News Physiol Sci 18: 130-136, 2003;
doi:10.1152/nips.01414.2002
1548-9213/03 $5.00
News in Physiological Sciences, Vol. 18, No. 3, 130-136,
June 2003
© 2003 Int. Union Physiol. Sci./Am. Physiol. Soc.
Chemotropic Molecules: Guides for Axonal Pathfinding and Cell Migration During CNS Development
Fernando de Castro
1 Neurobiología-Investigación, Hospital Ramón y Cajal, Madrid-28034; and
2 Instituto de Neurociencias de Castilla y León-INCYL, Universidad de Salamanca, Salamanca-37007, Spain
 |
Abstract
|
|---|
Different molecules (netrins, semaphorins, slits) with chemotropic functions and their receptors (neogenin, DCC, neuropilins, plexins, robos) have been identified that guide axons during development of the nervous system to establish the complex pattern of connections among a large number of neurons. These molecules have been recently identified to play a role in cell migration of the central nervous system during development.
|
Introduction
|
|---|
The neurons and the glial cells of the central nervous system (CNS) derive from a specialized region of the embryonic ectoderm called the neural plate. This region differentiates under the influence of inductive signals emanating from the mesendoderm and axial mesoderm, prechordal plate, and notochord and closes to form the neural tube. The neural tube is composed of a layer of neuroepithelial cells that leave a cavity in the middle (the future ventricular system of the brain). Regionalization mechanisms control differences in the proliferation and differentiation rates of the neuroepithelium. These differences result in the different specialized areas and the diversity of neural cells in the mature CNS (except for microglia). Neuronal progenitors mainly proliferate in the ventricular and subventricular zones of the neural tube and migrate to colonize the superficial zones of the tube (Fig. 1C
).
View larger version (31K):
[in this window]
[in a new window]
|
FIGURE 1. Growth cone navigation and cell migration during central nervous system (CNS) development. A: axons navigate long distances to contact far-away targets. Mitral cells (red) from the olfactory bulb (OB) project to different structures that form the olfactory cortex, all along the telencephalic vesicle. Axon collaterals of these cells (red arrows) specifically colonize the different targets. The schema represents a horizontal section of the brain of an embryonic day 15 (E15) rat embryo, OB being the anterior part.B: growth cone is the distal part of the axon. It has a handlike form and serves to explore and elongate the axon until it reaches the correct target (see text for details). C: migration of postmitotic neurons from the germinal subventricular zone (dotted gray) to colonize different structures (neocortex, striatum, piriform cortex). Arrows represent the different migration streams used by these cells on their way to their final positions in the brain, from where they will contact other neurons. The schema corresponds to a coronal section of an E15 rat embryo. D: radial migration of cells from their proliferation sites in the ventricular surface (PZ) to different areas in the cortex. Neuroblasts (blue) proliferate, and postmitotic neurons (red) migrate along radial glia (gray) toward the surface of the brain (radial migration) guided by their "leading process" (black arrows). Sometimes, neurons leave this highway to migrate in a tangential way (purple). The light blue areas represent the cerebral ventriculi in A,C, and D.
|
|
An extremely precise pattern of connections among a large number of neurons (~1012 neurons and 1015 of exact synaptic contacts in a human brain) is necessary for a mature nervous system to perform the functions of which it is capable. This precise connectivity is established during the embryonic and the early postnatal development, once postmitotic neurons have accomplished their migration from the proliferative sites (see below; Fig. 1C
). In general, the axon is the cell process through which a neuron sends messages to another/s. How can the axon of a callow postmitotic neuron select the correct targets, usually far away from its position in the brain (Fig. 1A)? One piece of this puzzle is the chemotropic molecules. (The terms "chemotropic" and "chemotactic" will be used interchangeably in this review.) Their discovery is, indeed, the corroboration of the first hypothesis launched to explain the formation of the complicated interneuronal connectivity in the CNS: Cajal’s chemotactic hypothesis (11). This hypothesis was based on Pfeffer’s chemotaxis, was probed in leucocytes by researches from Bordet to Metchnikoff, and was formed merely two years after the first description of the axonal growth cones (Fig. 1B; see legend for details) by Ramón y Cajal from experiments in the spinal cord of chicken embryos in 1891.
