Molecular and Cellular Neuroscience xxx (2009) xxx–xxx

YMCNE-02334; No. of pages: 11; 4C: 4, 7, 9

Contents lists available at ScienceDirect

Molecular and Cellular Neuroscience

j ourna l homepage: www.e lsev ie r.com/ locate /ymcne

ARTICLE IN PRESS

Differential roles of Netrin-1 and its receptor DCC in inferior olivary neuron migration

Séverine Marcos a,b, Stéphanie Backer a,b, Frédéric Causeret a,b,c,Marc Tessier-Lavigne d, Evelyne Bloch-Gallego a,b,⁎a Institut Cochin, Université Paris Descartes, CNRS (UMR 8104), Paris, Franceb Inserm, U567, Paris, Francec Institut Jacques Monod, Paris, Franced Division of Research, Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, USA

⁎ Corresponding author. Institut Cochin, Université8104), Paris, France. Fax: +33 1 44 41 24 21.

E-mail address: evelyne.bloch-gallego@inserm.fr (E.

1044-7431/$ – see front matter © 2009 Elsevier Inc. Aldoi:10.1016/j.mcn.2009.04.008

Please cite this article as: Marcos, S., et al.,Neurosci. (2009), doi:10.1016/j.mcn.2009.0

a b s t r a c t

a r t i c l e i n f o

Article history:Received 12 November 2008Revised 10 April 2009Accepted 23 April 2009Available online xxxx

Keywords:GuidancePathfindingFloor plateMicePrecerebellar neuronsOlivary neuronsRetrograde tracing

Netrin-1 was previously shown to be required for the tangential migration and survival of neurons that willform the inferior olivary nucleus (ION). Surprisingly, the compared analysis of mutant mice lacking eitherNetrin-1 or its major receptor DCC reveals striking phenotypic differences besides common features. Althoughectopic stops of ION cell bodies occur in the same positions along the migratory stream in bothmutants, the IONneurons' number is not affected by the lack of DCC whereas it is reduced in Netrin-1 mutant mice. Thus, celldeath results from the absence of Netrin-1 and not from neuron mis-routing, arguing for a role of Netrin-1 as asurvival factor in vivo. The secretion of Netrin-1 by the floor plate (FP) is strictly required –whereas DCC is not –to avoid ION axons' repulsion by the FP and allows them to cross it. Leading processes of neurons of other caudalprecerebellar nuclei (PCN) cannot cross the FP in either mutant mouse, suggesting differential sensitivity ormechanism of action of Netrin-1 for leading processes of ION and other PCN neurons.

© 2009 Elsevier Inc. All rights reserved.

Introduction

During the development of the caudal hindbrain, neurons that willform PCN migrate from their birthplace in the germinative neuroe-pithelium (rhombic lip) to the position they occupy in the maturebrain (Rakic, 1990). PCN neurons provide an interesting comparativemodel of neurophilic tangential migration, including the neurons thatwill form the external cuneatus nucleus (ECN), the lateral reticularnucleus (LRN) and the inferior olivary nucleus (ION) (Altman andBayer, 1987a,b). They differ by distinct migratory routes. ION neuronsmigrate through the submarginal stream (Altman and Bayer, 1987a,b;Bourrat and Sotelo, 1988, 1991), whereas neurons that will form theLRN and the ECN migrate through the marginal stream. All PCNneurons first extend a leading process that initiates the circumfer-ential migratory route and reaches the floor plate (FP) (Bourrat andSotelo, 1988), then their nuclei translocate inside the leading process(Gilthorpe et al., 2002; Causeret et al., 2002). Although the leadingprocesses of all caudal PCN neurons cross the FP, the behaviour of theirsoma differs. The cell bodies of ECN neurons and of LRN neurons crossthe FP and continue their translocation until reaching their properdestination in the contralateral rhombencephalon. In contrast, cellbodies of ION neurons stop before crossing the FP and aggregate to

Paris Descartes, CNRS (UMR

Bloch-Gallego).

l rights reserved.

Differential roles of Netrin-14.008

form club-shaped masses (Altman and Bayer, 1987a,b; Bourrat andSotelo, 1990). Thus, PCN cell bodies do not respond identically tosignals from the FP. Consequently, IONs develop a contralateralprojection to the cerebellumwhereas LRN/ECN develop an ipsilateralprojection.

The FP acts as an intermediate target and is a source of both contactand diffusible molecules that guide axons and cell body migration. Inprevious studies, we showed that Netrin-1, the first chemotropicfactor that has been discovered to be secreted by the FP, was involvedin vivo and in vitro in the migration of ION neurons (Bloch-Gallego etal., 1999; Causeret et al., 2002). Netrin-1 also attracts other PCNneurons in vitro, such as LRN or ECN neurons (Alcantara et al., 2000)and directs the tangential migration of pontine neurons (Yee et al.,1999). In the absence of Netrin-1, ION cell bodies are locatedectopically along their migratory pathway and ION axons mainlyproject ipsilaterally to their cerebellar target instead of contralaterallyas in wild type (WT) mice, which could be due to an abnormalcrossing of ION cell bodies or to the inability for axons to cross the FP(Bloch-Gallego et al., 1999). Additional experimental evidence wasrequired to support one of these two hypotheses. Although thereceptor Deleted in Colorectal Cancer (DCC)was implicated in pontineneurons responses to Netrin-1 gradient (Yee et al., 1999), theinvolvement of DCC in Netrin-1 responses for ION and LRN/ECNneurons had not been investigated so far.

In addition, in Netrin-1 mutant mice, the number of ION neuronswas observed to be reduced at birth (Bloch-Gallego et al., 1999) and

and its receptor DCC in inferior olivary neuron migration, Mol. Cell.

Fig. 1. DCC is required for ION neurons' attractive response to Netrin-1 in vitro. E11.5 rhombic lip explants containing ION neurons are faced with Netrin-1-secreting cells (net1),cultured for 3 days and then analyzed for axon outgrowth and nucleokinesis through Tuj1 and DAPI staining respectively. (A, B) Axon outgrowth fromWT and DCC−/− rhombic lipexplants. InWTexplants (n=18) (A), Netrin-1 induces oriented axon outgrowth in the proximal quadrant (p), but not in the distal one (d), whereas when explants fromDCCmutantembryos (n=29) (B) are used, axon outgrowth is no more promoted by Netrin-1-secreting cells. (C, D) Nucleokinesis in WT and DCC−/− rhombic lip explants. Netrin-1-secretingcells induce nuclei migration from WT explants (n=18) (C), but not from DCC−/− explants (n=15). (E) Bar graphs showing total axon outgrowth toward (proximal) and away(distal) from control cells or Netrin-1-secreting cells were obtained upon threshold quantifications of Tuj1 staining on WT, DCC–/– or anti-DCC blocking antibodies treated rhombiclip explants. (F) Bar graphs showing nucleokinesis toward (proximal) and away (distal) from control cells or Netrin-1-secreting cells were obtained upon threshold quantifications ofDAPI staining on WT, DCC–/– or anti-DCC blocking antibody treated rhombic lip explants. Quantification were performed using the Student's t-test, and ⁎ is for pb0.01.

2 S. Marcos et al. / Molecular and Cellular Neuroscience xxx (2009) xxx–xxx

ARTICLE IN PRESS

we had proposed that it might play the role of a survival factor via itsdependence receptors, members of the DCC and Unc5 families (Llambiet al., 2001).

