An intermediate target for axons leaving the cerebral cortex (corticothalamic projections) in embryonic mammals is the ganglionic eminence (GE), the embryonic precursor of the basal ganglia. After corticotalamic projections meet thalamocortical projections in the GE, early thalamic and cortical projections might be guided by interactions with each other. The cues that direct corticothalamic axons over the initial portion of their trajectory are not well understood, but could include both short-range and long-range attractants and repellents. In the present study, evidence is provided that corticofugal axons might be guided at least partly by a diffusible factor or factors originating in the lateral GE and the sulcus between the lateral and medial ridges of the GE (ISS), as well as evidence implicating the axonal chemoattractant netrin-1 in mediating these effects. Explants of lateral GE and ISS obtained from E12.5 and E13.5 mouse forebrain have a strong effect on both the outgrowth and orientation of corticofugal axons when cultured at a distance with explants of embryonic cortex in collagen gels. Netrin-1 mRNA is detected in these target tissues by in situ hybridization; both netrin-1 protein and heterologous cells secreting netrin-1 can mimic the outgrowth-promoting effect of these target tissues in vitro. The growth of corticofugal axons is oriented toward an ectopic source of netrin-1 in vitro; a function blocking anti-netrin-1 antiserum specifically abolishes the cortical axon outgrowth elicited by explants of lateral GE and the ISS in collagen gel cocultures. Taken together, these results suggest a role for netrin-1 in the attraction at a distance of early cortical axons by the GE. Thus in mammals -- as is also observed in nematodes -- the development of non-commissural projections in anterior regions of the embryo might be directed by mechanisms similar to those involved in directing the development of commissural projections in more posterior regions of the central nervous system (Metin, 1997).

During development, growing motor axons are excluded from the ventral midline of the neural tube by diffusible chemorepellents emanating from this region. Molecular candidates for this chemorepellent activity include semaphorin D and netrin-1; the latter is known to repel trochlear motor axons. Cell aggregates secreting netrin-1 or semaphorin D were cocultured at a distance from tissue explants containing different motor neuron subpopulations. Cranial motor axons that project dorsally in vivo such as those of the trigeminal, facial and glossopharyngeal nuclei are repelled by both netrin-1 and semaphorin D. In contrast, ventrally projecting spinal motor axons and abducens axons are not affected by netrin-1. Spinal and abducens motor neurons also respond to semaphorin D. The ventrally projecting axons of oculomotor neurons are not repelled by netrin-1 or semaphorin D. Differential responsiveness to netrin-1 and semaphorin D could thus contribute to the generation of dorsal and ventral motor axon pathways during development (Varela-Echavarria, 1997)

The Kidney and retinal defects (Krd) mouse carries a 7-cM transgene-induced deletion on chromosome 19 that includes the Pax2 locus. After closure of the embryonic fissure, Pax2 immunostaining disappears from the ventral retina, but persists in a cuff of cells encircling the developing optic disc, the site where ganglion cell axons exit the retina. In Krd/+ embryos, Pax2+ cells in the posterior optic cup and the optic stalk undergo abnormal morphogenetic movements and the embryonic fissure fails to form normally. This results in an abnormal organization of the Pax2+ cells and ganglion cell axons at the nascent optic disc. The abnormal morphogenetic movements of the Pax2+ cells in the embryonic retina and optic stalk and the initial misrouting of the ganglion cell axons give rise to retinal and optic disc defects observed in the adult Krd/+ mice. These results indicate a requirement for full diploid expression of Pax2 for normal morphogenesis of a portion of the embryonic fissure and the optic groove (Otteson, 1998).

Based on morphology and position, it is believed that the Pax2+ cells at the optic disc play a role in guiding ganglion cell axons out of the retina. The Pax2+ cells that encircle the optic disc are a subset of the channel-forming cells that create extracellular spaces or channels among their endfeet at the inner surface of the pia that are believed to play a role in guiding ganglion cell axons out of the retina into the optic nerve. Netrin-1, a secreted protein that is required for axon guidance of commissural neurons within the developing brain, is expressed by the Pax2+ cells at the optic disc and optic nerve; Netrin-1 mutants show significant axon pathfinding defects at the optic disc. In the Krd/+ embryos, the disorganization of the Pax2-expressing cells at the optic disc would be predicted to result in both an abnormal organization of the network of extracellular channels and a dispersion of the associated chemoattractant guidance cues to ectopic locations. Either or both of these events could give rise to the intraretinal defects in axon pathfinding observed in Krd/+ embryos (Otteson, 1998)