The chemotactic hypothesis was debunked and forgotten for almost a century, briefly dusted off by Sperry’s chemoaffinity hypothesis (for an historical overview, see Ref. 15), until Tessier-Lavigne identified the first family of chemotropic molecules secreted by the floor plate of the neural tube, the netrins, which guide commisural axons of the spinal cord to cross the midline (6). (The floor plate is a narrow part of the neural tube in the ventral midline. The roof plate is the corresponding region in the dorsal midline of the tube.) With them, Cajal’s hypothesis was confirmed. After the netrins, other molecular families with chemotropic activity were identified in the CNS: semaphorins, slits, and some growth factors (9). All of these are diffusible molecules that, secreted by intermediate and final targets, give rise to gradients that attract or repel growing axons along their way to their final targets. The orchestra of simultaneous chemotropic signals, in combination with contact-mediated mechanisms (laminin, cell adhesion molecules, etc.), allow axons to establish the proper connectivity of the CNS. Note, however, that chemotropism is a concept merely related to guidance: axons or cells are attracted or repelled from one region. It is also related to cell survival, which is promoted or avoided by trophic factors. To understand the physiological role of each of these molecules, we need to know the spatial distribution of the cells that secrete the molecule; in other words, we need to know the distribution pattern of cells generating the gradient of one chemotropic molecule. All of the data reviewed here showed a matching pattern of expression of the cells (that is, all cells secreted the cue in the original studies), which helps to explain the physiological effects of the molecules.
The role of the chemotropic molecules in the guidance of cell migration during CNS development is currently emerging. Postmitotic cells migrate from their sites of origin in the ventricular zone of the neural tube to find their final location (Fig. 1C). Cells (neurons, oligodendrocytes, astrocytes) travel relatively long distances, up to millimeters in some cases, and chemotropic molecules orient them in their migration (8). Cell migration and axonal pathfinding are two events in close relationship during CNS development, sharing molecular mechanisms, which include chemotropic molecules. For a comprehensive account of the effects of chemotactic molecules during CNS development known to date, we recommend two excellent reviews (8, 9).
|
Netrins and axonal guidance
|
|---|
Netrins are secreted proteins of ~65 kDa related to laminin, one of the most important contact-mediated stimulators of axonal growth. At the COOH-terminal end, netrins show a specific basic domain that allows them to bind extracellular matrix components and thereby modify their ability to diffuse. Four members of this family (netrin-1 through -4) have been identified in vertebrates, but there is no evidence to date to support the existence of a mammalian netrin-2 (6). Netrin-1 and -2 bind similarly to deleted in colorectal cancer (DCC), neogenin, and Unc5H-1, -2, and -3 (the known receptors for netrins in mammals), but netrin-3 shows a preferential binding to the Unc5H receptors (Fig. 2A). The binding properties of netrin-4 are as yet poorly understood.
View larger version (36K):
[in this window]
[in a new window]
|
FIGURE 2. Main chemotropic cues and their receptors. A: netrins bind receptors that mediate attractive effects [neogenin, deleted in colorectal cancer (DCC)] or repulsive effects (Unc5H1 through -3), with similar affinities, except for netrin-3, which binds better to Unc5H receptors (thicker arrows represent higher binding affinities). No reports of the binding properties of netrin-4 have been published. B: secreted class III semaphorins selectively bind to neuropilin-1 and -2. Thicker arrows represent higher binding affinities. Note that Sema 3A binds neuropilin-1 exclusively. Plexins (for example, plexin-A3) and vascular endotheilial growth factor receptor (VEGFR)-1 combine with neuropilins to form receptor complexes. Both Sema 3A and Sema 3F have also been reported to bind directly to plexin-A3. Plexins have been involved exclusively in repulsive effects of secreted semaphorins. C: every known Slit binds similarly to each receptor of the Robo family (thickness is similar for normal and doted arrows). The gray arrow shows binding that needs heparan sulfate to develop biological activity. Netrin-1 and Slit-2 bind each other with the same affinity with which Slit-2 binds Robos. Robos and Slit-2 have been reported to bind DCC and netrin-1, respectively. Given the variety of receptors for chemotropic molecules, and their diversity, it should be interesting to agree upon systematic nomenclature, as it was adapted for semaphorins by an ad hoc committee (Cell 97: 551–552, 1999).