Through comparative analysis of Netrin-1 and DCC mutant mice,we report observations that provide cues for understanding thedistinct involvement of Netrin-1 and DCC in ION survival and forvarious caudal PCN leading processes in crossing the FP during theirtangential migration.

Results

We have further investigated the role of Netrin-1 in neuronal PCNmigration, i.e. the positioning of cell bodies, axon outgrowth andguidance, and the development of the cerebellar projection. To thisend, we have compared the phenotypes of DCC and Netrin-1 mutantmice. Here we detail the differences and similarities that we observedduring PCN migratory process in ligand and receptor mutant mice.

DCC is required for ION axon attraction, promotion of axon outgrowthand permissivity of nuclear migration in response to Netrin-1 in vitro

To assess the role of DCC in axon outgrowth and neuronalmigration(nucleokinesis) for ION neurons in response to Netrin-1, in vitroexperiments were carried out. Rhombic lip explants from E11.5embryos were used since they are enriched in ION neurons due tothe peak of ION neurons' birthdate at this specific stage compared toother PCN populations (Causeret et al., 2002, 2004). They were takenfrom WT or DCC mutant embryos and were faced with either mock orNetrin-1 secreting 293-EBNA cells in a collagen matrix. ION axonoutgrowth and nucleokinesis were analyzed in these coculture assaysafter 3 days in culture. The explants were immunostained with class IIIβ-tubulin (Tuj1) antibodies and the cell nuclei were visualized withDAPI. In this type of assay, we have previously shown that a Netrin-1source promoted extensive and oriented axon outgrowth, as well asnucleokinesis from WT explants (pb0.01) (Causeret et al., 2002;

Please cite this article as: Marcos, S., et al., Differential roles of Netrin-1Neurosci. (2009), doi:10.1016/j.mcn.2009.04.008

Figs.1A, C and quantification in E and F)whereas only a slight degree ofaxon outgrowth and no cell body migration occurred when WTexplants were presented with control cells (not illustrated, quantifica-tion in Figs. 1E and F). The coculture assays were next performed withE11.5 rhombic lip explants taken from DCC mutant embryos (n=29).In this case, neuronal processes were much shorter, axon outgrowthwas greatly decreased and followed a random pattern of growth(proximal/distal ratio of 1.2±0.5; Fig.1B and quantification in Fig.1E).In addition, very few cell bodies left the explants (Fig. 1D and quan-tification in Fig. 1F). Similar results were observed when we addedblocking anti-DCC antibodies toWTexplants (n=13): axon outgrowthlost its directionality (proximal/distal of 1.1±0.3 — which is compar-able with the neuronal processes in WT explants faced with controlcells (quantification in Fig. 1E)) and nucleokinesis was impaired(quantification in Fig. 1F). As a control, we used an antibody directedagainst the intracellular domain of DCC: the addition of this antibodyhad no effect on axonal oriented outgrowth and nucleokinesis (notshown). These results indicate that DCC is required for the chemoat-tractive effect of Netrin-1 on IONaxons and cell bodies in vitro, and thatthe presence of both Netrin-1 and DCC promotes axon outgrowth.

Netrin-1 masks a repulsive activity of the floor plate

We next examined responses to FP explants, which provide asource of Netrin-1 but also of other potential chemotropic cues. Todetermine the role of Netrin-1 in responses to FP explants, we co-cultured rhombic lip explants taken from E11.5 WT embryos with FPexplants from either WT or Netrin-1 mutant embryos. When rhombiclip explants were confronted with FP explants dissected from E11.5WT embryos, both oriented outgrowth (Fig. 2A) and nucleokinesisoccurred (not shown) toward the FP explants (n=22; pb0.01, P/Dratio=3.60±0.80 for axon outgrowth) (quantification in Fig. 2E).However, both eventswere affectedwhen explantswere facedwith FPfrom Netrin-1 mutant embryos. Surprisingly, although no nuclearmigration occurred in the absence of Netrin-1 (not shown),

and its receptor DCC in inferior olivary neuron migration, Mol. Cell.

Fig. 2. Repulsion of ION axons by the floor plate of Netrin-1mutant mice. E11.5 rhombic lip explants containing ION neurons are faced with FP explants fromWT or Netrin-1 mutantembryos, cultured for 3 days, with or without blocking anti-DCC antibodies and analyzed for axon outgrowth through Tuj1 staining. FP explants(FP) from WT embryos (n=22)induce oriented axon outgrowth (A), whereas when FP explants from Netrin-1mutant embryos are used (n=12), axons are repelled by the FP and grow on the distal quadrant (B).When blocking anti-DCC antibodies are added to the culture medium, ION axon outgrowth, although very weak, is still oriented towardWT FP explants (n=22) (C). On the contrary,when Netrin-1 mutant FP explants are used (n=12), ION axons are still repulsed by the FP (D). (E) Bar graphs showing total axon outgrowth toward (proximal) and away (distal)from WT or Netrin−/− FP explants, with or without blocking anti-DCC antibodies, were obtained after Tuj1 staining. ⁎ indicates pb0.01.

3S. Marcos et al. / Molecular and Cellular Neuroscience xxx (2009) xxx–xxx

ARTICLE IN PRESS

confirming its permissive effects on nucleokinesis (Bloch-Gallego etal., 1999), axons were significantly repelled from the FP explants in theabsence of Netrin-1 (n=12; pb0.01, P/D ratio=0.25±0.05; Fig. 2Band quantification in Fig. 2E). These results confirm the attractive roleof Netrin-1 in the FP on axon outgrowth and cell body migration, butalso the ability of Netrin-1 to mask a repulsive activity of FP cells onION axons.

Netrin-1 protects ION axons from repulsion by the floor plateindependently of DCC

We next applied a blocking anti-DCC antibody in the culture me-dium of WT explants faced with explants of either WT or Netrin-1−/−

FP. As a control, normal mouse IgG (anti-GST mAb, clone sc-138,

Please cite this article as: Marcos, S., et al., Differential roles of Netrin-1Neurosci. (2009), doi:10.1016/j.mcn.2009.04.008

Santa Cruz Biotechnology) was added at the same concentration. Noeffect was observed. Despite the reduced axon outgrowth reportedabove after DCC blockade, a slight attraction toward WT FP explantswas still observed (n=22; P/D ratio=2.12±0.32). Conversely,when DCC function was blocked in the presence of Netrin-1−/− FP,axon outgrowth still strongly decreased but was still mainly orientedon the distal side of rhombic lip explants, as with Netrin-1−/− FPexplants (n=12; P/D ratio=0.58±0.13). The repulsive effects ofthe FP would result from cues that act independently of both DCCand Netrin-1. Since anti-DCC did not completely block the attractiveeffect of WT FP, nor the repulsive effect of Netrin-1−/− FP, weconclude that Netrin-1's effects can be mediated by a distinctreceptor, although we cannot exclude that the antibody did not blockDCC function completely.

and its receptor DCC in inferior olivary neuron migration, Mol. Cell.

