Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Mechanisms of Development 105 (2001) 47±56 www.elsevier.com/locate/modo Neuronal migration Catherine Lambert de Rouvroit, Andre M. Gof®net* Neurobiology Unit, University of Namur Medical School, 61 Rue de Bruxelles, B5000 Namur, Belgium Received 7 December 2000; accepted 22 January 2001 Abstract Like other motile cells, neurons migrate in three schematic steps, namely leading edge extension, nuclear translocation or nucleokinesis, and retraction of the trailing process. In addition, neurons are ordered into architectonic patterns at the end of migration. Leading edge extension can proceed at the extremity of the axon, by growth cone formation, or from the dendrites, by formation of dendritic tips. Among both categories of leading edges, variation seems to be related to the rate of extension of the leading process. Leading edge extension is directed by micro®lament polymerization following integration of extracellular cues and is regulated by Rho-type small GTPases. In humans, mutations of ®lamin, an actin-associated protein, result in heterotopic neurons, probably due to defective leading edge extension. The second event in neuron migration is nucleokinesis, a process which is critically dependent on the microtubule network, as shown in many cell types, from slime molds to vertebrates. In humans, mutations in the PAFAH1B1 gene (more commonly called LIS1) or in the doublecortin (DCX) gene result in type 1 lissencephalies that are most probably due to defective nucleokinesis. Both the Lis1 and doublecortin proteins interact with microtubules, and two Lis1-interacting proteins, Nudel and mammalian NudE, are components of the dynein motor complex and of microtubule organizing centers. In mice, mutations of Cdk5 or of its activators p35 and p39 result in a migration phenotype compatible with defective nucleokinesis, although an effect on leading edge formation is also likely. The formation of architectonic patterns at the end of migration requires the integrity of the Reelin signalling pathway. Other known components of the pathway include members of the lipoprotein receptor family, the intracellular adaptor Dab1, and possibly integrin alpha 3 beta 1. Defective Reelin leads to poor lamination and, in humans, to a lissencephaly phenotype different from type 1 lissencephaly. Although the action of Reelin is unknown, it may trigger some recognition-adhesion among target neurons. Finally, pattern formation requires the integrity of the external limiting membrane, defects of which lead to overmigration of neurons in meninges and to human type 2 lissencephaly. q 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Lissencephaly; PAFAH1B1; Lis1; Doublecortin; Cdk5; Nudel; p35; p39; Reeler; Reelin; Dab1; VLDLR; ApoER2; Filamin; Limiting membrane; Integrin 1. Introduction Cell migration occurs widely at all stages of embryonic development, yet it is nowhere more evident than during mammalian brain development. Like other cells, neurons migrate following three schematic steps (Fig. 1). First, the cell extends leading edges, preceded by ®lopodia and lamellipodia that explore the micro-environment. Leading edge extension proceeds following the integration of attractive and repulsive signals generated at the plasma membrane. This integration, thought to be regulated by small GTPases switches of the Rho family (Luo, 2000) in¯uences the peripheral micro®lament network by regulating actin polymerization. Secondly, cell migration requires that the nucleus moves into the leading process, a step referred to * Corresponding author. Tel.: 132-81-724-277; fax: 132-81-724-280. E-mail address: andre.gof®[email protected] (A.M. Gof®net). as `nucleokinesis' following Morris (1998; and references therein). Nucleokinesis is distinct and differentially regulated from leading edge extension, and critically dependent on the microtubule cytoskeleton. The third event in cell migration is retraction of the trailing process, an aspect of neuronal migration that is scarcely addressed in the literature. Unlike other cells, neurons form de®ned cell patterns (architectonics) at the end of migration, and this may be considered a `fourth step' of neuronal migration. Pattern formation is controlled by the Reelin signaling pathway and requires the integrity of the `limiting membrane', consisting of the sheet of glial end-feet and basal lamina that separate the brain from the meninges (Fig. 2). 2. Leading edge extension Neurons extend different leading edges, only some of which are followed by nucleokinesis and neuronal migra- 0925-4773/01/$ - see front matter q 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0925-477 3(01)00396-3 48 C. Lambert de Rouvroit, A.M. Gof®net / Mechanisms of Development 105 (2001) 47±56 Fig. 1. Schematic view of a migrating neuron. Leading edge extension occurs by polymerization of the actin meshwork, while nucleokinesis is dependent on microtubules organized with their minus end radiating from the microtubule organizing center (yellow dot). The retraction of the trailing process is generally not considered important. tion. By birth, neurons are epithelial bipolar cells (Bradke and Dotti, 2000). As a ®rst approximation, the axonal domain derives from the luminal, ventricular pole of the neuroepithelial precursor, whereas the somatodendritic membrane derives from the basolateral domain. Although common mechanisms probably regulate axonal and dendritic extension, this fundamental distinction between axonal and dendritic membrane domains is likely to be re¯ected at the level of the leading edge. Thus, a ®rst useful distinction may be between leading edge extension at the extremity of axons, that is growth cone formation, and at the level of dendrites. Axonal growth cone elongation is by far the most extensively investigated mode of process extension (reviewed in Tessier-Lavigne and Goodman, 1996; Mueller, 1999). In contrast, it is rarely considered in the context to neuronal Fig. 2. Schematic illustration of normal (A) and different abnormalities of cortical development (C±E). In the normal cortex, neurons proliferate in the ventricular zone and migrate along radial ®bers. Early postmigratory neurons settle horizontally in the marginal zone (red) and subplate (pink). Reelin produced in the marginal zone (red neurons) helps organize postmigratory neurons into layers. Early born cortical neurons (green) are found in the depth of the cortex and younger cells (blue) more super®cially, according to the so-called `inside-outside' gradient. When leading edge extension is defective (B) as in familial heterotopias, neurons cannot migrate from the ventricular zones and settle locally. In case of defective nucleokinesis (C), as occurs probably in type 1 lissencephaly, initial migration is satisfactory but many cells settle at subcortical level. In the absence of Reelin signaling (D), cortical neurons (green) settle obliquely in the cortex