|
|
Netrins are a paradigmatic case of double-function molecules that guide axonal growth cones: they are capable of both attracting and repelling axons (6). When neurons responding to netrins express DCC or neogenin, the result is an attraction of the axon toward the molecular source, but if they express receptors from the Unc5H family, netrins will repel them. In the case of coexpression of both types of receptors, DCC/neogenin and Unc5H, a repellent effect will predominate. Both types of receptors are used by netrin-1 to exert a survival-promoting (trophic) effect on neurons from the ventricular zone of the brain stem (7).
Netrin-1 has either attractive or repulsive effects, depending on the neuronal population and the expression of the cited receptors. It is attractive for the growth cones of cortical and hippocampal projection neurons, habenular neurons, thalamocortical axons, precerebellar neurons, retinal ganglion cells, and commisural interneurons of the spinal cord (6, 9). However, netrin-1 repels troclear axons, cranial motoneurons, and the parallel fibers of the cerebellar granule cells (9). In the absence of netrin-1 function, most of the main commissures of the brain (corpus callosum, anterior and hippocampal commissures, optic chiasm) fail to develop (9). In addition to these long-range effects, secreted netrins have a short-range function in both guiding neuritic outgrowth and in the maintenance of neurooligodendroglial interactions.
|
Semaphorins and axonal pathfinding
|
|---|
The large family of semaphorins (from the Greek semas, "signal for recognition," i.e., the flag signaling system "semaphores") comprise ~30 members that can be classified as either secreted or transmembrane semaphorins. The secreted ones are class II (invertebrates), III (vertebrates), and V (viral), whereas the other classes (I and IV–VII) are transmembrane. All of them have a common domain called a "sema" of ~500 amino acids with 12–16 cysteine residues, which confers the binding specificity of each semaphorin (12). The sema domain is necessary and sufficient to execute their activity. For the purpose of this review, class III semaphorins are the most interesting. They bind selectively to the two known members of the neuropilin family, neuropilin-1 and -2 (Fig. 2B), which in mammals are the biological receptors of secreted semaphorins. Neuropilins have a short cytoplasmatic domain (40 amino acids) that is not capable of intracellular transduction of the semaphorin signal. Plexins, transmembrane molecules that bind trasmembrane semaphorins, can also combine with neuropilins to form a receptor complex in which neuropilins represent the binding domain and plexins the transduction domain. Receptors for vascular endothelial growth factor (VEGF) have been reported to combine with neuropilins (and maybe with plexins) to form cell membrane complexes (Fig. 2B).
The effects described for secreted semaphorins on axonal pathfinding are almost exclusively repulsive (Fig. 3, A–D). In some cases, these molecules collapse growth cones, an effect directly related to repulsion, even promoting apoptosis of some neuronal populations. An opposite trophic effect has been recently observed in oligodendrocyte progenitors (see below). Sema 3A, the molecule that gives the rest of the family its name, is the most extensively studied, and in all cases it has been described as a repellent factor for axons, from sensorial neurons and spinal motoneurons to pyramidal neurons of the cortex (9). Sema 3A exclusively binds neuropilin-1 (Fig. 2B), and the main effects observed when Sema 3A is knocked out appear in the peripheral nervous system (mainly, lack of many groups of sensory neurons). To date, Sema 3A is the only chemotactic molecule with a demonstrated effect orienting the outgrowth of dendrites; interestingly, it attracts apical dendrites of pyramidal neurons toward the pial surface of the neocortex (10), although it repels axons and attracts dendrites in the pyramidal neurons of the cortex.