Specialized cells at the midline of the central nervous system have been implicated in controlling axon projections in both invertebrates and vertebrates. Analysis has been made of the requirement for ventral midline cells in the provision of cues to commissural axons using Gli2 mouse mutants, which specifically lack the floor plate and immediately adjacent interneurons. A specific enhancer drives tau-lacZ expression in a subpopulation of commissural axons (C-axons) and it has been found that C-axons project to the ventral midline in Gli2 minus embryos. Netrin1 mRNA expression is detected in Gli2 minus embryos and, although much weaker than in wild-type embryos, is found in a dorsally decreasing gradient. Netrin1 mRNA expression in the VZ in Gli2 -/- mutant embryos is likely to be sufficient to attract C-axons to the midline in these embryos. While the floor plate can serve as a source of long-range cues for C-axons in vitro, it is not required in vivo for the guidance of commissural axons to the ventral midline in the mouse spinal cord. After reaching the ventral midline, most commissural axons remain clustered in Gli2 minus embryos, although some are able to extend longitudinally. Interestingly, some of the longitudinally projecting axons in Gli2 minus embryos extend caudally and others rostrally at the ventral midline, in contrast to normal embryos in which virtually all commissural axons turn rostrally after crossing the midline. This finding indicates a critical role for ventral midline cells in regulating the rostral polarity choice made by commissural axons after they cross the midline. In rodents, the turning of C-axons into the longitudinal axis is correlated with a switch in adhesion molecule localization from TAG-1 to L1 on these axons. Evidence is provided that interactions between commissural axons and floor plate cells are required to modulate the localization of Nr-CAM and TAG-1 proteins on axons at the midline. Finally, it has been shown that the floor plate is not required for the early trajectory of motoneurons or axons of the posterior commissure, whose projections are directed away from the ventral midline in both WT and Gli2 minus embryos, although they are less well organized in Gli2 minus mutants (Matise, 1999).

During early development, multiple classes of interneurons are generated in the spinal cord including association interneurons that synapse with motor neurons and regulate their activity. Very little is known about the molecular mechanisms that generate these interneuron cell types, nor is it known how axons from association interneurons are guided toward somatic motor neurons.

EN1 is a prototypic cell-type-specific transcription factor that is expressed in a restricted population of early postmitotic ventral neurons that are located in two bilateral columns, dorsal to the motor neurons. The expression of EN1 in these ventral neurons is controlled by inductive signals that pattern the ventral neural tube. EN1 expression in the ventral spinal cord is dependent on the activity of the PAX6 transcription factor, which is expressed in ventral progenitors that give rise to EN1 interneurons, and EN1 is no longer expressed in the ventral spinal cord of Small eye (Pax6-) mutant embryos. The restricted expression of EN1 in early postmitotic neurons, together with the specific loss of these cells in Pax6 mutant mice has led to the hypothesis that EN1 marks a subclass of ventral interneurons and that this interneuron subclass may be specified in part by EN1. However, the interneurons in the embryonic spinal cord that express EN1 have not been characterized in detail, nor has the function of En1 in these neurons been determined. By targeting the axonal reporter gene tau-lacZ to the En1 locus, it has been shown that the cell-type-specific transcription factor Engrailed-1 (EN1) defines a population of association neurons that project locally to somatic motor neurons. These EN1 interneurons are born early and their axons pioneer an ipsilateral longitudinal projection in the ventral spinal cord. The EN1 interneurons extend axons in a stereotypic manner, first ventrally, then rostrally for one to two segments where their axons terminate close to motor neurons. The growth of EN1 axons along a ventrolateral pathway toward motor neurons is dependent on netrin-1 signaling. In addition, this study demonstrates that En1 regulates pathfinding and fasciculation during the second phase of EN1 axon growth in the ventrolateral funiculus (VLF); however, En1 is not required for the early specification of ventral interneuron cell types in the embryonic spinal cord (Saueressig, 1999).