and displace the early contingent (red and pink) `en bloc' in the marginal zone; the next generation (blue) cannot cross the ®rst one and the gradient is directed from outside to inside. Finally, when the limiting membrane is defective (E), as in type 2 lissencephaly, streams of neurons overmigrate in the marginal zone and in the meninges. C. Lambert de Rouvroit, A.M. Gof®net / Mechanisms of Development 105 (2001) 47±56 migration, because it is seldom followed by nucleokinesis. For example, motoneurones send long axons to peripheral muscles, but the cell body does not move out of the spinal cord. In contrast, sensory neurons derived from the neural crest extend an early axon capped by a growth cone; the nucleus then moves into this process for some length but stops in the dorsal root ganglion where it translocates out of the axon, leaving the well-known T-shaped sensory neuron. In the central nervous system (CNS), neuronal migration following growth cone elongation is referred to as `neuro(no)philic' (Rakic, 1990); it occurs tangentially and super®cially, close to the pial surface. The best studied examples are the subpial migration of external cerebellar granule cells from the rhombic lip over the cerebellar cortex (Hatten, 1999) and the circumferential migration of inferior olivary neurons from the rhombic lip around the brainstem (Bourrat and Sotelo, 1988). In the primate telencephalon, the stream formed by the subpial granular layer (SGL) seems to follow a similar migration mode (Meyer et al., 2000). Although the growth cone navigates in subpial position, it is worth pointing out that it does not come in direct contact with the basal lamina but is separated from the mesoderm by the end-feet of neuroepithelial-glial cells. The difference between growth cone elongation with and without cell migration is related to the control of nucleokinesis rather than growth cone navigation per se. The role of netrins during migration of pontine (Yee et al., 1999) and inferior olivary cells (Bloch-Gallego et al., 1999) suggests that the factors that control directional growth during neuronophilic migration are similar to those that control growth cone steering. In contrast to the leading process of inferior olivary neurons, indistinguishable from an axon (Bourrat and Sotelo, 1988; Bloch-Gallego et al., 1999), the long leading process that precedes migrating pontine neurons lacks many classical axonal markers, suggesting that there are intermediate differentiation states between long leading edges and fully differentiated axons (Yee et al., 1999). Other modes of leading edge extension proceed at the dendritic domain of the neuron and are morphologically described as `dendritic growth cones or tips'. Most often, the cell processes elongate along the radial dimension of the neuroepithelium and radial glial cells. This has been mostly studied in the cerebral cortex and cerebellum, and is referred to as `gliophilic migration' (for review, see Rakic, 1990; 2000a,b). It accounts for the majority of neuronal migration events all over the brain. The radial, gliophilic tip is elongate, closely apposed to the surface of radial cells that are thought to provide the substrate and guide its extension. Among the proteins implicated in neuronal-glial interactions are astrotactin (see Hatten, 1999, for review) and integrins, particularly integrin alpha3-beta1 that seems to mediate the interaction between the tip and substrate, at least in the cerebral cortex (Anton et al., 1999). Apart from this, our understanding of the determinants that govern dendritic tip steering and elongation remains frustratingly rudimentary. A reason may be that, contrary to axonal elon- 49 gation, migration by radial tip formation is not easily amenable to experimental analysis in vitro. It is worth pointing out that radial neuronal migration can occur without formation of a typical dendritic tip, particularly at early developmental stages when neurons have to travel very short distances. Recent observations suggest that, during very early cortical development, some neurons migrate radially without forming a leading process, simply by losing their attachment to the ventricle and undergoing nuclear translocation (Nadarajah et al., 1999). In addition, early neurons that migrate to the cerebral cortex do not have a bipolar but a stellate shape, as shown long ago with the Golgi method (e.g. Derer, 1994) and electron microscopic reconstruction (Shoukimas and Hinds, 1978). In comparison with later migrating cells, these stellate neurons seem to have more than one leading process and dendritic tip; they do not strictly follow radial cell extensions and their migration is slow. The presence of radially migrating cells that do not follow the classical gliophilic dendritic tip may explain the presence of an early, well formed cortical plate in the cerebral cortex of mice de®cient in cyclin-dependent kinase 5 (Cdk5) (see below). Clearly, migration by radial, dendritic tip formation obeys cues that are different from those that govern neuronophilic migration. In many instances, radial migration is orthogonal to subpial migration. Both processes can be activated sequentially in the same cell, for example in cerebellar granule cells, and lead to extension of axonal and dendritic processes in different directions. An important, unresolved question concerns the mechanisms that direct nucleokinesis either in a growing axon or in a growing dendrite. Until a few years ago, leading edge extension by formation of subpial axonal growth cones and radial dendritic tips was considered to account for all cases of neuronal migration. However, important examples have been discovered that depart from these two modes. First, as mentioned above, radial non-gliophilic migration is observed during early cortical development. A second case is the so-called `chain migration' by which neurons migrate from subventricular zones in the lateral ganglionic eminences to the olfactory bulb (and possibly other forebrain areas) by following a glial tunnel that extends through the tissue, mostly with a tangential orientation (Luskin, 1993; Wichterle et al., 1997; Law et al., 1999). The third example is the migration of GABAergic neurons from the medial ganglionic eminence to the cerebral cortex (Anderson et al., 1997, reviewed by Parnavelas, 2000). These neurons have an elongated shape with one or two prominent, horizontally elongated leading processes that extend through the intermediate zone by following the trajectory of corticofugal axons (Denaxa et al., 2000). Interestingly, they expressed calcium-permeant AMPA receptors whereas corticofugal axons are glutamatergic (MeÂtin et al., 2000). Some of these neurons even seem to migrate towards the cortical ventricular zone (Nadarajah et al., 2000). As far as we know, whether the leading edges from cortical immigrants 50 C. Lambert de Rouvroit, A.M. Gof®net / Mechanisms of Development 105 (2001) 47±56 from