View larger version (87K):
[in this window]
[in a new window]
|
FIGURE 3. Examples of chemoattractive and chemorepulsive effects. A: explant of E15 rat OB grown in control conditions for 2 days in vitro in rat-tail collagen. Axons outgrow radially. B: how to make quantitative assessments in this kind of experiment. Tissue explants are confronted at the source of chemotropic factors ("Test"; the source can be either transfected cells to secret the cue, one piece of tissue secreting the cue, or a bead embedded in the protein, for example). There must be a distance between the explant tested and the source of secreted molecule. After explant growth, we consider proximal (PQ) and distal quadrants (DQ), depending on their position relative to the source. Axonal outgrowth or cell migration must be compared between PQ and DQ. If axonal outgrowth (or number of migrating cells) is larger in PQ, there is an attraction, whereas it is considered repulsion when the DQ value is larger. C–D: OB axons are repelled by cells transfected to secrete Sema 3F (3F in C) and by an explant of olfactory epithelium (OE in D) that physiologically produces Sema 3F. Axonal outgrowth is significantly larger in the DQ than in the PQ, which means that a gradient of secreted cue established from the test (B) repels outgrowing axons. In both cases, the limits of the source of chemotropic molecules have been marked with dashed lines.E: oligodendrocyte progenitors migrating out from an explant of E16.5 mouse optic nerve (ON; the original perimeter of the explant is outlined by white dots) after 2 days in vitro cocultured with control cells (which do not secret any chemotropic substance; this source is outlined in white). The number of migrating oligodendrocyte progenitors is similar in PQ and DQ (counted as represented in B). On the contrary, Sema 3F attracts the oligodendrocyte progenitors (F), which are more numerous in the PQ (between the explant and the source) than in the DQ. Data are from unpublished observations of F. de Castro and A. Chédotal (A, C–D) or B. LeBras and F. de Castro (E–F).
|
|
Most of the class III semaphorins strongly repel axons of sympathetic neurons, as well as other populations in the cortex (for example the corticothalamic, entorhinal, hippocampal, and mitral cells; see Fig. 3C), the spinal cord, and the olfactory axons. Interestingly, Sema 3C attracts axons of cortical neurons and Sema 3B attracts axons from the mitral cell axons of the olfactory bulb (4); these are the only two known exceptions in which secreted semaphorins attract growing axons, and both neuronal populations are repelled by other secreted semaphorins, which implies a complex battery of physiological signals contributing to proper axonal guidance, some of these signals being closely related to each other (i.e., the axons of mitral cells are attracted by Sema 3B and repelled by Sema 3F and slit-2 but do not respond to the rest of secreted semaphorins nor to netrin-1, even when they express corresponding biological receptors; see Fig. 3C) (4).
|
Slits: the third big family of molecules guiding outgrowing axons
|
|---|
The third big family of chemotropic molecules is slit. In mammals known members of this family are slit-1, -2, and -3 (1, 9). These secreted molecules are 200 amino acids in length with an NH2-terminal leucine-rich region that is crucial for their biological functionality. Although not yet demonstrated in vertebrates, slits may be less diffusible than other secreted molecules because of the low diffusibility of their NH2-terminal fragments. In contrast to semaphorins, the biological receptors of slits, called robos (derived from "roundabout," the Drosophila mutation from which the cue was originally isolated, called this because commissural axons remain at the midline, crossing and recrossing, instead of leaving midline and projecting to the contralateral side) have a big cytoplasmatic domain that allows intracellular transduction. But direct proof of the involvement of robos as receptors for slit functions in mammals is still missing. In vertebrates, Robo-1, Robo-2, and Rig-1 are the three known receptors of the robo family. As happens in Drosophila, in mammals a promiscuity in the binding of slits to robos has been reported (every slit binds every robo with similar affinity; Fig. 2C), with the exception of Rig-1, whose relationship to slits is not completely clear. Remarkably, in invertebrates the interaction slit-2/netrin-1 (two secreted ligands) is as strong as slit-2/robo-2 (one ligand and its receptor), which complicates the scenario (Fig. 2C). Slits also bind glypican-1 (a heparan sulfate proteoglycan), and robos interact with other receptors such as DCC (a netrin receptor) and CXCR4 (a chemokine receptor). It has recently been demonstrated that slit-2/robo-1 interactions need cell-surface heparan sulfate to exert maximal repulsive activities (Fig. 2C).