Optic nerve formation in mouse involves interactions between netrin-1 at the optic disk and the netrin-1 receptor DCC (deleted in colorectal cancer) expressed on retinal ganglion cell (RGC) axons. Deficiency in either protein causes RGC pathfinding defects at the disk leading to optic nerve hypoplasia. Further along the visual pathway, RGC axons in netrin-1- or DCC-deficient mice grow in unusually angular trajectories within the ventral hypothalamus. In heterozygous Sey(neu) mice that also have a small optic nerve, RGC axon trajectories appear normal, indicating that the altered RGC axon trajectories in netrin-1 and DCC mutants are not secondarily caused by optic nerve hypoplasia. Intrinsic hypothalamic patterning is also affected in netrin-1 and DCC mutants, including a severe reduction in the posterior axon projections of gonadotropin-releasing hormone neurons. In addition to axon pathway defects, antidiuretic hormone and oxytocin neurons are found ectopically in the ventromedial hypothalamus, apparently no longer confined to the supraoptic nucleus in mutants. In summary, netrin-1 and DCC, presumably via direct interactions, govern both axon pathway formation and neuronal position during hypothalamic development, and loss of netrin-1 or DCC function affects both visual and neuroendocrine systems. Netrin protein localization also indicates that unlike the more caudal CNS, guidance about the hypothalamic ventral midline does not require midline expression of netrin (Deiner, 1999).

Diffusible factors, including netrins and semaphorins, are believed to be important cues for the formation of neural circuits in the forebrain. The role of netrin 1 in the development of hippocampal connections has been examined. netrin 1 and its receptor, Dcc, are expressed in the developing fimbria and in projection neurons, respectively, and netrin 1 promotes the outgrowth of hippocampal axons in vitro via DCC receptors. In the hippocampus at E14-E18, netrin 1 is highly expressed in the fimbria, which is the route followed by developing hippocampal axons to reach the hippocampal commissure at the midline. In the midline region, where the two telencephalic vesicles fuse, high expression is detected in the hippocampal commissure itself and in the ventral aspect of the cingulate cortex and in the septum, which are located just above and below the emerging corpus callosum and hippocampal commissure, respectively. In addition, low expression occurs at E16-E18 in the dentate gyrus, the pyramidal layer of the CA3-CA1 subfields, and in some interneurons in the stratum radiatum of the hippocampus. At early postnatal stages (P0-P5), overall netrin 1 expression decreases, but is still noticeable in the septum, dentate gyrus and in some hippocampal interneurons (Barallobre, 2000).

Dcc mRNA is detected in all the hippocampal fields originating axons growing towards the midline through the fimbria. Thus, the pyramidal neurons in the CA1 and the CA3 regions, and the entorhinal cortex (EC), are intensely labeled at E14-E18, with the expression decreasing after P0. The dentate gyrus, forming the mossy fibers, show intense Dcc hybridization at prenatal and postnatal stages. DCC immunostaining at E14-E18 labels many fiber fascicles in the white matter and fimbria of the developing hippocampus overlapping with netrin 1 expression. At E18-P0 some DCC-positive fascicles cross the midline, while other axons turn ventrally before crossing and descend towards the septum. This pattern of immunolabeling is consistent with the expression of the Dcc gene in the CA3 and CA1 pyramidal neurons, which form, respectively, the commissural and the hippocampo-septal pathway (which is mainly ipsilateral). It is concluded that the expression of netrin 1 along the route followed by hippocampal axons toward the midline, and the expression of its receptor Dcc in hippocampal projection neurons, are consistent with these genes having a role in axonal guidance in the hippocampal system (Barallobre, 2000).

The hippocampus of netrin 1-deficient mice shows a misorientation of fiber tracts and pathfinding errors, as detected with antibodies against the surface proteins TAG-1, L1 and DCC. DiI injections show that hippocampal commissural axons do not cross the midline in these mutants. Instead, when axons approach the midline, they turn ventrally and form a massive aberrant projection to the ipsilateral septum. In addition, both the ipsilateral entorhino-hippocampal and the CA3-to-CA1 associational projections show an altered pattern of layer-specific termination in netrin 1-deficient mice. Finally, optical recordings with the Ca 2+ indicator Fura 2-AM show that spontaneous neuronal activity is reduced in the septum of netrin 1-mutant mice. It is concluded that netrin 1 is required not only for the formation of crossed connections in the forebrain, but also for the appropriate layer-specific targeting of ipsilateral projections and for the control of

The morphogenetic development of the mammalian inner ear is a complex multistep process, the molecular and cellular details of which are only beginning to be unraveled. Mouse netrin 1, known to be involved in axon guidance and cell migration in the central nervous system, also plays a critical morphogenetic role during semicircular canal formation. netrin 1 is expressed at high levels in the otic epithelium, in cells that will come together to form a fusion plate, a prerequisite for the formation of semicircular canals. In netrin 1 mutant mice, fusion plate formation is severely affected, resulting in a reduced anterior semicircular canal and the complete lack of the posterior and lateral canals. These results suggest that netrin 1 facilitates semicircular canal formation through two different mechanisms: (1) it participates in the detachment of the fusion plate epithelia from the basement membrane, and (2) it stimulates proliferation of the periotic mesenchymal cells, which then push the epithelial cell walls together to form the fusion plate (Salminen, 2000).