the ganglionic eminences correspond to axonal growth cones or dendritic tips remains unclear. Similarly, the regulation of nuclear displacement in these cells is unknown. Such a wide variety of leading processes presumably re¯ects the diversity of microenvironmental cues and the diverse signaling pathways that direct their growth. In the face of such diversity, apart from the basic difference between axonal and dendritic cones, it is tempting to try and de®ne some common mechanisms. Morphological studies suggest that the shape of a leading process re¯ects its rate of extension, maximal for growth cones, less rapid for tangential intraparenchymatous extension and even slower for radial migration. In the case of axonal growth cones, it is well known that they are larger and more prominent when the elongation rate is slow, as the growth cone explores the environment, whereas growth cones that cap rapidly extending axons are thin and straight. Some observations, however, suggest that this view is oversimpli®ed and that there are different machineries for leading edge extension. As a ®rst example, X-linked periventricular heterotopias in humans have been linked to mutations in ®lamin, an actin-associated protein (see Fox et al., 1998; Fox and Walsh, 1999). This observation shows that, like axonal growth cones, leading tip elongation requires the contractility of actin micro®laments. Nonmuscle ®lamin is a large protein (265±280 kDa) with two actin-binding domains linked by a long ¯exible shaft, that cross links actin ®laments into a loose, plastic meshwork. Filamin can bind to the intracellular domain of integrins, caveolin-1, TNF alpha receptor, and other surface proteins, thus providing a link between the cell surface and the actin network. This coupling is regulated by the Ras-related small GTPases Cdc42, Rac, Rho and RalA which all bind ®lamin. In the case of radial neuronal migration, it is thought that ®lamin de®ciency blocks the elongation of the dendritic tip along radial cell extensions, for example by uncoupling integrins from micro®laments. As a result neurons cannot leave the ventricular zone and differentiate locally into nodular heterotopias. Interestingly, although detailed neuropathological studies are not available, it seems that this mutation does not affect leading edge extension by growth cone formation. Apparently, connections and even commissural tracts develop normally, and cerebellar granule cells migrate satisfactorily. This malformation may thus de®ne a speci®c, ®lamin-dependent mode of leading edge extension that appears to be used mainly for radial (gliophilic migration), an idea that could be checked by careful neuropathological studies and by analysis of relevant mouse models. Another example in which formation of different leading edges seems to be differentially affected in Cdk5 or p35 de®cient mice (Ohshima et al., 1996; Gilmore et al., 1998; Chae et al., 1997; Kwon and Tsai, 2000). Cdk 5 mutant mice harbor a wide defect of radial migration all over the brain. Subpial neuronophilic migration of cerebellar granule cells and inferior olivary neurons also appears defective, and there are signs of axonal degeneration. However, many tracts appear to develop normally, initial cortical plate formation in the cortex proceeds correctly and the tangential, intraparenchymatous migration of GABAergic neurons from ganglionic eminence to cortex seems relatively unaffected. 3. Nucleokinesis As mentioned above, leading edge extension is not consistently followed by nuclear translocation. However, it is mainly the translocation of the nucleus into the leading process that de®nes cell migration, and the factors that regulate this process are thus of paramount importance (Book and Morest, 1990). Two examples illustrate this point. The ®rst is again cerebellar granule cells. During tangential subpial migration, their nuclei move in the immature axon, forming the external granule cell layer. In response to a signal from Purkinje cells, probably Hedgehog (Wechsler-Reya and Scott, 1999), a dendritic tip extends from the cell body, and elongates radially along Bergmann glial cells. The nucleus then translocates in this process and settles in the depth of the cerebellar cortex, forming the inner granule cell layer. In this case, the axonal growth cone and dendritic tip are both followed by nuclear displacement, yet follow different cues, as they elongate in orthogonal directions. A second example is facial motoneurons. They are generated in the ventricular zone of the fourth ventricle and send an early axon laterally to form the facial nerve. Then they extend a dendritic process that extends radially, probably guided by radial cells, into which nucleokinesis proceeds. This movement of the cell body leaves the axon behind, resulting into the typical `hairpin' trajectory of facial nerve known as the `genu'. In our view, these examples illustrate a key question about neuronal migration: What are the factors that allow and regulate nuclear translocation into a growing immature process? As introduced above, the term `nucleokinesis' was proposed by Morris to describe mutations which disturb the regular spacing of the cell nuclei in Aspergillus nidulans, and are abbreviated `Nud' for nuclear distribution (note that, even though both processes depend on microtubules, `nucleokinesis' is different from `karyokinesis', which refers to the translocation of chromosomes during mitosis). Similar genes have been cloned in Neurospora crassa and abbreviated `ro' for `ropy', as well as in yeast (reviewed by Fischer, 1999). The relationships between these genes and their mammalian orthologs are summarized in Table 1. The similarity between NudF and mammalian PAFAH1B1/Lis1 (see below) suggests that nucleokinesis is quite independent from leading edge extension and is an essential, hitherto unrecognized component of neuronal migration that is more critically dependent on microtubules. A recent review of microtubule micromotors in neurons was published by Goldstein and Yang (2000), and mechanisms of nuclear positioning are discussed in Reinsch and GoÈnczy (1998). C. Lambert de Rouvroit, A.M. Gof®net / Mechanisms of Development 105 (2001) 47±56 51 Table 1 Relationship between genes or proteins involved in nuclear positioning in slime molds and higher eukaryotes a Cytoplasmic dynein components Dynactin components Other dynein pathway components a A. nidulans N. crassa Higher eukaryotes NUDA NUDI NUDG RO1 CD heavy chain CD intermediate chain CD 8-kDa light chain P150 glued ARP1 p70 Lis1 (PAFAH1b) NUDC Xenopus MP43, Nudel, mammalian NudE NUDK NUDF NUDC NUDE RO3 RO4 RO2 RO11 Adapted from Morris (2000). Disturbances of nucleokinesis are implicated in human brain malformations known as type 1 lissencephalies, which are characterized by defective migration to the cerebral cortex and are occasionally associated with disturbances of migration in other parts of the brain. In many cases, type 1 lissencephaly has a genetic origin and two genes have been cloned thus far, called lissencephaly 1 (PAFAH1B1 / Lis1) and doublecortin (DCX). Mutations of LIS1 result in autosomal dominant type 1 lissencephaly (Reiner et al., 1993; Hattori et al., 1994), whereas mutations of doublecortin are responsible for X-linked lissencephaly in males and for the double-cortex malformation in carrier females (des Portes et al., 1998; Gleeson et al., 1998). In addition, the migration defects in mice with null mutations in the cyclin-dependent kinase 5 (Cdk5) and its cofactor p35 are reminiscent of human lissencephalies (Ohshima et al., 1996; Chae et al., 1997) and will be summarized in this part. Other genes implicated in type 1 lissencephalies will most probably be isolated in the coming years. 