4 S. Marcos et al. / Molecular and Cellular Neuroscience xxx (2009) xxx–xxx

ARTICLE IN PRESS

Please cite this article as: Marcos, S., et al., Differential roles of Netrin-1 and its receptor DCC in inferior olivary neuron migration, Mol. Cell.Neurosci. (2009), doi:10.1016/j.mcn.2009.04.008

5S. Marcos et al. / Molecular and Cellular Neuroscience xxx (2009) xxx–xxx

ARTICLE IN PRESS

Stereotyped ectopic stops along the migratory stream of ION neurons inboth Netrin-1 and DCC mutant newborn mice

In order to further characterize the role of the DCC receptor in vivo,we compared the migration of ION and LRN/ECN neurons in Netrin-1and DCC mutant mice until birth. At E13.5, in WT embryos, IONneurons express Brn-3.b and form club-shaped masses on both sidesof the FP (Fig. 3A; Bloch-Gallego et al., 1999). In both Netrin-1 and DCCmutant embryos, ION neurons maintained Brn-3.b expression, butwere still located along the dorsal two thirds of the submarginalstream (Figs. 3B and C respectively), which could be due to a delayedinitiation of the migratory process or to the inability for ION neuronsto reach the ventral half of the hindbrain in the absence of Netrin-1/DCC signaling. At birth, WT ION was composed of two symmetricalventral masses on both sides of the FP that showed a characteristiclamellated organization, as revealed after in situ hybridization (ISH)with an Nr-CAM probe (Fig. 3D). Netrin-1 mutant mice showedinferior olivary migration defects with ectopic ION neurons locatedalong their migratory stream (Bloch-Gallego et al., 1999 and Fig. 3E).Two types of phenotypes were observed in the Netrin-1 mutant mice,with various degrees of alteration of the phenotype, previouslydescribed as type I and type II (Bloch-Gallego et al., 1999). Here, wefocused on Netrin-1 mutant mice carrying the more severe type IIphenotype. At birth, the extent of the ventrally located ION masseswas severely reduced in size. In addition, they showed a residuallamellation (Fig. 3E). Interestingly, at birth, DCC mutant mice showedmigratory and positioning defects that strongly resembled those ofNetrin-1 mutant mice. Ventrally located ION neurons organized inlamellae both in DCC and Netrin-1 mutant mice, and other patches ofION neurons organized as ectopic pools along the migratory stream(Figs. 3E, F).

To further analyze the position of ectopic ION neuron clusters, wecompared their distribution along the migratory pathway both inNetrin-1 and DCC mutant mice. We established their positionsprecisely by measuring the angles between the most-dorsal point ofthe brainstem and each ION cluster (Fig. 3G) that specificallymaintained its Nr-CAM expression at ectopic positions. Interestingly,ION neuron clusters stopped at four distinct sites along the migratorypathway (Figs. 3G and H), which we refer to as: a “dorsal pool”located around 45° (located in the most-rostral part, and notillustrated on the sections in Fig. 3), a “dorso-lateral pool” around75°, a “lateral pool” between 110° and 130° and a “ventral pool” at175° which corresponded to ION neurons that had reached theirnormal medio-ventral location close to the FP. Ectopic locations of IONcell bodies were similar in the migratory stream of both Netrin-1 andDCC mutant mice. Surprisingly, the ectopic lateral cluster formed byION neurons was coincident with the position of the facial motornucleus (fMN) in both mutant mice (inserts in Figs. 3E and F), asrevealed after fMN staining by ISH using Peripherin as a probe.Although the fMN migrates early on from rhombomere 4 torhombomere 6 (Studer, 2001), and the ION neurons from pseudo-rhombomeres 8–10 (Cambronero and Puelles, 2000) in WT embryos,we previously reported that the remaining ION neurons were

Fig. 3. In vivo analyses of ION neuronal migration in Netrin-1 and DCC mutant mice. (A–C) Attoward the ventral midline. InWTembryos, most ION neurons have reached their ventro-medNetrin-1 (B) and DCC (C) mutant embryos, they are still in the dorsal half of the hindbrain. Arvisualized through Nr-Cam expression. In WT mice, ION shows its characteristic lamellated salthough they seemmore affected in Netrin-1 (type II) than in DCCmutant mice (C). Ectopicaaxis of the hindbrain in both mutant mice. Localization of fMN on alternated sections is showION neurons in Netrin-1 and DCCmutant mice at birth. (G) shows a schematic representationand shapes of the ION clusters, 0° indicating a dorsally located cluster and 180° a ventrallyclusters along the dorso-ventral axis of the hindbrain in WT (pink), in type I (dark blue) andhalves of hindbrains were counted. ION clusters form 4 stereotyped pools whose shapes andorsal pool (green) around 45°, a dorso-lateral pool (light green) at 75°, and a bigger lateral pventrally (orange). (I) The total number of neurons was counted in type I and type II Netrin-1mdistribution in the 4 pools described above is represented in the bar graph. ⁎ indicates pb0

Please cite this article as: Marcos, S., et al., Differential roles of Netrin-1Neurosci. (2009), doi:10.1016/j.mcn.2009.04.008

rostralized in Netrin-1 mutant mice and located in the vicinity ofthe fMN (Bloch-Gallego et al., 1999). This similarity of abnormalmigratory behaviours in Netrin-1 and DCC mutant mice suggests thatDCC acts as a major receptor in directing the migration of ION neuronsin response to Netrin-1.

The number of ION neurons is not affected in DCC mutant micecompared with the dwindled population of neurons observed inNetrin-1 mutant mice at birth

Previous data showed that the total number of ION neurons in Ne-trin-1 mutant newborn mice was 60% of that in WT mice (Bloch-Gallego et al., 1999). Later on, we had correlated this decreased IONneuron number at birth with cell death that was observed upon TUNELstaining at E12.5 in DCC and Unc5H ventricular and subventricularexpressing zones in Netrin-1 mutant mice (Llambi et al., 2001). In thepresent work, we counted the total number of ION neurons using Nr-Cam as a specific marker for ION neurons at P0 (Backer et al., 2002)both in DCC and in Netrin-1 mutant mice. We found that 56.6% of Nr-Cam positive cells were maintained in type IINetrin-1mutant mice and76.2% in type I mutant mice compared to WT mice (pb0.01 andpb0.001 respectively using the Student's t-test) whereas 89.8% of IONneurons survive until birth in DCC mutant mice (p=0.31; table in Fig.3I), which confirms that Netrin-1 acts as a survival factor for IONneurons whereas the presence of DCC is not required for Netrin-1 toexert its survival effects. This observation is also in accordance with apossible role of DCC as a dependence receptor, since in this model theabsence of Netrin-1 often leads to apoptosis when DCC is expressed atthe cell surface but not when it is absent.

This analysis also allowed us to establish that the viability of theremaining ION neurons was not dependent on the position of the IONneurons pool since the ratios of ION neurons in the four pools wassimilar in Netrin-1 and DCCmutant mice (Fig. 3I). Thus, the decreasedsurvival observed in Netrin-1 mutant mice was not a function ofpreferential and distinct ectopic stop positions of the neurons, andboth ectopic and ventro-medially located ION neurons are alsosimilarly affected.