During semicircular duct formation, the epithelial walls destined to form fusion plates first become thinner, then the cells lose their epithelial morphology because of a local disruption of the underlying basement membrane. These areas of the epithelium produce secreted netrin 1 protein. In the absence of sufficient amounts of netrin 1, the loss of epithelial morphology does not occur normally, suggesting that netrin 1 is required for the local disruption of the basement membrane (Salminen, 2000).

The basement membrane is a specialized sheet-like extracellular matrix (ECM) structure. The ECM laminins are thought to play a key role in the attachment of cells to basement membranes. Laminin heterotrimers polymerize to form a two-dimensional network through interactions involving their globular VI domains in the ends of the cruciform structure. The fact that the netrin 1 N-terminal portion is related to laminin domain VI may enable the incorporation of netrin 1 molecules into the laminin network. High local amounts of netrin 1 protein could then interfere with laminin polymerization and directly disrupt the basement membrane. Alternatively, netrin 1 could be involved in a signaling cascade leading to a local production of matrix metalloproteinases to digest basement membrane proteins. These proteases are known to be required for the remodeling of the ECM during morphogenesis, angiogenesis, cell migration and axonal migration (Salminen, 2000).

Inner ear morphogenesis is known to depend on interactions between the otic epithelia and the periotic mesenchyme. How these interactions occur at the molecular level is still unknown. It is proposed here that the mouse netrin 1 protein, secreted by the fusion-plate-forming epithelium, participates in a signaling pathway that results in a local induction of proliferation in the periotic mesenchyme. This signaling pathway is likely to involve netrin 1 receptors. Several transmembrane proteins belonging to a subgroup of the immunoglobulin superfamily are known to bind netrins in vitro and, thus, thought to be components of netrin 1 receptors. Mouse mutants are available for two candidate genes coding for netrin 1 receptors. However, nothing has been reported about the potential role of these genes during inner ear development. Thus, the downstream components of netrin 1 signaling during inner ear morphogenesis remain to be elucidated (Salminen, 2000).

Fusion plate formation and the subsequent removal of the fused cells are critical events in semicircular canal morphogenesis. In addition to netrin 1, a set of transcription factors have been shown to play important roles in this multistep process. The semicircular canal formation starts by the outgrowth of the otic vesicle epithelium to form bilayered outpocketings. The Prx1 and Prx2 genes seem to be required for a normal epithelial outgrowth. Subsequently, in the middle of the outpocketings, the future fusion plate epithelium becomes thin and then detaches from the basement membrane. Nkx5-1 may be required to determine the area that will form a fusion plate and the detachment requires netrin-1 production by the fusion-plate-forming epithelium. In the next step, the two opposing epithelial walls have to come together to form a fusion plate. The expression of netrin 1 by the fusion-plate-forming epithelium is likely to be required to induce a local proliferation of the adjacent mesenchymal cells to push the epithelial walls together. After formation of the fusion plate, the fused cells have to be removed. This phenomenon does not occur efficiently in the Nkx5-1 mutant mice, suggesting a role for this transcription factor in the regulation of the fusion and/or removal of fusion plate cells (Salminen, 2000).

Modifying serotonin (5-HT) abundance in the embryonic mouse brain disrupts the precision of sensory maps formed by thalamocortical axons (TCAs), suggesting that 5-HT influences their growth. The mechanism by which 5-HT influences TCAs during development was investigated. 5-HT_1B and 5-HT_1D receptor expression in the fetal forebrain overlaps with that of the axon guidance receptors DCC and Unc5c. In coculture assays, axons originating from anterior and posterior halves of the embryonic day 14.5 dorsal thalamus responded differently to netrin-1, reflecting the patterns of DCC and Unc5c expression. 5-HT converts the attraction exerted by netrin-1 on posterior TCAs to repulsion. Pharmacological manipulation of 5-HT_1B/1D receptors and intracellular cAMP showed the signaling cascade through which this modulation occurs. An in vivo correlate of altered TCA pathfinding was obtained by transient manipulation of 5-HT_1B/1D receptor expression abundance in the dorsal thalamus by in utero electroporation. These data demonstrate that serotonergic signaling has a previously unrecognized role in the modulation of axonal responsiveness to a classic guidance cue (Bonnin, 2007).