In addition to this, expression patterns of slits and robos, complementary and/or overlapping in many structures of the CNS, have suggested to different scientists that they could also act in an autocrine fashion: a cell secretes the very cue to which it responds. This autocrine function has also been suggested for secreted semaphorins, and in both cases no direct evidence has been obtained to date either in invertebrates or vertebrates.
"Results published to date indicate that slits exclusively repel growth cones...."
Results published to date indicate that slits exclusively repel growth cones (1, 9). This is the case for motoneurons and for commissural neurons of the spinal cord (only after crossing midline; see below), retinal ganglion cells, mitral cells of the olfactory bulb, thalamic neurons projecting into the cortex, cortical axons projecting into the corpus callosum, and neurons projecting from the dentate gyrus. In some of these cases, slits also induced collapse of the growth cones, as we have previously described for semaphorins. It has also been suggested that slit-2 can induce axonal elongation and branching in neurons of the dorsal root ganglia from mammals (1).
|
Other chemotropic cues that guide axonal pathfinding
|
|---|
In addition to the three families of secreted cues reviewed above, diverse molecules have been described to have chemotropic effects on axonal pathfinding (for details about these molecules and their effects, see Refs. 3 and 9).
The hepatocytic growth factor/scatter factor (HGF/SF, originally described as a mitogen for adult hepatocytes and as a scatter factor for epithelial cells) attracts axons from spinal and cranial motoneurons and neocortical neurons; moreover, it promotes motility on these cells. Reelin (the protein responsible for the mutation "reeler," which results in a dramatic disorganization of the cytoarchitecture of the brain) is a protein secreted from Cajal-Retzius cells that is essential for corticogenesis (neuronal migration and development of cortical laminae). Nonetheless, it is still controversial whether reelin is directly involved in the formation of neuronal connections or whether its effects are indirect, due to cytoarchitectonic changes. The bone morphogenetic proteins (BMPs) were originally identified as morphogens for the bone tissue, but they also have important roles in the survival, induction, and differentiation of many neuron types. BMP-7 seems to be the repellent factor secreted by the roof plate of the neural tube that, in coordination with the attraction exerted by netrins secreted by the floor plate, orients growth cones of commissural axons in the spinal cord. BMP-7 also promotes a certain degree of growth cone collapse in these commissural neurons. Another interesting possibility is that, in coordination with BMP signals, another strong morphogen, called Sonic Hedgehog [a secreted protein related to Hedgehog, a gene that controls many events during embryonic development in Drosophila and was named after a character in a video game (Sonic), that plays a central role in different events of vertebrate development, such as specifying the identity of different cell types in the ventral neural tube, regulating cell proliferation (mainly in the CNS), and controlling finger formation in the limbs], guides axons of retinal ganglion cells during development and specifically regulates the collapse of these growth cones in their centripetal outgrowth, preventing them from leaving by way of the contralateral side through the optic chiasm. Finally, anosmin-1, a molecule defective in Kallmann syndrome [which is an inherited disease of anosmia (lack of olfaction) and hypogonadotropic hypogonadism; other satellite neurological symptoms can also appear], has been described as an attractant for olfactory bulb axons and, more importantly, to be a positive regulator for the formation of axon collaterals. Anosmin-1 and slit-2 are the only chemotropic cues demonstrated to play a role in this final phase of innervation (16).
Of course, many other molecules that play a chemotropic role may remain unidentified, including the cases in which one structure or organ has been identified as capable of attracting or repelling growing axons. To better focus this review, we did not cover this important field, which remains open for current and future research.
|
Organization and modulation of chemotropic effects on growing axons
|
|---|
As we have seen above, the growth cone of a growing axon is exposed to a great variety of different guiding molecules during its navigation toward the proper target. Many of these signals are due to secreted molecules, the different chemotropic molecules. These act by either attracting axons or repelling them. We can talk of a kind of orchestra of signals, so harmonious that the effect of one molecule cannot be achieved until the neuron has been previously exposed to other molecules. It has been demonstrated that these signals establish a kind of hierarchical organization of the receptors for the different guidance cues: at the same time that one repellent factor activates, the attractants that could be present are silenced through a combination of intracytoplasmic domains of their respective receptors, to prevent a tug-of-war of signals simultaneously presented to the growth cone (18).