Migration defects of LRN in the caudal hindbrain

Since both ION and LRN neurons are generated around the sametime and follow a tangential migration that is known to involveNetrin-1 (Bloch-Gallego et al., 1999; Alcantara et al., 2000), weanalyzed LRN neuron migration in both Netrin-1 and DCC mutantmice. The migratory pathway of LRN neurons was monitored throughTag1 expression (Backer et al., 2002). At E14.5, most WT LRN cellbodies had migrated dorso-ventrally and were crossing the midline,being mainly located in the ventral half of the hindbrain, crossing theventral midline (Fig. 4A), whereas in Netrin-1 and DCC mutantembryos, LRN cell bodies had only migrated over the half path, andmainly remained in the most-dorsal part of the hindbrain (Figs. 4Band C). Thus, the initiation of the migration of post-mitotic LRN

E13.5, Brn-3b transcripts allow the visualization of ION neurons during their migrationial location and begin to form a club-shapedmass on each side of the FP (A), whereas in

rows indicates the ventral limit reached by ION neurons. (D–F) At birth, ION neurons arehape near the FP (A). In both mutant mice, the ventral ION masses are reduced in size,lly located and highly stereotypic ION neuron clusters are found along the dorso-ventraln in inserts in B and C using Peripherin expression. (G–I) Dorso-ventral distribution ofof angles formed between the dorsal point of the midline and the center of ION clusterslocated one. (H) Graph showing the measured angles representing the location of IONtype II (light blue) Netrin-1mutant mice, as well as in DCC (purple) mutant mice. Bothd dorso-ventral distribution is shown in (G). Ectopic ION clusters form 3 pools: a smallool (light orange) between 110° and 135°. Normally located ION neurons are representedutant mice and in DCCmutant mice and compared to theWTone (table). Their relative

.01 and ⁎⁎ indicates pb0.001.

and its receptor DCC in inferior olivary neuron migration, Mol. Cell.

Fig. 4. In vivo analyses of LRN/ECN neuronal migration in Netrin-1 and DCC mutant mice. LRN/ECN cell bodies are visualized through Tag1 expression at E14.5 and P0 in WT andNetrin-1 and DCC mutant mice. (A–C) In WT embryos, at E14.5, cell bodies of neurons that will form the LRN and ECN are found in the ventral half of the hindbrain and most ofthem are migrating across the FP (A), whereas in Netrin-1 (B) and DCC (C) mutant embryos, they have barely reached the ventral half of the hindbrain by this time, most of them stillbeing located very dorsally. Arrows in A–C indicates the ventral limit reached by LRN/ECN neurons. (D–F) At birth, in WT mice, LRN form a lateral mass in the ventral half of thehindbrain (D). In Netrin-1 (E) and DCC (E) mutant mice, only a small cluster of Tag1 expressing cells is found at this location (arrow heads) and more rostrally than in WTmice. Mostof Tag1 positive neurons are located in the most-dorsal part of the hindbrain (arrows), under the cerebellum. Inserts in E and F show Pax6 expression by pontine neurons onalternated sections. The corresponding regions of the hindbrain are surrounded by dashed lines on Tag1 ISH.

6 S. Marcos et al. / Molecular and Cellular Neuroscience xxx (2009) xxx–xxx

ARTICLE IN PRESS

neurons, similarly to ION neuron migration, is affected in both Netrin-1 and DCC mutant embryos at E14.5.

At birth, the WT LRN was located in the caudal hindbrain, laterallyto the most-caudal MAO (Medial Accessory Olive) of the ION (Fig.4D). In Netrin-1 and DCC mutant mice, only a small cluster of Tag1expressing cells could be observed in the ventro-lateral presumptiveLRN domain (Figs. 4E and F). The expression of Peripherin onalternating sections (Figs. 3E, F), showed that the LRN cluster islocated close to the fMN, in a position similar to the ectopic lateralcluster of ION neurons. Indeed, most Tag1 positive cells were locateddorsally in the hindbrain (Figs. 4E and F). Since neurons that form thepontine nucleus in the rostral hindbrain also express Tag1, wedistinguished them from the LRN/ECN neurons through theirexpression of Pax6 (Yee et al., 1999; Engelkamp et al., 1999; seeinserts in Figs. 4E and F). This analysis indicated that Tag1 positiveneurons belong to both LRN and pontine nuclei that are intermingledin the alar part of the hindbrains in both Netrin-1 and DCC mutantmice at birth. Thus, the formation of LRN is similarly affected in theabsence of either Netrin-1 or DCC.

Taken together, these results suggest that Netrin-1 is required alsofor LRN neurons migration and that DCC acts as a major receptor indirecting the migration of both ION and LRN cell bodies in response toNetrin-1.

Inability of LRN axons to reach the floor plate in the absence of DCCor Netrin-1

To visualize growing LRN/ECN axons, we performed in totoimmunolabeling for Robo3/Rig1 using E13.5 hindbrains (Marillat etal., 2004). In WT E13.5 embryos, LRN/ECN Rig-1 immunolabeled

Please cite this article as: Marcos, S., et al., Differential roles of Netrin-1Neurosci. (2009), doi:10.1016/j.mcn.2009.04.008

axons extended from the rhombic lips to the FP, forming a largecommissure oriented perpendicular to the FP (Figs. 5A and D). Incontrast, no axonal commissure was visible upon Rig1 immunolabel-ing in Netrin-1 (Figs. 5B, E) and DCC (Figs. 5C, F) mutant hindbrains.The ventral part of the hindbrains was devoid of staining (Figs. 5B, E, Cand F). Indeed, most Rig1 positive fibers were observed on the dorsalhalf of Netrin-1 and DCC mutant hindbrains, growing longitudinallyfrom the rhombic lips to the ipsilateral cerebellum without reachingthe FP, and parallel to the latter, as shown in a lateral view (Figs. 5E, Fat low and high magnification in the lower part of the figure).

Thus, in the absence of Netrin-1/DCC signaling, LRN/ECN leadingprocesses were no longer attracted toward the FP and instead directlygrew toward the ipsilateral cerebellum without previously forming aventral commissure, responding to guidance cues from their ultimatetarget, the cerebellum.

Some ION axons can cross the floor plate in DCC but not Netrin-1mutant mice

We next analyzed axonal PCN projections from E14.5 to birththrough retrograde tracing after unilateral DiI injection in thecerebellum both in Netrin-1 and DCC mutant mice. At E14.5, E16.5and P0, ION neurons were labeled contralaterally to the injection sitein WT hindbrains (Figs. 6A, D and G). Labeling was more intense atE16.5 than at E14.5, consistent with the time course for ION neurons toreach their ventral position and for axons to reach the cerebellum. Theventral DiI labeling in Netrin-1mutant embryos was mainly ipsilateralalthough a slight contralateral labeling could also be observed untilbirth (Figs. 6B, E and H; Bloch-Gallego et al., 1999). In DCC mutantmice, retrograde labeling was observed on both sides of the FP at

and its receptor DCC in inferior olivary neuron migration, Mol. Cell.

Fig. 5. LRN/ECN axon pathways in Netrin-1 and DCC mutant embryos. LRN/ECN axons pathways were revealed by immunolabeling with an anti-Rig1 antibody on E13.5 hindbrainsfrom WT, Netrin-1 and DCC mutant embryos. (A–C) Ventral views of hindbrains show that LRN/ECN axons are crossing the ventral midline, perpendicularly in WT embryos (A),whereas in both Netrin-1 (B) and DCC (C) mutant embryos, no Rig1 staining is observed in the ventral part of the hindbrain. White rectangles indicates the level at which a highermagnification picture has been taken in D–F. (D–F) In lateral views, LRN/ECN axons grow straight and tangentially in the marginal migratory stream from the rhombic lip (right sideof the picture) toward the ventral midline (left) (D). In Netrin-1 (E) and DCC (F) mutant embryos, LRN/ECN axons are parallel to the ventral midline and grow toward the ipsilateralcerebellum (top of the picture) without entering the ventral half of the hindbrain. Higher magnifications of LRN/ECN axon trajectories are shown below.