3.1. Lis1 Homozygous null mutations of Lis1 are embryonic lethal in mice and presumably in humans as well. Heterozygosity for Lis1 mutations leads to type 1 lissencephaly, characterized by profuse migration defects in the brain (Reiner et al., 1993; Hattori et al., 1994). Mice with one inactive allele display cortical, hippocampal and olfactory bulb disorganization resulting from delayed neuronal migration by a cellautonomous neuronal pathway. Mice with further reduction of Lis1 activity display more severe brain disorganization as well as cerebellar defects (Hirotsune et al., 1998). The Lis1 protein, which binds to microtubules, is also a non-catalytic subunit of platelet-activating factor (PAF) acetyl hydrolase 1b (PAFAH1b), an enzyme that inactivates PAF (Hattori et al., 1994). Whether these two apparently different roles are related remains unknown. Interestingly, the composition of PAF acetyl hydrolase is developmentally regulated in the brain and the catalytic subunits of PAFAH compete with mammalian NudE (see below) for binding to Lis1 (Manya et al., 1998; Kitagawa et al., 2000). Since Morris ®rst noted the similarity between Lis1 and NudF, the link between Lis1 and the microtubular cytoskeleton has been actively studied (Morris, 2000). Lis1 protein coimmunoprecipitates with cytoplasmic dynein and components of the dynactin complex, and localizes to the cell cortex as well as to mitotic kinetochores, both sites of dynein binding (Smith et al., 2000). In cultured cells, overexpression of Lis1 leads to spindle misorientation and interferes with progression into mitosis, whereas interference with Lis1 function (e.g. with anti Lis1 antibodies) disturbs the arrangement of the metaphase plate (Faulkner et al., 2000). Lis1 also plays a role in dendritic elaboration and axonal transport, as shown in Drosophila and in cultured neurons (Liu et al., 2000). These experiments suggest that the lissencephaly that results from haploinsuf®ciency of Lis1 may be attributed to deregulation of cell cycle progression in the ventricular zone as well as to disturbances of the microtubule cytoskeleton that regulates nucleokinesis. Several groups have isolated two Lis1 interacting proteins that show similarity with the Aspergillus nidulans NudE protein (RO11 in Neurospora, MP43 in Xenopus). They have been named Nudel (for `NudE-like') and rNudE, mNudE, hNudE for rat/mouse/human NudE orthologs, which we will call mammalian NudE (Kitagawa et al., 2000; Feng et al., 2000; Niethammer et al., 2000; Sasaki et al., 2000). Mold NudE, Nudel and mammalian NudE are all coiled-coil proteins that have the capacity to bind Lis1 and provide an additional link between Lis1 and dynein. Nudel and mammalian NudE have a similar size and display 55% identity. However, the regulation of both genes is different, as mammalian NudE is ubiquitously expressed at early embryonic stages, whereas Nudel expression is restricted to brain and testes and is detected postnatally. The homology with Xenopus MP43, a protein that is phosphorylated in a cell-cycle dependent fashion, and the fact that Nudel is an in vitro substrate for Cdk5 suggest a role for phosphorylation in the regulation of NudE function (Niethammer et al., 2000; Sasaki et al., 2000). The coiled-coiled regions of both proteins interact with wild-type Lis1, but not with Lis1 proteins that harbor human lissencephaly mutations. In Xenopus embryos, dominant negative expression of 52 C. Lambert de Rouvroit, A.M. Gof®net / Mechanisms of Development 105 (2001) 47±56 mammalian NudE causes lamination defects in the retina, tectum and cortex. Both proteins are located in the centrosome or microtubule organizing center (MTOC) and interact with centrosomal components including gamma-tubulin, the dynein light chain and pericentrin (Feng et al., 2000). Thus, several lines of evidence indicate that mammalian NudE acts on the dynein motor complex and hence on nuclear migration through interaction with Lis1, re¯ecting the situation in lower organisms where NudE is located genetically between NudF and NudA (see Table 1). 3.2. Doublecortin Doublecortin (DCX) was characterized by positional cloning and is responsible for the doublecortex (subcortical laminar heterotopia) malformation in heterozygous carrier women, and for X-linked lissencephaly in males, a phenotype similar to type 1 lissencephaly (Gleeson et al., 1998; des Portes et al., 1998). The DCX gene encodes a 361 amino-acid protein that associates with and stabilizes microtubules (Francis et al., 1999; Gleeson et al., 1999) but shows no similarity to any other microtubule associated protein (Sapir et al., 1997; Taylor et al., 2000). An evolutionarily conserved domain is tandemly repeated in the N-terminal part of the protein and named DC repeat. Interestingly, the majority of human missense mutations cluster in these repeats. Each repeat alone is able to bind tubulin, but neither is suf®cient to mediate microtubule stabilization. Some human point mutations lead to impaired polymerization or disruption of microtubules, thereby altering the morphology of transfected cells. Biochemical and genetic evidence thus point to a role for DCX in crucial microtubule-based events in neuronal migration. However, which step of the migration process is affected in DCX mutants remains to be elucidated. In addition to the similarity between LIS1 and DCX mutant phenotypes, a physical interaction between Lis1 and DCX has been demonstrated both in vitro and in vivo (Horesh et al., 1999; Caspi et al., 2000). This interaction is not altered in DCX mutant proteins that fail to bind microtubules. As mentioned above, both LIS1 and DCX associate with microtubules and enhance their polymerization, but the precise mode of interaction between the three partners, and how these events relate to the lissencephaly phenotype remains unknown. Sequence similarity searches identi®ed two additional mammalian proteins with DC repeats, namely DCLK (doublecortin-like kinase) also called KIAA0369 or DCAMKL1, and ORP1 (retinitis pigmentosa 1). The DCLK gene maps to human chromosome 13q12.3 where no neurological disorders have been located so far. In addition to a N-terminal DC motif, the DCLK protein contains a transmembrane domain and a region of similarity with Ca 21-calmodulin-dependent protein kinases. The expression pattern of the DCLK mRNA in the embryonic brain suggests that this gene may be involved in cortical devel- opment (Sossey-Alaoui and Srivastava, 1999; Burgess and Reiner, 2000; Matsumoto et al., 1999). Moreover, recent data show that DCLK is associated with microtubules in neuronal growth cones (Burgess and Reiner, 2000). 