guidance cues can be modulated by a variety of intrinsic and extrinsic factors that ultimately regulate growth cone sensitivity and adaptation during development. Thus, by controlling the response of growth cones in a precise spatiotemporal fashion and by integrating and prioritizing coincidently encountered signals, the incredibly complex connections of the nervous system can be generated using a small repertoire of guidance molecules.
In recent years, our understanding of the development and function of the CNS has expanded significantly. With respect to axon guidance, one of the most important advances in our knowledge has been the identification of the main molecular families responsible for navigation, patterning, and target innervation: the netrin-1s, semaphorins, Slits and ephrins. Remarkably, despite millions of years of evolutionary divergence, most of the key molecules that mediate axon guidance during formation of the nervous system in vertebrates and invertebrates are highly conserved, as are the intracellular signaling pathways they activate. For example, the netrins act as critical determinants controlling attraction of commissural axons toward the midline in both flies and mammals, despite an evolutionary separation exceeding 600 million years. The way in which pathfinding axons navigate to distant targets by sensing the molecular characteristics of the local environment, thereby breaking their journey into small segments, is also very similar across a broad range of species. For instance, to find their correct target, the pioneer Ti1 axons of the grasshopper limb are guided by local cues or guidepost cells, that are spaced short distances apart, while many developing vertebrate axon tracts, such as the retinal pathway, have multiple points at which axons receive directional information to simplify their navigation en route to their target. Such pathway subdivision is an extremely efficient method of enabling axons to navigate in an error-free way to their targets, and also endows the pathway with multiple regulatory checkpoints that can be altered during development to enable differing guidance decisions to be made over time.
However, there are a number of features of axon guidance that do reflect the evolutionary divergence of the vertebrate and invertebrate lineages. One example of this lies in the number of ligand and receptor molecules acting to control axon navigation. Vertebrates generally possess a greater number of axon guidance molecules, perhaps due to the genome and chromosomal duplications that have occurred in the vertebrate lineage. For instance, the Eph and ephrins form large, highly homologous families within the vertebrate phyla, suggesting that a major expansion of these genes has occurred. This expansion may have occurred in tandem with the evolution of the vertebrate lineage, and could have been a critical factor facilitating the construction of increasingly complex brains. The broad range of molecular components that can comprise semaphorin receptor complexes also provides an example of strong evolutionary divergence, as these complexes differ widely between different species.
Was this article helpful?