7S. Marcos et al. / Molecular and Cellular Neuroscience xxx (2009) xxx–xxx

ARTICLE IN PRESS

E14.5. At E16.5, the extent of the contralateral labeling was larger thanthe ipsilateral one but the degree of ispsilateral labelled ION neuronsthen increased until birth such that ipsi- and contralateral projectionsare equally represented in newborn mice upon tracing (Fig. 6I). Theseobservations suggest that ION axons that develop the remainingcontralateral projection in DCC mutant mice reach their cerebellartarget faster than axons that do not succeed in crossing the FP andfurther develop the abnormal ipsilateral projection between E16.5 andbirth. Thus, half the olivo-cerebellar projections remain contralateral,and the other half, although initially delayed, is identical to that in theNetrin-1 mutant embryos, where the vast majority of ION neuronsdeveloped an ipsilateral projection (Fig. 6H).

Two hypotheses could explain the development of ipsilateralprojections: (i) an abnormal crossing through the FP by ION cellbodies in the absence of Netrin-1 signaling or (ii) an inability of IONgrowth cones to cross the FP. Interestingly, we found that the olivarycommissure formed by olivary fibers crossing the FP was absent (Figs.6B, E) in Netrin-1 mutant mice, or much thinner in DCC mutant mice(Figs. 6C, F) than in WT brainstems from E14.5 to birth (Figs. 6A, D),indicating that ION leading processes are prevented from crossing theFP in the absence of Netrin-1 signaling. In addition, we never observedany cell bodies crossing the FP at any of the stages analyzed in eitherNetrin-1 or DCC mutant embryos, neither upon retrograde tracing norby ISH with specific markers.

In Netrin-1 mutant mice, the ION projection is almost exclusivelyipsilateral at birth. In DCCmutant mice, all the ectopically located IONneurons (about 55% of the neurons) (see Fig. 3I) also projectipsilaterally since they are labeled on the ipsilateral side of DiIinjection (see white arrow in Fig. 6I). Finally, among ventro-mediallylocated ION neurons (accounting for 45% of them)— half of them alsodevelop ipsilateral projections. Thus, the global rate of ipsilateralprojecting ION neurons is about 80% in DCC and over 90% in Netrin-1mutant mice. Therefore, DCC is involved in Netrin-1 responsesmediating the ability of axons to cross the FP. Taken together, theseobservations suggest that in the absence of Netrin-1/DCC signaling,

Please cite this article as: Marcos, S., et al., Differential roles of Netrin-1Neurosci. (2009), doi:10.1016/j.mcn.2009.04.008

ION axons lose their ability to cross the FP; this inability is partial inDCCmutant mice and complete in Netrin-1mutant mice. We concludethat the ipsilateral olivary projections observed at P0 in both Netrin-1and DCCmutant mice are due to axon guidance defects, with Netrin-1being strictly required for ION leading processes to efficiently cross theFP while DCC is not.

DiI retrograde labeling was also observed laterally, at a positioncoincident with both LRN and the lateral ectopic pool of ION neuronsfor Netrin-1 and DCC mutant mice (arrows in Figs. 6G–I). Since thesize of the LRN is quite reduced (Figs. 4E and F) compared with theextent of the DiI positive mass, the retrograde DiI labeling observedlaterally could result from both LRN and lateral ectopic ION neuronalprojections.

Discussion

Netrin-1 acts as a survival factor for ION neurons in vivo

We demonstrate here that the absence of the ligand Netrin-1 leadsto a decrease by half of the ION neuron number whereas only 10% ofION neurons are lost in the absence of DCC. Netrin-1 could affectneuronal survival as a consequence of mis-positioning of neurons, asproposed for pontine neurons (Yee et al., 1999) or through receptors ofthe DCC and/or Unc5 families, inducing apoptosis in the absence oftheir ligand in vitro for spinal commissural neurons (Furne et al.,2008) or in vivo (Llambi et al., 2001). Interestingly, migratory defectsof ION neurons are identical in both Netrin-1 and DCC mutant mice,and ectopic ION neurons succeed in reaching their cerebellar target.This suggests that the loss of ION neurons in Netrin-1 mutant miceresults from an early requirement of the survival effects of Netrin-1 invivo, and not from an inability of ION to survive and reach their targetwhen mislocalized. To determine the relative importance of DCC andUnc5 as dependence receptors, it would be necessary to analyze thedegree of ION survival in Netrin-1/DCC mutant mice compared withNetrin-1/Unc-5A-D mutant mice. Other authors reported that in

and its receptor DCC in inferior olivary neuron migration, Mol. Cell.

Fig. 6. Development of olivo-cerebellar projections in Netrin-1 and DCC mutant embryos. Olivo-cerebellar is observed through unilateral DiI injections in the cerebellum andretrograde labeling of precerebellar nuclei. Inserts in A to F show higher magnifications of the olivary commissure at the ventral midline. (A–C) At E14.5, in WT embryos, both IONneurons and some LRN neurons have reached their locations and developed a contralateral and ipsilateral projection to the cerebellum respectively (A). In contrast, in Netrin-1mutant embryos, only a poor staining was observed ventrally and ipsilaterally to the injection site, while no labeling can be observed laterally (B). In DCC mutant mice (C), ventrallabeled masses of similar sizes are found ipsilaterally and contralaterally and a small commissure has formed. In both mutant mice, no cell body can be observed in the FP, neither inthe marginal nor in the submarginal migratory stream (B, C). (D–F) At E16.5, the ION has started to acquire its lamellated shape, the LRN appears bigger than at E14.5 and a dorsalmass corresponding to the ECN is visible in WT brainstems (D). In Netrin-1 mutant embryos, most ventral labeling is ipsilateral, with only a small mass being retrogradely labeledcontralaterally and an ipsilateral mass is labeled more laterally (E). In DCCmutant mice, a lateral mass is also labeled ipsilaterally to the injection site. Ventral masses are found bothipsilaterally and controlaterally although the latter seems bigger (F). In Netrin-1 mutant embryos, at both stages, the FP is almost devoid of fibers (B, E), whereas a small ventralcommissure has formed in DCCmutant embryos (C, F), compared toWT (A, D). (G–I) At birth, the ION, contralaterally, and the LRN (white arrow), ipsilaterally, show strong labeling(G). In Netrin-1 (H) and DCC (I) mutant newborn mice, ventral and lateral (white arrows) retrogradely labeled masses are smaller than in WT. Netrin-1 mutant mice only show anipsilateral ventral mass (asterisks), while DCC mutant mice show labeling on both the ipsilateral side (asterisks) and the contralateral side.

8 S. Marcos et al. / Molecular and Cellular Neuroscience xxx (2009) xxx–xxx

ARTICLE IN PRESS

Netrin-1 mutants, loss of Netrin-1 in vivo did not increase the rate ofapoptosis, neither for cervical motoneurons, nor in hindbrain, butobserved a reduced cell death in Unc-5Amutant embryos at E12.5 andE14.5 (Williams et al., 2006). Unfortunately, the absence of any reportof numbering of remaining motoneurons or ION neurons in Netrin-1mutants makes the comparison between our and their study difficult.

Netrin-1 participates in the initial oriented extension of the leadingprocess and FP crossing

In vivo, we showed that the initial migration of neurons from therhombic lip was delayed in the absence of Netrin-1 and DCC. In vitro,the FP deprived of Netrin-1 lost its attractive effects on leadingprocesses. Interestingly, in Netrin-1 mutant mice, some of the ION cellbodies still succeeded in positioning ventrally close to the FP,suggesting that their leading processes had reached it. Thus, othercues than Netrin-1 could be able to direct ION neuron migrationtoward the FP in vivo.