3.3. Cyclin-dependent kinase 5 (Cdk5) and its cofactors p35/p39 Cdk5 was named because of its similarity to Cdc2. However, the protein is not primarily involved in the regulation of the cell cycle. It is expressed in all postmitotic neurons and is thought to regulate many developmental processes during neuronal differentiation. The range of Cdk5 substrates is not fully de®ned, but one of the main targets is the microtubule associated protein `tau'. In addition, Nudel is also a substrate of Cdk5, and may be a link between Cdk5/p35 and neuronal migration (Sasaki et al., 2000; Niethammer et al., 2000). The Cdk5 kinase activity requires the formation of a complex with one of two activators named p35 and p39. A role for these proteins in neuronal migration was demonstrated when inactivation of Cdk5 in mice was observed to result in profuse migration defects (Ohshima et al., 1996). Inactivation of p35 results in a more subtle neuronal migration disorder, probably due to some redundancy between p35 and p39 (Chae et al., 1997). Cdk5-de®cient mice have profuse neuronal migration defects in the cerebral cortex, hippocampus, cerebellum, as well as in several other parts of the CNS (Gilmore et al., 1998; Kwon and Tsai, 2000). In addition, some axon bundles are defective with evidence of axon swelling. Interestingly, as already pointed out above, the initial formation of the telencephalic preplate and early cortical plate proceeds normally, and GABAergic neurons from the ganglionic eminence seem to migrate in the cortex of Cdk5de®cient mice, suggesting that some neuron migration is Cdk5-independent. The initial migration in the telencephalic preplate and cortical plate is short range and morphologically different from later radial migration, suggesting that Cdk5 is needed only for long-range gliophilic migration. Even though Cdk5 has been shown to in¯uence growth cone formation (Nikolic et al., 1998), morphological observations of mutant mice suggest that the main action of Cdk5/p35/p39 during neuronal migration is via the control of nucleokinesis. 4. Patterning after migration. The Reelin signaling pathway and type 2 lissencephalies When neurons reach their destination, they stop migrating and order themselves into speci®c `architectonic' patterns. Observations in reeler mice show that the normal function of the Reelin signaling pathway is essential for proper radial organization and layering of postmigratory neurons at all levels of the CNS (reviewed by Lambert de Rouvroit and Gof®net, 1998a). Strong genetic evidence demonstrates that Reelin, the very low density lipoprotein receptor (VLDLR), C. Lambert de Rouvroit, A.M. Gof®net / Mechanisms of Development 105 (2001) 47±56 ApoE receptor type 2 (ApoER2), and the Disabled 1 (Dab1) tyrosine kinase adaptor belong to the same pathway. Integrin alpha3 beta1 also binds Reelin and could be another Reelin receptor (see below). Observations of type 2 lissencephaly in humans, together with some mouse mutations, suggest that the integrity of the external limiting membrane is also necessary for normal architectonic formation. Reelin, the protein defective in reeler mutant mice, is a large (about 400 kDa) glycoprotein secreted by several neurons in the extracellular micro-environment (D'Arcangelo et al., 1995; reviewed in Lambert de Rouvroit and Gof®net, 1998b). In the embryonic cerebral cortex, hippocampus and cerebellum, Reelin is expressed in the marginal zone and governs the disposition and dendritic deployment of target cells, that correspond respectively to cortical plate cells, pyramidal cells and Purkinje cells. The response of target cells to the Reelin signal involves the expression of membrane receptors and intracellular signaling proteins. Mice doubly de®cient in two lipoprotein receptors, namely very low density lipoprotein receptor (VLDLR) and apolipoprotein E receptor type 2 (ApoER2) have a phenotype indistinguishable from reeler (Trommsdorff et al., 1999). These proteins are expressed in Reelin target cells and bind Reelin in a speci®c manner (Hiesberger et al., 1999; D'Arcangelo et al., 1999). VLDLR and ApoER2 have a short cytoplasmic tail that interacts with the tyrosine kinase adapter Dab1 (Disabled 1) via a NPXY sequence involved in lipoprotein receptormediated endocytosis. Mutations of Dab1 also generate a reeler phenotype (Howell et al., 1997; Sheldon et al., 1997; Ware et al., 1997). Mammalian Dab1 is expressed in the cytoplasm of migrating cortical neurons as a 555 aminoacid protein comprising an N-terminal phosphotyrosine binding (PTB) domain responsible for the interaction with the lipoprotein receptors. Binding of Reelin to target cells induces Dab1 phosphorylation in vivo and in vitro (Howell et al., 1999). Five tyrosine phosphorylation sites have been mapped to the Dab1 sequence. The crucial importance of Dab1 phosphorylation in neuronal positioning is demonstrated by the generation of Dab1 `knock-in' animals in which the ®ve tyrosine residues have been replaced by phenylalanine. These mice display a phenotype similar to Dab1 null mutants (Howell et al., 2000). The kinase(s) involved in the tyrosine phosphorylation of Dab1 have not yet been identi®ed. Dab1 was identi®ed as a Src-binding protein in a two-hybrid screen, and is an in vitro substrate for tyrosine kinases of the Src family. However, none of the null mutants for Src, Abl or Fyn has a phenotype evocative of reeler. Reelin and presumably the other partners of the pathway are present and conserved in all vertebrates, suggesting a role in brain evolution (Bar et al., 2000). As mentioned above, integrin alpha 3 beta1 is implicated in the interaction between radially migrating cortical neurons and radial glial ®bers (Anton et al., 1999). Interestingly, biochemical studies show that this integrin has many properties of a Reelin receptor and participates in the formation of a supramolecular complex with Reelin and lipopro- 53 tein receptors (Dulabon et al., 2000). The formation of such a complex on target cells means that alpha 3 beta 1 integrin could be downregulated by endocytosis as soon as migrating neurons reach a Reelin-rich area such as the border of the marginal zone. This would detach immature neurons from radial glial ®bers and trigger their laminar organization in the cortical plate. Other mechanisms are considered to explain the action of Reelin. Reelin may provide a stop signal to migrating cells (Pearlman and Sheppard, 1996; Pearlman et al., 1998). Reelin cannot stop