Regardless of this complex hierarchy, the history of physicochemical stimuli experienced by a growth cone is crucial for it to reach a target or not: two sequential stimuli would have very different results depending on the order of exposure (5). This is the case cited above: spinal commisural axons respond to Sema 3B, Sema 3F and slit-2 but only after crossing the midline, toward which they have been attracted by netrin-1. Therefore, to cross or not to cross the midline is much more than a Hamlet-like rhetorical question for growing axons.
But one chemotropic molecule can be modulated in some other ways, not only by the order that it occupies in the history of stimuli of a growth cone. The addition or presence of other molecules modifies the biological effects of exposure to some chemotropic molecules: neurotrophins modulate the response of sensory axons to secreted semaphorins; some secreted semaphorins and VEGF can antagonize each other by blocking the binding sites of their receptors in the growth cone; molecules from extracellular matrix can dramatically modulate the effects of chemotropic molecules (i.e., laminin turns one secreted semaphorin from attractive to repulsive; we have previously described the effects of heparan sulfate on slit/robo interactions); even chemical inhibitors of metalloproteases potentiate chemoattractant effects of netrin-1. Interestingly, a netrin-synergizing protein (25–35 kDa in size) has been characterized that opens the door to putative synergizing activities for the other chemotropic molecules, which could explain differences in activity even when similar expression of ligands and receptors happen.
The metabolic state of the navigating growth cone is also very important. Either electrical stimulation or changes in the cytoplasmatic levels of calcium or cyclic nucleotides such as cAMP and cGMP modulate the effects of secreted cues, even turning netrins and semaphorins from attractive to repulsive or vice versa (14). Of course, down- and upregulation of receptors would also modulate the final effect of one of these cues. All of these facts give us an important perspective: effects of different cues important for axonal pathfinding can be actively and passively modulated by many other factors. This could explain how, with a relatively low number of molecules (we have reviewed here fewer than 20 chemotropic molecules), a huge number of neurons can establish a proper pattern of interconnections and result in a normal-functioning nervous system, capable of processing sensorial information and elaborating motor answers and thoughts.
|
Chemotropic molecules and cell migration
|
|---|
During CNS development, cell migration is a very important process. In general, cells generate in proliferative areas from the neural tube and migrate to occupy their final position (Fig. 1C). Most of the cells migrate in a radial way, along radial glia, and the mechanisms involved in this gliophilic or radial migration are mainly contact mediated. Contrary to the classical idea, radial migration is not the only significant way for neurons to occupy the entire nervous system. A considerable number of cells migrate independently from the radial glia, in a neurophilic or tangential migration (8). This second form of cell migration in the CNS suggests a parallelism with axonal pathfinding (Fig. 1D). The growth cone (more frequently referred to as the "leading process" in migration) suffers cytoskeleton changes resulting in cell motility, which reflects the most important difference between cell migration and axonal pathfinding: the movement of the soma in the migration process, which usually remains still in axonal guidance. In general, the radially migrating neurons are projection neurons, whereas tangentially migrating ones are interneurons.
Although tangentially migrating neurons have been described as closely related to corticofugal axons (the names neurophilic and axonophilic migration derive from this fact), it is not clear whether a direct interaction between the migrating cells and the preexisting axonal tracts exists. The pattern of tangential cell migration suggests that, besides contact-mediated mechanisms, long-range mechanisms such as chemotropic molecules could have a relevant role in this migration process. In fact, and leaving Reelin aside (the role of Reelin in neuronal migration has been the focus of many complete reviews; given the controversial role of reelin as chemotropic molecule, we will not review its effects here; see above), in the first half of 1999 slit-2 was the first chemotactic molecule that was shown to have roles in cell migration in the mammalian CNS (8). It repels migrating motoneurons, GABAergic interneurons migrating from the ganglionic eminence into the neocortex, and those migrating along the rostral migratory stream from the subventricular zone of the forebrain toward the olfactory bulb [GABAergic interneurons (granular and periglomerular cells)]. This last type of cells has been recently demonstrated to be induced to migrate by an astrocyte-derived, migration-induced activity that seems to be less strong than slit-2. It is very remarkable that the same functional domain of slit proteins is involved in guiding growth cone navigation and in neuronal migration, which confirms the idea of the degree of conservation and sharing of molecular mechanisms in both developmental events.