In the absence of Netrin-1, we had observed that ION mainlydeveloped an ipsilateral projection toward the cerebellum (Bloch-Gallego et al., 1999). The detailed analysis of the development of IONprojections revealed that the ability of axons to cross the FPwas totallyimpaired in Netrin-1mutant mice, demonstrating a strict requirementfor Netrin-1 to allow axons crossing through the FP (schematized inFig. 7). It has been recently reported that in the absence of Slit1/2, aswell as its receptors Robo1/2, ION cell bodies crossed the FP (DiMeglio et al., 2008). Conversely, although ION cell bodies still succeed

Please cite this article as: Marcos, S., et al., Differential roles of Netrin-1Neurosci. (2009), doi:10.1016/j.mcn.2009.04.008

in reaching their ventro-medial location, an inability of axons of bothION and LRN/ECN to cross the midline had also been reported in theabsence of Robo3/Rig1, another Slit receptor proposed to inhibit therepulsive effect of Slit before midline crossing (Sabatier et al., 2004;Marillat et al., 2004). In vitro, Slit is able to mask the attractive effectsof Netrin-1 (Causeret et al., 2002), and we show in the present workthat the absence of Netrin-1 unmasks a repulsive activity of the FP onaxons that is likely to be due to Slit.

Together, these results show that Netrin-1 is first involved in theattraction of PCN leading processes to the FP. Then, once PCN growthcones have reached the FP, Netrin-1 is required to allow them to crossthe ventral midline by counteracting the effects of a repulsive signal.

The ventral attractive effect of Netrin-1 on ION cell bodies requires DCCbut its role in floor plate crossing may rely on other partners

Through in vitro assays, we report that DCC is involved in theattraction of axons toward a Netrin-1 source. Our comparative in vivoanalyses of Netrin-1 and DCC mutant mice have revealed that bothmutant mice exhibit very similar defects of PCN neuronal migration.However, the analysis of the developing projections across the FPtoward the cerebellum in both Netrin-1 and DCC mutant micedemonstrates that a contingent of ION axonal fibers still succeeds incrossing the FP in the absence of DCC, suggesting that anotherreceptor could act with Netrin-1 to allow a partial axon crossing orthat Netrin-1 alone is sufficient to protect axons from FP repulsion,possibly through a direct binding to Slit (Brose et al., 1999). Our

and its receptor DCC in inferior olivary neuron migration, Mol. Cell.

Fig. 7. Schematic representation of the pathways of the leading processes and of the final locations of cell bodies of ION and LRN in WT and Netrin-1 mutant embryos. (A) In WTembryos, ION leading processes grow from the rhombic lip toward the FP, and cross it to reach the contralateral cerebellum, while ION cell bodies stop their migration before crossingthe FP. In Netrin-1 mutant embryos, only some of the leading processes can reach the FP but none can cross it. They are directly attracted toward the ipsilateral cerebellum. As aconsequence, cell bodies stop their migration at ectopic locations all along the dorso-ventral axis of the hindbrain. (B) InWTembryos, LRN leading processes and cell bodies cross theFP. In Netrin-1mutant embryos, leading processes are unable to reach the vicinity of the FP and are rapidly reoriented toward the ipsilateral cerebellum. Only few cell bodies succeedin leaving the rhombic lip and stop in the ipsilateral side, laterally, and most of them stay in the rhombic lip, dorsally. FP: floor plate; RL: rhombic lip; CP: cerebellar plate.

9S. Marcos et al. / Molecular and Cellular Neuroscience xxx (2009) xxx–xxx

ARTICLE IN PRESS

previous detailed analysis of the expression of other Netrin-1receptors (Bloch-Gallego et al., 1999) had revealed that at E11.5 andE12.5, there is no overlap in the expression domains of Unc-5H2 andDCC: DCC is expressed in post-mitotic neurons and all along themigratory stream, whereas Unc-5H1, Unc-5H2, and neogenin colocalizein the alar part of the ventricular zone. At E13.5, axons of ventrallypacked ION neurons have crossed themidline through the interolivarycommissure and reached the inferior cerebellar peduncle. Neitherneogenin nor Unc-5HmRNAs are expressed in the club-shaped massesof ION neurons, whereas DCC mRNA is now expressed by IO neuronslocated close to the midline. These descriptive data do not favor otherNetrin-1 receptors to allow partial FP crossing and compensate theabsence of DCC, but rather the formation of a complex includingNetrin-1 and Slit, as well as DCC and Robo3/Rig1 that are all expressedby ION at E12.5–E13.5 (Marillat et al., 2004), that could be required forPCN neurons leading processes to cross the FP. In Xenopus, activationof the Slit receptor Roundabout (Robo1) had been proposed to silencethe attractive effect of Netrin-1 on spinal neurons (Stein and Tessier-Lavigne, 2001). Whatever the mechanism, the interaction betweenboth ligands could inhibit or modulate their respective effects duringPCN development.

Taken together, in vivo and in vitro analyses confirm that most ofNetrin-1 effects during PCN neurophilic migration, i.e. the attraction ofthe leading process, the translocation of the cell bodies, as well as cellbodies and axons ventral midline crossing, aremainly mediated by thereceptor DCC but ION axons can also cross it independently of DCC.

Recently, another receptor for Netrin-1 as been described: theDown's syndrome Cell Adhesion Molecule (DSCAM) has been shownto bind Netrin-1 and to be necessary for commissural axons fromXenopus to grow toward and across the midline (Ly et al., 2008) andfor chick commissural axon guidance (Liu et al., 2009). In Drosophila,Dscam proteins function as attractive receptors for Netrin and also actin parallel to Frazzled/DCC (Andrews et al., 2008). These recent data

Please cite this article as: Marcos, S., et al., Differential roles of Netrin-1Neurosci. (2009), doi:10.1016/j.mcn.2009.04.008

strengthen the possible interest of analyzing DSCAM expressionduring PCN development and its role as a candidate among theunexpected complexity in receptors that mediate vertebrate Netrin-1signaling and axon guidance and midline crossing.

PCN axons are prematurely attracted by the cerebellum when ventralattraction by Netrin-1 is defective

We found that ION neurons succeed in developing projectionstoward the cerebellum when located ventro-medially, but alsoectopically along the migratory stream. Similarly, from the compar-ison between axonal pathways detected by Robo3/Rig1 immunola-beling and LRN cell bodies migration revealed by ISH with Tag1, weshow that instead of growing straight toward and across the FP toreach the contralateral cerebellum, other PCN leading processes firstorient ventrally and turn back early to the ipsilateral cerebellum atvarious sites along the migratory route. Thus, the position of ectopicION cell bodies stops could correspond to the turning back points oftheir leading processes.

It has been shown that ectopic cerebellar grafts can be attractivefor LRN neurons (Taniguchi et al., 2002) and that the sensitivity of IONaxons to cerebellar cues is increased upon FP crossing (Zhu et al.,2003). In the absence of Netrin-1/DCC signaling from the FP, IONaxons would answer earlier to cerebellar cues. Although the identityof PCN attractive cerebellar cues has not yet been characterized, it hasbeen shown that the cerebellum secretes neurotrophins from E13.5,while TrkB and TrkC receptors are expressed from E11.5 in thehindbrain (Sherrard and Bower, 2002). Thus, the cerebellum could bea potent source of trophic and tropic cues, partly responsible for thepremature attraction toward the cerebellum in the absence of Netrin-1/DCC signaling. It is noteworthy that a differential sensitivity ofvarious tangential migrating neurons could be established, withpontine neurons which barely initiate they migration, being the most

and its receptor DCC in inferior olivary neuron migration, Mol. Cell.