Axon guidance cues trigger rapid changes in protein dynamics in retinal growth cones: netrin-1 stimulates both protein synthesis and degradation, while Sema3A elicits synthesis, and LPA induces degradation. What signaling pathways are involved? These studies confirm that p42/44 MAPK mediates netrin-1 responses and further show that inhibiting its activity blocks cue-induced protein synthesis. Unexpectedly, p38 MAPK is also activated by netrin-1 in retinal growth cones and is required for chemotropic responses and translation. Sema3A- and LPA-induced responses, by contrast, require a single MAPK, p42/p44 and p38, respectively. In addition, caspase-3, an apoptotic protease, is rapidly activated by netrin-1 and LPA in a proteasome- and p38-dependent manner and is required for chemotropic responses. These findings suggest that the apoptotic pathway may be used locally to control protein levels in growth cones and that the differential activation of MAPK pathways may underlie cue-directed migration (Campbell, 2003).

These data provide evidence for the presence of caspases in growth cones and identify caspase-3 as a potential target of p38 signaling for mediating both netrin-1-induced turning and LPA-induced growth cone collapse. This suggests that, in addition to their roles in apoptosis, caspase-induced protein degradation may play a role in growth cone guidance. Previous studies identified the netrin receptor DCC in regulating cell survival via the activation of caspase-3 by caspase-9 in the absence of netrin-1 in human embryonic kidney 293T cells. By contrast, in Xenopus retinal growth cones, netrin-1 and LPA induce the rapid activation of caspase-3 independent of caspase-9 via the MAPK- and proteasome-mediated proteolysis pathways. The activation of caspase-3 in the confined cellular compartment of the growth cone might not lead to activation of the full apoptotic cascade and cell death but rather to transient, localized changes in specific proteins. The p42/p44 and PI-3 kinase pathways identified in netrin-1 signaling are known to play roles in mediating cell survival and may ensure tight regulation of caspase activity in the growth cone. A role has been identified for caspases in synaptic plasticity independent of their roles in cell death (Campbell, 2003 and references therein).

Since caspases are proteases, a key question asks which proteins do caspases degrade? Candidate proteins include known caspase substrates, such as actin, actin binding proteins, and signal transduction pathway components. For example, gelsolin, an actin severing protein, is present in growth cones and is activated by caspase-3-mediated cleavage. Netrin-1 and LPA stimulate the rapid caspase-3-dependent cleavage of PARP. In addition to its role in maintaining genomic stability, PARP is able to interact with and activate proteasome-mediated proteolysis. Cleavage of PARP may inactivate itself, providing a possible mechanism by which proteasome-mediated proteolysis may be regulated in the case of netrin-1 and LPA. The netrin-1 receptor DCC is itself a substrate of caspase-3, and caspase-mediated cleavage of DCC may potentially be involved in mediating netrin-1-induced chemotropic responses. During apoptosis, caspase-3 is also able to cleave eukaryotic initiation factor 4G (eIF-4G), a crucial protein required for binding cellular mRNA to ribosomes. This may decrease the rate of translation and provide a possible mechanism for negative regulation of netrin-1-stimulated protein synthesis in growth cones. Since the chemotropic responses of growth cones elicited by netrin-1 and LPA are essentially blocked by inhibition of caspase-3, it is likely that of the caspases, caspase-3 plays a major role in these processes (Campbell, 2003 and references therein).

The ubiquitin-proteasome system is critically involved in apoptosis and in mediating chemotropic responses of growth cones. In neuronal cells, proteasome inhibitors protect against apoptosis by acting upstream of caspase activation. These results have revealed a parallel in retinal growth cones where the activation or cleavage of caspase-3 in response to netrin-1 and LPA requires proteasome function, suggesting that caspase-mediated protein degradation lies downstream of proteasome/ubiquitin-mediated proteolysis. Candidate proteins to undergo proteasome/ubiquitin-mediated proteolysis include the inhibitor of apoptosis (IAP) family of proteins, degradation of which can result in caspase activation. IAPs can also target caspase-3 itself for proteasome/ubiquitin-mediated proteolysis, suggesting a possible mechanism for the transient and localized nature of caspase-3 activation in growth cones (Campbell, 2003 and references therein).