Netrins were identified to play a role in neuronal migration, again attracting or repelling cells depending on the population studied (8). Midline netrin-1 attracts tangentially migrating neurons that will form the inferior oliva, being this antagonized by slit (2). Netrin-1 repels migrating hypothalamic neurons, postnatally generated cerebellar granule cells, and, surprisingly, late-born striatal projection neurons, which migrate in radial paths from the ventricular surface within the striatal primordium. The relevance of DCC (and thus maybe of netrins) in the migration of luteinizing hormone-releasing hormone neurons from the nasal pit, where they are generated, to their final destination in the hypothalamus has also been described (8).
"...effects of different cues important for axonal pathfinding can be actively and passively modulated by many other factors."
Although the first clue about the role of semaphorins in cell migration was described in the migration of the neural crest cells toward their peripheral targets in the chicken embryo, recently it has been demonstrated (8) that a differential response to secreted semaphorins via neuropilins sorts migrating cells from the ganglionic eminences to invade the striatum, the paleocortex, and the neocortex (Fig. 1C). This population of migrating cells responds to the three largest known families of chemotropic molecules (slits, netrins, and semaphorins), even to HGF/SF.
Interestingly, a similar case of multiple response in migrating cells in the CNS has been observed for oligodendrocyte progenitors that colonize the optic nerve: they are selectively repelled by Sema 3A and attracted by Sema 3F (Fig. 3, E–F) and netrin-1 (17, 19). All of these effects are produced via the same receptors that have been described for axonal migration and that we have reviewed above (Fig. 2, A–B). Sema 3F exerts a double effect; in fact, it is also a chemotrophic (survival-promoting) factor for the same population (17). Early postnatal upregulation of Unc5H receptors reverses netrin-1 attraction into repulsion for these cells.
In summary, chemotropic molecules involved in axonal guidance are also useful in orienting cell migration in the CNS, a closely related event. Furthermore, such a system is not only valid for migrating neurons but also for cells of oligodendroglial lineage and, maybe, for other types of CNS cells. The fact that one of these molecules present in one structure could serve at the same time for growth cone navigation of one cell population and for migration of a different one could be due to the specific localization of molecular isoforms. Nevertheless, differences in the isoforms involved in the intracellular transduction and effector machinery may also participate.
|
Summary
|
|---|
One century after Ramón y Cajal raised his chemotactic hypothesis to explain how interconnections between 1012 neurons (the estimated figure for the brain of a primate) can be established accurately, chemotropic molecules (netrins, semaphorins, and slits) surfaced as the fulfillment of this prescient hypothesis. Together with contact-mediated mechanisms (laminin, cell adhesion molecules, etc.), a relatively low number of molecules result in a combination of spatially and temporally ordered signals that determine every single neuron’s connections along the CNS. Axonal growth cones read and interpret this symphony of cues present in the tissue in a moment, growing toward or away from structures to finally follow a very particular path and reach their targets. Interactions between different molecules modulate their activity. A second important point of modulation is the intracellular transductional and effector machineries, as well as the metabolic state of the cell.
These chemotropic molecules also orient postmitotic cells in their migration within the CNS. In this case, the "leading process" acts as a growth cone. Both events, cell migration and formation of connections between them, share molecular mechanisms. These overlappings and interweavings offer us a fascinating field of work for the future. The relevance of chemotropic molecules in neuropathology is increasingly recognized and an important subject of research (13). Their role in neuroregeneration might also be manipulated (up- or downregulation of factors and their receptors, pharmacological blocking or modulation, etc.) to design future successful therapies.
|
Acknowledgments
|
|---|
I am indebted to Dr. Ángel Nadal (Universidad Miguel Hernández, Alicante, Spain) and Dr. Santiago Canals (Hospital Ramón y Cajal) for their useful comments and suggestions.