10 S. Marcos et al. / Molecular and Cellular Neuroscience xxx (2009) xxx–xxx

ARTICLE IN PRESS

sensitive (Yee et al., 1999; Supplemental Fig. S1), followed by LRN andthen ION neurons. Yee et al. suggested that the nature of the leadingprocess – axon versus leading process – could be responsible for thevarious sensitivities to the lack of Netrin/DCC signaling (Yee et al.,1999). Alternative hypothesis could involve the birthdates of each PCNpopulation – ION neurons being generated earlier than LRN andpontine neurons – that parallel the time course of maturation of theirultimate cerebellar target, as well as their birthplace, pontine neuronsbeing generated closer to the cerebellum.

In the absence of Netrin-1, the balance between factors involved inneuron positioning during hindbrain development is disturbed. Thus,when this major ventral attractive cue is lacking, axons appear toanswer prematurely to cerebellar cues and project to the ipsilateralcerebellum.

Other hindbrain cues provide alternative stop signals for positioningectopic ION pools in the hindbrain

We have reported three stereotyped ectopic stops for migratingION neurons that were identically positioned in the absence of Netrin-1 or of DCC. Theymight reflect the presence of intermediate targets forION neurons migration in the dorso-ventral hindbrain that couldpartially compensate for the defect in Netrin-1 signaling. Indeed, thesepools seem to have stopped their migration at sites where distincthindbrain nuclei settle: the most-dorsal location could correspond toneurons that have failed initiating their migration and are unable toleave the rhombic lips, and also corresponds to the dorsal position ofthe External Cuneatus Nucleus (ECN). The two lateral clustersbelonging to the ION lateral pool, and the remaining lateral clusterof LRN neurons correspond to the location of fMN neurons, whichexpressed adhesion molecules similar to ION neurons (Backer et al.,2007). Cues responsible for these stops along the migratory pathwayremain to be characterized.

Experimental methods

Embryo processing

Mouse embryos were obtained from timed mating of Netrin-1 andDCC heterozygous mice or Swiss mice (Janvier, le Genest St-Isle,France). Producing, breeding and genotyping of the Netrin-1 hetero-zygous mice were carried out as described in Serafini et al., 1996 andDCC heterozygousmice as described in Fazeli et al., 1997. For all cultureassays, hindbrains from E11.5 or E12.5 embryos were dissected inGBSS (Sigma), supplemented with 0.5% glucose.

Collagen co-cultures

Collagen assays were performed as previously described (Causeretet al., 2002), using rhombic lip explants from E11.5 WT or DCC mouseembryos that are enriched in ION neurons. Rhombic lip explants facemock or Netrin-1-secreting EBNA cells (Kennedy et al., 1994), or FPexplants from WT or Netrin-1 mutant embryos. After 2 days in vitro(DIV) in a 5% CO2, 37 °C, 95% humidity incubator, explants were fixedfor 20 min in a 4% paraformaldehyde at 37%. A mouse anti-DCCantibody (clone AF5, Oncogene Science Inc.) was used to block DCCfunction at 5 μg/ml. A normal mouse IgG (R and D System) was addedat the same concentration for control experiments.

Immunofluorescence and antibodies

After permeabilization, cultures were processed for anti-class III β-tubulin antibody (TUJ1; 1:4000; Jackson ImmunoResearch) asdescribed in Causeret et al., 2002. DAPI (1 μg/ml, Vector) was usedto visualize nuclei.

Please cite this article as: Marcos, S., et al., Differential roles of Netrin-1Neurosci. (2009), doi:10.1016/j.mcn.2009.04.008

Quantification of axon outgrowth and nucleokinesis

Picture acquisition and processing for axon outgrowth andnucleokinesis quantification were performed as previously described(Causeret et al., 2004). Differences between calculated averages fordistal and proximal axon outgrowth and nucleokinesis were con-sidered significant when Pb0.01 (⁎ in bar graphs) and highlysignificant when Pb0.001 (⁎⁎ in bar graphs) using a Student's t-test.

In situ hybridization

ISH was carried out according to Causeret et al., 2004. Brn-3.b, Tag1and Nr-Cam probes were used and have been described previously(Backer et al., 2002; Bloch-Gallego et al., 1999). Peripherin probe wasused as described in Backer et al., 2007. We also used a mouse Pax6probe, kindly provided by S. Saule from the Pasteur Institute, Lille.

ION neurons counting and dorso-ventral ION clusters localization

ION neurons were counted at P0 using acquisitions overlays ofDAPI staining and ISH of Nr-Cam probe and the angle between thedorsal part of themedullamidline and the center of eachNr-Campositivecluster was measured at the midpoint of the midline (see Fig. 3A).

In toto immunohistochemistry

In toto Rig-1 labeling was performed as described in Marillat et al.,2004, using a rabbit polyclonal anti-Rig1 antibody (a kind gift fromDr. Murakami, 1:2000) overnight at 4 °C in PBS, 0.2% gelatin, 0.25%Triton X-100.

Retrograde labeling of precerebellar neurons

Injections of DiI carbocyanine (D282, Molecular Probes) wereperformed both in Netrin-1 and DCC mutant mice as described inBloch-Gallego et al. (1999), between E13.5 and birth.

Acknowledgments

We are grateful to Drs. Sonia Garel and Matias Hidalgo-Sanchez fortheir critical reading of the manuscript. Annie Goldman is thanked forproof-reading in English. This work was initiated in Dr. ConstantinoSotelo's lab (INSERM U. 106, Paris) that we acknowledge for fruitfuldiscussions. It was partially funded by the Association pour laRecherche sur le Cancer (ARC 3697), the Association Française pourla recherche contre les Myopathies (AFM) and INSERM.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.mcn.2009.04.008.

References

Alcantara, S., Ruiz, M., De Castro, F., Soriano, E., Sotelo, C., 2000. Netrin 1 acts as anattractive or as a repulsive cue for distinct migrating neurons during thedevelopment of the cerebellar system. Development 127, 1359–1372.

Altman, J., Bayer, S.A., 1987a. Development of the precerebellar nuclei in the rat: II. Theintramural olivary migratory stream and the neurogenetic organization of theinferior olive. J. Comp. Neurol. 257, 490–512.

Altman, J., Bayer, S.A., 1987b. Development of the precerebellar nuclei in the rat: III. Theposterior precerebellar extramural migratory stream and the lateral reticular andexternal cuneate nuclei. J. Comp. Neurol. 257, 513–528.

Andrews, G.L., Tanglao, S., Farmer, W.T., Morin, S., Brotman, S., Berberoglu, M.A., Price,H., Fernandez, G.C., Mastick, G.S., Charron, F., Kidd, T., 2008. Dscam guidesembryonic axons by Netrin-dependent and -independent functions. Development135 (23), 3839–3848.

Backer, S., Sakurai, T., Grumet, M., Sotelo, C., Bloch-Gallego, E., 2002. Nr-CAM and TAG-1are expressed in distinct populations of developing precerebellar and cerebellarneurons. Neuroscience 113, 743–748.

and its receptor DCC in inferior olivary neuron migration, Mol. Cell.