leading edge extension, as the presence of Reelin is permissive to the deployment of dendritic tips, for example in the marginal zone. On the other hand, cell bodies of target cells tend to avoid Reelin-rich areas, so that Reelin may act as a stop signal for nucleokinesis (Walsh and Gof®net, 2000). However, recent observations (Jensen et al., 2000) show that stopping nucleokinesis cannot explain all actions of Reelin. A possibility, already proposed long ago (Gof®net, 1979) is that Reelin could promote recognitionadhesion among postmigratory cells, thereby stabilizing early architectonic patterns. These and possibly other unrecognized actions of Reelin are not mutually exclusive. Surely, the complexity and size of the protein is compatible with engagement of different receptors, activation of different intracellular signaling pathways, and complex phenotypic response of target cells. In humans, Reelin is expressed during brain development (Meyer et al., 2000). Two families with recessive lissencephaly have been shown to carry reelin mutations (Hong et al., 2000). The human DAB1 gene maps to chromosome 1p31 (Lambert de Rouvroit and Gof®net, 1998b); mutations of this gene are predicted to yield a phenotype similar to mutations in reelin. The observation that defective Reelin signaling in humans is incompatible with normal cortical lamination and folding further demonstrates the key importance of this pathway for normal brain development. 4.1. The external limiting membrane In contrast to the Reelin pathway, which seems to play an `instructive' role during pattern formation, the role of the limiting membrane may be viewed more as a mechanical barrier to migration, and thus permissive to cortical lamination. The limiting membrane is composed of the end-feet of radial neuroepithelioglial cells and the basal lamina to which they adhere. Meningeal cells are the mesodermal component of the border between neural and mesodermal tissue. Defects in any of these three components is predicted to result in radial `overmigration', with invasion of the marginal zone and formation of neuronal streams in the meninges. This phenotype is observed in mice de®cient in presenilin-1 (Hartmann et al., 1999) and in alpha 6 integrins (Georges-Labouesse et al., 1998), as well as in MARCKS (Stumpo et al., 1995), a prominent neural substrate for protein kinase C (although the exencephaly phenotype in this mutant is more dramatic). These observations suggest 54 C. Lambert de Rouvroit, A.M. Gof®net / Mechanisms of Development 105 (2001) 47±56 that neuronal overmigration may result from defects in the basal lamina but also in proteoglycans in the marginal zone (discussed by Hartmann et al., 1999), and could also be related to mesodermal, particularly vascular anomalies. In humans, focal gliomeningeal heterotopias with overmigration are found sporadically, but are particularly prominent and considered typical of type 2 lissencephalies (Friede, 1989; Barkovich and Kjos, 1992). Genetic type 2 lissencephaly is associated with Fukuyama muscular dystrophy (FCMD), due to mutations in the so-called `fukutin' gene of unknown function, with merosin (laminin alpha-2, LAMA2) de®ciency, with Walker±Warburg syndrome (WWS) and with muscle-eye-brain (MEB) disease. As far as we know, the genes mutated in WWS and MEB have not yet been identi®ed. Interestingly, the basal lamina of skeletal muscle ®bers is thinned and disrupted in muscle biopsies of patients with FCMD, LAMA2 and WWS, further emphasizing the importance of the basal lamina as a key component of the limiting membrane. 4.2. Conclusion The data summarized above show that our understanding of neuronal migration is progressing at an accelerated pace. As always, new knowledge leads to new questions. As a conclusion, we would like to address some of the problems that we tend to consider as important and timely, bearing in mind that this is a most subjective selection. As we underscored in part one, it is important to de®ne the various classes of neuronal leading edges on a cell biological basis. In particular, we need to understand better the signals, receptors and postreceptors mechanisms that govern dendritic tip extension. Second, about nucleokinesis, the role of PAF metabolism needs to be clari®ed and the mechanism by which microtubule direct the position of the nucleus in the cell should be addressed. For example, is it simply the density of microtubules that allows or prohibit nuclear translocation or, as we think, are they more subtle features? The role on nucleokinesis of Cdk5, its cofactors and the recently characterized Nudel and NudE remains to be determined; human mutations for these genes are likely to be found in families with type 1 lissencephaly. Thirdly, about endmigration event and architectonic patterns, we still do not understand in molecular and even in descriptive terms what Reelin really does to end the migration of neurons. The role of the limiting membrane should be progressively unraveled with the analysis of different mouse and human mutations. Finally, all these themes would bene®t enormously from the development of a reliable in vitro system in which genuine neuronal migration and some degree of pattern formation can be observed and studied experimentally. Acknowledgements We wish to thank L.-H. Tsai, C. Walsh and A. WynshawBoris for sharing prepublication data and we apologize to colleagues whose work is not cited because of limitation constraints. Our laboratory is supported by grants FRSM 3.4533.95, FRFC 2.4544.01, ARC 99/03-248 and the Fondation Reine Elisabeth, all from Belgium. References Anderson, S.A., Eisenstat, D.D., Shi, L., Rubenstein, J.L., 1997. Interneuron migration from basal forebrain to neocortex: dependence on Dlx genes. Science 278, 474±476. Anton, E.S., Kreidberg, J.A., Rakic, P., 1999. Distinct functions of alpha3 and alpha(v) integrin receptors in neuronal migration and laminar organization of the cerebral cortex. Neuron 22, 277±289. Bar, I., Lambert de Rouvroit, C., Gof®net, A.M., 2000. The evolution of the cerebral cortex: an hypothesis based on the role of the reelin pathway. Trends Neurosci. 23, 633±638. Barkovich, A.J., Kjos, B., 1992. Gray matter heterotopias: MR characteristics and correlation with developmental and neurological manifestations. Radiology 182, 483±499. Bloch-Gallego, E., Ezan, F., Tessier-Lavigne, M., Sotelo, C., 1999. Floor plate and netrin-1 are involved in the migration and survival of inferior olivary neurons. J. Neurosci. 19, 4407±4420. Book, K.J., Morest, K., 1990. Migration of neuroblasts by perikaryal translocation. J. Comp. Neurol. 297, 55±76. Bourrat, F., Sotelo, C., 1988. Migratory pathways and neuritic differentiation of inferior olivary neurons in the rat embryo. Axonal tracing study using the in vitro slab technique. Brain Res. 467, 19±37. Bradke, F., Dotti, C.G., 2000. Establishment of neuronal polarity: lessons from cultured hippocampal neurons. Curr. Opin. Neurobiol. 