I gratefully recognize the support of FIS-Ministerio de Sanidad (Spain) 01/3136 and PI020768.
|
References
|
|---|
- Brose K and Tessier-Lavigne M. Slit proteins: key regulators of axon guidance, axonal branching and cell migration. Curr Opin Neurobiol 10: 95–102, 2000.[ISI][Medline]
- Causeret F, Danne F, Ezan F, Sotelo C, and Bloch-Gallego E. Slit antagonizes Netrin-1 attractive effects during the migration of the inferior olivary neurons. Dev Biol 246: 429–440, 2002.[ISI][Medline]
- De Castro F. Moléculas quimiotrópicas como mecanismo de orientación del crecimiento axonal y de la migración neuronal durante el desarrollo del sistema nervioso de los mamíferos. Rev Neurol 33: 54–68, 2001.[Medline]
- De Castro F, Hu L, Drabkin H, Sotelo C, and Chédotal A. Chemoattraction and chemorepulsion of olfactory bulb axons by different secreted semaphorins. J Neurosci 19: 4428–4436, 1999.[Abstract/Free Full Text]
- Diefenbach TJ, Guthrie PB, and Kater SB. Stimulus story alters behavioural responses of neuronal growth cones. J Neurosci 20: 1484–1494, 2000.[Abstract/Free Full Text]
- Kennedy TE. Cellular mechanisms of netrin function: long-range and short-range actions. Biochem Cell Biol 78: 569–575, 2000.[ISI][Medline]
- Llambi F, Causeret F, Bloch-Gallego E, and Mehlen P. Netrin-1 acts as a survival factor via its receptors UNC5H and DCC. EMBO J 20: 2715–2722, 2001.[ISI][Medline]
- Marín O and Rubenstein JLR. A long, remarkable journey: tangential migration in the telencephalon. Nat Neurosci Rev 2: 1–11, 2001.
- Mueller BK. Growth cone guidance: first steps towards a deeper understanding. Annu Rev Neurosci 22: 351–388, 1999.[ISI][Medline]
- Polleux F, Morrow T, and Ghosh A. Semaphorin 3A is a chemoattractant for cortical apical dendrites. Nature 404: 567–573, 2000.[Medline]
- Ramón y Cajal S. La retine des vertébrés. La Cellule 9: 119–258, 1893.
- Raper JA. Semaphorins and their receptors in vertebrates and invertebrates. Curr Opin Neurobiol 10: 88–94, 2000.[ISI][Medline]
- Sarnat HB. Molecular genetic classification of central nervous system malformations. J Child Neurol 15: 675–687, 2000.[Abstract/Free Full Text]
- Song H and Poo M. The cell biology of neuronal navigation. Nat Cell Biol 3: E81–E88, 2001.[ISI][Medline]
- Sotelo C. From Cajal’s chemotaxis to the molecular biology of axon guidance. Brain Res Bull 50: 395–396, 1999.[Medline]
- Soussi-Yanicostas N, de Castro F, Julliard K, Perfettini I, Chédotal A, and Petit C. Anosmin-1, defective in X-linked form of Kallman syndrome, promotes axonal branch formation from olfactory bulb output neurons. Cell 109: 217–228, 2002.[ISI][Medline]
- Spassky N, de Castro F, Le Bras B, Heydon K, Quéraud-LeSaux F, Bloch-Gallego E, Chédotal A, Zalc B, and Thomas J-L. Directional guidance of oligodendroglial migration by class 3 semaphorins and netrin-1. J Neurosci 22: 5992–6004, 2002.[Abstract/Free Full Text]
- Stein E and Tessier-Lavigne M. Hierarchical organization of guidance receptors: silencing of netrin attraction by slit through a Robo/DCC receptor complex. Science 291: 1928–1938, 2001.[Abstract/Free Full Text]
- Sugimoto Y, Taniguchi M, Yagi T, Akagi Y, Nojyo Y, and Tamamaki N. Guidance of glial precursor cell migration by secreted cues in the developing optic nerve. Development 128: 3321–3330, 2001.[Abstract/Free Full Text]
Copyright © 2003 by the Int. Union Physiol. Sci./Am. Physiol. Soc.
Top
Introduction
Netrins and axonal guidance