11S. Marcos et al. / Molecular and Cellular Neuroscience xxx (2009) xxx–xxx

ARTICLE IN PRESS

Backer, S., Hidalgo-Sanchez, M., Offner, N., Portales-Casamar, E., Debant, A., Fort, P.,Gauthier-Rouviere, C., Bloch-Gallego, E., 2007. Trio controls the mature organiza-tion of neuronal clusters in the hindbrain. J. Neurosci. 27, 10323–10332.

Bloch-Gallego, E., Ezan, F., Tessier-Lavigne, M., Sotelo, C., 1999. Floor plate and netrin-1are involved in themigration and survival of inferior olivary neurons. J. Neurosci.19,4407–4420.

Bourrat, F., Sotelo, C., 1988. Migratory pathways and neuritic differentiation of inferiorolivary neurons in the rat embryo. Axonal tracing study using the in vitro slabtechnique. Brain Res. 467, 19–37.

Bourrat, F., Sotelo, C., 1990. Early development of the rat precerebellar system: migratoryroutes, selective aggregation and neuritic differentiation of the inferior olive andlateral reticular nucleus neurons. An overview. Arch. Ital. Biol. 128, 151–170.

Bourrat, F., Sotelo, C., 1991. Relationships between neuronal birthdates and cytoarch-itecture in the rat inferior olivary complex. J. Comp. Neurol. 313, 509–521.

Brose, K., Bland, K.S., Wang, K.H., Arnott, D., Henzel, W., Goodman, C.S., Tessier-Lavigne,M., Kidd, T., 1999. Slit proteins bind Robo receptors and have an evolutionarilyconserved role in repulsive axon guidance. Cell 96, 795–806.

Cambronero, F., Puelles, L., 2000. Rostrocaudal nuclear relationships in the avianmedulla oblongata: a fate map with quail chick chimeras. J. Comp. Neurol. 427,522–545.

Causeret, F., Danne, F., Ezan, F., Sotelo, C., Bloch-Gallego, E., 2002. Slit antagonizesnetrin-1 attractive effects during the migration of inferior olivary neurons. Dev.Biol. 246, 429–440.

Causeret, F., Hidalgo-Sanchez, M., Fort, P., Backer, S., Popoff, M.R., Gauthier-Rouviere,C., Bloch-Gallego, E., 2004. Distinct roles of Rac1/Cdc42 and Rho/Rock foraxon outgrowth and nucleokinesis of precerebellar neurons toward netrin 1.Development 131, 2841–2852.

Di Meglio, T., Nguyen-Ba-Charvet, K.T., Tessier-Lavigne, M., Sotelo, C., Chédotal, A., 2008.Molecular mechanisms controlling midline crossing by precerebellar neurons.J. Neurosci. 28, 6285–6294.

Fazeli, A., Dickinson, S.L., Hermiston, M.L., Tighe, R.V., Steen, R.G., Small, C.G., Stoeckli,E.T., Keino-Masu, K., Masu, M., Rayburn, H., Simons, J., Bronson, R.T., Gordon, J.I.,Tessier-Lavigne, M., Weinberg, R.A., 1997. Phenotype of mice lacking functionalDeleted in colorectal cancer (Dcc) gene. Nature 386, 796–804.

Furne, C., Rama, N., Corset, V., Chédotal, A., Mehlen, P., 2008. Netrin-1 is a survival factorduring commissural neuron navigation. Proc. Natl. Acad. Sci. U. S. A. 105 (38),14465–14470.

Engelkamp, D., Rashbass, P., Seawright, A., van Heyningen, V., 1999. Role of Pax6 indevelopment of the cerebellar system. Development 126, 3585–3596.

Gilthorpe, J.D., Papantoniou, E.K., Chedotal, A., Lumsden, A., Wingate, R.J., 2002. Themigration of cerebellar rhombic lip derivatives. Development 129, 4719–4728.

Please cite this article as: Marcos, S., et al., Differential roles of Netrin-1Neurosci. (2009), doi:10.1016/j.mcn.2009.04.008

Kennedy, T.E., Serafini, T., de la Torre, J.R., Tessier-Lavigne, M.,1994. Netrins are diffusiblechemotropic factors for commissural axons in the embryonic spinal cord. Cell 78,425–435.

Llambi, F., Causeret, F., Bloch-Gallego, E., Mehlen, P., 2001. Netrin-1 acts as a survivalfactor via its receptors UNC5H and DCC. EMBO J. 20, 2715–2722.

Liu, G., Li, W., Wang, L., Kar, A., Guan, K.L., Rao, Y., Wu, J.Y., 2009. DSCAM functions as anetrin receptor in commissural axon pathfinding. Proc. Natl. Acad. Sci. U. S. A. 106(8), 2951–2956.

Ly, A., Nikolaev, A., Suresh, G., Zheng, Y., Tessier-Lavigne, M., Stein, E., 2008. DSCAM is anetrin receptor that collaborates with DCC in mediating turning responses tonetrin-1. Cell 133 (7), 1241–1254.

Marillat, V., Sabatier, C., Failli, V., Matsunaga, E., Sotelo, C., Tessier-Lavigne, M., Chedotal,A., 2004. The slit receptor Rig-1/Robo3 controls midline crossing by hindbrainprecerebellar neurons and axons. Neuron 43, 69–79.

Rakic, P., 1990. Principles of neural cell migration. Experientia 46, 882–891.Sabatier, C., Plump, A.S., Ma, Le, Brose, K., Tamada, A., Murakami, F., Lee, E.Y., Tessier-

Lavigne, M., 2004. The divergent Robo family protein rig-1/Robo3 is a negativeregulator of slit responsiveness required for midline crossing by commissuralaxons. Cell 117 (2), 157–169.

Serafini, T., Colamarino, S.A., Leonardo, E.D., Wang, H., Beddington, R., Skarnes, W.C.,Tessier-Lavigne, M.,1996. Netrin-1 is required for commissural axon guidance in thedeveloping vertebrate nervous system. Cell 87, 1001–1014.

Sherrard, R.M., Bower, A.J., 2002. Climbing fiber development: do neurotrophins have apart to play? Cerebellum 1, 265–275.

Stein, E., Tessier-Lavigne, M., 2001. Hierarchical organization of guidance receptors:silencing of netrin attraction by slit through a Robo/DCC receptor complex. Science291, 1928–1938.

Studer, M., 2001. Initiation of facial motoneurone migration is dependent onrhombomeres 5 and 6. Development 128, 3707–3716.

Taniguchi, H., Tamada, A., Kennedy, T.E., Murakami, F., 2002. Crossing the ventralmidline causes neurons to change their response to floor plate and alar plateattractive cues during transmedian migration. Dev. Biol. 249, 321–332.

Williams, M.E., Lu, X., McKenna, W.L., Washington, R., Boyette, A., Strickland, P., Dillon,A., Kaprielian, Z., Tessier-Lavigne, M., Hinck, L., 2006. UNC5A promotes neuronalapoptosis during spinal cord development independent of netrin-1. Nat. Neurosci.9, 996–998.

Yee, K.T., Simon, H.H., Tessier-Lavigne, M., O'Leary, D.M., 1999. Extension of long leadingprocesses and neuronal migration in the mammalian brain directed by thechemoattractant netrin-1. Neuron 24, 607–622.

Zhu, Y., Khan, K., Guthrie, S., 2003. Signals from the cerebellum guide the pathfinding ofinferior olivary axons. Dev. Biol. 257, 233–248.

and its receptor DCC in inferior olivary neuron migration, Mol. Cell.