10, 574± 581. Burgess, H.A., Reiner, O., 2000. Doublecortin-like kinase is associated with microtubules in neuronal growth cones. Mol. Cell. Neurosci. 16, 529±541. Caspi, M., Atlas, R., Kantor, A., Sapir, T., Reiner, O., 2000. Interaction between LIS1 and doublecortin, two lissencephaly gene products. Hum. Mol. Genet. 9, 2205±2213. Chae, T., Kwo, Y.T., Bronson, R., Dikkes, P., Li, E., Tsai, L.-H., 1997. Mice lacking p35, a neuronal speci®c activator of Cdk5, display cortical lamination defects, seizures, and adult lethality. Neuron 18, 29±42. D'Arcangelo, G., Miao, G.G., Chen, S.C., Soares, H.D., Morgan, J.I., Curran, T., 1995. A protein related to extracellular matrix proteins deleted in the mouse mutant reeler. Nature 374, 719±723. D'Arcangelo, G., Homayouni, R., Keshvara, L., Rice, D.S., Sheldon, M., Curran, T., 1999. Reelin is a ligand for lipoprotein receptors. Neuron 24, 471±479. Denaxa, M., Chan, C.H., Parnavelas, J.G., Karagogeos, D., 2000. The adhesion molecule TAG-1 mediates the migration of cortical interneurons from the ganglionic eminence along the axons of corticofugal ®bers. Soc. Neurosci. Abstr. 26, 569. Derer, P., 1994. HistogeneÁse du neÂocortex du Rat albinos durant la peÂriode fútale et neÂonatale. J. Hirnforsch. 15, 1±34. des Portes, V., Pinard, J., Billuart, P., Vinet, M., Koulakoff, A., Carrie, A., Gelot, A., Dupuis, E., Motte, J., Berwald-Netter, Y., Catala, M., Kahn, A., Beldjord, C., Chelly, J., 1998. A novel CNS gene required for neuronal migration and involved in X-linked subcortical laminar heterotopia and lissencephaly syndrome. Cell 92, 51±61. Dulabon, L., Olson, E.C., Taglienti, M.G., Eisenhuth, S., MacGrath, B., Walsh, C.A., Kreidberg, J.A., Anton, E.S., 2000. Reelin binds alpha3beta1 integrin and inhibits neuronal migration. Neuron 27, 33±44. Faulkner, N.E., Dujardin, D.L., Tai, C.Y., Vaughan, K.T., O'Connell, C.B., Wang, Y., Vallee, R.B., 2000. A role for the lissencephaly gene LIS1 in mitosis and cytoplasmic dynein function. Nat. Cell Biol. 2, E201±E202. Feng, Y., Olson, E., Todd Stukenberg, P., Flanagan, L.A., Kirscher, M.W., Walsh, C.A., 2000. LIS1 regulates CNS lamination by interacting with mNudE, a central component of the centrosome. Neuron 28, 665±679. C. Lambert de Rouvroit, A.M. Gof®net / Mechanisms of Development 105 (2001) 47±56 Fischer, R., 1999. Nuclear movement in ®lamentous fungi. FEMS Microbiol. Rev. 23, 39±68. Fox, J.W., Walsh, C.A., 1999. Periventricular heterotopia and the genetics of neuronal migration in the cerebral cortex. Am. J. Hum. Genet. 65, 19±24. Fox, J.W., Lamperti, E.D., Eksioglu, Y.Z., Hong, S.E., Feng, Y., Graham, D.A., Scheffer, I.E., Dobyns, W.B., Hirsch, B.A., Radtke, R.A., Berkovic, S.F., Huttenlocher, P.R., Walsh, C.A., 1998. Mutations in ®lamin 1 prevent migration of cerebral cortical neurons in human periventricular heterotopia. Neuron 21, 1315±1325. Francis, F., Koulakoff, A., Boucher, D., Chafey, P., Schaar, B., Vinet, M.C., Friocourt, G., McDonnell, N., Reiner, O., Kahn, A., McConnell, S.K., Berwald-Netter, Y., Denoulet, P., Chelly, J., 1999. Doublecortin is a developmentally regulated, microtubule-associated protein expressed in migrating and differentiating neurons. Neuron 23, 247±256. Friede, R.L., 1989. Developmental Neuropathology, Springer Verlag. Georges-Labouesse, E., Mark, M., Messaddeq, N., Gansmuller, A., 1998. Essential role of alpha 6 integrins in cortical and retinal lamination. Curr. Biol. 8, 983±986. Gilmore, E.C., Ohshima, T., Gof®net, A.M., Kulkarni, A.B., Herrup, K., 1998. Cyclin-dependent kinase 5-de®cient mice demonstrate novel developmental arrest in cerebral cortex. J. Neurosci. 18, 6370±6377. Gleeson, J.G., Allen, K.M., Fox, J.W., Lamperti, E.D., Berkovic, S., Scheffer, I., Cooper, E.C., Dobyns, W.B., Minnerath, S.R., Ross, M.E., Walsh, C.A., 1998. Doublecortin, a brain-speci®c gene mutated in human X-linked lissencephaly and double cortex syndrome, encodes a putative signaling protein. Cell 92, 63±72. Gleeson, J.G., Lin, P.T., Flanagan, L.A., Walsh, C.A., 1999. Doublecortin is a microtubule-associated protein and is expressed widely by migrating neurons. Neuron 23, 257±271. Gof®net, A.M., 1979. An early development defect in the cerebral cortex of the reeler mouse. A morphological study leading to a hypothesis concerning the action of the mutant gene. Anat. Embryol. 157, 205± 216. Goldstein, L.S.B., Yang, Z., 2000. Microtubule-based transport systems in neurons. Annu. Rev. Neurosci. 23, 39±71. Hartmann, D., De Strooper, B., Saftig, P., 1999. Presenilin-1 de®ciency leads to loss of Cajal-Retzius neurons and cortical dysplasia similar to human type 2 lissencephaly. Curr. Biol. 9, 719±727. Hatten, M.E., 1999. Central nervous system neuronal migration. Annu. Rev. Neurosci. 22, 511±539. Hattori, M., Adachi, H., Tsujimoto, M., Arai, H., Inoue, K., 1994. The catalytic subunit of bovine brain platelet-activating factor acetylhydrolase is a novel type of serine esterase. J. Biol. Chem. 269, 23150±23155. Hiesberger, T., Trommsdorff, M., Howell, B.W., Gof®net, A.M., Mumby, M.C., Cooper, J.A., Herz, J., 1999. Direct binding of Reelin to VLDL receptor and ApoE receptor 2 induces tyrosine phosphorylation of disabled-1 and modulates tau phosphorylation. Neuron 24, 481±489. Hirotsune, S., Fleck, M.W., Gambello, M.J., Bix, G.J., Chen, A., Clark, G.D., Ledbetter, D.H., McBain, C.J., Wynshaw-Boris, A., 1998. Graded reduction of Pafah1b1 (Lis1) activity results in neuronal migration defects and early embryonic lethality. Nat. Genet. 19, 333±339. Hong, S.E., Shugart, Y.Y., Huang, Y.Y., Shahwan, S.A., Grant, P.E., Hourihane, J.O., Martin, N.D., Walsh, C.A., 2000. Autosomal recessive lissencephaly with cerebellar hypoplasia is associated with human RELN mutations. Nat. Genet. 26, 93±96. Horesh, D., Sapir, T., Francis, F., Wolf, S.G., Caspi, M., Elbaum, M., Chelly, J., Reiner, O., 1999. Doublecortin, a stabilizer of microtubules. Hum. Mol. Genet. 8, 1599±1610. Howell, B.W., Hawkes, R., Soriano, P., Cooper, J.A., 1997. Neuronal position in the developing brain is regulated by mouse disabled-1. Nature 389, 733±737. Howell, B.W., Herrick, T.M., Cooper, J.A., 1999. Reelin-induced tyrosine phosphorylation of disabled 1 during neuronal positioning. Genes Dev. 13, 643±648. Howell, B.W., Herrick, T.M., Hildebrand, J.D., Zhang, Y., Cooper, J.A., 55 2000. Dab1 tyrosine phosphorylation sites relay positional signals during mouse brain development. Curr. Biol. 10, 877±885. Jensen, P., Zoghbi, H.Y., Goldowitz, D., 2000. Dissection of the cellular and molecular events that direct cerebellar development. Soc. Neurosci. Abstr. 26, 58. Kitagawa, M., Umezu, M., Aoki, J., Koizumi, H., Arai, H., Inoue, K., 2000. Direct association of LIS1, the lissencephaly gene product, with a mammalian homologue of a fungal nuclear distribution protein, rNUDE. FEBS Lett. 479, 57±62. Kwon, Y.T., Tsai, L.H., 2000. The role of the p35/cdk5 kinase in cortical development. Results Probl. Cell Differ. 30, 241±253. Lambert de Rouvroit, C., Gof®net, A.M., 1998a. The reeler mouse as a model of brain development. Adv. Anat. Embryol. Cell Biol. 150, 1± 106. Lambert de Rouvroit, C., Gof®net, A.M., 1998b. Cloning of human DAB1 and mapping to chromosome 1p31-p32. Genomics 53, 246±247. Law, A.K., Pencea, V., Buck, C.R., Luskin, M.B., 1999. Neurogenesis and neuronal migration in the neonatal rat forebrain anterior subventricular zone do not require GFAP-positive astrocytes. Dev. Biol. 216, 622±634. Liu, Z., Steward, R., Luo, L., 2000. Drosophila Lis1 is required for neuroblast proliferation, dendritic elaboration and axonal tranport. Nat. Cell Biol. 2, 776±783. Luo, L., 2000. Rho GTPases in neuronal morphogenesis. Nat. Rev. Neurosci. 1, 173±180. Luskin, M.B., 1993. Restricted proliferation and migration of postnatally generated neurons derived from the forebrain subventricular zone. Neuron 11, 173±189. Manya, H., Aoki, J., Watanabe, M., Adachi, T., Asou, H., Inoue, Y., Arai, H., Inoue, K., 1998. Switching of platelet-activating factor acetylhydrolase catalytic subunits in developing rat brain. J. Biol. Chem. 273, 18567±18572. Matsumoto, N., Pilz, D.T., Ledbetter, D.H., 1999. Genomic structure, chromosomal mapping, and expression pattern of human DCAMKL1 (KIAA0369), a homologue of DCX (XLIS). Genomics 56, 179±183. MeÂtin, C., Denizot, J.P., Ropert, N., 2000. Intermediate zone cells express calcium-permeable AMPA receptors and establish close contact with growing axons. J. Neurosci. 20, 696±708. Meyer, G., Schaaps, J.P., Moreau, L., Gof®net, A.M., 2000. Embryonic and early fetal development of the human neocortex. J. Neurosci. 20, 1858± 1868. Morris, N.R., E®mov, V.P., Xiang, X., 1998. Nuclear migration, nucleokinesis and lissencephaly. Trends Cell Biol. 8, 467±470. Morris, R., 2000. A rough guide to a smooth brain. Nat. Cell Biol. 2, E201± E202. Mueller, B.K., 1999. Growth cone guidance. Annu. Rev. Neurosci. 22, 351±388. Nadarajah, B., Brunstrom, J.E., Wong, R.O.L., Pearlman, A.L., 1999. Somal translocation during early layer formation in developing neocortex. Soc. Neurosci. Abstr. 25, 504. Nadarajah, B., Wong, R.O.L., Parnavelas, J.G., 2000. Neuronal migration directed towards the ventricular zone of the developing cortex. Soc. Neurosci. Abstr. 26, 569. Niethammer, M., Smith, D.S., Ayala, R., Peng, J., KO, M.-S., Morabito, M., Tsai, L.-H., 2000. NUDEL is a novel Cdk5 substrate that associates with Lis1 and cytoplasmic dynein. Neuron 28, 697±711. Nikolic, M., Chou, M.M., Lu, W., Mayer, B.J., Tsai, L.-H., 1998. The p35/ Cdk5 kinase is a neuron-speci®c Rac effector that inhibits Pak1 activity. Nature 395, 194±198. Ohshima, T., Ward, J.M., Huh, C.G., Longenecker, G., Veeranna, Pant, H.C., Brady, R.O., Martin, L.J., Kulkarni, A.B., 1996. Targeted disruption of the cyclin-dependent kinase 5 gene results in abnormal corticogenesis, neuronal pathology and perinatal death. Proc. Natl. Acad. Sci. USA 93, 11173±11178. Parnavelas, J., 2000. The origin and migration of cortical neurons: New vistas. Trends Neurosci. 23, 126±131. Pearlman, A.L., Sheppard, A.M., 1996. Extracellular matrix in early cortical development. Prog. Brain Res. 108, 117±134. 56 C. Lambert de Rouvroit, A.M. Gof®net / Mechanisms of Development 105 (2001) 47±56 Pearlman, A.M., Faust, P.L., Hatten, M.E., Brunstrom, J.E., 1998. New directions for neuronal migration. Curr. Opin. Neurobiol. 8, 45±54. Rakic, P., 1990. Principles of neural cell migration. Experientia 46, 882± 891. Rakic, P., 2000a. Molecular and cellular mechanisms of neuronal migration: relevance to cortical epilepsies. Adv. Neurol. 84, 1±14. Rakic, P., 2000b. Radial unit hypothesis of neocortical expansion. In: Bock, G.R., Cardew, G. (Eds.). `Evolutionray Developmental Biology of the Cerebral Cortex', Novartis foundation Symp, pp. 30±45. Reiner, O., Carrozzo, R., Shen, Y., Wehnert, M., Faustinella, F., Dobyns, W.B., Caskey, C.T., Ledbetter, D.H., 1993. Isolation of a Miller-Dieker lissencephaly gene containing G protein beta-subunit-like repeats. Nature 364, 717±721. Reinsch, S., GoÈnczy, P., 1998. Mechanisms of nuclear positioning. J. Cell Sci. 111, 2283±2295. Sapir, T., Elbaum, M., Reiner, O., 1997. Reduction of microtubule catastrophe events by LIS1, platelet-activating factor acetylhydrolase subunit. EMBO J. 16, 6977±6984. Sasaki, S., Shionoya, A., Ishida, M., Gambello, M., Yingling, J., WynshawBoris, A., Hirotsune, S., 2000. A LIS1/NUDEL/cytoplasmic dynein heavy chain complex in the developing and adult nervous system. Neuron 28, 681±696. Sheldon, M., Rice, D.S., D'Arcangelo, G., Yoneshima, H., Nakajima, K., Mikoshiba, K., Howell, B.W., Cooper, J.A., Goldowitz, D., Curran, T., 1997. Scrambler and yotari disrupt the disabled gene and produce a reeler- like phenotype in mice. Nature 389, 730±733. Shoukimas, G.M., Hinds, J.W., 1978. The development of the cerebral cortex in the embryonic mouse: an electron microscopic serial section analysis. J. Comp. Neurol. 179, 795±830. Smith, D.S., Niethammer, M., Ayala, R., Zhou, Y., Gambello, M.J., Wynshaw-Boris, A., Tsai, L.H., 2000. Regulation of cytoplasmic dynein behaviour and microtubule organization by mammalian Lis1. Nat. Cell Biol. 2, 767±775. Sossey-Alaoui, K., Srivastava, A.K., 1999. DCAMKL1, a brain-speci®c transmembrane protein on 13q12.3 that is similar to doublecortin (DCX). Genomics 56, 121±126. Stumpo, D.J., Bock, C.B., Tuttle, J.S., Blackshear, P.J., 1995. MARCKS de®ciency in mice leads to abnormal brain development and perinatal death. Proc. Natl. Acad. Sci. USA 92, 944±948. Taylor, K.R., Hallzer, A.K., Bazan, J.F., Walsh, C.A., Gleeson, J.G., 2000. Patient mutations in doublecortin de®ne a repeated tubulin-binding domain. J. Biol. Chem. 275, 34442±34450. Tessier-Lavigne, M., Goodman, C., 1996. The molecular biology of axon guidance. Science 274, 1123±1133. Trommsdorff, M., Gotthardt, M., Hiesberger, T., Shelton, J., Stockinger, W., Nimpf, J., Hammer, R.E., Richardson, J.A., Herz, J., 1999. Reeler/ Disabled-like disruption of neuronal migration in knockout mice lacking the VLDL receptor and ApoE receptor 2. Cell 97, 689±701. Walsh, C.A., Gof®net, A.M., 2000. Potential mechanisms of mutations that affect neuronal migration in man and mouse. Curr. Opin. Genet. Dev. 10, 270±274. Ware, M.L., Fox, J.W., Gonzalez, J.L., Davis, N.M., Lambert de Rouvroit, C., Russo, C.J., Chua Jr, S.C., Gof®net, A.M., Walsh, C.A., 1997. Aberrant splicing of a mouse disabled homolog, mdab1, in the scrambler mouse. Neuron 19, 239±249. Wechsler-Reya, R.J., Scott, M.P., 1999. Control of neuronal precursor proliferation in the cerebellum by Sonic Hedgehog. Neuron 22, 103± 114. Wichterle, H., Garcia-Verdugo, J.M., Alvarez-Buylla, A., 1997. Direct evidence for homotypic, glia-independent neuronal migration. Neuron 18, 779±791. Yee, K.T., Simon, H.H., Tessier-Lavigne, M., O'Leary, D.M., 1999. Extension of long leading processes and neuronal migration in the mammalian brain directed by the chemoattractant netrin-1. Neuron 24, 607± 622.